ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.121 by root, Mon Jan 28 12:13:54 2008 UTC vs.
Revision 1.462 by root, Wed Jan 22 13:06:53 2020 UTC

		1	=encoding utf-8
		2
1	=head1 NAME	3	=head1 NAME
2		4
3	libev - a high performance full-featured event loop written in C	5	libev - a high performance full-featured event loop written in C
4		6
5	=head1 SYNOPSIS	7	=head1 SYNOPSIS
6		8
7	#include <ev.h>	9	#include <ev.h>
8		10
9	=head2 EXAMPLE PROGRAM	11	=head2 EXAMPLE PROGRAM
10		12
		13	// a single header file is required
11	#include <ev.h>	14	#include <ev.h>
12		15
		16	#include <stdio.h> // for puts
		17
		18	// every watcher type has its own typedef'd struct
		19	// with the name ev_TYPE
13	ev_io stdin_watcher;	20	ev_io stdin_watcher;
14	ev_timer timeout_watcher;	21	ev_timer timeout_watcher;
15		22
16	/* called when data readable on stdin */	23	// all watcher callbacks have a similar signature
		24	// this callback is called when data is readable on stdin
17	static void	25	static void
18	stdin_cb (EV_P_ struct ev_io *w, int revents)	26	stdin_cb (EV_P_ ev_io *w, int revents)
19	{	27	{
20	/* puts ("stdin ready"); */	28	puts ("stdin ready");
21	ev_io_stop (EV_A_ w); /* just a syntax example */	29	// for one-shot events, one must manually stop the watcher
22	ev_unloop (EV_A_ EVUNLOOP_ALL); /* leave all loop calls */	30	// with its corresponding stop function.
		31	ev_io_stop (EV_A_ w);
		32
		33	// this causes all nested ev_run's to stop iterating
		34	ev_break (EV_A_ EVBREAK_ALL);
23	}	35	}
24		36
		37	// another callback, this time for a time-out
25	static void	38	static void
26	timeout_cb (EV_P_ struct ev_timer *w, int revents)	39	timeout_cb (EV_P_ ev_timer *w, int revents)
27	{	40	{
28	/* puts ("timeout"); */	41	puts ("timeout");
29	ev_unloop (EV_A_ EVUNLOOP_ONE); /* leave one loop call */	42	// this causes the innermost ev_run to stop iterating
		43	ev_break (EV_A_ EVBREAK_ONE);
30	}	44	}
31		45
32	int	46	int
33	main (void)	47	main (void)
34	{	48	{
35	struct ev_loop *loop = ev_default_loop (0);	49	// use the default event loop unless you have special needs
		50	struct ev_loop *loop = EV_DEFAULT;
36		51
37	/* initialise an io watcher, then start it */	52	// initialise an io watcher, then start it
		53	// this one will watch for stdin to become readable
38	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);	54	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
39	ev_io_start (loop, &stdin_watcher);	55	ev_io_start (loop, &stdin_watcher);
40		56
		57	// initialise a timer watcher, then start it
41	/* simple non-repeating 5.5 second timeout */	58	// simple non-repeating 5.5 second timeout
42	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);	59	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
43	ev_timer_start (loop, &timeout_watcher);	60	ev_timer_start (loop, &timeout_watcher);
44		61
45	/* loop till timeout or data ready */	62	// now wait for events to arrive
46	ev_loop (loop, 0);	63	ev_run (loop, 0);
47		64
		65	// break was called, so exit
48	return 0;	66	return 0;
49	}	67	}
50		68
51	=head1 DESCRIPTION	69	=head1 ABOUT THIS DOCUMENT
52		70
		71	This document documents the libev software package.
		72
53	The newest version of this document is also available as a html-formatted	73	The newest version of this document is also available as an html-formatted
54	web page you might find easier to navigate when reading it for the first	74	web page you might find easier to navigate when reading it for the first
55	time: L<http://cvs.schmorp.de/libev/ev.html>.	75	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		76
		77	While this document tries to be as complete as possible in documenting
		78	libev, its usage and the rationale behind its design, it is not a tutorial
		79	on event-based programming, nor will it introduce event-based programming
		80	with libev.
		81
		82	Familiarity with event based programming techniques in general is assumed
		83	throughout this document.
		84
		85	=head1 WHAT TO READ WHEN IN A HURRY
		86
		87	This manual tries to be very detailed, but unfortunately, this also makes
		88	it very long. If you just want to know the basics of libev, I suggest
		89	reading L</ANATOMY OF A WATCHER>, then the L</EXAMPLE PROGRAM> above and
		90	look up the missing functions in L</GLOBAL FUNCTIONS> and the C<ev_io> and
		91	C<ev_timer> sections in L</WATCHER TYPES>.
		92
		93	=head1 ABOUT LIBEV
56		94
57	Libev is an event loop: you register interest in certain events (such as a	95	Libev is an event loop: you register interest in certain events (such as a
58	file descriptor being readable or a timeout occurring), and it will manage	96	file descriptor being readable or a timeout occurring), and it will manage
59	these event sources and provide your program with events.	97	these event sources and provide your program with events.
60		98
…		…
67	details of the event, and then hand it over to libev by I<starting> the	105	details of the event, and then hand it over to libev by I<starting> the
68	watcher.	106	watcher.
69		107
70	=head2 FEATURES	108	=head2 FEATURES
71		109
72	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	110	Libev supports C<select>, C<poll>, the Linux-specific aio and C<epoll>
73	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	111	interfaces, the BSD-specific C<kqueue> and the Solaris-specific event port
74	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	112	mechanisms for file descriptor events (C<ev_io>), the Linux C<inotify>
75	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	113	interface (for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
76	with customised rescheduling (C<ev_periodic>), synchronous signals	114	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
77	(C<ev_signal>), process status change events (C<ev_child>), and event	115	timers (C<ev_timer>), absolute timers with customised rescheduling
78	watchers dealing with the event loop mechanism itself (C<ev_idle>,	116	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
79	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as	117	change events (C<ev_child>), and event watchers dealing with the event
80	file watchers (C<ev_stat>) and even limited support for fork events	118	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
81	(C<ev_fork>).	119	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		120	limited support for fork events (C<ev_fork>).
82		121
83	It also is quite fast (see this	122	It also is quite fast (see this
84	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	123	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
85	for example).	124	for example).
86		125
87	=head2 CONVENTIONS	126	=head2 CONVENTIONS
88		127
89	Libev is very configurable. In this manual the default configuration will	128	Libev is very configurable. In this manual the default (and most common)
90	be described, which supports multiple event loops. For more info about	129	configuration will be described, which supports multiple event loops. For
91	various configuration options please have a look at B<EMBED> section in	130	more info about various configuration options please have a look at
92	this manual. If libev was configured without support for multiple event	131	B<EMBED> section in this manual. If libev was configured without support
93	loops, then all functions taking an initial argument of name C<loop>	132	for multiple event loops, then all functions taking an initial argument of
94	(which is always of type C<struct ev_loop *>) will not have this argument.	133	name C<loop> (which is always of type C<struct ev_loop *>) will not have
		134	this argument.
95		135
96	=head2 TIME REPRESENTATION	136	=head2 TIME REPRESENTATION
97		137
98	Libev represents time as a single floating point number, representing the	138	Libev represents time as a single floating point number, representing
99	(fractional) number of seconds since the (POSIX) epoch (somewhere near	139	the (fractional) number of seconds since the (POSIX) epoch (in practice
100	the beginning of 1970, details are complicated, don't ask). This type is	140	somewhere near the beginning of 1970, details are complicated, don't
101	called C<ev_tstamp>, which is what you should use too. It usually aliases	141	ask). This type is called C<ev_tstamp>, which is what you should use
102	to the C<double> type in C, and when you need to do any calculations on	142	too. It usually aliases to the C<double> type in C. When you need to do
103	it, you should treat it as some floatingpoint value. Unlike the name	143	any calculations on it, you should treat it as some floating point value.
		144
104	component C<stamp> might indicate, it is also used for time differences	145	Unlike the name component C<stamp> might indicate, it is also used for
105	throughout libev.	146	time differences (e.g. delays) throughout libev.
		147
		148	=head1 ERROR HANDLING
		149
		150	Libev knows three classes of errors: operating system errors, usage errors
		151	and internal errors (bugs).
		152
		153	When libev catches an operating system error it cannot handle (for example
		154	a system call indicating a condition libev cannot fix), it calls the callback
		155	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		156	abort. The default is to print a diagnostic message and to call C<abort
		157	()>.
		158
		159	When libev detects a usage error such as a negative timer interval, then
		160	it will print a diagnostic message and abort (via the C<assert> mechanism,
		161	so C<NDEBUG> will disable this checking): these are programming errors in
		162	the libev caller and need to be fixed there.
		163
		164	Via the C<EV_FREQUENT> macro you can compile in and/or enable extensive
		165	consistency checking code inside libev that can be used to check for
		166	internal inconsistencies, suually caused by application bugs.
		167
		168	Libev also has a few internal error-checking C<assert>ions. These do not
		169	trigger under normal circumstances, as they indicate either a bug in libev
		170	or worse.
		171
106		172
107	=head1 GLOBAL FUNCTIONS	173	=head1 GLOBAL FUNCTIONS
108		174
109	These functions can be called anytime, even before initialising the	175	These functions can be called anytime, even before initialising the
110	library in any way.	176	library in any way.
…		…
113		179
114	=item ev_tstamp ev_time ()	180	=item ev_tstamp ev_time ()
115		181
116	Returns the current time as libev would use it. Please note that the	182	Returns the current time as libev would use it. Please note that the
117	C<ev_now> function is usually faster and also often returns the timestamp	183	C<ev_now> function is usually faster and also often returns the timestamp
118	you actually want to know.	184	you actually want to know. Also interesting is the combination of
		185	C<ev_now_update> and C<ev_now>.
119		186
120	=item ev_sleep (ev_tstamp interval)	187	=item ev_sleep (ev_tstamp interval)
121		188
122	Sleep for the given interval: The current thread will be blocked until	189	Sleep for the given interval: The current thread will be blocked
123	either it is interrupted or the given time interval has passed. Basically	190	until either it is interrupted or the given time interval has
		191	passed (approximately - it might return a bit earlier even if not
		192	interrupted). Returns immediately if C<< interval <= 0 >>.
		193
124	this is a subsecond-resolution C<sleep ()>.	194	Basically this is a sub-second-resolution C<sleep ()>.
		195
		196	The range of the C<interval> is limited - libev only guarantees to work
		197	with sleep times of up to one day (C<< interval <= 86400 >>).
125		198
126	=item int ev_version_major ()	199	=item int ev_version_major ()
127		200
128	=item int ev_version_minor ()	201	=item int ev_version_minor ()
129		202
…		…
140	as this indicates an incompatible change. Minor versions are usually	213	as this indicates an incompatible change. Minor versions are usually
141	compatible to older versions, so a larger minor version alone is usually	214	compatible to older versions, so a larger minor version alone is usually
142	not a problem.	215	not a problem.
143		216
144	Example: Make sure we haven't accidentally been linked against the wrong	217	Example: Make sure we haven't accidentally been linked against the wrong
145	version.	218	version (note, however, that this will not detect other ABI mismatches,
		219	such as LFS or reentrancy).
146		220
147	assert (("libev version mismatch",	221	assert (("libev version mismatch",
148	ev_version_major () == EV_VERSION_MAJOR	222	ev_version_major () == EV_VERSION_MAJOR
149	&& ev_version_minor () >= EV_VERSION_MINOR));	223	&& ev_version_minor () >= EV_VERSION_MINOR));
150		224
151	=item unsigned int ev_supported_backends ()	225	=item unsigned int ev_supported_backends ()
152		226
153	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	227	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
154	value) compiled into this binary of libev (independent of their	228	value) compiled into this binary of libev (independent of their
…		…
156	a description of the set values.	230	a description of the set values.
157		231
158	Example: make sure we have the epoll method, because yeah this is cool and	232	Example: make sure we have the epoll method, because yeah this is cool and
159	a must have and can we have a torrent of it please!!!11	233	a must have and can we have a torrent of it please!!!11
160		234
161	assert (("sorry, no epoll, no sex",	235	assert (("sorry, no epoll, no sex",
162	ev_supported_backends () & EVBACKEND_EPOLL));	236	ev_supported_backends () & EVBACKEND_EPOLL));
163		237
164	=item unsigned int ev_recommended_backends ()	238	=item unsigned int ev_recommended_backends ()
165		239
166	Return the set of all backends compiled into this binary of libev and also	240	Return the set of all backends compiled into this binary of libev and
167	recommended for this platform. This set is often smaller than the one	241	also recommended for this platform, meaning it will work for most file
		242	descriptor types. This set is often smaller than the one returned by
168	returned by C<ev_supported_backends>, as for example kqueue is broken on	243	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
169	most BSDs and will not be autodetected unless you explicitly request it	244	and will not be auto-detected unless you explicitly request it (assuming
170	(assuming you know what you are doing). This is the set of backends that	245	you know what you are doing). This is the set of backends that libev will
171	libev will probe for if you specify no backends explicitly.	246	probe for if you specify no backends explicitly.
172		247
173	=item unsigned int ev_embeddable_backends ()	248	=item unsigned int ev_embeddable_backends ()
174		249
175	Returns the set of backends that are embeddable in other event loops. This	250	Returns the set of backends that are embeddable in other event loops. This
176	is the theoretical, all-platform, value. To find which backends	251	value is platform-specific but can include backends not available on the
177	might be supported on the current system, you would need to look at	252	current system. To find which embeddable backends might be supported on
178	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	253	the current system, you would need to look at C<ev_embeddable_backends ()
179	recommended ones.	254	& ev_supported_backends ()>, likewise for recommended ones.
180		255
181	See the description of C<ev_embed> watchers for more info.	256	See the description of C<ev_embed> watchers for more info.
182		257
183	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	258	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
184		259
185	Sets the allocation function to use (the prototype is similar - the	260	Sets the allocation function to use (the prototype is similar - the
186	semantics is identical - to the realloc C function). It is used to	261	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
187	allocate and free memory (no surprises here). If it returns zero when	262	used to allocate and free memory (no surprises here). If it returns zero
188	memory needs to be allocated, the library might abort or take some	263	when memory needs to be allocated (C<size != 0>), the library might abort
189	potentially destructive action. The default is your system realloc	264	or take some potentially destructive action.
190	function.	265
		266	Since some systems (at least OpenBSD and Darwin) fail to implement
		267	correct C<realloc> semantics, libev will use a wrapper around the system
		268	C<realloc> and C<free> functions by default.
191		269
192	You could override this function in high-availability programs to, say,	270	You could override this function in high-availability programs to, say,
193	free some memory if it cannot allocate memory, to use a special allocator,	271	free some memory if it cannot allocate memory, to use a special allocator,
194	or even to sleep a while and retry until some memory is available.	272	or even to sleep a while and retry until some memory is available.
195		273
		274	Example: The following is the C<realloc> function that libev itself uses
		275	which should work with C<realloc> and C<free> functions of all kinds and
		276	is probably a good basis for your own implementation.
		277
		278	static void *
		279	ev_realloc_emul (void *ptr, long size) EV_NOEXCEPT
		280	{
		281	if (size)
		282	return realloc (ptr, size);
		283
		284	free (ptr);
		285	return 0;
		286	}
		287
196	Example: Replace the libev allocator with one that waits a bit and then	288	Example: Replace the libev allocator with one that waits a bit and then
197	retries).	289	retries.
198		290
199	static void *	291	static void *
200	persistent_realloc (void *ptr, size_t size)	292	persistent_realloc (void *ptr, size_t size)
201	{	293	{
		294	if (!size)
		295	{
		296	free (ptr);
		297	return 0;
		298	}
		299
202	for (;;)	300	for (;;)
203	{	301	{
204	void *newptr = realloc (ptr, size);	302	void *newptr = realloc (ptr, size);
205		303
206	if (newptr)	304	if (newptr)
…		…
211	}	309	}
212		310
213	...	311	...
214	ev_set_allocator (persistent_realloc);	312	ev_set_allocator (persistent_realloc);
215		313
216	=item ev_set_syserr_cb (void (*cb)(const char *msg));	314	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
217		315
218	Set the callback function to call on a retryable syscall error (such	316	Set the callback function to call on a retryable system call error (such
219	as failed select, poll, epoll_wait). The message is a printable string	317	as failed select, poll, epoll_wait). The message is a printable string
220	indicating the system call or subsystem causing the problem. If this	318	indicating the system call or subsystem causing the problem. If this
221	callback is set, then libev will expect it to remedy the sitution, no	319	callback is set, then libev will expect it to remedy the situation, no
222	matter what, when it returns. That is, libev will generally retry the	320	matter what, when it returns. That is, libev will generally retry the
223	requested operation, or, if the condition doesn't go away, do bad stuff	321	requested operation, or, if the condition doesn't go away, do bad stuff
224	(such as abort).	322	(such as abort).
225		323
226	Example: This is basically the same thing that libev does internally, too.	324	Example: This is basically the same thing that libev does internally, too.
…		…
233	}	331	}
234		332
235	...	333	...
236	ev_set_syserr_cb (fatal_error);	334	ev_set_syserr_cb (fatal_error);
237		335
		336	=item ev_feed_signal (int signum)
		337
		338	This function can be used to "simulate" a signal receive. It is completely
		339	safe to call this function at any time, from any context, including signal
		340	handlers or random threads.
		341
		342	Its main use is to customise signal handling in your process, especially
		343	in the presence of threads. For example, you could block signals
		344	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		345	creating any loops), and in one thread, use C<sigwait> or any other
		346	mechanism to wait for signals, then "deliver" them to libev by calling
		347	C<ev_feed_signal>.
		348
238	=back	349	=back
239		350
240	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	351	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
241		352
242	An event loop is described by a C<struct ev_loop *>. The library knows two	353	An event loop is described by a C<struct ev_loop *> (the C<struct> is
243	types of such loops, the I<default> loop, which supports signals and child	354	I<not> optional in this case unless libev 3 compatibility is disabled, as
244	events, and dynamically created loops which do not.	355	libev 3 had an C<ev_loop> function colliding with the struct name).
245		356
246	If you use threads, a common model is to run the default event loop	357	The library knows two types of such loops, the I<default> loop, which
247	in your main thread (or in a separate thread) and for each thread you	358	supports child process events, and dynamically created event loops which
248	create, you also create another event loop. Libev itself does no locking	359	do not.
249	whatsoever, so if you mix calls to the same event loop in different
250	threads, make sure you lock (this is usually a bad idea, though, even if
251	done correctly, because it's hideous and inefficient).
252		360
253	=over 4	361	=over 4
254		362
255	=item struct ev_loop *ev_default_loop (unsigned int flags)	363	=item struct ev_loop *ev_default_loop (unsigned int flags)
256		364
257	This will initialise the default event loop if it hasn't been initialised	365	This returns the "default" event loop object, which is what you should
258	yet and return it. If the default loop could not be initialised, returns	366	normally use when you just need "the event loop". Event loop objects and
259	false. If it already was initialised it simply returns it (and ignores the	367	the C<flags> parameter are described in more detail in the entry for
260	flags. If that is troubling you, check C<ev_backend ()> afterwards).	368	C<ev_loop_new>.
		369
		370	If the default loop is already initialised then this function simply
		371	returns it (and ignores the flags. If that is troubling you, check
		372	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		373	flags, which should almost always be C<0>, unless the caller is also the
		374	one calling C<ev_run> or otherwise qualifies as "the main program".
261		375
262	If you don't know what event loop to use, use the one returned from this	376	If you don't know what event loop to use, use the one returned from this
263	function.	377	function (or via the C<EV_DEFAULT> macro).
264		378
		379	Note that this function is I<not> thread-safe, so if you want to use it
		380	from multiple threads, you have to employ some kind of mutex (note also
		381	that this case is unlikely, as loops cannot be shared easily between
		382	threads anyway).
		383
265	The default loop is the only loop that can handle C<ev_signal> and	384	The default loop is the only loop that can handle C<ev_child> watchers,
266	C<ev_child> watchers, and to do this, it always registers a handler	385	and to do this, it always registers a handler for C<SIGCHLD>. If this is
267	for C<SIGCHLD>. If this is a problem for your app you can either	386	a problem for your application you can either create a dynamic loop with
268	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	387	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
269	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	388	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
270	C<ev_default_init>.	389
		390	Example: This is the most typical usage.
		391
		392	if (!ev_default_loop (0))
		393	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		394
		395	Example: Restrict libev to the select and poll backends, and do not allow
		396	environment settings to be taken into account:
		397
		398	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		399
		400	=item struct ev_loop *ev_loop_new (unsigned int flags)
		401
		402	This will create and initialise a new event loop object. If the loop
		403	could not be initialised, returns false.
		404
		405	This function is thread-safe, and one common way to use libev with
		406	threads is indeed to create one loop per thread, and using the default
		407	loop in the "main" or "initial" thread.
271		408
272	The flags argument can be used to specify special behaviour or specific	409	The flags argument can be used to specify special behaviour or specific
273	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	410	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
274		411
275	The following flags are supported:	412	The following flags are supported:
…		…
281	The default flags value. Use this if you have no clue (it's the right	418	The default flags value. Use this if you have no clue (it's the right
282	thing, believe me).	419	thing, believe me).
283		420
284	=item C<EVFLAG_NOENV>	421	=item C<EVFLAG_NOENV>
285		422
286	If this flag bit is ored into the flag value (or the program runs setuid	423	If this flag bit is or'ed into the flag value (or the program runs setuid
287	or setgid) then libev will I<not> look at the environment variable	424	or setgid) then libev will I<not> look at the environment variable
288	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	425	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
289	override the flags completely if it is found in the environment. This is	426	override the flags completely if it is found in the environment. This is
290	useful to try out specific backends to test their performance, or to work	427	useful to try out specific backends to test their performance, to work
291	around bugs.	428	around bugs, or to make libev threadsafe (accessing environment variables
		429	cannot be done in a threadsafe way, but usually it works if no other
		430	thread modifies them).
292		431
293	=item C<EVFLAG_FORKCHECK>	432	=item C<EVFLAG_FORKCHECK>
294		433
295	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	434	Instead of calling C<ev_loop_fork> manually after a fork, you can also
296	a fork, you can also make libev check for a fork in each iteration by	435	make libev check for a fork in each iteration by enabling this flag.
297	enabling this flag.
298		436
299	This works by calling C<getpid ()> on every iteration of the loop,	437	This works by calling C<getpid ()> on every iteration of the loop,
300	and thus this might slow down your event loop if you do a lot of loop	438	and thus this might slow down your event loop if you do a lot of loop
301	iterations and little real work, but is usually not noticeable (on my	439	iterations and little real work, but is usually not noticeable (on my
302	Linux system for example, C<getpid> is actually a simple 5-insn sequence	440	GNU/Linux system for example, C<getpid> is actually a simple 5-insn
303	without a syscall and thus I<very> fast, but my Linux system also has	441	sequence without a system call and thus I<very> fast, but my GNU/Linux
304	C<pthread_atfork> which is even faster).	442	system also has C<pthread_atfork> which is even faster). (Update: glibc
		443	versions 2.25 apparently removed the C<getpid> optimisation again).
305		444
306	The big advantage of this flag is that you can forget about fork (and	445	The big advantage of this flag is that you can forget about fork (and
307	forget about forgetting to tell libev about forking) when you use this	446	forget about forgetting to tell libev about forking, although you still
308	flag.	447	have to ignore C<SIGPIPE>) when you use this flag.
309		448
310	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>	449	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
311	environment variable.	450	environment variable.
		451
		452	=item C<EVFLAG_NOINOTIFY>
		453
		454	When this flag is specified, then libev will not attempt to use the
		455	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
		456	testing, this flag can be useful to conserve inotify file descriptors, as
		457	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		458
		459	=item C<EVFLAG_SIGNALFD>
		460
		461	When this flag is specified, then libev will attempt to use the
		462	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
		463	delivers signals synchronously, which makes it both faster and might make
		464	it possible to get the queued signal data. It can also simplify signal
		465	handling with threads, as long as you properly block signals in your
		466	threads that are not interested in handling them.
		467
		468	Signalfd will not be used by default as this changes your signal mask, and
		469	there are a lot of shoddy libraries and programs (glib's threadpool for
		470	example) that can't properly initialise their signal masks.
		471
		472	=item C<EVFLAG_NOSIGMASK>
		473
		474	When this flag is specified, then libev will avoid to modify the signal
		475	mask. Specifically, this means you have to make sure signals are unblocked
		476	when you want to receive them.
		477
		478	This behaviour is useful when you want to do your own signal handling, or
		479	want to handle signals only in specific threads and want to avoid libev
		480	unblocking the signals.
		481
		482	It's also required by POSIX in a threaded program, as libev calls
		483	C<sigprocmask>, whose behaviour is officially unspecified.
		484
		485	=item C<EVFLAG_NOTIMERFD>
		486
		487	When this flag is specified, the libev will avoid using a C<timerfd> to
		488	detect time jumps. It will still be able to detect time jumps, but takes
		489	longer and has a lower accuracy in doing so, but saves a file descriptor
		490	per loop.
		491
		492	The current implementation only tries to use a C<timerfd> when the first
		493	C<ev_periodic> watcher is started and falls back on other methods if it
		494	cannot be created, but this behaviour might change in the future.
312		495
313	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	496	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
314		497
315	This is your standard select(2) backend. Not I<completely> standard, as	498	This is your standard select(2) backend. Not I<completely> standard, as
316	libev tries to roll its own fd_set with no limits on the number of fds,	499	libev tries to roll its own fd_set with no limits on the number of fds,
317	but if that fails, expect a fairly low limit on the number of fds when	500	but if that fails, expect a fairly low limit on the number of fds when
318	using this backend. It doesn't scale too well (O(highest_fd)), but its	501	using this backend. It doesn't scale too well (O(highest_fd)), but its
319	usually the fastest backend for a low number of (low-numbered :) fds.	502	usually the fastest backend for a low number of (low-numbered :) fds.
320		503
321	To get good performance out of this backend you need a high amount of	504	To get good performance out of this backend you need a high amount of
322	parallelity (most of the file descriptors should be busy). If you are	505	parallelism (most of the file descriptors should be busy). If you are
323	writing a server, you should C<accept ()> in a loop to accept as many	506	writing a server, you should C<accept ()> in a loop to accept as many
324	connections as possible during one iteration. You might also want to have	507	connections as possible during one iteration. You might also want to have
325	a look at C<ev_set_io_collect_interval ()> to increase the amount of	508	a look at C<ev_set_io_collect_interval ()> to increase the amount of
326	readyness notifications you get per iteration.	509	readiness notifications you get per iteration.
		510
		511	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		512	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		513	C<exceptfds> set on that platform).
327		514
328	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	515	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
329		516
330	And this is your standard poll(2) backend. It's more complicated	517	And this is your standard poll(2) backend. It's more complicated
331	than select, but handles sparse fds better and has no artificial	518	than select, but handles sparse fds better and has no artificial
332	limit on the number of fds you can use (except it will slow down	519	limit on the number of fds you can use (except it will slow down
333	considerably with a lot of inactive fds). It scales similarly to select,	520	considerably with a lot of inactive fds). It scales similarly to select,
334	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	521	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
335	performance tips.	522	performance tips.
336		523
		524	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		525	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		526
337	=item C<EVBACKEND_EPOLL> (value 4, Linux)	527	=item C<EVBACKEND_EPOLL> (value 4, Linux)
338		528
		529	Use the Linux-specific epoll(7) interface (for both pre- and post-2.6.9
		530	kernels).
		531
339	For few fds, this backend is a bit little slower than poll and select,	532	For few fds, this backend is a bit little slower than poll and select, but
340	but it scales phenomenally better. While poll and select usually scale	533	it scales phenomenally better. While poll and select usually scale like
341	like O(total_fds) where n is the total number of fds (or the highest fd),	534	O(total_fds) where total_fds is the total number of fds (or the highest
342	epoll scales either O(1) or O(active_fds). The epoll design has a number	535	fd), epoll scales either O(1) or O(active_fds).
343	of shortcomings, such as silently dropping events in some hard-to-detect	536
344	cases and rewiring a syscall per fd change, no fork support and bad	537	The epoll mechanism deserves honorable mention as the most misdesigned
345	support for dup.	538	of the more advanced event mechanisms: mere annoyances include silently
		539	dropping file descriptors, requiring a system call per change per file
		540	descriptor (and unnecessary guessing of parameters), problems with dup,
		541	returning before the timeout value, resulting in additional iterations
		542	(and only giving 5ms accuracy while select on the same platform gives
		543	0.1ms) and so on. The biggest issue is fork races, however - if a program
		544	forks then I<both> parent and child process have to recreate the epoll
		545	set, which can take considerable time (one syscall per file descriptor)
		546	and is of course hard to detect.
		547
		548	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
		549	but of course I<doesn't>, and epoll just loves to report events for
		550	totally I<different> file descriptors (even already closed ones, so
		551	one cannot even remove them from the set) than registered in the set
		552	(especially on SMP systems). Libev tries to counter these spurious
		553	notifications by employing an additional generation counter and comparing
		554	that against the events to filter out spurious ones, recreating the set
		555	when required. Epoll also erroneously rounds down timeouts, but gives you
		556	no way to know when and by how much, so sometimes you have to busy-wait
		557	because epoll returns immediately despite a nonzero timeout. And last
		558	not least, it also refuses to work with some file descriptors which work
		559	perfectly fine with C<select> (files, many character devices...).
		560
		561	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		562	cobbled together in a hurry, no thought to design or interaction with
		563	others. Oh, the pain, will it ever stop...
346		564
347	While stopping, setting and starting an I/O watcher in the same iteration	565	While stopping, setting and starting an I/O watcher in the same iteration
348	will result in some caching, there is still a syscall per such incident	566	will result in some caching, there is still a system call per such
349	(because the fd could point to a different file description now), so its	567	incident (because the same I<file descriptor> could point to a different
350	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	568	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
351	very well if you register events for both fds.	569	file descriptors might not work very well if you register events for both
352		570	file descriptors.
353	Please note that epoll sometimes generates spurious notifications, so you
354	need to use non-blocking I/O or other means to avoid blocking when no data
355	(or space) is available.
356		571
357	Best performance from this backend is achieved by not unregistering all	572	Best performance from this backend is achieved by not unregistering all
358	watchers for a file descriptor until it has been closed, if possible, i.e.	573	watchers for a file descriptor until it has been closed, if possible,
359	keep at least one watcher active per fd at all times.	574	i.e. keep at least one watcher active per fd at all times. Stopping and
		575	starting a watcher (without re-setting it) also usually doesn't cause
		576	extra overhead. A fork can both result in spurious notifications as well
		577	as in libev having to destroy and recreate the epoll object, which can
		578	take considerable time and thus should be avoided.
360		579
		580	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		581	faster than epoll for maybe up to a hundred file descriptors, depending on
		582	the usage. So sad.
		583
361	While nominally embeddeble in other event loops, this feature is broken in	584	While nominally embeddable in other event loops, this feature is broken in
362	all kernel versions tested so far.	585	a lot of kernel revisions, but probably(!) works in current versions.
		586
		587	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		588	C<EVBACKEND_POLL>.
		589
		590	=item C<EVBACKEND_LINUXAIO> (value 64, Linux)
		591
		592	Use the Linux-specific Linux AIO (I<not> C<< aio(7) >> but C<<
		593	io_submit(2) >>) event interface available in post-4.18 kernels (but libev
		594	only tries to use it in 4.19+).
		595
		596	This is another Linux train wreck of an event interface.
		597
		598	If this backend works for you (as of this writing, it was very
		599	experimental), it is the best event interface available on Linux and might
		600	be well worth enabling it - if it isn't available in your kernel this will
		601	be detected and this backend will be skipped.
		602
		603	This backend can batch oneshot requests and supports a user-space ring
		604	buffer to receive events. It also doesn't suffer from most of the design
		605	problems of epoll (such as not being able to remove event sources from
		606	the epoll set), and generally sounds too good to be true. Because, this
		607	being the Linux kernel, of course it suffers from a whole new set of
		608	limitations, forcing you to fall back to epoll, inheriting all its design
		609	issues.
		610
		611	For one, it is not easily embeddable (but probably could be done using
		612	an event fd at some extra overhead). It also is subject to a system wide
		613	limit that can be configured in F</proc/sys/fs/aio-max-nr>. If no AIO
		614	requests are left, this backend will be skipped during initialisation, and
		615	will switch to epoll when the loop is active.
		616
		617	Most problematic in practice, however, is that not all file descriptors
		618	work with it. For example, in Linux 5.1, TCP sockets, pipes, event fds,
		619	files, F</dev/null> and many others are supported, but ttys do not work
		620	properly (a known bug that the kernel developers don't care about, see
		621	L<https://lore.kernel.org/patchwork/patch/1047453/>), so this is not
		622	(yet?) a generic event polling interface.
		623
		624	Overall, it seems the Linux developers just don't want it to have a
		625	generic event handling mechanism other than C<select> or C<poll>.
		626
		627	To work around all these problem, the current version of libev uses its
		628	epoll backend as a fallback for file descriptor types that do not work. Or
		629	falls back completely to epoll if the kernel acts up.
		630
		631	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		632	C<EVBACKEND_POLL>.
363		633
364	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	634	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
365		635
366	Kqueue deserves special mention, as at the time of this writing, it	636	Kqueue deserves special mention, as at the time this backend was
367	was broken on all BSDs except NetBSD (usually it doesn't work reliably	637	implemented, it was broken on all BSDs except NetBSD (usually it doesn't
368	with anything but sockets and pipes, except on Darwin, where of course	638	work reliably with anything but sockets and pipes, except on Darwin,
369	it's completely useless). For this reason it's not being "autodetected"	639	where of course it's completely useless). Unlike epoll, however, whose
370	unless you explicitly specify it explicitly in the flags (i.e. using	640	brokenness is by design, these kqueue bugs can be (and mostly have been)
371	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	641	fixed without API changes to existing programs. For this reason it's not
372	system like NetBSD.	642	being "auto-detected" on all platforms unless you explicitly specify it
		643	in the flags (i.e. using C<EVBACKEND_KQUEUE>) or libev was compiled on a
		644	known-to-be-good (-enough) system like NetBSD.
373		645
374	You still can embed kqueue into a normal poll or select backend and use it	646	You still can embed kqueue into a normal poll or select backend and use it
375	only for sockets (after having made sure that sockets work with kqueue on	647	only for sockets (after having made sure that sockets work with kqueue on
376	the target platform). See C<ev_embed> watchers for more info.	648	the target platform). See C<ev_embed> watchers for more info.
377		649
378	It scales in the same way as the epoll backend, but the interface to the	650	It scales in the same way as the epoll backend, but the interface to the
379	kernel is more efficient (which says nothing about its actual speed, of	651	kernel is more efficient (which says nothing about its actual speed, of
380	course). While stopping, setting and starting an I/O watcher does never	652	course). While stopping, setting and starting an I/O watcher does never
381	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to	653	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
382	two event changes per incident, support for C<fork ()> is very bad and it	654	two event changes per incident. Support for C<fork ()> is very bad (you
		655	might have to leak fds on fork, but it's more sane than epoll) and it
383	drops fds silently in similarly hard-to-detect cases.	656	drops fds silently in similarly hard-to-detect cases.
384		657
385	This backend usually performs well under most conditions.	658	This backend usually performs well under most conditions.
386		659
387	While nominally embeddable in other event loops, this doesn't work	660	While nominally embeddable in other event loops, this doesn't work
388	everywhere, so you might need to test for this. And since it is broken	661	everywhere, so you might need to test for this. And since it is broken
389	almost everywhere, you should only use it when you have a lot of sockets	662	almost everywhere, you should only use it when you have a lot of sockets
390	(for which it usually works), by embedding it into another event loop	663	(for which it usually works), by embedding it into another event loop
391	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	664	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
392	sockets.	665	also broken on OS X)) and, did I mention it, using it only for sockets.
		666
		667	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		668	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		669	C<NOTE_EOF>.
393		670
394	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	671	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
395		672
396	This is not implemented yet (and might never be, unless you send me an	673	This is not implemented yet (and might never be, unless you send me an
397	implementation). According to reports, C</dev/poll> only supports sockets	674	implementation). According to reports, C</dev/poll> only supports sockets
…		…
401	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	678	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
402		679
403	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	680	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
404	it's really slow, but it still scales very well (O(active_fds)).	681	it's really slow, but it still scales very well (O(active_fds)).
405		682
406	Please note that solaris event ports can deliver a lot of spurious
407	notifications, so you need to use non-blocking I/O or other means to avoid
408	blocking when no data (or space) is available.
409
410	While this backend scales well, it requires one system call per active	683	While this backend scales well, it requires one system call per active
411	file descriptor per loop iteration. For small and medium numbers of file	684	file descriptor per loop iteration. For small and medium numbers of file
412	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	685	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
413	might perform better.	686	might perform better.
414		687
415	On the positive side, ignoring the spurious readyness notifications, this	688	On the positive side, this backend actually performed fully to
416	backend actually performed to specification in all tests and is fully	689	specification in all tests and is fully embeddable, which is a rare feat
417	embeddable, which is a rare feat among the OS-specific backends.	690	among the OS-specific backends (I vastly prefer correctness over speed
		691	hacks).
		692
		693	On the negative side, the interface is I<bizarre> - so bizarre that
		694	even sun itself gets it wrong in their code examples: The event polling
		695	function sometimes returns events to the caller even though an error
		696	occurred, but with no indication whether it has done so or not (yes, it's
		697	even documented that way) - deadly for edge-triggered interfaces where you
		698	absolutely have to know whether an event occurred or not because you have
		699	to re-arm the watcher.
		700
		701	Fortunately libev seems to be able to work around these idiocies.
		702
		703	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		704	C<EVBACKEND_POLL>.
418		705
419	=item C<EVBACKEND_ALL>	706	=item C<EVBACKEND_ALL>
420		707
421	Try all backends (even potentially broken ones that wouldn't be tried	708	Try all backends (even potentially broken ones that wouldn't be tried
422	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	709	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
423	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	710	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
424		711
425	It is definitely not recommended to use this flag.	712	It is definitely not recommended to use this flag, use whatever
		713	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		714	at all.
		715
		716	=item C<EVBACKEND_MASK>
		717
		718	Not a backend at all, but a mask to select all backend bits from a
		719	C<flags> value, in case you want to mask out any backends from a flags
		720	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
426		721
427	=back	722	=back
428		723
429	If one or more of these are ored into the flags value, then only these	724	If one or more of the backend flags are or'ed into the flags value,
430	backends will be tried (in the reverse order as listed here). If none are	725	then only these backends will be tried (in the reverse order as listed
431	specified, all backends in C<ev_recommended_backends ()> will be tried.	726	here). If none are specified, all backends in C<ev_recommended_backends
432		727	()> will be tried.
433	The most typical usage is like this:
434
435	if (!ev_default_loop (0))
436	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
437
438	Restrict libev to the select and poll backends, and do not allow
439	environment settings to be taken into account:
440
441	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
442
443	Use whatever libev has to offer, but make sure that kqueue is used if
444	available (warning, breaks stuff, best use only with your own private
445	event loop and only if you know the OS supports your types of fds):
446
447	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
448
449	=item struct ev_loop *ev_loop_new (unsigned int flags)
450
451	Similar to C<ev_default_loop>, but always creates a new event loop that is
452	always distinct from the default loop. Unlike the default loop, it cannot
453	handle signal and child watchers, and attempts to do so will be greeted by
454	undefined behaviour (or a failed assertion if assertions are enabled).
455		728
456	Example: Try to create a event loop that uses epoll and nothing else.	729	Example: Try to create a event loop that uses epoll and nothing else.
457		730
458	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	731	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
459	if (!epoller)	732	if (!epoller)
460	fatal ("no epoll found here, maybe it hides under your chair");	733	fatal ("no epoll found here, maybe it hides under your chair");
461		734
		735	Example: Use whatever libev has to offer, but make sure that kqueue is
		736	used if available.
		737
		738	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		739
		740	Example: Similarly, on linux, you mgiht want to take advantage of the
		741	linux aio backend if possible, but fall back to something else if that
		742	isn't available.
		743
		744	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_LINUXAIO);
		745
462	=item ev_default_destroy ()	746	=item ev_loop_destroy (loop)
463		747
464	Destroys the default loop again (frees all memory and kernel state	748	Destroys an event loop object (frees all memory and kernel state
465	etc.). None of the active event watchers will be stopped in the normal	749	etc.). None of the active event watchers will be stopped in the normal
466	sense, so e.g. C<ev_is_active> might still return true. It is your	750	sense, so e.g. C<ev_is_active> might still return true. It is your
467	responsibility to either stop all watchers cleanly yoursef I<before>	751	responsibility to either stop all watchers cleanly yourself I<before>
468	calling this function, or cope with the fact afterwards (which is usually	752	calling this function, or cope with the fact afterwards (which is usually
469	the easiest thing, you can just ignore the watchers and/or C<free ()> them	753	the easiest thing, you can just ignore the watchers and/or C<free ()> them
470	for example).	754	for example).
471		755
472	Note that certain global state, such as signal state, will not be freed by	756	Note that certain global state, such as signal state (and installed signal
473	this function, and related watchers (such as signal and child watchers)	757	handlers), will not be freed by this function, and related watchers (such
474	would need to be stopped manually.	758	as signal and child watchers) would need to be stopped manually.
475		759
476	In general it is not advisable to call this function except in the	760	This function is normally used on loop objects allocated by
477	rare occasion where you really need to free e.g. the signal handling	761	C<ev_loop_new>, but it can also be used on the default loop returned by
		762	C<ev_default_loop>, in which case it is not thread-safe.
		763
		764	Note that it is not advisable to call this function on the default loop
		765	except in the rare occasion where you really need to free its resources.
478	pipe fds. If you need dynamically allocated loops it is better to use	766	If you need dynamically allocated loops it is better to use C<ev_loop_new>
479	C<ev_loop_new> and C<ev_loop_destroy>).	767	and C<ev_loop_destroy>.
480		768
481	=item ev_loop_destroy (loop)	769	=item ev_loop_fork (loop)
482		770
483	Like C<ev_default_destroy>, but destroys an event loop created by an
484	earlier call to C<ev_loop_new>.
485
486	=item ev_default_fork ()
487
488	This function sets a flag that causes subsequent C<ev_loop> iterations	771	This function sets a flag that causes subsequent C<ev_run> iterations
489	to reinitialise the kernel state for backends that have one. Despite the	772	to reinitialise the kernel state for backends that have one. Despite
490	name, you can call it anytime, but it makes most sense after forking, in	773	the name, you can call it anytime you are allowed to start or stop
491	the child process (or both child and parent, but that again makes little	774	watchers (except inside an C<ev_prepare> callback), but it makes most
492	sense). You I<must> call it in the child before using any of the libev	775	sense after forking, in the child process. You I<must> call it (or use
493	functions, and it will only take effect at the next C<ev_loop> iteration.	776	C<EVFLAG_FORKCHECK>) in the child before resuming or calling C<ev_run>.
		777
		778	In addition, if you want to reuse a loop (via this function or
		779	C<EVFLAG_FORKCHECK>), you I<also> have to ignore C<SIGPIPE>.
		780
		781	Again, you I<have> to call it on I<any> loop that you want to re-use after
		782	a fork, I<even if you do not plan to use the loop in the parent>. This is
		783	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		784	during fork.
494		785
495	On the other hand, you only need to call this function in the child	786	On the other hand, you only need to call this function in the child
496	process if and only if you want to use the event library in the child. If	787	process if and only if you want to use the event loop in the child. If
497	you just fork+exec, you don't have to call it at all.	788	you just fork+exec or create a new loop in the child, you don't have to
		789	call it at all (in fact, C<epoll> is so badly broken that it makes a
		790	difference, but libev will usually detect this case on its own and do a
		791	costly reset of the backend).
498		792
499	The function itself is quite fast and it's usually not a problem to call	793	The function itself is quite fast and it's usually not a problem to call
500	it just in case after a fork. To make this easy, the function will fit in	794	it just in case after a fork.
501	quite nicely into a call to C<pthread_atfork>:
502		795
		796	Example: Automate calling C<ev_loop_fork> on the default loop when
		797	using pthreads.
		798
		799	static void
		800	post_fork_child (void)
		801	{
		802	ev_loop_fork (EV_DEFAULT);
		803	}
		804
		805	...
503	pthread_atfork (0, 0, ev_default_fork);	806	pthread_atfork (0, 0, post_fork_child);
504		807
505	=item ev_loop_fork (loop)	808	=item int ev_is_default_loop (loop)
506		809
507	Like C<ev_default_fork>, but acts on an event loop created by	810	Returns true when the given loop is, in fact, the default loop, and false
508	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	811	otherwise.
509	after fork, and how you do this is entirely your own problem.
510		812
511	=item unsigned int ev_loop_count (loop)	813	=item unsigned int ev_iteration (loop)
512		814
513	Returns the count of loop iterations for the loop, which is identical to	815	Returns the current iteration count for the event loop, which is identical
514	the number of times libev did poll for new events. It starts at C<0> and	816	to the number of times libev did poll for new events. It starts at C<0>
515	happily wraps around with enough iterations.	817	and happily wraps around with enough iterations.
516		818
517	This value can sometimes be useful as a generation counter of sorts (it	819	This value can sometimes be useful as a generation counter of sorts (it
518	"ticks" the number of loop iterations), as it roughly corresponds with	820	"ticks" the number of loop iterations), as it roughly corresponds with
519	C<ev_prepare> and C<ev_check> calls.	821	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		822	prepare and check phases.
		823
		824	=item unsigned int ev_depth (loop)
		825
		826	Returns the number of times C<ev_run> was entered minus the number of
		827	times C<ev_run> was exited normally, in other words, the recursion depth.
		828
		829	Outside C<ev_run>, this number is zero. In a callback, this number is
		830	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
		831	in which case it is higher.
		832
		833	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
		834	throwing an exception etc.), doesn't count as "exit" - consider this
		835	as a hint to avoid such ungentleman-like behaviour unless it's really
		836	convenient, in which case it is fully supported.
520		837
521	=item unsigned int ev_backend (loop)	838	=item unsigned int ev_backend (loop)
522		839
523	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	840	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
524	use.	841	use.
…		…
529	received events and started processing them. This timestamp does not	846	received events and started processing them. This timestamp does not
530	change as long as callbacks are being processed, and this is also the base	847	change as long as callbacks are being processed, and this is also the base
531	time used for relative timers. You can treat it as the timestamp of the	848	time used for relative timers. You can treat it as the timestamp of the
532	event occurring (or more correctly, libev finding out about it).	849	event occurring (or more correctly, libev finding out about it).
533		850
		851	=item ev_now_update (loop)
		852
		853	Establishes the current time by querying the kernel, updating the time
		854	returned by C<ev_now ()> in the progress. This is a costly operation and
		855	is usually done automatically within C<ev_run ()>.
		856
		857	This function is rarely useful, but when some event callback runs for a
		858	very long time without entering the event loop, updating libev's idea of
		859	the current time is a good idea.
		860
		861	See also L</The special problem of time updates> in the C<ev_timer> section.
		862
		863	=item ev_suspend (loop)
		864
		865	=item ev_resume (loop)
		866
		867	These two functions suspend and resume an event loop, for use when the
		868	loop is not used for a while and timeouts should not be processed.
		869
		870	A typical use case would be an interactive program such as a game: When
		871	the user presses C<^Z> to suspend the game and resumes it an hour later it
		872	would be best to handle timeouts as if no time had actually passed while
		873	the program was suspended. This can be achieved by calling C<ev_suspend>
		874	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		875	C<ev_resume> directly afterwards to resume timer processing.
		876
		877	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		878	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		879	will be rescheduled (that is, they will lose any events that would have
		880	occurred while suspended).
		881
		882	After calling C<ev_suspend> you B<must not> call I<any> function on the
		883	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		884	without a previous call to C<ev_suspend>.
		885
		886	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		887	event loop time (see C<ev_now_update>).
		888
534	=item ev_loop (loop, int flags)	889	=item bool ev_run (loop, int flags)
535		890
536	Finally, this is it, the event handler. This function usually is called	891	Finally, this is it, the event handler. This function usually is called
537	after you initialised all your watchers and you want to start handling	892	after you have initialised all your watchers and you want to start
538	events.	893	handling events. It will ask the operating system for any new events, call
		894	the watcher callbacks, and then repeat the whole process indefinitely: This
		895	is why event loops are called I<loops>.
539		896
540	If the flags argument is specified as C<0>, it will not return until	897	If the flags argument is specified as C<0>, it will keep handling events
541	either no event watchers are active anymore or C<ev_unloop> was called.	898	until either no event watchers are active anymore or C<ev_break> was
		899	called.
542		900
		901	The return value is false if there are no more active watchers (which
		902	usually means "all jobs done" or "deadlock"), and true in all other cases
		903	(which usually means " you should call C<ev_run> again").
		904
543	Please note that an explicit C<ev_unloop> is usually better than	905	Please note that an explicit C<ev_break> is usually better than
544	relying on all watchers to be stopped when deciding when a program has	906	relying on all watchers to be stopped when deciding when a program has
545	finished (especially in interactive programs), but having a program that	907	finished (especially in interactive programs), but having a program
546	automatically loops as long as it has to and no longer by virtue of	908	that automatically loops as long as it has to and no longer by virtue
547	relying on its watchers stopping correctly is a thing of beauty.	909	of relying on its watchers stopping correctly, that is truly a thing of
		910	beauty.
548		911
		912	This function is I<mostly> exception-safe - you can break out of a
		913	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		914	exception and so on. This does not decrement the C<ev_depth> value, nor
		915	will it clear any outstanding C<EVBREAK_ONE> breaks.
		916
549	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	917	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
550	those events and any outstanding ones, but will not block your process in	918	those events and any already outstanding ones, but will not wait and
551	case there are no events and will return after one iteration of the loop.	919	block your process in case there are no events and will return after one
		920	iteration of the loop. This is sometimes useful to poll and handle new
		921	events while doing lengthy calculations, to keep the program responsive.
552		922
553	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	923	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
554	neccessary) and will handle those and any outstanding ones. It will block	924	necessary) and will handle those and any already outstanding ones. It
555	your process until at least one new event arrives, and will return after	925	will block your process until at least one new event arrives (which could
556	one iteration of the loop. This is useful if you are waiting for some	926	be an event internal to libev itself, so there is no guarantee that a
557	external event in conjunction with something not expressible using other	927	user-registered callback will be called), and will return after one
		928	iteration of the loop.
		929
		930	This is useful if you are waiting for some external event in conjunction
		931	with something not expressible using other libev watchers (i.e. "roll your
558	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	932	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
559	usually a better approach for this kind of thing.	933	usually a better approach for this kind of thing.
560		934
561	Here are the gory details of what C<ev_loop> does:	935	Here are the gory details of what C<ev_run> does (this is for your
		936	understanding, not a guarantee that things will work exactly like this in
		937	future versions):
562		938
		939	- Increment loop depth.
		940	- Reset the ev_break status.
563	- Before the first iteration, call any pending watchers.	941	- Before the first iteration, call any pending watchers.
		942	LOOP:
564	* If EVFLAG_FORKCHECK was used, check for a fork.	943	- If EVFLAG_FORKCHECK was used, check for a fork.
565	- If a fork was detected, queue and call all fork watchers.	944	- If a fork was detected (by any means), queue and call all fork watchers.
566	- Queue and call all prepare watchers.	945	- Queue and call all prepare watchers.
		946	- If ev_break was called, goto FINISH.
567	- If we have been forked, recreate the kernel state.	947	- If we have been forked, detach and recreate the kernel state
		948	as to not disturb the other process.
568	- Update the kernel state with all outstanding changes.	949	- Update the kernel state with all outstanding changes.
569	- Update the "event loop time".	950	- Update the "event loop time" (ev_now ()).
570	- Calculate for how long to sleep or block, if at all	951	- Calculate for how long to sleep or block, if at all
571	(active idle watchers, EVLOOP_NONBLOCK or not having	952	(active idle watchers, EVRUN_NOWAIT or not having
572	any active watchers at all will result in not sleeping).	953	any active watchers at all will result in not sleeping).
573	- Sleep if the I/O and timer collect interval say so.	954	- Sleep if the I/O and timer collect interval say so.
		955	- Increment loop iteration counter.
574	- Block the process, waiting for any events.	956	- Block the process, waiting for any events.
575	- Queue all outstanding I/O (fd) events.	957	- Queue all outstanding I/O (fd) events.
576	- Update the "event loop time" and do time jump handling.	958	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
577	- Queue all outstanding timers.	959	- Queue all expired timers.
578	- Queue all outstanding periodics.	960	- Queue all expired periodics.
579	- If no events are pending now, queue all idle watchers.	961	- Queue all idle watchers with priority higher than that of pending events.
580	- Queue all check watchers.	962	- Queue all check watchers.
581	- Call all queued watchers in reverse order (i.e. check watchers first).	963	- Call all queued watchers in reverse order (i.e. check watchers first).
582	Signals and child watchers are implemented as I/O watchers, and will	964	Signals and child watchers are implemented as I/O watchers, and will
583	be handled here by queueing them when their watcher gets executed.	965	be handled here by queueing them when their watcher gets executed.
584	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	966	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
585	were used, or there are no active watchers, return, otherwise	967	were used, or there are no active watchers, goto FINISH, otherwise
586	continue with step *.	968	continue with step LOOP.
		969	FINISH:
		970	- Reset the ev_break status iff it was EVBREAK_ONE.
		971	- Decrement the loop depth.
		972	- Return.
587		973
588	Example: Queue some jobs and then loop until no events are outstanding	974	Example: Queue some jobs and then loop until no events are outstanding
589	anymore.	975	anymore.
590		976
591	... queue jobs here, make sure they register event watchers as long	977	... queue jobs here, make sure they register event watchers as long
592	... as they still have work to do (even an idle watcher will do..)	978	... as they still have work to do (even an idle watcher will do..)
593	ev_loop (my_loop, 0);	979	ev_run (my_loop, 0);
594	... jobs done. yeah!	980	... jobs done or somebody called break. yeah!
595		981
596	=item ev_unloop (loop, how)	982	=item ev_break (loop, how)
597		983
598	Can be used to make a call to C<ev_loop> return early (but only after it	984	Can be used to make a call to C<ev_run> return early (but only after it
599	has processed all outstanding events). The C<how> argument must be either	985	has processed all outstanding events). The C<how> argument must be either
600	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	986	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
601	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	987	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
602		988
603	This "unloop state" will be cleared when entering C<ev_loop> again.	989	This "break state" will be cleared on the next call to C<ev_run>.
		990
		991	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		992	which case it will have no effect.
604		993
605	=item ev_ref (loop)	994	=item ev_ref (loop)
606		995
607	=item ev_unref (loop)	996	=item ev_unref (loop)
608		997
609	Ref/unref can be used to add or remove a reference count on the event	998	Ref/unref can be used to add or remove a reference count on the event
610	loop: Every watcher keeps one reference, and as long as the reference	999	loop: Every watcher keeps one reference, and as long as the reference
611	count is nonzero, C<ev_loop> will not return on its own. If you have	1000	count is nonzero, C<ev_run> will not return on its own.
612	a watcher you never unregister that should not keep C<ev_loop> from	1001
613	returning, ev_unref() after starting, and ev_ref() before stopping it. For	1002	This is useful when you have a watcher that you never intend to
		1003	unregister, but that nevertheless should not keep C<ev_run> from
		1004	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		1005	before stopping it.
		1006
614	example, libev itself uses this for its internal signal pipe: It is not	1007	As an example, libev itself uses this for its internal signal pipe: It
615	visible to the libev user and should not keep C<ev_loop> from exiting if	1008	is not visible to the libev user and should not keep C<ev_run> from
616	no event watchers registered by it are active. It is also an excellent	1009	exiting if no event watchers registered by it are active. It is also an
617	way to do this for generic recurring timers or from within third-party	1010	excellent way to do this for generic recurring timers or from within
618	libraries. Just remember to I<unref after start> and I<ref before stop>	1011	third-party libraries. Just remember to I<unref after start> and I<ref
619	(but only if the watcher wasn't active before, or was active before,	1012	before stop> (but only if the watcher wasn't active before, or was active
620	respectively).	1013	before, respectively. Note also that libev might stop watchers itself
		1014	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		1015	in the callback).
621		1016
622	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	1017	Example: Create a signal watcher, but keep it from keeping C<ev_run>
623	running when nothing else is active.	1018	running when nothing else is active.
624		1019
625	struct ev_signal exitsig;	1020	ev_signal exitsig;
626	ev_signal_init (&exitsig, sig_cb, SIGINT);	1021	ev_signal_init (&exitsig, sig_cb, SIGINT);
627	ev_signal_start (loop, &exitsig);	1022	ev_signal_start (loop, &exitsig);
628	evf_unref (loop);	1023	ev_unref (loop);
629		1024
630	Example: For some weird reason, unregister the above signal handler again.	1025	Example: For some weird reason, unregister the above signal handler again.
631		1026
632	ev_ref (loop);	1027	ev_ref (loop);
633	ev_signal_stop (loop, &exitsig);	1028	ev_signal_stop (loop, &exitsig);
634		1029
635	=item ev_set_io_collect_interval (loop, ev_tstamp interval)	1030	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
636		1031
637	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)	1032	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
638		1033
639	These advanced functions influence the time that libev will spend waiting	1034	These advanced functions influence the time that libev will spend waiting
640	for events. Both are by default C<0>, meaning that libev will try to	1035	for events. Both time intervals are by default C<0>, meaning that libev
641	invoke timer/periodic callbacks and I/O callbacks with minimum latency.	1036	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		1037	latency.
642		1038
643	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	1039	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
644	allows libev to delay invocation of I/O and timer/periodic callbacks to	1040	allows libev to delay invocation of I/O and timer/periodic callbacks
645	increase efficiency of loop iterations.	1041	to increase efficiency of loop iterations (or to increase power-saving
		1042	opportunities).
646		1043
647	The background is that sometimes your program runs just fast enough to	1044	The idea is that sometimes your program runs just fast enough to handle
648	handle one (or very few) event(s) per loop iteration. While this makes	1045	one (or very few) event(s) per loop iteration. While this makes the
649	the program responsive, it also wastes a lot of CPU time to poll for new	1046	program responsive, it also wastes a lot of CPU time to poll for new
650	events, especially with backends like C<select ()> which have a high	1047	events, especially with backends like C<select ()> which have a high
651	overhead for the actual polling but can deliver many events at once.	1048	overhead for the actual polling but can deliver many events at once.
652		1049
653	By setting a higher I<io collect interval> you allow libev to spend more	1050	By setting a higher I<io collect interval> you allow libev to spend more
654	time collecting I/O events, so you can handle more events per iteration,	1051	time collecting I/O events, so you can handle more events per iteration,
655	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	1052	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
656	C<ev_timer>) will be not affected. Setting this to a non-null value will	1053	C<ev_timer>) will not be affected. Setting this to a non-null value will
657	introduce an additional C<ev_sleep ()> call into most loop iterations.	1054	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		1055	sleep time ensures that libev will not poll for I/O events more often then
		1056	once per this interval, on average (as long as the host time resolution is
		1057	good enough).
658		1058
659	Likewise, by setting a higher I<timeout collect interval> you allow libev	1059	Likewise, by setting a higher I<timeout collect interval> you allow libev
660	to spend more time collecting timeouts, at the expense of increased	1060	to spend more time collecting timeouts, at the expense of increased
661	latency (the watcher callback will be called later). C<ev_io> watchers	1061	latency/jitter/inexactness (the watcher callback will be called
662	will not be affected. Setting this to a non-null value will not introduce	1062	later). C<ev_io> watchers will not be affected. Setting this to a non-null
663	any overhead in libev.	1063	value will not introduce any overhead in libev.
664		1064
665	Many (busy) programs can usually benefit by setting the io collect	1065	Many (busy) programs can usually benefit by setting the I/O collect
666	interval to a value near C<0.1> or so, which is often enough for	1066	interval to a value near C<0.1> or so, which is often enough for
667	interactive servers (of course not for games), likewise for timeouts. It	1067	interactive servers (of course not for games), likewise for timeouts. It
668	usually doesn't make much sense to set it to a lower value than C<0.01>,	1068	usually doesn't make much sense to set it to a lower value than C<0.01>,
669	as this approsaches the timing granularity of most systems.	1069	as this approaches the timing granularity of most systems. Note that if
		1070	you do transactions with the outside world and you can't increase the
		1071	parallelity, then this setting will limit your transaction rate (if you
		1072	need to poll once per transaction and the I/O collect interval is 0.01,
		1073	then you can't do more than 100 transactions per second).
		1074
		1075	Setting the I<timeout collect interval> can improve the opportunity for
		1076	saving power, as the program will "bundle" timer callback invocations that
		1077	are "near" in time together, by delaying some, thus reducing the number of
		1078	times the process sleeps and wakes up again. Another useful technique to
		1079	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		1080	they fire on, say, one-second boundaries only.
		1081
		1082	Example: we only need 0.1s timeout granularity, and we wish not to poll
		1083	more often than 100 times per second:
		1084
		1085	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		1086	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		1087
		1088	=item ev_invoke_pending (loop)
		1089
		1090	This call will simply invoke all pending watchers while resetting their
		1091	pending state. Normally, C<ev_run> does this automatically when required,
		1092	but when overriding the invoke callback this call comes handy. This
		1093	function can be invoked from a watcher - this can be useful for example
		1094	when you want to do some lengthy calculation and want to pass further
		1095	event handling to another thread (you still have to make sure only one
		1096	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
		1097
		1098	=item int ev_pending_count (loop)
		1099
		1100	Returns the number of pending watchers - zero indicates that no watchers
		1101	are pending.
		1102
		1103	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		1104
		1105	This overrides the invoke pending functionality of the loop: Instead of
		1106	invoking all pending watchers when there are any, C<ev_run> will call
		1107	this callback instead. This is useful, for example, when you want to
		1108	invoke the actual watchers inside another context (another thread etc.).
		1109
		1110	If you want to reset the callback, use C<ev_invoke_pending> as new
		1111	callback.
		1112
		1113	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
		1114
		1115	Sometimes you want to share the same loop between multiple threads. This
		1116	can be done relatively simply by putting mutex_lock/unlock calls around
		1117	each call to a libev function.
		1118
		1119	However, C<ev_run> can run an indefinite time, so it is not feasible
		1120	to wait for it to return. One way around this is to wake up the event
		1121	loop via C<ev_break> and C<ev_async_send>, another way is to set these
		1122	I<release> and I<acquire> callbacks on the loop.
		1123
		1124	When set, then C<release> will be called just before the thread is
		1125	suspended waiting for new events, and C<acquire> is called just
		1126	afterwards.
		1127
		1128	Ideally, C<release> will just call your mutex_unlock function, and
		1129	C<acquire> will just call the mutex_lock function again.
		1130
		1131	While event loop modifications are allowed between invocations of
		1132	C<release> and C<acquire> (that's their only purpose after all), no
		1133	modifications done will affect the event loop, i.e. adding watchers will
		1134	have no effect on the set of file descriptors being watched, or the time
		1135	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
		1136	to take note of any changes you made.
		1137
		1138	In theory, threads executing C<ev_run> will be async-cancel safe between
		1139	invocations of C<release> and C<acquire>.
		1140
		1141	See also the locking example in the C<THREADS> section later in this
		1142	document.
		1143
		1144	=item ev_set_userdata (loop, void *data)
		1145
		1146	=item void *ev_userdata (loop)
		1147
		1148	Set and retrieve a single C<void *> associated with a loop. When
		1149	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		1150	C<0>.
		1151
		1152	These two functions can be used to associate arbitrary data with a loop,
		1153	and are intended solely for the C<invoke_pending_cb>, C<release> and
		1154	C<acquire> callbacks described above, but of course can be (ab-)used for
		1155	any other purpose as well.
		1156
		1157	=item ev_verify (loop)
		1158
		1159	This function only does something when C<EV_VERIFY> support has been
		1160	compiled in, which is the default for non-minimal builds. It tries to go
		1161	through all internal structures and checks them for validity. If anything
		1162	is found to be inconsistent, it will print an error message to standard
		1163	error and call C<abort ()>.
		1164
		1165	This can be used to catch bugs inside libev itself: under normal
		1166	circumstances, this function will never abort as of course libev keeps its
		1167	data structures consistent.
670		1168
671	=back	1169	=back
672		1170
673		1171
674	=head1 ANATOMY OF A WATCHER	1172	=head1 ANATOMY OF A WATCHER
675		1173
		1174	In the following description, uppercase C<TYPE> in names stands for the
		1175	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		1176	watchers and C<ev_io_start> for I/O watchers.
		1177
676	A watcher is a structure that you create and register to record your	1178	A watcher is an opaque structure that you allocate and register to record
677	interest in some event. For instance, if you want to wait for STDIN to	1179	your interest in some event. To make a concrete example, imagine you want
678	become readable, you would create an C<ev_io> watcher for that:	1180	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1181	for that:
679		1182
680	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1183	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
681	{	1184	{
682	ev_io_stop (w);	1185	ev_io_stop (w);
683	ev_unloop (loop, EVUNLOOP_ALL);	1186	ev_break (loop, EVBREAK_ALL);
684	}	1187	}
685		1188
686	struct ev_loop *loop = ev_default_loop (0);	1189	struct ev_loop *loop = ev_default_loop (0);
		1190
687	struct ev_io stdin_watcher;	1191	ev_io stdin_watcher;
		1192
688	ev_init (&stdin_watcher, my_cb);	1193	ev_init (&stdin_watcher, my_cb);
689	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1194	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
690	ev_io_start (loop, &stdin_watcher);	1195	ev_io_start (loop, &stdin_watcher);
		1196
691	ev_loop (loop, 0);	1197	ev_run (loop, 0);
692		1198
693	As you can see, you are responsible for allocating the memory for your	1199	As you can see, you are responsible for allocating the memory for your
694	watcher structures (and it is usually a bad idea to do this on the stack,	1200	watcher structures (and it is I<usually> a bad idea to do this on the
695	although this can sometimes be quite valid).	1201	stack).
696		1202
		1203	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1204	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
		1205
697	Each watcher structure must be initialised by a call to C<ev_init	1206	Each watcher structure must be initialised by a call to C<ev_init (watcher
698	(watcher *, callback)>, which expects a callback to be provided. This	1207	*, callback)>, which expects a callback to be provided. This callback is
699	callback gets invoked each time the event occurs (or, in the case of io	1208	invoked each time the event occurs (or, in the case of I/O watchers, each
700	watchers, each time the event loop detects that the file descriptor given	1209	time the event loop detects that the file descriptor given is readable
701	is readable and/or writable).	1210	and/or writable).
702		1211
703	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1212	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
704	with arguments specific to this watcher type. There is also a macro	1213	macro to configure it, with arguments specific to the watcher type. There
705	to combine initialisation and setting in one call: C<< ev_<type>_init	1214	is also a macro to combine initialisation and setting in one call: C<<
706	(watcher *, callback, ...) >>.	1215	ev_TYPE_init (watcher *, callback, ...) >>.
707		1216
708	To make the watcher actually watch out for events, you have to start it	1217	To make the watcher actually watch out for events, you have to start it
709	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1218	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
710	*) >>), and you can stop watching for events at any time by calling the	1219	*) >>), and you can stop watching for events at any time by calling the
711	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1220	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
712		1221
713	As long as your watcher is active (has been started but not stopped) you	1222	As long as your watcher is active (has been started but not stopped) you
714	must not touch the values stored in it. Most specifically you must never	1223	must not touch the values stored in it except when explicitly documented
715	reinitialise it or call its C<set> macro.	1224	otherwise. Most specifically you must never reinitialise it or call its
		1225	C<ev_TYPE_set> macro.
716		1226
717	Each and every callback receives the event loop pointer as first, the	1227	Each and every callback receives the event loop pointer as first, the
718	registered watcher structure as second, and a bitset of received events as	1228	registered watcher structure as second, and a bitset of received events as
719	third argument.	1229	third argument.
720		1230
…		…
729	=item C<EV_WRITE>	1239	=item C<EV_WRITE>
730		1240
731	The file descriptor in the C<ev_io> watcher has become readable and/or	1241	The file descriptor in the C<ev_io> watcher has become readable and/or
732	writable.	1242	writable.
733		1243
734	=item C<EV_TIMEOUT>	1244	=item C<EV_TIMER>
735		1245
736	The C<ev_timer> watcher has timed out.	1246	The C<ev_timer> watcher has timed out.
737		1247
738	=item C<EV_PERIODIC>	1248	=item C<EV_PERIODIC>
739		1249
…		…
757		1267
758	=item C<EV_PREPARE>	1268	=item C<EV_PREPARE>
759		1269
760	=item C<EV_CHECK>	1270	=item C<EV_CHECK>
761		1271
762	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1272	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
763	to gather new events, and all C<ev_check> watchers are invoked just after	1273	gather new events, and all C<ev_check> watchers are queued (not invoked)
764	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1274	just after C<ev_run> has gathered them, but before it queues any callbacks
		1275	for any received events. That means C<ev_prepare> watchers are the last
		1276	watchers invoked before the event loop sleeps or polls for new events, and
		1277	C<ev_check> watchers will be invoked before any other watchers of the same
		1278	or lower priority within an event loop iteration.
		1279
765	received events. Callbacks of both watcher types can start and stop as	1280	Callbacks of both watcher types can start and stop as many watchers as
766	many watchers as they want, and all of them will be taken into account	1281	they want, and all of them will be taken into account (for example, a
767	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1282	C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
768	C<ev_loop> from blocking).	1283	blocking).
769		1284
770	=item C<EV_EMBED>	1285	=item C<EV_EMBED>
771		1286
772	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1287	The embedded event loop specified in the C<ev_embed> watcher needs attention.
773		1288
774	=item C<EV_FORK>	1289	=item C<EV_FORK>
775		1290
776	The event loop has been resumed in the child process after fork (see	1291	The event loop has been resumed in the child process after fork (see
777	C<ev_fork>).	1292	C<ev_fork>).
778		1293
		1294	=item C<EV_CLEANUP>
		1295
		1296	The event loop is about to be destroyed (see C<ev_cleanup>).
		1297
		1298	=item C<EV_ASYNC>
		1299
		1300	The given async watcher has been asynchronously notified (see C<ev_async>).
		1301
		1302	=item C<EV_CUSTOM>
		1303
		1304	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1305	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1306
779	=item C<EV_ERROR>	1307	=item C<EV_ERROR>
780		1308
781	An unspecified error has occured, the watcher has been stopped. This might	1309	An unspecified error has occurred, the watcher has been stopped. This might
782	happen because the watcher could not be properly started because libev	1310	happen because the watcher could not be properly started because libev
783	ran out of memory, a file descriptor was found to be closed or any other	1311	ran out of memory, a file descriptor was found to be closed or any other
		1312	problem. Libev considers these application bugs.
		1313
784	problem. You best act on it by reporting the problem and somehow coping	1314	You best act on it by reporting the problem and somehow coping with the
785	with the watcher being stopped.	1315	watcher being stopped. Note that well-written programs should not receive
		1316	an error ever, so when your watcher receives it, this usually indicates a
		1317	bug in your program.
786		1318
787	Libev will usually signal a few "dummy" events together with an error,	1319	Libev will usually signal a few "dummy" events together with an error, for
788	for example it might indicate that a fd is readable or writable, and if	1320	example it might indicate that a fd is readable or writable, and if your
789	your callbacks is well-written it can just attempt the operation and cope	1321	callbacks is well-written it can just attempt the operation and cope with
790	with the error from read() or write(). This will not work in multithreaded	1322	the error from read() or write(). This will not work in multi-threaded
791	programs, though, so beware.	1323	programs, though, as the fd could already be closed and reused for another
		1324	thing, so beware.
792		1325
793	=back	1326	=back
794		1327
795	=head2 GENERIC WATCHER FUNCTIONS	1328	=head2 GENERIC WATCHER FUNCTIONS
796
797	In the following description, C<TYPE> stands for the watcher type,
798	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
799		1329
800	=over 4	1330	=over 4
801		1331
802	=item C<ev_init> (ev_TYPE *watcher, callback)	1332	=item C<ev_init> (ev_TYPE *watcher, callback)
803		1333
…		…
809	which rolls both calls into one.	1339	which rolls both calls into one.
810		1340
811	You can reinitialise a watcher at any time as long as it has been stopped	1341	You can reinitialise a watcher at any time as long as it has been stopped
812	(or never started) and there are no pending events outstanding.	1342	(or never started) and there are no pending events outstanding.
813		1343
814	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1344	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
815	int revents)>.	1345	int revents)>.
816		1346
		1347	Example: Initialise an C<ev_io> watcher in two steps.
		1348
		1349	ev_io w;
		1350	ev_init (&w, my_cb);
		1351	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1352
817	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1353	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
818		1354
819	This macro initialises the type-specific parts of a watcher. You need to	1355	This macro initialises the type-specific parts of a watcher. You need to
820	call C<ev_init> at least once before you call this macro, but you can	1356	call C<ev_init> at least once before you call this macro, but you can
821	call C<ev_TYPE_set> any number of times. You must not, however, call this	1357	call C<ev_TYPE_set> any number of times. You must not, however, call this
822	macro on a watcher that is active (it can be pending, however, which is a	1358	macro on a watcher that is active (it can be pending, however, which is a
823	difference to the C<ev_init> macro).	1359	difference to the C<ev_init> macro).
824		1360
825	Although some watcher types do not have type-specific arguments	1361	Although some watcher types do not have type-specific arguments
826	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1362	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
827		1363
		1364	See C<ev_init>, above, for an example.
		1365
828	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1366	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
829		1367
830	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1368	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
831	calls into a single call. This is the most convinient method to initialise	1369	calls into a single call. This is the most convenient method to initialise
832	a watcher. The same limitations apply, of course.	1370	a watcher. The same limitations apply, of course.
833		1371
		1372	Example: Initialise and set an C<ev_io> watcher in one step.
		1373
		1374	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1375
834	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1376	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
835		1377
836	Starts (activates) the given watcher. Only active watchers will receive	1378	Starts (activates) the given watcher. Only active watchers will receive
837	events. If the watcher is already active nothing will happen.	1379	events. If the watcher is already active nothing will happen.
838		1380
		1381	Example: Start the C<ev_io> watcher that is being abused as example in this
		1382	whole section.
		1383
		1384	ev_io_start (EV_DEFAULT_UC, &w);
		1385
839	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1386	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
840		1387
841	Stops the given watcher again (if active) and clears the pending	1388	Stops the given watcher if active, and clears the pending status (whether
		1389	the watcher was active or not).
		1390
842	status. It is possible that stopped watchers are pending (for example,	1391	It is possible that stopped watchers are pending - for example,
843	non-repeating timers are being stopped when they become pending), but	1392	non-repeating timers are being stopped when they become pending - but
844	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1393	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
845	you want to free or reuse the memory used by the watcher it is therefore a	1394	pending. If you want to free or reuse the memory used by the watcher it is
846	good idea to always call its C<ev_TYPE_stop> function.	1395	therefore a good idea to always call its C<ev_TYPE_stop> function.
847		1396
848	=item bool ev_is_active (ev_TYPE *watcher)	1397	=item bool ev_is_active (ev_TYPE *watcher)
849		1398
850	Returns a true value iff the watcher is active (i.e. it has been started	1399	Returns a true value iff the watcher is active (i.e. it has been started
851	and not yet been stopped). As long as a watcher is active you must not modify	1400	and not yet been stopped). As long as a watcher is active you must not modify
…		…
862		1411
863	=item callback ev_cb (ev_TYPE *watcher)	1412	=item callback ev_cb (ev_TYPE *watcher)
864		1413
865	Returns the callback currently set on the watcher.	1414	Returns the callback currently set on the watcher.
866		1415
867	=item ev_cb_set (ev_TYPE *watcher, callback)	1416	=item ev_set_cb (ev_TYPE *watcher, callback)
868		1417
869	Change the callback. You can change the callback at virtually any time	1418	Change the callback. You can change the callback at virtually any time
870	(modulo threads).	1419	(modulo threads).
871		1420
872	=item ev_set_priority (ev_TYPE *watcher, priority)	1421	=item ev_set_priority (ev_TYPE *watcher, int priority)
873		1422
874	=item int ev_priority (ev_TYPE *watcher)	1423	=item int ev_priority (ev_TYPE *watcher)
875		1424
876	Set and query the priority of the watcher. The priority is a small	1425	Set and query the priority of the watcher. The priority is a small
877	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1426	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
878	(default: C<-2>). Pending watchers with higher priority will be invoked	1427	(default: C<-2>). Pending watchers with higher priority will be invoked
879	before watchers with lower priority, but priority will not keep watchers	1428	before watchers with lower priority, but priority will not keep watchers
880	from being executed (except for C<ev_idle> watchers).	1429	from being executed (except for C<ev_idle> watchers).
881		1430
882	This means that priorities are I<only> used for ordering callback
883	invocation after new events have been received. This is useful, for
884	example, to reduce latency after idling, or more often, to bind two
885	watchers on the same event and make sure one is called first.
886
887	If you need to suppress invocation when higher priority events are pending	1431	If you need to suppress invocation when higher priority events are pending
888	you need to look at C<ev_idle> watchers, which provide this functionality.	1432	you need to look at C<ev_idle> watchers, which provide this functionality.
889		1433
890	You I<must not> change the priority of a watcher as long as it is active or	1434	You I<must not> change the priority of a watcher as long as it is active or
891	pending.	1435	pending.
892		1436
		1437	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1438	fine, as long as you do not mind that the priority value you query might
		1439	or might not have been clamped to the valid range.
		1440
893	The default priority used by watchers when no priority has been set is	1441	The default priority used by watchers when no priority has been set is
894	always C<0>, which is supposed to not be too high and not be too low :).	1442	always C<0>, which is supposed to not be too high and not be too low :).
895		1443
896	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1444	See L</WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
897	fine, as long as you do not mind that the priority value you query might	1445	priorities.
898	or might not have been adjusted to be within valid range.
899		1446
900	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1447	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
901		1448
902	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1449	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
903	C<loop> nor C<revents> need to be valid as long as the watcher callback	1450	C<loop> nor C<revents> need to be valid as long as the watcher callback
904	can deal with that fact.	1451	can deal with that fact, as both are simply passed through to the
		1452	callback.
905		1453
906	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1454	=item int ev_clear_pending (loop, ev_TYPE *watcher)
907		1455
908	If the watcher is pending, this function returns clears its pending status	1456	If the watcher is pending, this function clears its pending status and
909	and returns its C<revents> bitset (as if its callback was invoked). If the	1457	returns its C<revents> bitset (as if its callback was invoked). If the
910	watcher isn't pending it does nothing and returns C<0>.	1458	watcher isn't pending it does nothing and returns C<0>.
911		1459
		1460	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1461	callback to be invoked, which can be accomplished with this function.
		1462
		1463	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1464
		1465	Feeds the given event set into the event loop, as if the specified event
		1466	had happened for the specified watcher (which must be a pointer to an
		1467	initialised but not necessarily started event watcher). Obviously you must
		1468	not free the watcher as long as it has pending events.
		1469
		1470	Stopping the watcher, letting libev invoke it, or calling
		1471	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1472	not started in the first place.
		1473
		1474	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1475	functions that do not need a watcher.
		1476
912	=back	1477	=back
913		1478
		1479	See also the L</ASSOCIATING CUSTOM DATA WITH A WATCHER> and L</BUILDING YOUR
		1480	OWN COMPOSITE WATCHERS> idioms.
914		1481
915	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1482	=head2 WATCHER STATES
916		1483
917	Each watcher has, by default, a member C<void *data> that you can change	1484	There are various watcher states mentioned throughout this manual -
918	and read at any time, libev will completely ignore it. This can be used	1485	active, pending and so on. In this section these states and the rules to
919	to associate arbitrary data with your watcher. If you need more data and	1486	transition between them will be described in more detail - and while these
920	don't want to allocate memory and store a pointer to it in that data	1487	rules might look complicated, they usually do "the right thing".
921	member, you can also "subclass" the watcher type and provide your own
922	data:
923		1488
924	struct my_io	1489	=over 4
925	{
926	struct ev_io io;
927	int otherfd;
928	void *somedata;
929	struct whatever *mostinteresting;
930	}
931		1490
932	And since your callback will be called with a pointer to the watcher, you	1491	=item initialised
933	can cast it back to your own type:
934		1492
935	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1493	Before a watcher can be registered with the event loop it has to be
936	{	1494	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
937	struct my_io *w = (struct my_io *)w_;	1495	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
938	...
939	}
940		1496
941	More interesting and less C-conformant ways of casting your callback type	1497	In this state it is simply some block of memory that is suitable for
942	instead have been omitted.	1498	use in an event loop. It can be moved around, freed, reused etc. at
		1499	will - as long as you either keep the memory contents intact, or call
		1500	C<ev_TYPE_init> again.
943		1501
944	Another common scenario is having some data structure with multiple	1502	=item started/running/active
945	watchers:
946		1503
947	struct my_biggy	1504	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
948	{	1505	property of the event loop, and is actively waiting for events. While in
949	int some_data;	1506	this state it cannot be accessed (except in a few documented ways), moved,
950	ev_timer t1;	1507	freed or anything else - the only legal thing is to keep a pointer to it,
951	ev_timer t2;	1508	and call libev functions on it that are documented to work on active watchers.
952	}
953		1509
954	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1510	=item pending
955	you need to use C<offsetof>:
956		1511
957	#include <stddef.h>	1512	If a watcher is active and libev determines that an event it is interested
		1513	in has occurred (such as a timer expiring), it will become pending. It will
		1514	stay in this pending state until either it is stopped or its callback is
		1515	about to be invoked, so it is not normally pending inside the watcher
		1516	callback.
958		1517
		1518	The watcher might or might not be active while it is pending (for example,
		1519	an expired non-repeating timer can be pending but no longer active). If it
		1520	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1521	but it is still property of the event loop at this time, so cannot be
		1522	moved, freed or reused. And if it is active the rules described in the
		1523	previous item still apply.
		1524
		1525	It is also possible to feed an event on a watcher that is not active (e.g.
		1526	via C<ev_feed_event>), in which case it becomes pending without being
		1527	active.
		1528
		1529	=item stopped
		1530
		1531	A watcher can be stopped implicitly by libev (in which case it might still
		1532	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1533	latter will clear any pending state the watcher might be in, regardless
		1534	of whether it was active or not, so stopping a watcher explicitly before
		1535	freeing it is often a good idea.
		1536
		1537	While stopped (and not pending) the watcher is essentially in the
		1538	initialised state, that is, it can be reused, moved, modified in any way
		1539	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1540	it again).
		1541
		1542	=back
		1543
		1544	=head2 WATCHER PRIORITY MODELS
		1545
		1546	Many event loops support I<watcher priorities>, which are usually small
		1547	integers that influence the ordering of event callback invocation
		1548	between watchers in some way, all else being equal.
		1549
		1550	In libev, watcher priorities can be set using C<ev_set_priority>. See its
		1551	description for the more technical details such as the actual priority
		1552	range.
		1553
		1554	There are two common ways how these these priorities are being interpreted
		1555	by event loops:
		1556
		1557	In the more common lock-out model, higher priorities "lock out" invocation
		1558	of lower priority watchers, which means as long as higher priority
		1559	watchers receive events, lower priority watchers are not being invoked.
		1560
		1561	The less common only-for-ordering model uses priorities solely to order
		1562	callback invocation within a single event loop iteration: Higher priority
		1563	watchers are invoked before lower priority ones, but they all get invoked
		1564	before polling for new events.
		1565
		1566	Libev uses the second (only-for-ordering) model for all its watchers
		1567	except for idle watchers (which use the lock-out model).
		1568
		1569	The rationale behind this is that implementing the lock-out model for
		1570	watchers is not well supported by most kernel interfaces, and most event
		1571	libraries will just poll for the same events again and again as long as
		1572	their callbacks have not been executed, which is very inefficient in the
		1573	common case of one high-priority watcher locking out a mass of lower
		1574	priority ones.
		1575
		1576	Static (ordering) priorities are most useful when you have two or more
		1577	watchers handling the same resource: a typical usage example is having an
		1578	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1579	timeouts. Under load, data might be received while the program handles
		1580	other jobs, but since timers normally get invoked first, the timeout
		1581	handler will be executed before checking for data. In that case, giving
		1582	the timer a lower priority than the I/O watcher ensures that I/O will be
		1583	handled first even under adverse conditions (which is usually, but not
		1584	always, what you want).
		1585
		1586	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1587	will only be executed when no same or higher priority watchers have
		1588	received events, they can be used to implement the "lock-out" model when
		1589	required.
		1590
		1591	For example, to emulate how many other event libraries handle priorities,
		1592	you can associate an C<ev_idle> watcher to each such watcher, and in
		1593	the normal watcher callback, you just start the idle watcher. The real
		1594	processing is done in the idle watcher callback. This causes libev to
		1595	continuously poll and process kernel event data for the watcher, but when
		1596	the lock-out case is known to be rare (which in turn is rare :), this is
		1597	workable.
		1598
		1599	Usually, however, the lock-out model implemented that way will perform
		1600	miserably under the type of load it was designed to handle. In that case,
		1601	it might be preferable to stop the real watcher before starting the
		1602	idle watcher, so the kernel will not have to process the event in case
		1603	the actual processing will be delayed for considerable time.
		1604
		1605	Here is an example of an I/O watcher that should run at a strictly lower
		1606	priority than the default, and which should only process data when no
		1607	other events are pending:
		1608
		1609	ev_idle idle; // actual processing watcher
		1610	ev_io io; // actual event watcher
		1611
959	static void	1612	static void
960	t1_cb (EV_P_ struct ev_timer *w, int revents)	1613	io_cb (EV_P_ ev_io *w, int revents)
961	{	1614	{
962	struct my_biggy big = (struct my_biggy *	1615	// stop the I/O watcher, we received the event, but
963	(((char *)w) - offsetof (struct my_biggy, t1));	1616	// are not yet ready to handle it.
		1617	ev_io_stop (EV_A_ w);
		1618
		1619	// start the idle watcher to handle the actual event.
		1620	// it will not be executed as long as other watchers
		1621	// with the default priority are receiving events.
		1622	ev_idle_start (EV_A_ &idle);
964	}	1623	}
965		1624
966	static void	1625	static void
967	t2_cb (EV_P_ struct ev_timer *w, int revents)	1626	idle_cb (EV_P_ ev_idle *w, int revents)
968	{	1627	{
969	struct my_biggy big = (struct my_biggy *	1628	// actual processing
970	(((char *)w) - offsetof (struct my_biggy, t2));	1629	read (STDIN_FILENO, ...);
		1630
		1631	// have to start the I/O watcher again, as
		1632	// we have handled the event
		1633	ev_io_start (EV_P_ &io);
971	}	1634	}
		1635
		1636	// initialisation
		1637	ev_idle_init (&idle, idle_cb);
		1638	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1639	ev_io_start (EV_DEFAULT_ &io);
		1640
		1641	In the "real" world, it might also be beneficial to start a timer, so that
		1642	low-priority connections can not be locked out forever under load. This
		1643	enables your program to keep a lower latency for important connections
		1644	during short periods of high load, while not completely locking out less
		1645	important ones.
972		1646
973		1647
974	=head1 WATCHER TYPES	1648	=head1 WATCHER TYPES
975		1649
976	This section describes each watcher in detail, but will not repeat	1650	This section describes each watcher in detail, but will not repeat
977	information given in the last section. Any initialisation/set macros,	1651	information given in the last section. Any initialisation/set macros,
978	functions and members specific to the watcher type are explained.	1652	functions and members specific to the watcher type are explained.
979		1653
980	Members are additionally marked with either I<[read-only]>, meaning that,	1654	Most members are additionally marked with either I<[read-only]>, meaning
981	while the watcher is active, you can look at the member and expect some	1655	that, while the watcher is active, you can look at the member and expect
982	sensible content, but you must not modify it (you can modify it while the	1656	some sensible content, but you must not modify it (you can modify it while
983	watcher is stopped to your hearts content), or I<[read-write]>, which	1657	the watcher is stopped to your hearts content), or I<[read-write]>, which
984	means you can expect it to have some sensible content while the watcher	1658	means you can expect it to have some sensible content while the watcher
985	is active, but you can also modify it. Modifying it may not do something	1659	is active, but you can also modify it. Modifying it may not do something
986	sensible or take immediate effect (or do anything at all), but libev will	1660	sensible or take immediate effect (or do anything at all), but libev will
987	not crash or malfunction in any way.	1661	not crash or malfunction in any way.
988		1662
		1663	In any case, the documentation for each member will explain what the
		1664	effects are, and if there are any additional access restrictions.
989		1665
990	=head2 C<ev_io> - is this file descriptor readable or writable?	1666	=head2 C<ev_io> - is this file descriptor readable or writable?
991		1667
992	I/O watchers check whether a file descriptor is readable or writable	1668	I/O watchers check whether a file descriptor is readable or writable
993	in each iteration of the event loop, or, more precisely, when reading	1669	in each iteration of the event loop, or, more precisely, when reading
…		…
1000	In general you can register as many read and/or write event watchers per	1676	In general you can register as many read and/or write event watchers per
1001	fd as you want (as long as you don't confuse yourself). Setting all file	1677	fd as you want (as long as you don't confuse yourself). Setting all file
1002	descriptors to non-blocking mode is also usually a good idea (but not	1678	descriptors to non-blocking mode is also usually a good idea (but not
1003	required if you know what you are doing).	1679	required if you know what you are doing).
1004		1680
1005	If you must do this, then force the use of a known-to-be-good backend
1006	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
1007	C<EVBACKEND_POLL>).
1008
1009	Another thing you have to watch out for is that it is quite easy to	1681	Another thing you have to watch out for is that it is quite easy to
1010	receive "spurious" readyness notifications, that is your callback might	1682	receive "spurious" readiness notifications, that is, your callback might
1011	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1683	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1012	because there is no data. Not only are some backends known to create a	1684	because there is no data. It is very easy to get into this situation even
1013	lot of those (for example solaris ports), it is very easy to get into	1685	with a relatively standard program structure. Thus it is best to always
1014	this situation even with a relatively standard program structure. Thus	1686	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1015	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1016	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1687	preferable to a program hanging until some data arrives.
1017		1688
1018	If you cannot run the fd in non-blocking mode (for example you should not	1689	If you cannot run the fd in non-blocking mode (for example you should
1019	play around with an Xlib connection), then you have to seperately re-test	1690	not play around with an Xlib connection), then you have to separately
1020	whether a file descriptor is really ready with a known-to-be good interface	1691	re-test whether a file descriptor is really ready with a known-to-be good
1021	such as poll (fortunately in our Xlib example, Xlib already does this on	1692	interface such as poll (fortunately in the case of Xlib, it already does
1022	its own, so its quite safe to use).	1693	this on its own, so its quite safe to use). Some people additionally
		1694	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1695	indefinitely.
		1696
		1697	But really, best use non-blocking mode.
1023		1698
1024	=head3 The special problem of disappearing file descriptors	1699	=head3 The special problem of disappearing file descriptors
1025		1700
1026	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1701	Some backends (e.g. kqueue, epoll, linuxaio) need to be told about closing
1027	descriptor (either by calling C<close> explicitly or by any other means,	1702	a file descriptor (either due to calling C<close> explicitly or any other
1028	such as C<dup>). The reason is that you register interest in some file	1703	means, such as C<dup2>). The reason is that you register interest in some
1029	descriptor, but when it goes away, the operating system will silently drop	1704	file descriptor, but when it goes away, the operating system will silently
1030	this interest. If another file descriptor with the same number then is	1705	drop this interest. If another file descriptor with the same number then
1031	registered with libev, there is no efficient way to see that this is, in	1706	is registered with libev, there is no efficient way to see that this is,
1032	fact, a different file descriptor.	1707	in fact, a different file descriptor.
1033		1708
1034	To avoid having to explicitly tell libev about such cases, libev follows	1709	To avoid having to explicitly tell libev about such cases, libev follows
1035	the following policy: Each time C<ev_io_set> is being called, libev	1710	the following policy: Each time C<ev_io_set> is being called, libev
1036	will assume that this is potentially a new file descriptor, otherwise	1711	will assume that this is potentially a new file descriptor, otherwise
1037	it is assumed that the file descriptor stays the same. That means that	1712	it is assumed that the file descriptor stays the same. That means that
…		…
1051		1726
1052	There is no workaround possible except not registering events	1727	There is no workaround possible except not registering events
1053	for potentially C<dup ()>'ed file descriptors, or to resort to	1728	for potentially C<dup ()>'ed file descriptors, or to resort to
1054	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1729	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1055		1730
		1731	=head3 The special problem of files
		1732
		1733	Many people try to use C<select> (or libev) on file descriptors
		1734	representing files, and expect it to become ready when their program
		1735	doesn't block on disk accesses (which can take a long time on their own).
		1736
		1737	However, this cannot ever work in the "expected" way - you get a readiness
		1738	notification as soon as the kernel knows whether and how much data is
		1739	there, and in the case of open files, that's always the case, so you
		1740	always get a readiness notification instantly, and your read (or possibly
		1741	write) will still block on the disk I/O.
		1742
		1743	Another way to view it is that in the case of sockets, pipes, character
		1744	devices and so on, there is another party (the sender) that delivers data
		1745	on its own, but in the case of files, there is no such thing: the disk
		1746	will not send data on its own, simply because it doesn't know what you
		1747	wish to read - you would first have to request some data.
		1748
		1749	Since files are typically not-so-well supported by advanced notification
		1750	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1751	to files, even though you should not use it. The reason for this is
		1752	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1753	usually a tty, often a pipe, but also sometimes files or special devices
		1754	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1755	F</dev/urandom>), and even though the file might better be served with
		1756	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1757	it "just works" instead of freezing.
		1758
		1759	So avoid file descriptors pointing to files when you know it (e.g. use
		1760	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1761	when you rarely read from a file instead of from a socket, and want to
		1762	reuse the same code path.
		1763
1056	=head3 The special problem of fork	1764	=head3 The special problem of fork
1057		1765
1058	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1766	Some backends (epoll, kqueue, linuxaio, iouring) do not support C<fork ()>
1059	useless behaviour. Libev fully supports fork, but needs to be told about	1767	at all or exhibit useless behaviour. Libev fully supports fork, but needs
1060	it in the child.	1768	to be told about it in the child if you want to continue to use it in the
		1769	child.
1061		1770
1062	To support fork in your programs, you either have to call	1771	To support fork in your child processes, you have to call C<ev_loop_fork
1063	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1772	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1064	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1773	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1065	C<EVBACKEND_POLL>.
1066		1774
		1775	=head3 The special problem of SIGPIPE
		1776
		1777	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1778	when writing to a pipe whose other end has been closed, your program gets
		1779	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1780	this is sensible behaviour, for daemons, this is usually undesirable.
		1781
		1782	So when you encounter spurious, unexplained daemon exits, make sure you
		1783	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1784	somewhere, as that would have given you a big clue).
		1785
		1786	=head3 The special problem of accept()ing when you can't
		1787
		1788	Many implementations of the POSIX C<accept> function (for example,
		1789	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1790	connection from the pending queue in all error cases.
		1791
		1792	For example, larger servers often run out of file descriptors (because
		1793	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1794	rejecting the connection, leading to libev signalling readiness on
		1795	the next iteration again (the connection still exists after all), and
		1796	typically causing the program to loop at 100% CPU usage.
		1797
		1798	Unfortunately, the set of errors that cause this issue differs between
		1799	operating systems, there is usually little the app can do to remedy the
		1800	situation, and no known thread-safe method of removing the connection to
		1801	cope with overload is known (to me).
		1802
		1803	One of the easiest ways to handle this situation is to just ignore it
		1804	- when the program encounters an overload, it will just loop until the
		1805	situation is over. While this is a form of busy waiting, no OS offers an
		1806	event-based way to handle this situation, so it's the best one can do.
		1807
		1808	A better way to handle the situation is to log any errors other than
		1809	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1810	messages, and continue as usual, which at least gives the user an idea of
		1811	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1812	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1813	usage.
		1814
		1815	If your program is single-threaded, then you could also keep a dummy file
		1816	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1817	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1818	close that fd, and create a new dummy fd. This will gracefully refuse
		1819	clients under typical overload conditions.
		1820
		1821	The last way to handle it is to simply log the error and C<exit>, as
		1822	is often done with C<malloc> failures, but this results in an easy
		1823	opportunity for a DoS attack.
1067		1824
1068	=head3 Watcher-Specific Functions	1825	=head3 Watcher-Specific Functions
1069		1826
1070	=over 4	1827	=over 4
1071		1828
1072	=item ev_io_init (ev_io *, callback, int fd, int events)	1829	=item ev_io_init (ev_io *, callback, int fd, int events)
1073		1830
1074	=item ev_io_set (ev_io *, int fd, int events)	1831	=item ev_io_set (ev_io *, int fd, int events)
1075		1832
1076	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1833	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1077	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or	1834	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE>, both
1078	C<EV_READ | EV_WRITE> to receive the given events.	1835	C<EV_READ | EV_WRITE> or C<0>, to express the desire to receive the given
		1836	events.
1079		1837
1080	=item int fd [read-only]	1838	Note that setting the C<events> to C<0> and starting the watcher is
		1839	supported, but not specially optimized - if your program sometimes happens
		1840	to generate this combination this is fine, but if it is easy to avoid
		1841	starting an io watcher watching for no events you should do so.
1081		1842
1082	The file descriptor being watched.	1843	=item ev_io_modify (ev_io *, int events)
1083		1844
		1845	Similar to C<ev_io_set>, but only changes the event mask. Using this might
		1846	be faster with some backends, as libev can assume that the C<fd> still
		1847	refers to the same underlying file description, something it cannot do
		1848	when using C<ev_io_set>.
		1849
		1850	=item int fd [no-modify]
		1851
		1852	The file descriptor being watched. While it can be read at any time, you
		1853	must not modify this member even when the watcher is stopped - always use
		1854	C<ev_io_set> for that.
		1855
1084	=item int events [read-only]	1856	=item int events [no-modify]
1085		1857
1086	The events being watched.	1858	The set of events the fd is being watched for, among other flags. Remember
		1859	that this is a bit set - to test for C<EV_READ>, use C<< w->events &
		1860	EV_READ >>, and similarly for C<EV_WRITE>.
		1861
		1862	As with C<fd>, you must not modify this member even when the watcher is
		1863	stopped, always use C<ev_io_set> or C<ev_io_modify> for that.
1087		1864
1088	=back	1865	=back
1089		1866
1090	=head3 Examples	1867	=head3 Examples
1091		1868
1092	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1869	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1093	readable, but only once. Since it is likely line-buffered, you could	1870	readable, but only once. Since it is likely line-buffered, you could
1094	attempt to read a whole line in the callback.	1871	attempt to read a whole line in the callback.
1095		1872
1096	static void	1873	static void
1097	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1874	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1098	{	1875	{
1099	ev_io_stop (loop, w);	1876	ev_io_stop (loop, w);
1100	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1877	.. read from stdin here (or from w->fd) and handle any I/O errors
1101	}	1878	}
1102		1879
1103	...	1880	...
1104	struct ev_loop *loop = ev_default_init (0);	1881	struct ev_loop *loop = ev_default_init (0);
1105	struct ev_io stdin_readable;	1882	ev_io stdin_readable;
1106	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1883	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1107	ev_io_start (loop, &stdin_readable);	1884	ev_io_start (loop, &stdin_readable);
1108	ev_loop (loop, 0);	1885	ev_run (loop, 0);
1109		1886
1110		1887
1111	=head2 C<ev_timer> - relative and optionally repeating timeouts	1888	=head2 C<ev_timer> - relative and optionally repeating timeouts
1112		1889
1113	Timer watchers are simple relative timers that generate an event after a	1890	Timer watchers are simple relative timers that generate an event after a
1114	given time, and optionally repeating in regular intervals after that.	1891	given time, and optionally repeating in regular intervals after that.
1115		1892
1116	The timers are based on real time, that is, if you register an event that	1893	The timers are based on real time, that is, if you register an event that
1117	times out after an hour and you reset your system clock to last years	1894	times out after an hour and you reset your system clock to January last
1118	time, it will still time out after (roughly) and hour. "Roughly" because	1895	year, it will still time out after (roughly) one hour. "Roughly" because
1119	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1896	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1120	monotonic clock option helps a lot here).	1897	monotonic clock option helps a lot here).
		1898
		1899	The callback is guaranteed to be invoked only I<after> its timeout has
		1900	passed (not I<at>, so on systems with very low-resolution clocks this
		1901	might introduce a small delay, see "the special problem of being too
		1902	early", below). If multiple timers become ready during the same loop
		1903	iteration then the ones with earlier time-out values are invoked before
		1904	ones of the same priority with later time-out values (but this is no
		1905	longer true when a callback calls C<ev_run> recursively).
		1906
		1907	=head3 Be smart about timeouts
		1908
		1909	Many real-world problems involve some kind of timeout, usually for error
		1910	recovery. A typical example is an HTTP request - if the other side hangs,
		1911	you want to raise some error after a while.
		1912
		1913	What follows are some ways to handle this problem, from obvious and
		1914	inefficient to smart and efficient.
		1915
		1916	In the following, a 60 second activity timeout is assumed - a timeout that
		1917	gets reset to 60 seconds each time there is activity (e.g. each time some
		1918	data or other life sign was received).
		1919
		1920	=over 4
		1921
		1922	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1923
		1924	This is the most obvious, but not the most simple way: In the beginning,
		1925	start the watcher:
		1926
		1927	ev_timer_init (timer, callback, 60., 0.);
		1928	ev_timer_start (loop, timer);
		1929
		1930	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1931	and start it again:
		1932
		1933	ev_timer_stop (loop, timer);
		1934	ev_timer_set (timer, 60., 0.);
		1935	ev_timer_start (loop, timer);
		1936
		1937	This is relatively simple to implement, but means that each time there is
		1938	some activity, libev will first have to remove the timer from its internal
		1939	data structure and then add it again. Libev tries to be fast, but it's
		1940	still not a constant-time operation.
		1941
		1942	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1943
		1944	This is the easiest way, and involves using C<ev_timer_again> instead of
		1945	C<ev_timer_start>.
		1946
		1947	To implement this, configure an C<ev_timer> with a C<repeat> value
		1948	of C<60> and then call C<ev_timer_again> at start and each time you
		1949	successfully read or write some data. If you go into an idle state where
		1950	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1951	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1952
		1953	That means you can ignore both the C<ev_timer_start> function and the
		1954	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1955	member and C<ev_timer_again>.
		1956
		1957	At start:
		1958
		1959	ev_init (timer, callback);
		1960	timer->repeat = 60.;
		1961	ev_timer_again (loop, timer);
		1962
		1963	Each time there is some activity:
		1964
		1965	ev_timer_again (loop, timer);
		1966
		1967	It is even possible to change the time-out on the fly, regardless of
		1968	whether the watcher is active or not:
		1969
		1970	timer->repeat = 30.;
		1971	ev_timer_again (loop, timer);
		1972
		1973	This is slightly more efficient then stopping/starting the timer each time
		1974	you want to modify its timeout value, as libev does not have to completely
		1975	remove and re-insert the timer from/into its internal data structure.
		1976
		1977	It is, however, even simpler than the "obvious" way to do it.
		1978
		1979	=item 3. Let the timer time out, but then re-arm it as required.
		1980
		1981	This method is more tricky, but usually most efficient: Most timeouts are
		1982	relatively long compared to the intervals between other activity - in
		1983	our example, within 60 seconds, there are usually many I/O events with
		1984	associated activity resets.
		1985
		1986	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1987	but remember the time of last activity, and check for a real timeout only
		1988	within the callback:
		1989
		1990	ev_tstamp timeout = 60.;
		1991	ev_tstamp last_activity; // time of last activity
		1992	ev_timer timer;
		1993
		1994	static void
		1995	callback (EV_P_ ev_timer *w, int revents)
		1996	{
		1997	// calculate when the timeout would happen
		1998	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
		1999
		2000	// if negative, it means we the timeout already occurred
		2001	if (after < 0.)
		2002	{
		2003	// timeout occurred, take action
		2004	}
		2005	else
		2006	{
		2007	// callback was invoked, but there was some recent
		2008	// activity. simply restart the timer to time out
		2009	// after "after" seconds, which is the earliest time
		2010	// the timeout can occur.
		2011	ev_timer_set (w, after, 0.);
		2012	ev_timer_start (EV_A_ w);
		2013	}
		2014	}
		2015
		2016	To summarise the callback: first calculate in how many seconds the
		2017	timeout will occur (by calculating the absolute time when it would occur,
		2018	C<last_activity + timeout>, and subtracting the current time, C<ev_now
		2019	(EV_A)> from that).
		2020
		2021	If this value is negative, then we are already past the timeout, i.e. we
		2022	timed out, and need to do whatever is needed in this case.
		2023
		2024	Otherwise, we now the earliest time at which the timeout would trigger,
		2025	and simply start the timer with this timeout value.
		2026
		2027	In other words, each time the callback is invoked it will check whether
		2028	the timeout occurred. If not, it will simply reschedule itself to check
		2029	again at the earliest time it could time out. Rinse. Repeat.
		2030
		2031	This scheme causes more callback invocations (about one every 60 seconds
		2032	minus half the average time between activity), but virtually no calls to
		2033	libev to change the timeout.
		2034
		2035	To start the machinery, simply initialise the watcher and set
		2036	C<last_activity> to the current time (meaning there was some activity just
		2037	now), then call the callback, which will "do the right thing" and start
		2038	the timer:
		2039
		2040	last_activity = ev_now (EV_A);
		2041	ev_init (&timer, callback);
		2042	callback (EV_A_ &timer, 0);
		2043
		2044	When there is some activity, simply store the current time in
		2045	C<last_activity>, no libev calls at all:
		2046
		2047	if (activity detected)
		2048	last_activity = ev_now (EV_A);
		2049
		2050	When your timeout value changes, then the timeout can be changed by simply
		2051	providing a new value, stopping the timer and calling the callback, which
		2052	will again do the right thing (for example, time out immediately :).
		2053
		2054	timeout = new_value;
		2055	ev_timer_stop (EV_A_ &timer);
		2056	callback (EV_A_ &timer, 0);
		2057
		2058	This technique is slightly more complex, but in most cases where the
		2059	time-out is unlikely to be triggered, much more efficient.
		2060
		2061	=item 4. Wee, just use a double-linked list for your timeouts.
		2062
		2063	If there is not one request, but many thousands (millions...), all
		2064	employing some kind of timeout with the same timeout value, then one can
		2065	do even better:
		2066
		2067	When starting the timeout, calculate the timeout value and put the timeout
		2068	at the I<end> of the list.
		2069
		2070	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		2071	the list is expected to fire (for example, using the technique #3).
		2072
		2073	When there is some activity, remove the timer from the list, recalculate
		2074	the timeout, append it to the end of the list again, and make sure to
		2075	update the C<ev_timer> if it was taken from the beginning of the list.
		2076
		2077	This way, one can manage an unlimited number of timeouts in O(1) time for
		2078	starting, stopping and updating the timers, at the expense of a major
		2079	complication, and having to use a constant timeout. The constant timeout
		2080	ensures that the list stays sorted.
		2081
		2082	=back
		2083
		2084	So which method the best?
		2085
		2086	Method #2 is a simple no-brain-required solution that is adequate in most
		2087	situations. Method #3 requires a bit more thinking, but handles many cases
		2088	better, and isn't very complicated either. In most case, choosing either
		2089	one is fine, with #3 being better in typical situations.
		2090
		2091	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		2092	rather complicated, but extremely efficient, something that really pays
		2093	off after the first million or so of active timers, i.e. it's usually
		2094	overkill :)
		2095
		2096	=head3 The special problem of being too early
		2097
		2098	If you ask a timer to call your callback after three seconds, then
		2099	you expect it to be invoked after three seconds - but of course, this
		2100	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		2101	guaranteed to any precision by libev - imagine somebody suspending the
		2102	process with a STOP signal for a few hours for example.
		2103
		2104	So, libev tries to invoke your callback as soon as possible I<after> the
		2105	delay has occurred, but cannot guarantee this.
		2106
		2107	A less obvious failure mode is calling your callback too early: many event
		2108	loops compare timestamps with a "elapsed delay >= requested delay", but
		2109	this can cause your callback to be invoked much earlier than you would
		2110	expect.
		2111
		2112	To see why, imagine a system with a clock that only offers full second
		2113	resolution (think windows if you can't come up with a broken enough OS
		2114	yourself). If you schedule a one-second timer at the time 500.9, then the
		2115	event loop will schedule your timeout to elapse at a system time of 500
		2116	(500.9 truncated to the resolution) + 1, or 501.
		2117
		2118	If an event library looks at the timeout 0.1s later, it will see "501 >=
		2119	501" and invoke the callback 0.1s after it was started, even though a
		2120	one-second delay was requested - this is being "too early", despite best
		2121	intentions.
		2122
		2123	This is the reason why libev will never invoke the callback if the elapsed
		2124	delay equals the requested delay, but only when the elapsed delay is
		2125	larger than the requested delay. In the example above, libev would only invoke
		2126	the callback at system time 502, or 1.1s after the timer was started.
		2127
		2128	So, while libev cannot guarantee that your callback will be invoked
		2129	exactly when requested, it I<can> and I<does> guarantee that the requested
		2130	delay has actually elapsed, or in other words, it always errs on the "too
		2131	late" side of things.
		2132
		2133	=head3 The special problem of time updates
		2134
		2135	Establishing the current time is a costly operation (it usually takes
		2136	at least one system call): EV therefore updates its idea of the current
		2137	time only before and after C<ev_run> collects new events, which causes a
		2138	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		2139	lots of events in one iteration.
1121		2140
1122	The relative timeouts are calculated relative to the C<ev_now ()>	2141	The relative timeouts are calculated relative to the C<ev_now ()>
1123	time. This is usually the right thing as this timestamp refers to the time	2142	time. This is usually the right thing as this timestamp refers to the time
1124	of the event triggering whatever timeout you are modifying/starting. If	2143	of the event triggering whatever timeout you are modifying/starting. If
1125	you suspect event processing to be delayed and you I<need> to base the timeout	2144	you suspect event processing to be delayed and you I<need> to base the
1126	on the current time, use something like this to adjust for this:	2145	timeout on the current time, use something like the following to adjust
		2146	for it:
1127		2147
1128	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2148	ev_timer_set (&timer, after + (ev_time () - ev_now ()), 0.);
1129		2149
1130	The callback is guarenteed to be invoked only when its timeout has passed,	2150	If the event loop is suspended for a long time, you can also force an
1131	but if multiple timers become ready during the same loop iteration then	2151	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1132	order of execution is undefined.	2152	()>, although that will push the event time of all outstanding events
		2153	further into the future.
		2154
		2155	=head3 The special problem of unsynchronised clocks
		2156
		2157	Modern systems have a variety of clocks - libev itself uses the normal
		2158	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2159	jumps).
		2160
		2161	Neither of these clocks is synchronised with each other or any other clock
		2162	on the system, so C<ev_time ()> might return a considerably different time
		2163	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2164	a call to C<gettimeofday> might return a second count that is one higher
		2165	than a directly following call to C<time>.
		2166
		2167	The moral of this is to only compare libev-related timestamps with
		2168	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2169	a second or so.
		2170
		2171	One more problem arises due to this lack of synchronisation: if libev uses
		2172	the system monotonic clock and you compare timestamps from C<ev_time>
		2173	or C<ev_now> from when you started your timer and when your callback is
		2174	invoked, you will find that sometimes the callback is a bit "early".
		2175
		2176	This is because C<ev_timer>s work in real time, not wall clock time, so
		2177	libev makes sure your callback is not invoked before the delay happened,
		2178	I<measured according to the real time>, not the system clock.
		2179
		2180	If your timeouts are based on a physical timescale (e.g. "time out this
		2181	connection after 100 seconds") then this shouldn't bother you as it is
		2182	exactly the right behaviour.
		2183
		2184	If you want to compare wall clock/system timestamps to your timers, then
		2185	you need to use C<ev_periodic>s, as these are based on the wall clock
		2186	time, where your comparisons will always generate correct results.
		2187
		2188	=head3 The special problems of suspended animation
		2189
		2190	When you leave the server world it is quite customary to hit machines that
		2191	can suspend/hibernate - what happens to the clocks during such a suspend?
		2192
		2193	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		2194	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		2195	to run until the system is suspended, but they will not advance while the
		2196	system is suspended. That means, on resume, it will be as if the program
		2197	was frozen for a few seconds, but the suspend time will not be counted
		2198	towards C<ev_timer> when a monotonic clock source is used. The real time
		2199	clock advanced as expected, but if it is used as sole clocksource, then a
		2200	long suspend would be detected as a time jump by libev, and timers would
		2201	be adjusted accordingly.
		2202
		2203	I would not be surprised to see different behaviour in different between
		2204	operating systems, OS versions or even different hardware.
		2205
		2206	The other form of suspend (job control, or sending a SIGSTOP) will see a
		2207	time jump in the monotonic clocks and the realtime clock. If the program
		2208	is suspended for a very long time, and monotonic clock sources are in use,
		2209	then you can expect C<ev_timer>s to expire as the full suspension time
		2210	will be counted towards the timers. When no monotonic clock source is in
		2211	use, then libev will again assume a timejump and adjust accordingly.
		2212
		2213	It might be beneficial for this latter case to call C<ev_suspend>
		2214	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		2215	deterministic behaviour in this case (you can do nothing against
		2216	C<SIGSTOP>).
1133		2217
1134	=head3 Watcher-Specific Functions and Data Members	2218	=head3 Watcher-Specific Functions and Data Members
1135		2219
1136	=over 4	2220	=over 4
1137		2221
1138	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2222	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1139		2223
1140	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2224	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1141		2225
1142	Configure the timer to trigger after C<after> seconds. If C<repeat> is	2226	Configure the timer to trigger after C<after> seconds (fractional and
1143	C<0.>, then it will automatically be stopped. If it is positive, then the	2227	negative values are supported). If C<repeat> is C<0.>, then it will
		2228	automatically be stopped once the timeout is reached. If it is positive,
1144	timer will automatically be configured to trigger again C<repeat> seconds	2229	then the timer will automatically be configured to trigger again C<repeat>
1145	later, again, and again, until stopped manually.	2230	seconds later, again, and again, until stopped manually.
1146		2231
1147	The timer itself will do a best-effort at avoiding drift, that is, if you	2232	The timer itself will do a best-effort at avoiding drift, that is, if
1148	configure a timer to trigger every 10 seconds, then it will trigger at	2233	you configure a timer to trigger every 10 seconds, then it will normally
1149	exactly 10 second intervals. If, however, your program cannot keep up with	2234	trigger at exactly 10 second intervals. If, however, your program cannot
1150	the timer (because it takes longer than those 10 seconds to do stuff) the	2235	keep up with the timer (because it takes longer than those 10 seconds to
1151	timer will not fire more than once per event loop iteration.	2236	do stuff) the timer will not fire more than once per event loop iteration.
1152		2237
1153	=item ev_timer_again (loop)	2238	=item ev_timer_again (loop, ev_timer *)
1154		2239
1155	This will act as if the timer timed out and restart it again if it is	2240	This will act as if the timer timed out, and restarts it again if it is
1156	repeating. The exact semantics are:	2241	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2242	timeout to the C<repeat> value and calling C<ev_timer_start>.
1157		2243
		2244	The exact semantics are as in the following rules, all of which will be
		2245	applied to the watcher:
		2246
		2247	=over 4
		2248
1158	If the timer is pending, its pending status is cleared.	2249	=item If the timer is pending, the pending status is always cleared.
1159		2250
1160	If the timer is started but nonrepeating, stop it (as if it timed out).	2251	=item If the timer is started but non-repeating, stop it (as if it timed
		2252	out, without invoking it).
1161		2253
1162	If the timer is repeating, either start it if necessary (with the	2254	=item If the timer is repeating, make the C<repeat> value the new timeout
1163	C<repeat> value), or reset the running timer to the C<repeat> value.	2255	and start the timer, if necessary.
1164		2256
1165	This sounds a bit complicated, but here is a useful and typical	2257	=back
1166	example: Imagine you have a tcp connection and you want a so-called idle
1167	timeout, that is, you want to be called when there have been, say, 60
1168	seconds of inactivity on the socket. The easiest way to do this is to
1169	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1170	C<ev_timer_again> each time you successfully read or write some data. If
1171	you go into an idle state where you do not expect data to travel on the
1172	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1173	automatically restart it if need be.
1174		2258
1175	That means you can ignore the C<after> value and C<ev_timer_start>	2259	This sounds a bit complicated, see L</Be smart about timeouts>, above, for a
1176	altogether and only ever use the C<repeat> value and C<ev_timer_again>:	2260	usage example.
1177		2261
1178	ev_timer_init (timer, callback, 0., 5.);	2262	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1179	ev_timer_again (loop, timer);
1180	...
1181	timer->again = 17.;
1182	ev_timer_again (loop, timer);
1183	...
1184	timer->again = 10.;
1185	ev_timer_again (loop, timer);
1186		2263
1187	This is more slightly efficient then stopping/starting the timer each time	2264	Returns the remaining time until a timer fires. If the timer is active,
1188	you want to modify its timeout value.	2265	then this time is relative to the current event loop time, otherwise it's
		2266	the timeout value currently configured.
		2267
		2268	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
		2269	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		2270	will return C<4>. When the timer expires and is restarted, it will return
		2271	roughly C<7> (likely slightly less as callback invocation takes some time,
		2272	too), and so on.
1189		2273
1190	=item ev_tstamp repeat [read-write]	2274	=item ev_tstamp repeat [read-write]
1191		2275
1192	The current C<repeat> value. Will be used each time the watcher times out	2276	The current C<repeat> value. Will be used each time the watcher times out
1193	or C<ev_timer_again> is called and determines the next timeout (if any),	2277	or C<ev_timer_again> is called, and determines the next timeout (if any),
1194	which is also when any modifications are taken into account.	2278	which is also when any modifications are taken into account.
1195		2279
1196	=back	2280	=back
1197		2281
1198	=head3 Examples	2282	=head3 Examples
1199		2283
1200	Example: Create a timer that fires after 60 seconds.	2284	Example: Create a timer that fires after 60 seconds.
1201		2285
1202	static void	2286	static void
1203	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2287	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1204	{	2288	{
1205	.. one minute over, w is actually stopped right here	2289	.. one minute over, w is actually stopped right here
1206	}	2290	}
1207		2291
1208	struct ev_timer mytimer;	2292	ev_timer mytimer;
1209	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	2293	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1210	ev_timer_start (loop, &mytimer);	2294	ev_timer_start (loop, &mytimer);
1211		2295
1212	Example: Create a timeout timer that times out after 10 seconds of	2296	Example: Create a timeout timer that times out after 10 seconds of
1213	inactivity.	2297	inactivity.
1214		2298
1215	static void	2299	static void
1216	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2300	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1217	{	2301	{
1218	.. ten seconds without any activity	2302	.. ten seconds without any activity
1219	}	2303	}
1220		2304
1221	struct ev_timer mytimer;	2305	ev_timer mytimer;
1222	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2306	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1223	ev_timer_again (&mytimer); /* start timer */	2307	ev_timer_again (&mytimer); /* start timer */
1224	ev_loop (loop, 0);	2308	ev_run (loop, 0);
1225		2309
1226	// and in some piece of code that gets executed on any "activity":	2310	// and in some piece of code that gets executed on any "activity":
1227	// reset the timeout to start ticking again at 10 seconds	2311	// reset the timeout to start ticking again at 10 seconds
1228	ev_timer_again (&mytimer);	2312	ev_timer_again (&mytimer);
1229		2313
1230		2314
1231	=head2 C<ev_periodic> - to cron or not to cron?	2315	=head2 C<ev_periodic> - to cron or not to cron?
1232		2316
1233	Periodic watchers are also timers of a kind, but they are very versatile	2317	Periodic watchers are also timers of a kind, but they are very versatile
1234	(and unfortunately a bit complex).	2318	(and unfortunately a bit complex).
1235		2319
1236	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	2320	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1237	but on wallclock time (absolute time). You can tell a periodic watcher	2321	relative time, the physical time that passes) but on wall clock time
1238	to trigger "at" some specific point in time. For example, if you tell a	2322	(absolute time, the thing you can read on your calendar or clock). The
1239	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	2323	difference is that wall clock time can run faster or slower than real
1240	+ 10.>) and then reset your system clock to the last year, then it will	2324	time, and time jumps are not uncommon (e.g. when you adjust your
1241	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	2325	wrist-watch).
1242	roughly 10 seconds later).
1243		2326
1244	They can also be used to implement vastly more complex timers, such as	2327	You can tell a periodic watcher to trigger after some specific point
1245	triggering an event on each midnight, local time or other, complicated,	2328	in time: for example, if you tell a periodic watcher to trigger "in 10
1246	rules.	2329	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		2330	not a delay) and then reset your system clock to January of the previous
		2331	year, then it will take a year or more to trigger the event (unlike an
		2332	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		2333	it, as it uses a relative timeout).
1247		2334
		2335	C<ev_periodic> watchers can also be used to implement vastly more complex
		2336	timers, such as triggering an event on each "midnight, local time", or
		2337	other complicated rules. This cannot easily be done with C<ev_timer>
		2338	watchers, as those cannot react to time jumps.
		2339
1248	As with timers, the callback is guarenteed to be invoked only when the	2340	As with timers, the callback is guaranteed to be invoked only when the
1249	time (C<at>) has been passed, but if multiple periodic timers become ready	2341	point in time where it is supposed to trigger has passed. If multiple
1250	during the same loop iteration then order of execution is undefined.	2342	timers become ready during the same loop iteration then the ones with
		2343	earlier time-out values are invoked before ones with later time-out values
		2344	(but this is no longer true when a callback calls C<ev_run> recursively).
1251		2345
1252	=head3 Watcher-Specific Functions and Data Members	2346	=head3 Watcher-Specific Functions and Data Members
1253		2347
1254	=over 4	2348	=over 4
1255		2349
1256	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	2350	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1257		2351
1258	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2352	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1259		2353
1260	Lots of arguments, lets sort it out... There are basically three modes of	2354	Lots of arguments, let's sort it out... There are basically three modes of
1261	operation, and we will explain them from simplest to complex:	2355	operation, and we will explain them from simplest to most complex:
1262		2356
1263	=over 4	2357	=over 4
1264		2358
1265	=item * absolute timer (at = time, interval = reschedule_cb = 0)	2359	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1266		2360
1267	In this configuration the watcher triggers an event at the wallclock time	2361	In this configuration the watcher triggers an event after the wall clock
1268	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	2362	time C<offset> has passed. It will not repeat and will not adjust when a
1269	that is, if it is to be run at January 1st 2011 then it will run when the	2363	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1270	system time reaches or surpasses this time.	2364	will be stopped and invoked when the system clock reaches or surpasses
		2365	this point in time.
1271		2366
1272	=item * non-repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	2367	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1273		2368
1274	In this mode the watcher will always be scheduled to time out at the next	2369	In this mode the watcher will always be scheduled to time out at the next
1275	C<at + N * interval> time (for some integer N, which can also be negative)	2370	C<offset + N * interval> time (for some integer N, which can also be
1276	and then repeat, regardless of any time jumps.	2371	negative) and then repeat, regardless of any time jumps. The C<offset>
		2372	argument is merely an offset into the C<interval> periods.
1277		2373
1278	This can be used to create timers that do not drift with respect to system	2374	This can be used to create timers that do not drift with respect to the
1279	time:	2375	system clock, for example, here is an C<ev_periodic> that triggers each
		2376	hour, on the hour (with respect to UTC):
1280		2377
1281	ev_periodic_set (&periodic, 0., 3600., 0);	2378	ev_periodic_set (&periodic, 0., 3600., 0);
1282		2379
1283	This doesn't mean there will always be 3600 seconds in between triggers,	2380	This doesn't mean there will always be 3600 seconds in between triggers,
1284	but only that the the callback will be called when the system time shows a	2381	but only that the callback will be called when the system time shows a
1285	full hour (UTC), or more correctly, when the system time is evenly divisible	2382	full hour (UTC), or more correctly, when the system time is evenly divisible
1286	by 3600.	2383	by 3600.
1287		2384
1288	Another way to think about it (for the mathematically inclined) is that	2385	Another way to think about it (for the mathematically inclined) is that
1289	C<ev_periodic> will try to run the callback in this mode at the next possible	2386	C<ev_periodic> will try to run the callback in this mode at the next possible
1290	time where C<time = at (mod interval)>, regardless of any time jumps.	2387	time where C<time = offset (mod interval)>, regardless of any time jumps.
1291		2388
1292	For numerical stability it is preferable that the C<at> value is near	2389	The C<interval> I<MUST> be positive, and for numerical stability, the
1293	C<ev_now ()> (the current time), but there is no range requirement for	2390	interval value should be higher than C<1/8192> (which is around 100
1294	this value.	2391	microseconds) and C<offset> should be higher than C<0> and should have
		2392	at most a similar magnitude as the current time (say, within a factor of
		2393	ten). Typical values for offset are, in fact, C<0> or something between
		2394	C<0> and C<interval>, which is also the recommended range.
1295		2395
		2396	Note also that there is an upper limit to how often a timer can fire (CPU
		2397	speed for example), so if C<interval> is very small then timing stability
		2398	will of course deteriorate. Libev itself tries to be exact to be about one
		2399	millisecond (if the OS supports it and the machine is fast enough).
		2400
1296	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2401	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1297		2402
1298	In this mode the values for C<interval> and C<at> are both being	2403	In this mode the values for C<interval> and C<offset> are both being
1299	ignored. Instead, each time the periodic watcher gets scheduled, the	2404	ignored. Instead, each time the periodic watcher gets scheduled, the
1300	reschedule callback will be called with the watcher as first, and the	2405	reschedule callback will be called with the watcher as first, and the
1301	current time as second argument.	2406	current time as second argument.
1302		2407
1303	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2408	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1304	ever, or make any event loop modifications>. If you need to stop it,	2409	or make ANY other event loop modifications whatsoever, unless explicitly
1305	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	2410	allowed by documentation here>.
1306	starting an C<ev_prepare> watcher, which is legal).
1307		2411
		2412	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		2413	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		2414	only event loop modification you are allowed to do).
		2415
1308	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	2416	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1309	ev_tstamp now)>, e.g.:	2417	*w, ev_tstamp now)>, e.g.:
1310		2418
		2419	static ev_tstamp
1311	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2420	my_rescheduler (ev_periodic *w, ev_tstamp now)
1312	{	2421	{
1313	return now + 60.;	2422	return now + 60.;
1314	}	2423	}
1315		2424
1316	It must return the next time to trigger, based on the passed time value	2425	It must return the next time to trigger, based on the passed time value
1317	(that is, the lowest time value larger than to the second argument). It	2426	(that is, the lowest time value larger than to the second argument). It
1318	will usually be called just before the callback will be triggered, but	2427	will usually be called just before the callback will be triggered, but
1319	might be called at other times, too.	2428	might be called at other times, too.
1320		2429
1321	NOTE: I<< This callback must always return a time that is later than the	2430	NOTE: I<< This callback must always return a time that is higher than or
1322	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	2431	equal to the passed C<now> value >>.
1323		2432
1324	This can be used to create very complex timers, such as a timer that	2433	This can be used to create very complex timers, such as a timer that
1325	triggers on each midnight, local time. To do this, you would calculate the	2434	triggers on "next midnight, local time". To do this, you would calculate
1326	next midnight after C<now> and return the timestamp value for this. How	2435	the next midnight after C<now> and return the timestamp value for
1327	you do this is, again, up to you (but it is not trivial, which is the main	2436	this. Here is a (completely untested, no error checking) example on how to
1328	reason I omitted it as an example).	2437	do this:
		2438
		2439	#include <time.h>
		2440
		2441	static ev_tstamp
		2442	my_rescheduler (ev_periodic *w, ev_tstamp now)
		2443	{
		2444	time_t tnow = (time_t)now;
		2445	struct tm tm;
		2446	localtime_r (&tnow, &tm);
		2447
		2448	tm.tm_sec = tm.tm_min = tm.tm_hour = 0; // midnight current day
		2449	++tm.tm_mday; // midnight next day
		2450
		2451	return mktime (&tm);
		2452	}
		2453
		2454	Note: this code might run into trouble on days that have more then two
		2455	midnights (beginning and end).
1329		2456
1330	=back	2457	=back
1331		2458
1332	=item ev_periodic_again (loop, ev_periodic *)	2459	=item ev_periodic_again (loop, ev_periodic *)
1333		2460
1334	Simply stops and restarts the periodic watcher again. This is only useful	2461	Simply stops and restarts the periodic watcher again. This is only useful
1335	when you changed some parameters or the reschedule callback would return	2462	when you changed some parameters or the reschedule callback would return
1336	a different time than the last time it was called (e.g. in a crond like	2463	a different time than the last time it was called (e.g. in a crond like
1337	program when the crontabs have changed).	2464	program when the crontabs have changed).
1338		2465
		2466	=item ev_tstamp ev_periodic_at (ev_periodic *)
		2467
		2468	When active, returns the absolute time that the watcher is supposed
		2469	to trigger next. This is not the same as the C<offset> argument to
		2470	C<ev_periodic_set>, but indeed works even in interval and manual
		2471	rescheduling modes.
		2472
1339	=item ev_tstamp offset [read-write]	2473	=item ev_tstamp offset [read-write]
1340		2474
1341	When repeating, this contains the offset value, otherwise this is the	2475	When repeating, this contains the offset value, otherwise this is the
1342	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2476	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2477	although libev might modify this value for better numerical stability).
1343		2478
1344	Can be modified any time, but changes only take effect when the periodic	2479	Can be modified any time, but changes only take effect when the periodic
1345	timer fires or C<ev_periodic_again> is being called.	2480	timer fires or C<ev_periodic_again> is being called.
1346		2481
1347	=item ev_tstamp interval [read-write]	2482	=item ev_tstamp interval [read-write]
1348		2483
1349	The current interval value. Can be modified any time, but changes only	2484	The current interval value. Can be modified any time, but changes only
1350	take effect when the periodic timer fires or C<ev_periodic_again> is being	2485	take effect when the periodic timer fires or C<ev_periodic_again> is being
1351	called.	2486	called.
1352		2487
1353	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2488	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1354		2489
1355	The current reschedule callback, or C<0>, if this functionality is	2490	The current reschedule callback, or C<0>, if this functionality is
1356	switched off. Can be changed any time, but changes only take effect when	2491	switched off. Can be changed any time, but changes only take effect when
1357	the periodic timer fires or C<ev_periodic_again> is being called.	2492	the periodic timer fires or C<ev_periodic_again> is being called.
1358		2493
1359	=item ev_tstamp at [read-only]
1360
1361	When active, contains the absolute time that the watcher is supposed to
1362	trigger next.
1363
1364	=back	2494	=back
1365		2495
1366	=head3 Examples	2496	=head3 Examples
1367		2497
1368	Example: Call a callback every hour, or, more precisely, whenever the	2498	Example: Call a callback every hour, or, more precisely, whenever the
1369	system clock is divisible by 3600. The callback invocation times have	2499	system time is divisible by 3600. The callback invocation times have
1370	potentially a lot of jittering, but good long-term stability.	2500	potentially a lot of jitter, but good long-term stability.
1371		2501
1372	static void	2502	static void
1373	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2503	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
1374	{	2504	{
1375	... its now a full hour (UTC, or TAI or whatever your clock follows)	2505	... its now a full hour (UTC, or TAI or whatever your clock follows)
1376	}	2506	}
1377		2507
1378	struct ev_periodic hourly_tick;	2508	ev_periodic hourly_tick;
1379	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2509	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1380	ev_periodic_start (loop, &hourly_tick);	2510	ev_periodic_start (loop, &hourly_tick);
1381		2511
1382	Example: The same as above, but use a reschedule callback to do it:	2512	Example: The same as above, but use a reschedule callback to do it:
1383		2513
1384	#include <math.h>	2514	#include <math.h>
1385		2515
1386	static ev_tstamp	2516	static ev_tstamp
1387	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2517	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1388	{	2518	{
1389	return fmod (now, 3600.) + 3600.;	2519	return now + (3600. - fmod (now, 3600.));
1390	}	2520	}
1391		2521
1392	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2522	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1393		2523
1394	Example: Call a callback every hour, starting now:	2524	Example: Call a callback every hour, starting now:
1395		2525
1396	struct ev_periodic hourly_tick;	2526	ev_periodic hourly_tick;
1397	ev_periodic_init (&hourly_tick, clock_cb,	2527	ev_periodic_init (&hourly_tick, clock_cb,
1398	fmod (ev_now (loop), 3600.), 3600., 0);	2528	fmod (ev_now (loop), 3600.), 3600., 0);
1399	ev_periodic_start (loop, &hourly_tick);	2529	ev_periodic_start (loop, &hourly_tick);
1400		2530
1401		2531
1402	=head2 C<ev_signal> - signal me when a signal gets signalled!	2532	=head2 C<ev_signal> - signal me when a signal gets signalled!
1403		2533
1404	Signal watchers will trigger an event when the process receives a specific	2534	Signal watchers will trigger an event when the process receives a specific
1405	signal one or more times. Even though signals are very asynchronous, libev	2535	signal one or more times. Even though signals are very asynchronous, libev
1406	will try it's best to deliver signals synchronously, i.e. as part of the	2536	will try its best to deliver signals synchronously, i.e. as part of the
1407	normal event processing, like any other event.	2537	normal event processing, like any other event.
1408		2538
		2539	If you want signals to be delivered truly asynchronously, just use
		2540	C<sigaction> as you would do without libev and forget about sharing
		2541	the signal. You can even use C<ev_async> from a signal handler to
		2542	synchronously wake up an event loop.
		2543
1409	You can configure as many watchers as you like per signal. Only when the	2544	You can configure as many watchers as you like for the same signal, but
1410	first watcher gets started will libev actually register a signal watcher	2545	only within the same loop, i.e. you can watch for C<SIGINT> in your
1411	with the kernel (thus it coexists with your own signal handlers as long	2546	default loop and for C<SIGIO> in another loop, but you cannot watch for
1412	as you don't register any with libev). Similarly, when the last signal	2547	C<SIGINT> in both the default loop and another loop at the same time. At
1413	watcher for a signal is stopped libev will reset the signal handler to	2548	the moment, C<SIGCHLD> is permanently tied to the default loop.
1414	SIG_DFL (regardless of what it was set to before).	2549
		2550	Only after the first watcher for a signal is started will libev actually
		2551	register something with the kernel. It thus coexists with your own signal
		2552	handlers as long as you don't register any with libev for the same signal.
		2553
		2554	If possible and supported, libev will install its handlers with
		2555	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
		2556	not be unduly interrupted. If you have a problem with system calls getting
		2557	interrupted by signals you can block all signals in an C<ev_check> watcher
		2558	and unblock them in an C<ev_prepare> watcher.
		2559
		2560	=head3 The special problem of inheritance over fork/execve/pthread_create
		2561
		2562	Both the signal mask (C<sigprocmask>) and the signal disposition
		2563	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2564	stopping it again), that is, libev might or might not block the signal,
		2565	and might or might not set or restore the installed signal handler (but
		2566	see C<EVFLAG_NOSIGMASK>).
		2567
		2568	While this does not matter for the signal disposition (libev never
		2569	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2570	C<execve>), this matters for the signal mask: many programs do not expect
		2571	certain signals to be blocked.
		2572
		2573	This means that before calling C<exec> (from the child) you should reset
		2574	the signal mask to whatever "default" you expect (all clear is a good
		2575	choice usually).
		2576
		2577	The simplest way to ensure that the signal mask is reset in the child is
		2578	to install a fork handler with C<pthread_atfork> that resets it. That will
		2579	catch fork calls done by libraries (such as the libc) as well.
		2580
		2581	In current versions of libev, the signal will not be blocked indefinitely
		2582	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2583	the window of opportunity for problems, it will not go away, as libev
		2584	I<has> to modify the signal mask, at least temporarily.
		2585
		2586	So I can't stress this enough: I<If you do not reset your signal mask when
		2587	you expect it to be empty, you have a race condition in your code>. This
		2588	is not a libev-specific thing, this is true for most event libraries.
		2589
		2590	=head3 The special problem of threads signal handling
		2591
		2592	POSIX threads has problematic signal handling semantics, specifically,
		2593	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2594	threads in a process block signals, which is hard to achieve.
		2595
		2596	When you want to use sigwait (or mix libev signal handling with your own
		2597	for the same signals), you can tackle this problem by globally blocking
		2598	all signals before creating any threads (or creating them with a fully set
		2599	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2600	loops. Then designate one thread as "signal receiver thread" which handles
		2601	these signals. You can pass on any signals that libev might be interested
		2602	in by calling C<ev_feed_signal>.
1415		2603
1416	=head3 Watcher-Specific Functions and Data Members	2604	=head3 Watcher-Specific Functions and Data Members
1417		2605
1418	=over 4	2606	=over 4
1419		2607
…		…
1428		2616
1429	The signal the watcher watches out for.	2617	The signal the watcher watches out for.
1430		2618
1431	=back	2619	=back
1432		2620
		2621	=head3 Examples
		2622
		2623	Example: Try to exit cleanly on SIGINT.
		2624
		2625	static void
		2626	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		2627	{
		2628	ev_break (loop, EVBREAK_ALL);
		2629	}
		2630
		2631	ev_signal signal_watcher;
		2632	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		2633	ev_signal_start (loop, &signal_watcher);
		2634
1433		2635
1434	=head2 C<ev_child> - watch out for process status changes	2636	=head2 C<ev_child> - watch out for process status changes
1435		2637
1436	Child watchers trigger when your process receives a SIGCHLD in response to	2638	Child watchers trigger when your process receives a SIGCHLD in response to
1437	some child status changes (most typically when a child of yours dies).	2639	some child status changes (most typically when a child of yours dies or
		2640	exits). It is permissible to install a child watcher I<after> the child
		2641	has been forked (which implies it might have already exited), as long
		2642	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2643	forking and then immediately registering a watcher for the child is fine,
		2644	but forking and registering a watcher a few event loop iterations later or
		2645	in the next callback invocation is not.
		2646
		2647	Only the default event loop is capable of handling signals, and therefore
		2648	you can only register child watchers in the default event loop.
		2649
		2650	Due to some design glitches inside libev, child watchers will always be
		2651	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2652	libev)
		2653
		2654	=head3 Process Interaction
		2655
		2656	Libev grabs C<SIGCHLD> as soon as the default event loop is
		2657	initialised. This is necessary to guarantee proper behaviour even if the
		2658	first child watcher is started after the child exits. The occurrence
		2659	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		2660	synchronously as part of the event loop processing. Libev always reaps all
		2661	children, even ones not watched.
		2662
		2663	=head3 Overriding the Built-In Processing
		2664
		2665	Libev offers no special support for overriding the built-in child
		2666	processing, but if your application collides with libev's default child
		2667	handler, you can override it easily by installing your own handler for
		2668	C<SIGCHLD> after initialising the default loop, and making sure the
		2669	default loop never gets destroyed. You are encouraged, however, to use an
		2670	event-based approach to child reaping and thus use libev's support for
		2671	that, so other libev users can use C<ev_child> watchers freely.
		2672
		2673	=head3 Stopping the Child Watcher
		2674
		2675	Currently, the child watcher never gets stopped, even when the
		2676	child terminates, so normally one needs to stop the watcher in the
		2677	callback. Future versions of libev might stop the watcher automatically
		2678	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2679	problem).
1438		2680
1439	=head3 Watcher-Specific Functions and Data Members	2681	=head3 Watcher-Specific Functions and Data Members
1440		2682
1441	=over 4	2683	=over 4
1442		2684
…		…
1468		2710
1469	=back	2711	=back
1470		2712
1471	=head3 Examples	2713	=head3 Examples
1472		2714
1473	Example: Try to exit cleanly on SIGINT and SIGTERM.	2715	Example: C<fork()> a new process and install a child handler to wait for
		2716	its completion.
1474		2717
		2718	ev_child cw;
		2719
1475	static void	2720	static void
1476	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2721	child_cb (EV_P_ ev_child *w, int revents)
1477	{	2722	{
1478	ev_unloop (loop, EVUNLOOP_ALL);	2723	ev_child_stop (EV_A_ w);
		2724	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1479	}	2725	}
1480		2726
1481	struct ev_signal signal_watcher;	2727	pid_t pid = fork ();
1482	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2728
1483	ev_signal_start (loop, &sigint_cb);	2729	if (pid < 0)
		2730	// error
		2731	else if (pid == 0)
		2732	{
		2733	// the forked child executes here
		2734	exit (1);
		2735	}
		2736	else
		2737	{
		2738	ev_child_init (&cw, child_cb, pid, 0);
		2739	ev_child_start (EV_DEFAULT_ &cw);
		2740	}
1484		2741
1485		2742
1486	=head2 C<ev_stat> - did the file attributes just change?	2743	=head2 C<ev_stat> - did the file attributes just change?
1487		2744
1488	This watches a filesystem path for attribute changes. That is, it calls	2745	This watches a file system path for attribute changes. That is, it calls
1489	C<stat> regularly (or when the OS says it changed) and sees if it changed	2746	C<stat> on that path in regular intervals (or when the OS says it changed)
1490	compared to the last time, invoking the callback if it did.	2747	and sees if it changed compared to the last time, invoking the callback
		2748	if it did. Starting the watcher C<stat>'s the file, so only changes that
		2749	happen after the watcher has been started will be reported.
1491		2750
1492	The path does not need to exist: changing from "path exists" to "path does	2751	The path does not need to exist: changing from "path exists" to "path does
1493	not exist" is a status change like any other. The condition "path does	2752	not exist" is a status change like any other. The condition "path does not
1494	not exist" is signified by the C<st_nlink> field being zero (which is	2753	exist" (or more correctly "path cannot be stat'ed") is signified by the
1495	otherwise always forced to be at least one) and all the other fields of	2754	C<st_nlink> field being zero (which is otherwise always forced to be at
1496	the stat buffer having unspecified contents.	2755	least one) and all the other fields of the stat buffer having unspecified
		2756	contents.
1497		2757
1498	The path I<should> be absolute and I<must not> end in a slash. If it is	2758	The path I<must not> end in a slash or contain special components such as
		2759	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1499	relative and your working directory changes, the behaviour is undefined.	2760	your working directory changes, then the behaviour is undefined.
1500		2761
1501	Since there is no standard to do this, the portable implementation simply	2762	Since there is no portable change notification interface available, the
1502	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2763	portable implementation simply calls C<stat(2)> regularly on the path
1503	can specify a recommended polling interval for this case. If you specify	2764	to see if it changed somehow. You can specify a recommended polling
1504	a polling interval of C<0> (highly recommended!) then a I<suitable,	2765	interval for this case. If you specify a polling interval of C<0> (highly
1505	unspecified default> value will be used (which you can expect to be around	2766	recommended!) then a I<suitable, unspecified default> value will be used
1506	five seconds, although this might change dynamically). Libev will also	2767	(which you can expect to be around five seconds, although this might
1507	impose a minimum interval which is currently around C<0.1>, but thats	2768	change dynamically). Libev will also impose a minimum interval which is
1508	usually overkill.	2769	currently around C<0.1>, but that's usually overkill.
1509		2770
1510	This watcher type is not meant for massive numbers of stat watchers,	2771	This watcher type is not meant for massive numbers of stat watchers,
1511	as even with OS-supported change notifications, this can be	2772	as even with OS-supported change notifications, this can be
1512	resource-intensive.	2773	resource-intensive.
1513		2774
1514	At the time of this writing, only the Linux inotify interface is	2775	At the time of this writing, the only OS-specific interface implemented
1515	implemented (implementing kqueue support is left as an exercise for the	2776	is the Linux inotify interface (implementing kqueue support is left as an
1516	reader). Inotify will be used to give hints only and should not change the	2777	exercise for the reader. Note, however, that the author sees no way of
1517	semantics of C<ev_stat> watchers, which means that libev sometimes needs	2778	implementing C<ev_stat> semantics with kqueue, except as a hint).
1518	to fall back to regular polling again even with inotify, but changes are
1519	usually detected immediately, and if the file exists there will be no
1520	polling.
1521		2779
1522	=head3 Inotify	2780	=head3 ABI Issues (Largefile Support)
1523		2781
		2782	Libev by default (unless the user overrides this) uses the default
		2783	compilation environment, which means that on systems with large file
		2784	support disabled by default, you get the 32 bit version of the stat
		2785	structure. When using the library from programs that change the ABI to
		2786	use 64 bit file offsets the programs will fail. In that case you have to
		2787	compile libev with the same flags to get binary compatibility. This is
		2788	obviously the case with any flags that change the ABI, but the problem is
		2789	most noticeably displayed with ev_stat and large file support.
		2790
		2791	The solution for this is to lobby your distribution maker to make large
		2792	file interfaces available by default (as e.g. FreeBSD does) and not
		2793	optional. Libev cannot simply switch on large file support because it has
		2794	to exchange stat structures with application programs compiled using the
		2795	default compilation environment.
		2796
		2797	=head3 Inotify and Kqueue
		2798
1524	When C<inotify (7)> support has been compiled into libev (generally only	2799	When C<inotify (7)> support has been compiled into libev and present at
1525	available on Linux) and present at runtime, it will be used to speed up	2800	runtime, it will be used to speed up change detection where possible. The
1526	change detection where possible. The inotify descriptor will be created lazily	2801	inotify descriptor will be created lazily when the first C<ev_stat>
1527	when the first C<ev_stat> watcher is being started.	2802	watcher is being started.
1528		2803
1529	Inotify presense does not change the semantics of C<ev_stat> watchers	2804	Inotify presence does not change the semantics of C<ev_stat> watchers
1530	except that changes might be detected earlier, and in some cases, to avoid	2805	except that changes might be detected earlier, and in some cases, to avoid
1531	making regular C<stat> calls. Even in the presense of inotify support	2806	making regular C<stat> calls. Even in the presence of inotify support
1532	there are many cases where libev has to resort to regular C<stat> polling.	2807	there are many cases where libev has to resort to regular C<stat> polling,
		2808	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2809	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2810	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2811	xfs are fully working) libev usually gets away without polling.
1533		2812
1534	(There is no support for kqueue, as apparently it cannot be used to	2813	There is no support for kqueue, as apparently it cannot be used to
1535	implement this functionality, due to the requirement of having a file	2814	implement this functionality, due to the requirement of having a file
1536	descriptor open on the object at all times).	2815	descriptor open on the object at all times, and detecting renames, unlinks
		2816	etc. is difficult.
		2817
		2818	=head3 C<stat ()> is a synchronous operation
		2819
		2820	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2821	the process. The exception are C<ev_stat> watchers - those call C<stat
		2822	()>, which is a synchronous operation.
		2823
		2824	For local paths, this usually doesn't matter: unless the system is very
		2825	busy or the intervals between stat's are large, a stat call will be fast,
		2826	as the path data is usually in memory already (except when starting the
		2827	watcher).
		2828
		2829	For networked file systems, calling C<stat ()> can block an indefinite
		2830	time due to network issues, and even under good conditions, a stat call
		2831	often takes multiple milliseconds.
		2832
		2833	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2834	paths, although this is fully supported by libev.
1537		2835
1538	=head3 The special problem of stat time resolution	2836	=head3 The special problem of stat time resolution
1539		2837
1540	The C<stat ()> syscall only supports full-second resolution portably, and	2838	The C<stat ()> system call only supports full-second resolution portably,
1541	even on systems where the resolution is higher, many filesystems still	2839	and even on systems where the resolution is higher, most file systems
1542	only support whole seconds.	2840	still only support whole seconds.
1543		2841
1544	That means that, if the time is the only thing that changes, you might	2842	That means that, if the time is the only thing that changes, you can
1545	miss updates: on the first update, C<ev_stat> detects a change and calls	2843	easily miss updates: on the first update, C<ev_stat> detects a change and
1546	your callback, which does something. When there is another update within	2844	calls your callback, which does something. When there is another update
1547	the same second, C<ev_stat> will be unable to detect it.	2845	within the same second, C<ev_stat> will be unable to detect unless the
		2846	stat data does change in other ways (e.g. file size).
1548		2847
1549	The solution to this is to delay acting on a change for a second (or till	2848	The solution to this is to delay acting on a change for slightly more
1550	the next second boundary), using a roughly one-second delay C<ev_timer>	2849	than a second (or till slightly after the next full second boundary), using
1551	(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>	2850	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1552	is added to work around small timing inconsistencies of some operating	2851	ev_timer_again (loop, w)>).
1553	systems.	2852
		2853	The C<.02> offset is added to work around small timing inconsistencies
		2854	of some operating systems (where the second counter of the current time
		2855	might be be delayed. One such system is the Linux kernel, where a call to
		2856	C<gettimeofday> might return a timestamp with a full second later than
		2857	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2858	update file times then there will be a small window where the kernel uses
		2859	the previous second to update file times but libev might already execute
		2860	the timer callback).
1554		2861
1555	=head3 Watcher-Specific Functions and Data Members	2862	=head3 Watcher-Specific Functions and Data Members
1556		2863
1557	=over 4	2864	=over 4
1558		2865
…		…
1564	C<path>. The C<interval> is a hint on how quickly a change is expected to	2871	C<path>. The C<interval> is a hint on how quickly a change is expected to
1565	be detected and should normally be specified as C<0> to let libev choose	2872	be detected and should normally be specified as C<0> to let libev choose
1566	a suitable value. The memory pointed to by C<path> must point to the same	2873	a suitable value. The memory pointed to by C<path> must point to the same
1567	path for as long as the watcher is active.	2874	path for as long as the watcher is active.
1568		2875
1569	The callback will be receive C<EV_STAT> when a change was detected,	2876	The callback will receive an C<EV_STAT> event when a change was detected,
1570	relative to the attributes at the time the watcher was started (or the	2877	relative to the attributes at the time the watcher was started (or the
1571	last change was detected).	2878	last change was detected).
1572		2879
1573	=item ev_stat_stat (ev_stat *)	2880	=item ev_stat_stat (loop, ev_stat *)
1574		2881
1575	Updates the stat buffer immediately with new values. If you change the	2882	Updates the stat buffer immediately with new values. If you change the
1576	watched path in your callback, you could call this fucntion to avoid	2883	watched path in your callback, you could call this function to avoid
1577	detecting this change (while introducing a race condition). Can also be	2884	detecting this change (while introducing a race condition if you are not
1578	useful simply to find out the new values.	2885	the only one changing the path). Can also be useful simply to find out the
		2886	new values.
1579		2887
1580	=item ev_statdata attr [read-only]	2888	=item ev_statdata attr [read-only]
1581		2889
1582	The most-recently detected attributes of the file. Although the type is of	2890	The most-recently detected attributes of the file. Although the type is
1583	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types	2891	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
1584	suitable for your system. If the C<st_nlink> member is C<0>, then there	2892	suitable for your system, but you can only rely on the POSIX-standardised
		2893	members to be present. If the C<st_nlink> member is C<0>, then there was
1585	was some error while C<stat>ing the file.	2894	some error while C<stat>ing the file.
1586		2895
1587	=item ev_statdata prev [read-only]	2896	=item ev_statdata prev [read-only]
1588		2897
1589	The previous attributes of the file. The callback gets invoked whenever	2898	The previous attributes of the file. The callback gets invoked whenever
1590	C<prev> != C<attr>.	2899	C<prev> != C<attr>, or, more precisely, one or more of these members
		2900	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2901	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1591		2902
1592	=item ev_tstamp interval [read-only]	2903	=item ev_tstamp interval [read-only]
1593		2904
1594	The specified interval.	2905	The specified interval.
1595		2906
1596	=item const char *path [read-only]	2907	=item const char *path [read-only]
1597		2908
1598	The filesystem path that is being watched.	2909	The file system path that is being watched.
1599		2910
1600	=back	2911	=back
1601		2912
1602	=head3 Examples	2913	=head3 Examples
1603		2914
1604	Example: Watch C</etc/passwd> for attribute changes.	2915	Example: Watch C</etc/passwd> for attribute changes.
1605		2916
1606	static void	2917	static void
1607	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	2918	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1608	{	2919	{
1609	/* /etc/passwd changed in some way */	2920	/* /etc/passwd changed in some way */
1610	if (w->attr.st_nlink)	2921	if (w->attr.st_nlink)
1611	{	2922	{
1612	printf ("passwd current size %ld\n", (long)w->attr.st_size);	2923	printf ("passwd current size %ld\n", (long)w->attr.st_size);
1613	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);	2924	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
1614	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);	2925	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
1615	}	2926	}
1616	else	2927	else
1617	/* you shalt not abuse printf for puts */	2928	/* you shalt not abuse printf for puts */
1618	puts ("wow, /etc/passwd is not there, expect problems. "	2929	puts ("wow, /etc/passwd is not there, expect problems. "
1619	"if this is windows, they already arrived\n");	2930	"if this is windows, they already arrived\n");
1620	}	2931	}
1621		2932
1622	...	2933	...
1623	ev_stat passwd;	2934	ev_stat passwd;
1624		2935
1625	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);	2936	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1626	ev_stat_start (loop, &passwd);	2937	ev_stat_start (loop, &passwd);
1627		2938
1628	Example: Like above, but additionally use a one-second delay so we do not	2939	Example: Like above, but additionally use a one-second delay so we do not
1629	miss updates (however, frequent updates will delay processing, too, so	2940	miss updates (however, frequent updates will delay processing, too, so
1630	one might do the work both on C<ev_stat> callback invocation I<and> on	2941	one might do the work both on C<ev_stat> callback invocation I<and> on
1631	C<ev_timer> callback invocation).	2942	C<ev_timer> callback invocation).
1632		2943
1633	static ev_stat passwd;	2944	static ev_stat passwd;
1634	static ev_timer timer;	2945	static ev_timer timer;
1635		2946
1636	static void	2947	static void
1637	timer_cb (EV_P_ ev_timer *w, int revents)	2948	timer_cb (EV_P_ ev_timer *w, int revents)
1638	{	2949	{
1639	ev_timer_stop (EV_A_ w);	2950	ev_timer_stop (EV_A_ w);
1640		2951
1641	/* now it's one second after the most recent passwd change */	2952	/* now it's one second after the most recent passwd change */
1642	}	2953	}
1643		2954
1644	static void	2955	static void
1645	stat_cb (EV_P_ ev_stat *w, int revents)	2956	stat_cb (EV_P_ ev_stat *w, int revents)
1646	{	2957	{
1647	/* reset the one-second timer */	2958	/* reset the one-second timer */
1648	ev_timer_again (EV_A_ &timer);	2959	ev_timer_again (EV_A_ &timer);
1649	}	2960	}
1650		2961
1651	...	2962	...
1652	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	2963	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1653	ev_stat_start (loop, &passwd);	2964	ev_stat_start (loop, &passwd);
1654	ev_timer_init (&timer, timer_cb, 0., 1.01);	2965	ev_timer_init (&timer, timer_cb, 0., 1.02);
1655		2966
1656		2967
1657	=head2 C<ev_idle> - when you've got nothing better to do...	2968	=head2 C<ev_idle> - when you've got nothing better to do...
1658		2969
1659	Idle watchers trigger events when no other events of the same or higher	2970	Idle watchers trigger events when no other events of the same or higher
1660	priority are pending (prepare, check and other idle watchers do not	2971	priority are pending (prepare, check and other idle watchers do not count
1661	count).	2972	as receiving "events").
1662		2973
1663	That is, as long as your process is busy handling sockets or timeouts	2974	That is, as long as your process is busy handling sockets or timeouts
1664	(or even signals, imagine) of the same or higher priority it will not be	2975	(or even signals, imagine) of the same or higher priority it will not be
1665	triggered. But when your process is idle (or only lower-priority watchers	2976	triggered. But when your process is idle (or only lower-priority watchers
1666	are pending), the idle watchers are being called once per event loop	2977	are pending), the idle watchers are being called once per event loop
…		…
1673	Apart from keeping your process non-blocking (which is a useful	2984	Apart from keeping your process non-blocking (which is a useful
1674	effect on its own sometimes), idle watchers are a good place to do	2985	effect on its own sometimes), idle watchers are a good place to do
1675	"pseudo-background processing", or delay processing stuff to after the	2986	"pseudo-background processing", or delay processing stuff to after the
1676	event loop has handled all outstanding events.	2987	event loop has handled all outstanding events.
1677		2988
		2989	=head3 Abusing an C<ev_idle> watcher for its side-effect
		2990
		2991	As long as there is at least one active idle watcher, libev will never
		2992	sleep unnecessarily. Or in other words, it will loop as fast as possible.
		2993	For this to work, the idle watcher doesn't need to be invoked at all - the
		2994	lowest priority will do.
		2995
		2996	This mode of operation can be useful together with an C<ev_check> watcher,
		2997	to do something on each event loop iteration - for example to balance load
		2998	between different connections.
		2999
		3000	See L</Abusing an ev_check watcher for its side-effect> for a longer
		3001	example.
		3002
1678	=head3 Watcher-Specific Functions and Data Members	3003	=head3 Watcher-Specific Functions and Data Members
1679		3004
1680	=over 4	3005	=over 4
1681		3006
1682	=item ev_idle_init (ev_signal *, callback)	3007	=item ev_idle_init (ev_idle *, callback)
1683		3008
1684	Initialises and configures the idle watcher - it has no parameters of any	3009	Initialises and configures the idle watcher - it has no parameters of any
1685	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	3010	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1686	believe me.	3011	believe me.
1687		3012
…		…
1690	=head3 Examples	3015	=head3 Examples
1691		3016
1692	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	3017	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1693	callback, free it. Also, use no error checking, as usual.	3018	callback, free it. Also, use no error checking, as usual.
1694		3019
1695	static void	3020	static void
1696	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	3021	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1697	{	3022	{
		3023	// stop the watcher
		3024	ev_idle_stop (loop, w);
		3025
		3026	// now we can free it
1698	free (w);	3027	free (w);
		3028
1699	// now do something you wanted to do when the program has	3029	// now do something you wanted to do when the program has
1700	// no longer anything immediate to do.	3030	// no longer anything immediate to do.
1701	}	3031	}
1702		3032
1703	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	3033	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1704	ev_idle_init (idle_watcher, idle_cb);	3034	ev_idle_init (idle_watcher, idle_cb);
1705	ev_idle_start (loop, idle_cb);	3035	ev_idle_start (loop, idle_watcher);
1706		3036
1707		3037
1708	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	3038	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1709		3039
1710	Prepare and check watchers are usually (but not always) used in tandem:	3040	Prepare and check watchers are often (but not always) used in pairs:
1711	prepare watchers get invoked before the process blocks and check watchers	3041	prepare watchers get invoked before the process blocks and check watchers
1712	afterwards.	3042	afterwards.
1713		3043
1714	You I<must not> call C<ev_loop> or similar functions that enter	3044	You I<must not> call C<ev_run> (or similar functions that enter the
1715	the current event loop from either C<ev_prepare> or C<ev_check>	3045	current event loop) or C<ev_loop_fork> from either C<ev_prepare> or
1716	watchers. Other loops than the current one are fine, however. The	3046	C<ev_check> watchers. Other loops than the current one are fine,
1717	rationale behind this is that you do not need to check for recursion in	3047	however. The rationale behind this is that you do not need to check
1718	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	3048	for recursion in those watchers, i.e. the sequence will always be
1719	C<ev_check> so if you have one watcher of each kind they will always be	3049	C<ev_prepare>, blocking, C<ev_check> so if you have one watcher of each
1720	called in pairs bracketing the blocking call.	3050	kind they will always be called in pairs bracketing the blocking call.
1721		3051
1722	Their main purpose is to integrate other event mechanisms into libev and	3052	Their main purpose is to integrate other event mechanisms into libev and
1723	their use is somewhat advanced. This could be used, for example, to track	3053	their use is somewhat advanced. They could be used, for example, to track
1724	variable changes, implement your own watchers, integrate net-snmp or a	3054	variable changes, implement your own watchers, integrate net-snmp or a
1725	coroutine library and lots more. They are also occasionally useful if	3055	coroutine library and lots more. They are also occasionally useful if
1726	you cache some data and want to flush it before blocking (for example,	3056	you cache some data and want to flush it before blocking (for example,
1727	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	3057	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1728	watcher).	3058	watcher).
1729		3059
1730	This is done by examining in each prepare call which file descriptors need	3060	This is done by examining in each prepare call which file descriptors
1731	to be watched by the other library, registering C<ev_io> watchers for	3061	need to be watched by the other library, registering C<ev_io> watchers
1732	them and starting an C<ev_timer> watcher for any timeouts (many libraries	3062	for them and starting an C<ev_timer> watcher for any timeouts (many
1733	provide just this functionality). Then, in the check watcher you check for	3063	libraries provide exactly this functionality). Then, in the check watcher,
1734	any events that occured (by checking the pending status of all watchers	3064	you check for any events that occurred (by checking the pending status
1735	and stopping them) and call back into the library. The I/O and timer	3065	of all watchers and stopping them) and call back into the library. The
1736	callbacks will never actually be called (but must be valid nevertheless,	3066	I/O and timer callbacks will never actually be called (but must be valid
1737	because you never know, you know?).	3067	nevertheless, because you never know, you know?).
1738		3068
1739	As another example, the Perl Coro module uses these hooks to integrate	3069	As another example, the Perl Coro module uses these hooks to integrate
1740	coroutines into libev programs, by yielding to other active coroutines	3070	coroutines into libev programs, by yielding to other active coroutines
1741	during each prepare and only letting the process block if no coroutines	3071	during each prepare and only letting the process block if no coroutines
1742	are ready to run (it's actually more complicated: it only runs coroutines	3072	are ready to run (it's actually more complicated: it only runs coroutines
1743	with priority higher than or equal to the event loop and one coroutine	3073	with priority higher than or equal to the event loop and one coroutine
1744	of lower priority, but only once, using idle watchers to keep the event	3074	of lower priority, but only once, using idle watchers to keep the event
1745	loop from blocking if lower-priority coroutines are active, thus mapping	3075	loop from blocking if lower-priority coroutines are active, thus mapping
1746	low-priority coroutines to idle/background tasks).	3076	low-priority coroutines to idle/background tasks).
1747		3077
1748	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	3078	When used for this purpose, it is recommended to give C<ev_check> watchers
1749	priority, to ensure that they are being run before any other watchers	3079	highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
		3080	any other watchers after the poll (this doesn't matter for C<ev_prepare>
		3081	watchers).
		3082
1750	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	3083	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1751	too) should not activate ("feed") events into libev. While libev fully	3084	activate ("feed") events into libev. While libev fully supports this, they
1752	supports this, they will be called before other C<ev_check> watchers	3085	might get executed before other C<ev_check> watchers did their job. As
1753	did their job. As C<ev_check> watchers are often used to embed other	3086	C<ev_check> watchers are often used to embed other (non-libev) event
1754	(non-libev) event loops those other event loops might be in an unusable	3087	loops those other event loops might be in an unusable state until their
1755	state until their C<ev_check> watcher ran (always remind yourself to	3088	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1756	coexist peacefully with others).	3089	others).
		3090
		3091	=head3 Abusing an C<ev_check> watcher for its side-effect
		3092
		3093	C<ev_check> (and less often also C<ev_prepare>) watchers can also be
		3094	useful because they are called once per event loop iteration. For
		3095	example, if you want to handle a large number of connections fairly, you
		3096	normally only do a bit of work for each active connection, and if there
		3097	is more work to do, you wait for the next event loop iteration, so other
		3098	connections have a chance of making progress.
		3099
		3100	Using an C<ev_check> watcher is almost enough: it will be called on the
		3101	next event loop iteration. However, that isn't as soon as possible -
		3102	without external events, your C<ev_check> watcher will not be invoked.
		3103
		3104	This is where C<ev_idle> watchers come in handy - all you need is a
		3105	single global idle watcher that is active as long as you have one active
		3106	C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
		3107	will not sleep, and the C<ev_check> watcher makes sure a callback gets
		3108	invoked. Neither watcher alone can do that.
1757		3109
1758	=head3 Watcher-Specific Functions and Data Members	3110	=head3 Watcher-Specific Functions and Data Members
1759		3111
1760	=over 4	3112	=over 4
1761		3113
…		…
1763		3115
1764	=item ev_check_init (ev_check *, callback)	3116	=item ev_check_init (ev_check *, callback)
1765		3117
1766	Initialises and configures the prepare or check watcher - they have no	3118	Initialises and configures the prepare or check watcher - they have no
1767	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	3119	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1768	macros, but using them is utterly, utterly and completely pointless.	3120	macros, but using them is utterly, utterly, utterly and completely
		3121	pointless.
1769		3122
1770	=back	3123	=back
1771		3124
1772	=head3 Examples	3125	=head3 Examples
1773		3126
1774	There are a number of principal ways to embed other event loops or modules	3127	There are a number of principal ways to embed other event loops or modules
1775	into libev. Here are some ideas on how to include libadns into libev	3128	into libev. Here are some ideas on how to include libadns into libev
1776	(there is a Perl module named C<EV::ADNS> that does this, which you could	3129	(there is a Perl module named C<EV::ADNS> that does this, which you could
1777	use for an actually working example. Another Perl module named C<EV::Glib>	3130	use as a working example. Another Perl module named C<EV::Glib> embeds a
1778	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV	3131	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
1779	into the Glib event loop).	3132	Glib event loop).
1780		3133
1781	Method 1: Add IO watchers and a timeout watcher in a prepare handler,	3134	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1782	and in a check watcher, destroy them and call into libadns. What follows	3135	and in a check watcher, destroy them and call into libadns. What follows
1783	is pseudo-code only of course. This requires you to either use a low	3136	is pseudo-code only of course. This requires you to either use a low
1784	priority for the check watcher or use C<ev_clear_pending> explicitly, as	3137	priority for the check watcher or use C<ev_clear_pending> explicitly, as
1785	the callbacks for the IO/timeout watchers might not have been called yet.	3138	the callbacks for the IO/timeout watchers might not have been called yet.
1786		3139
1787	static ev_io iow [nfd];	3140	static ev_io iow [nfd];
1788	static ev_timer tw;	3141	static ev_timer tw;
1789		3142
1790	static void	3143	static void
1791	io_cb (ev_loop *loop, ev_io *w, int revents)	3144	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1792	{	3145	{
1793	}	3146	}
1794		3147
1795	// create io watchers for each fd and a timer before blocking	3148	// create io watchers for each fd and a timer before blocking
1796	static void	3149	static void
1797	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	3150	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1798	{	3151	{
1799	int timeout = 3600000;	3152	int timeout = 3600000;
1800	struct pollfd fds [nfd];	3153	struct pollfd fds [nfd];
1801	// actual code will need to loop here and realloc etc.	3154	// actual code will need to loop here and realloc etc.
1802	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	3155	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1803		3156
1804	/* the callback is illegal, but won't be called as we stop during check */	3157	/* the callback is illegal, but won't be called as we stop during check */
1805	ev_timer_init (&tw, 0, timeout * 1e-3);	3158	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
1806	ev_timer_start (loop, &tw);	3159	ev_timer_start (loop, &tw);
1807		3160
1808	// create one ev_io per pollfd	3161	// create one ev_io per pollfd
1809	for (int i = 0; i < nfd; ++i)	3162	for (int i = 0; i < nfd; ++i)
1810	{	3163	{
1811	ev_io_init (iow + i, io_cb, fds [i].fd,	3164	ev_io_init (iow + i, io_cb, fds [i].fd,
1812	((fds [i].events & POLLIN ? EV_READ : 0)	3165	((fds [i].events & POLLIN ? EV_READ : 0)
1813	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	3166	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1814		3167
1815	fds [i].revents = 0;	3168	fds [i].revents = 0;
1816	ev_io_start (loop, iow + i);	3169	ev_io_start (loop, iow + i);
1817	}	3170	}
1818	}	3171	}
1819		3172
1820	// stop all watchers after blocking	3173	// stop all watchers after blocking
1821	static void	3174	static void
1822	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	3175	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
1823	{	3176	{
1824	ev_timer_stop (loop, &tw);	3177	ev_timer_stop (loop, &tw);
1825		3178
1826	for (int i = 0; i < nfd; ++i)	3179	for (int i = 0; i < nfd; ++i)
1827	{	3180	{
1828	// set the relevant poll flags	3181	// set the relevant poll flags
1829	// could also call adns_processreadable etc. here	3182	// could also call adns_processreadable etc. here
1830	struct pollfd *fd = fds + i;	3183	struct pollfd *fd = fds + i;
1831	int revents = ev_clear_pending (iow + i);	3184	int revents = ev_clear_pending (iow + i);
1832	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;	3185	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1833	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;	3186	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1834		3187
1835	// now stop the watcher	3188	// now stop the watcher
1836	ev_io_stop (loop, iow + i);	3189	ev_io_stop (loop, iow + i);
1837	}	3190	}
1838		3191
1839	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	3192	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1840	}	3193	}
1841		3194
1842	Method 2: This would be just like method 1, but you run C<adns_afterpoll>	3195	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
1843	in the prepare watcher and would dispose of the check watcher.	3196	in the prepare watcher and would dispose of the check watcher.
1844		3197
1845	Method 3: If the module to be embedded supports explicit event	3198	Method 3: If the module to be embedded supports explicit event
1846	notification (adns does), you can also make use of the actual watcher	3199	notification (libadns does), you can also make use of the actual watcher
1847	callbacks, and only destroy/create the watchers in the prepare watcher.	3200	callbacks, and only destroy/create the watchers in the prepare watcher.
1848		3201
1849	static void	3202	static void
1850	timer_cb (EV_P_ ev_timer *w, int revents)	3203	timer_cb (EV_P_ ev_timer *w, int revents)
1851	{	3204	{
1852	adns_state ads = (adns_state)w->data;	3205	adns_state ads = (adns_state)w->data;
1853	update_now (EV_A);	3206	update_now (EV_A);
1854		3207
1855	adns_processtimeouts (ads, &tv_now);	3208	adns_processtimeouts (ads, &tv_now);
1856	}	3209	}
1857		3210
1858	static void	3211	static void
1859	io_cb (EV_P_ ev_io *w, int revents)	3212	io_cb (EV_P_ ev_io *w, int revents)
1860	{	3213	{
1861	adns_state ads = (adns_state)w->data;	3214	adns_state ads = (adns_state)w->data;
1862	update_now (EV_A);	3215	update_now (EV_A);
1863		3216
1864	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);	3217	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
1865	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);	3218	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
1866	}	3219	}
1867		3220
1868	// do not ever call adns_afterpoll	3221	// do not ever call adns_afterpoll
1869		3222
1870	Method 4: Do not use a prepare or check watcher because the module you	3223	Method 4: Do not use a prepare or check watcher because the module you
1871	want to embed is too inflexible to support it. Instead, youc na override	3224	want to embed is not flexible enough to support it. Instead, you can
1872	their poll function. The drawback with this solution is that the main	3225	override their poll function. The drawback with this solution is that the
1873	loop is now no longer controllable by EV. The C<Glib::EV> module does	3226	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
1874	this.	3227	this approach, effectively embedding EV as a client into the horrible
		3228	libglib event loop.
1875		3229
1876	static gint	3230	static gint
1877	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	3231	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
1878	{	3232	{
1879	int got_events = 0;	3233	int got_events = 0;
1880		3234
1881	for (n = 0; n < nfds; ++n)	3235	for (n = 0; n < nfds; ++n)
1882	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events	3236	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
1883		3237
1884	if (timeout >= 0)	3238	if (timeout >= 0)
1885	// create/start timer	3239	// create/start timer
1886		3240
1887	// poll	3241	// poll
1888	ev_loop (EV_A_ 0);	3242	ev_run (EV_A_ 0);
1889		3243
1890	// stop timer again	3244	// stop timer again
1891	if (timeout >= 0)	3245	if (timeout >= 0)
1892	ev_timer_stop (EV_A_ &to);	3246	ev_timer_stop (EV_A_ &to);
1893		3247
1894	// stop io watchers again - their callbacks should have set	3248	// stop io watchers again - their callbacks should have set
1895	for (n = 0; n < nfds; ++n)	3249	for (n = 0; n < nfds; ++n)
1896	ev_io_stop (EV_A_ iow [n]);	3250	ev_io_stop (EV_A_ iow [n]);
1897		3251
1898	return got_events;	3252	return got_events;
1899	}	3253	}
1900		3254
1901		3255
1902	=head2 C<ev_embed> - when one backend isn't enough...	3256	=head2 C<ev_embed> - when one backend isn't enough...
1903		3257
1904	This is a rather advanced watcher type that lets you embed one event loop	3258	This is a rather advanced watcher type that lets you embed one event loop
…		…
1910	prioritise I/O.	3264	prioritise I/O.
1911		3265
1912	As an example for a bug workaround, the kqueue backend might only support	3266	As an example for a bug workaround, the kqueue backend might only support
1913	sockets on some platform, so it is unusable as generic backend, but you	3267	sockets on some platform, so it is unusable as generic backend, but you
1914	still want to make use of it because you have many sockets and it scales	3268	still want to make use of it because you have many sockets and it scales
1915	so nicely. In this case, you would create a kqueue-based loop and embed it	3269	so nicely. In this case, you would create a kqueue-based loop and embed
1916	into your default loop (which might use e.g. poll). Overall operation will	3270	it into your default loop (which might use e.g. poll). Overall operation
1917	be a bit slower because first libev has to poll and then call kevent, but	3271	will be a bit slower because first libev has to call C<poll> and then
1918	at least you can use both at what they are best.	3272	C<kevent>, but at least you can use both mechanisms for what they are
		3273	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1919		3274
1920	As for prioritising I/O: rarely you have the case where some fds have	3275	As for prioritising I/O: under rare circumstances you have the case where
1921	to be watched and handled very quickly (with low latency), and even	3276	some fds have to be watched and handled very quickly (with low latency),
1922	priorities and idle watchers might have too much overhead. In this case	3277	and even priorities and idle watchers might have too much overhead. In
1923	you would put all the high priority stuff in one loop and all the rest in	3278	this case you would put all the high priority stuff in one loop and all
1924	a second one, and embed the second one in the first.	3279	the rest in a second one, and embed the second one in the first.
1925		3280
1926	As long as the watcher is active, the callback will be invoked every time	3281	As long as the watcher is active, the callback will be invoked every
1927	there might be events pending in the embedded loop. The callback must then	3282	time there might be events pending in the embedded loop. The callback
1928	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	3283	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
1929	their callbacks (you could also start an idle watcher to give the embedded	3284	sweep and invoke their callbacks (the callback doesn't need to invoke the
1930	loop strictly lower priority for example). You can also set the callback	3285	C<ev_embed_sweep> function directly, it could also start an idle watcher
1931	to C<0>, in which case the embed watcher will automatically execute the	3286	to give the embedded loop strictly lower priority for example).
1932	embedded loop sweep.
1933		3287
1934	As long as the watcher is started it will automatically handle events. The	3288	You can also set the callback to C<0>, in which case the embed watcher
1935	callback will be invoked whenever some events have been handled. You can	3289	will automatically execute the embedded loop sweep whenever necessary.
1936	set the callback to C<0> to avoid having to specify one if you are not
1937	interested in that.
1938		3290
1939	Also, there have not currently been made special provisions for forking:	3291	Fork detection will be handled transparently while the C<ev_embed> watcher
1940	when you fork, you not only have to call C<ev_loop_fork> on both loops,	3292	is active, i.e., the embedded loop will automatically be forked when the
1941	but you will also have to stop and restart any C<ev_embed> watchers	3293	embedding loop forks. In other cases, the user is responsible for calling
1942	yourself.	3294	C<ev_loop_fork> on the embedded loop.
1943		3295
1944	Unfortunately, not all backends are embeddable, only the ones returned by	3296	Unfortunately, not all backends are embeddable: only the ones returned by
1945	C<ev_embeddable_backends> are, which, unfortunately, does not include any	3297	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1946	portable one.	3298	portable one.
1947		3299
1948	So when you want to use this feature you will always have to be prepared	3300	So when you want to use this feature you will always have to be prepared
1949	that you cannot get an embeddable loop. The recommended way to get around	3301	that you cannot get an embeddable loop. The recommended way to get around
1950	this is to have a separate variables for your embeddable loop, try to	3302	this is to have a separate variables for your embeddable loop, try to
1951	create it, and if that fails, use the normal loop for everything.	3303	create it, and if that fails, use the normal loop for everything.
1952		3304
		3305	=head3 C<ev_embed> and fork
		3306
		3307	While the C<ev_embed> watcher is running, forks in the embedding loop will
		3308	automatically be applied to the embedded loop as well, so no special
		3309	fork handling is required in that case. When the watcher is not running,
		3310	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		3311	as applicable.
		3312
1953	=head3 Watcher-Specific Functions and Data Members	3313	=head3 Watcher-Specific Functions and Data Members
1954		3314
1955	=over 4	3315	=over 4
1956		3316
1957	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	3317	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
1958		3318
1959	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	3319	=item ev_embed_set (ev_embed *, struct ev_loop *embedded_loop)
1960		3320
1961	Configures the watcher to embed the given loop, which must be	3321	Configures the watcher to embed the given loop, which must be
1962	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3322	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1963	invoked automatically, otherwise it is the responsibility of the callback	3323	invoked automatically, otherwise it is the responsibility of the callback
1964	to invoke it (it will continue to be called until the sweep has been done,	3324	to invoke it (it will continue to be called until the sweep has been done,
1965	if you do not want thta, you need to temporarily stop the embed watcher).	3325	if you do not want that, you need to temporarily stop the embed watcher).
1966		3326
1967	=item ev_embed_sweep (loop, ev_embed *)	3327	=item ev_embed_sweep (loop, ev_embed *)
1968		3328
1969	Make a single, non-blocking sweep over the embedded loop. This works	3329	Make a single, non-blocking sweep over the embedded loop. This works
1970	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3330	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
1971	apropriate way for embedded loops.	3331	appropriate way for embedded loops.
1972		3332
1973	=item struct ev_loop *other [read-only]	3333	=item struct ev_loop *other [read-only]
1974		3334
1975	The embedded event loop.	3335	The embedded event loop.
1976		3336
…		…
1978		3338
1979	=head3 Examples	3339	=head3 Examples
1980		3340
1981	Example: Try to get an embeddable event loop and embed it into the default	3341	Example: Try to get an embeddable event loop and embed it into the default
1982	event loop. If that is not possible, use the default loop. The default	3342	event loop. If that is not possible, use the default loop. The default
1983	loop is stored in C<loop_hi>, while the mebeddable loop is stored in	3343	loop is stored in C<loop_hi>, while the embeddable loop is stored in
1984	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be	3344	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
1985	used).	3345	used).
1986		3346
1987	struct ev_loop *loop_hi = ev_default_init (0);	3347	struct ev_loop *loop_hi = ev_default_init (0);
1988	struct ev_loop *loop_lo = 0;	3348	struct ev_loop *loop_lo = 0;
1989	struct ev_embed embed;	3349	ev_embed embed;
1990		3350
1991	// see if there is a chance of getting one that works	3351	// see if there is a chance of getting one that works
1992	// (remember that a flags value of 0 means autodetection)	3352	// (remember that a flags value of 0 means autodetection)
1993	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	3353	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1994	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	3354	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1995	: 0;	3355	: 0;
1996		3356
1997	// if we got one, then embed it, otherwise default to loop_hi	3357	// if we got one, then embed it, otherwise default to loop_hi
1998	if (loop_lo)	3358	if (loop_lo)
1999	{	3359	{
2000	ev_embed_init (&embed, 0, loop_lo);	3360	ev_embed_init (&embed, 0, loop_lo);
2001	ev_embed_start (loop_hi, &embed);	3361	ev_embed_start (loop_hi, &embed);
2002	}	3362	}
2003	else	3363	else
2004	loop_lo = loop_hi;	3364	loop_lo = loop_hi;
2005		3365
2006	Example: Check if kqueue is available but not recommended and create	3366	Example: Check if kqueue is available but not recommended and create
2007	a kqueue backend for use with sockets (which usually work with any	3367	a kqueue backend for use with sockets (which usually work with any
2008	kqueue implementation). Store the kqueue/socket-only event loop in	3368	kqueue implementation). Store the kqueue/socket-only event loop in
2009	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	3369	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2010		3370
2011	struct ev_loop *loop = ev_default_init (0);	3371	struct ev_loop *loop = ev_default_init (0);
2012	struct ev_loop *loop_socket = 0;	3372	struct ev_loop *loop_socket = 0;
2013	struct ev_embed embed;	3373	ev_embed embed;
2014		3374
2015	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	3375	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2016	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	3376	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2017	{	3377	{
2018	ev_embed_init (&embed, 0, loop_socket);	3378	ev_embed_init (&embed, 0, loop_socket);
2019	ev_embed_start (loop, &embed);	3379	ev_embed_start (loop, &embed);
2020	}	3380	}
2021		3381
2022	if (!loop_socket)	3382	if (!loop_socket)
2023	loop_socket = loop;	3383	loop_socket = loop;
2024		3384
2025	// now use loop_socket for all sockets, and loop for everything else	3385	// now use loop_socket for all sockets, and loop for everything else
2026		3386
2027		3387
2028	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	3388	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
2029		3389
2030	Fork watchers are called when a C<fork ()> was detected (usually because	3390	Fork watchers are called when a C<fork ()> was detected (usually because
2031	whoever is a good citizen cared to tell libev about it by calling	3391	whoever is a good citizen cared to tell libev about it by calling
2032	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the	3392	C<ev_loop_fork>). The invocation is done before the event loop blocks next
2033	event loop blocks next and before C<ev_check> watchers are being called,	3393	and before C<ev_check> watchers are being called, and only in the child
2034	and only in the child after the fork. If whoever good citizen calling	3394	after the fork. If whoever good citizen calling C<ev_default_fork> cheats
2035	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3395	and calls it in the wrong process, the fork handlers will be invoked, too,
2036	handlers will be invoked, too, of course.	3396	of course.
		3397
		3398	=head3 The special problem of life after fork - how is it possible?
		3399
		3400	Most uses of C<fork ()> consist of forking, then some simple calls to set
		3401	up/change the process environment, followed by a call to C<exec()>. This
		3402	sequence should be handled by libev without any problems.
		3403
		3404	This changes when the application actually wants to do event handling
		3405	in the child, or both parent in child, in effect "continuing" after the
		3406	fork.
		3407
		3408	The default mode of operation (for libev, with application help to detect
		3409	forks) is to duplicate all the state in the child, as would be expected
		3410	when I<either> the parent I<or> the child process continues.
		3411
		3412	When both processes want to continue using libev, then this is usually the
		3413	wrong result. In that case, usually one process (typically the parent) is
		3414	supposed to continue with all watchers in place as before, while the other
		3415	process typically wants to start fresh, i.e. without any active watchers.
		3416
		3417	The cleanest and most efficient way to achieve that with libev is to
		3418	simply create a new event loop, which of course will be "empty", and
		3419	use that for new watchers. This has the advantage of not touching more
		3420	memory than necessary, and thus avoiding the copy-on-write, and the
		3421	disadvantage of having to use multiple event loops (which do not support
		3422	signal watchers).
		3423
		3424	When this is not possible, or you want to use the default loop for
		3425	other reasons, then in the process that wants to start "fresh", call
		3426	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
		3427	Destroying the default loop will "orphan" (not stop) all registered
		3428	watchers, so you have to be careful not to execute code that modifies
		3429	those watchers. Note also that in that case, you have to re-register any
		3430	signal watchers.
2037		3431
2038	=head3 Watcher-Specific Functions and Data Members	3432	=head3 Watcher-Specific Functions and Data Members
2039		3433
2040	=over 4	3434	=over 4
2041		3435
2042	=item ev_fork_init (ev_signal *, callback)	3436	=item ev_fork_init (ev_fork *, callback)
2043		3437
2044	Initialises and configures the fork watcher - it has no parameters of any	3438	Initialises and configures the fork watcher - it has no parameters of any
2045	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3439	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
2046	believe me.	3440	really.
2047		3441
2048	=back	3442	=back
2049		3443
2050		3444
		3445	=head2 C<ev_cleanup> - even the best things end
		3446
		3447	Cleanup watchers are called just before the event loop is being destroyed
		3448	by a call to C<ev_loop_destroy>.
		3449
		3450	While there is no guarantee that the event loop gets destroyed, cleanup
		3451	watchers provide a convenient method to install cleanup hooks for your
		3452	program, worker threads and so on - you just to make sure to destroy the
		3453	loop when you want them to be invoked.
		3454
		3455	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3456	all other watchers, they do not keep a reference to the event loop (which
		3457	makes a lot of sense if you think about it). Like all other watchers, you
		3458	can call libev functions in the callback, except C<ev_cleanup_start>.
		3459
		3460	=head3 Watcher-Specific Functions and Data Members
		3461
		3462	=over 4
		3463
		3464	=item ev_cleanup_init (ev_cleanup *, callback)
		3465
		3466	Initialises and configures the cleanup watcher - it has no parameters of
		3467	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3468	pointless, I assure you.
		3469
		3470	=back
		3471
		3472	Example: Register an atexit handler to destroy the default loop, so any
		3473	cleanup functions are called.
		3474
		3475	static void
		3476	program_exits (void)
		3477	{
		3478	ev_loop_destroy (EV_DEFAULT_UC);
		3479	}
		3480
		3481	...
		3482	atexit (program_exits);
		3483
		3484
		3485	=head2 C<ev_async> - how to wake up an event loop
		3486
		3487	In general, you cannot use an C<ev_loop> from multiple threads or other
		3488	asynchronous sources such as signal handlers (as opposed to multiple event
		3489	loops - those are of course safe to use in different threads).
		3490
		3491	Sometimes, however, you need to wake up an event loop you do not control,
		3492	for example because it belongs to another thread. This is what C<ev_async>
		3493	watchers do: as long as the C<ev_async> watcher is active, you can signal
		3494	it by calling C<ev_async_send>, which is thread- and signal safe.
		3495
		3496	This functionality is very similar to C<ev_signal> watchers, as signals,
		3497	too, are asynchronous in nature, and signals, too, will be compressed
		3498	(i.e. the number of callback invocations may be less than the number of
		3499	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
		3500	of "global async watchers" by using a watcher on an otherwise unused
		3501	signal, and C<ev_feed_signal> to signal this watcher from another thread,
		3502	even without knowing which loop owns the signal.
		3503
		3504	=head3 Queueing
		3505
		3506	C<ev_async> does not support queueing of data in any way. The reason
		3507	is that the author does not know of a simple (or any) algorithm for a
		3508	multiple-writer-single-reader queue that works in all cases and doesn't
		3509	need elaborate support such as pthreads or unportable memory access
		3510	semantics.
		3511
		3512	That means that if you want to queue data, you have to provide your own
		3513	queue. But at least I can tell you how to implement locking around your
		3514	queue:
		3515
		3516	=over 4
		3517
		3518	=item queueing from a signal handler context
		3519
		3520	To implement race-free queueing, you simply add to the queue in the signal
		3521	handler but you block the signal handler in the watcher callback. Here is
		3522	an example that does that for some fictitious SIGUSR1 handler:
		3523
		3524	static ev_async mysig;
		3525
		3526	static void
		3527	sigusr1_handler (void)
		3528	{
		3529	sometype data;
		3530
		3531	// no locking etc.
		3532	queue_put (data);
		3533	ev_async_send (EV_DEFAULT_ &mysig);
		3534	}
		3535
		3536	static void
		3537	mysig_cb (EV_P_ ev_async *w, int revents)
		3538	{
		3539	sometype data;
		3540	sigset_t block, prev;
		3541
		3542	sigemptyset (&block);
		3543	sigaddset (&block, SIGUSR1);
		3544	sigprocmask (SIG_BLOCK, &block, &prev);
		3545
		3546	while (queue_get (&data))
		3547	process (data);
		3548
		3549	if (sigismember (&prev, SIGUSR1)
		3550	sigprocmask (SIG_UNBLOCK, &block, 0);
		3551	}
		3552
		3553	(Note: pthreads in theory requires you to use C<pthread_setmask>
		3554	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		3555	either...).
		3556
		3557	=item queueing from a thread context
		3558
		3559	The strategy for threads is different, as you cannot (easily) block
		3560	threads but you can easily preempt them, so to queue safely you need to
		3561	employ a traditional mutex lock, such as in this pthread example:
		3562
		3563	static ev_async mysig;
		3564	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		3565
		3566	static void
		3567	otherthread (void)
		3568	{
		3569	// only need to lock the actual queueing operation
		3570	pthread_mutex_lock (&mymutex);
		3571	queue_put (data);
		3572	pthread_mutex_unlock (&mymutex);
		3573
		3574	ev_async_send (EV_DEFAULT_ &mysig);
		3575	}
		3576
		3577	static void
		3578	mysig_cb (EV_P_ ev_async *w, int revents)
		3579	{
		3580	pthread_mutex_lock (&mymutex);
		3581
		3582	while (queue_get (&data))
		3583	process (data);
		3584
		3585	pthread_mutex_unlock (&mymutex);
		3586	}
		3587
		3588	=back
		3589
		3590
		3591	=head3 Watcher-Specific Functions and Data Members
		3592
		3593	=over 4
		3594
		3595	=item ev_async_init (ev_async *, callback)
		3596
		3597	Initialises and configures the async watcher - it has no parameters of any
		3598	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
		3599	trust me.
		3600
		3601	=item ev_async_send (loop, ev_async *)
		3602
		3603	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		3604	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3605	returns.
		3606
		3607	Unlike C<ev_feed_event>, this call is safe to do from other threads,
		3608	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
		3609	embedding section below on what exactly this means).
		3610
		3611	Note that, as with other watchers in libev, multiple events might get
		3612	compressed into a single callback invocation (another way to look at
		3613	this is that C<ev_async> watchers are level-triggered: they are set on
		3614	C<ev_async_send>, reset when the event loop detects that).
		3615
		3616	This call incurs the overhead of at most one extra system call per event
		3617	loop iteration, if the event loop is blocked, and no syscall at all if
		3618	the event loop (or your program) is processing events. That means that
		3619	repeated calls are basically free (there is no need to avoid calls for
		3620	performance reasons) and that the overhead becomes smaller (typically
		3621	zero) under load.
		3622
		3623	=item bool = ev_async_pending (ev_async *)
		3624
		3625	Returns a non-zero value when C<ev_async_send> has been called on the
		3626	watcher but the event has not yet been processed (or even noted) by the
		3627	event loop.
		3628
		3629	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		3630	the loop iterates next and checks for the watcher to have become active,
		3631	it will reset the flag again. C<ev_async_pending> can be used to very
		3632	quickly check whether invoking the loop might be a good idea.
		3633
		3634	Not that this does I<not> check whether the watcher itself is pending,
		3635	only whether it has been requested to make this watcher pending: there
		3636	is a time window between the event loop checking and resetting the async
		3637	notification, and the callback being invoked.
		3638
		3639	=back
		3640
		3641
2051	=head1 OTHER FUNCTIONS	3642	=head1 OTHER FUNCTIONS
2052		3643
2053	There are some other functions of possible interest. Described. Here. Now.	3644	There are some other functions of possible interest. Described. Here. Now.
2054		3645
2055	=over 4	3646	=over 4
2056		3647
2057	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3648	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback, arg)
2058		3649
2059	This function combines a simple timer and an I/O watcher, calls your	3650	This function combines a simple timer and an I/O watcher, calls your
2060	callback on whichever event happens first and automatically stop both	3651	callback on whichever event happens first and automatically stops both
2061	watchers. This is useful if you want to wait for a single event on an fd	3652	watchers. This is useful if you want to wait for a single event on an fd
2062	or timeout without having to allocate/configure/start/stop/free one or	3653	or timeout without having to allocate/configure/start/stop/free one or
2063	more watchers yourself.	3654	more watchers yourself.
2064		3655
2065	If C<fd> is less than 0, then no I/O watcher will be started and events	3656	If C<fd> is less than 0, then no I/O watcher will be started and the
2066	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3657	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2067	C<events> set will be craeted and started.	3658	the given C<fd> and C<events> set will be created and started.
2068		3659
2069	If C<timeout> is less than 0, then no timeout watcher will be	3660	If C<timeout> is less than 0, then no timeout watcher will be
2070	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3661	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2071	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3662	repeat = 0) will be started. C<0> is a valid timeout.
2072	dubious value.
2073		3663
2074	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3664	The callback has the type C<void (*cb)(int revents, void *arg)> and is
2075	passed an C<revents> set like normal event callbacks (a combination of	3665	passed an C<revents> set like normal event callbacks (a combination of
2076	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3666	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
2077	value passed to C<ev_once>:	3667	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3668	a timeout and an io event at the same time - you probably should give io
		3669	events precedence.
2078		3670
		3671	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3672
2079	static void stdin_ready (int revents, void *arg)	3673	static void stdin_ready (int revents, void *arg)
		3674	{
		3675	if (revents & EV_READ)
		3676	/* stdin might have data for us, joy! */;
		3677	else if (revents & EV_TIMER)
		3678	/* doh, nothing entered */;
		3679	}
		3680
		3681	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
		3682
		3683	=item ev_feed_fd_event (loop, int fd, int revents)
		3684
		3685	Feed an event on the given fd, as if a file descriptor backend detected
		3686	the given events.
		3687
		3688	=item ev_feed_signal_event (loop, int signum)
		3689
		3690	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
		3691	which is async-safe.
		3692
		3693	=back
		3694
		3695
		3696	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3697
		3698	This section explains some common idioms that are not immediately
		3699	obvious. Note that examples are sprinkled over the whole manual, and this
		3700	section only contains stuff that wouldn't fit anywhere else.
		3701
		3702	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3703
		3704	Each watcher has, by default, a C<void *data> member that you can read
		3705	or modify at any time: libev will completely ignore it. This can be used
		3706	to associate arbitrary data with your watcher. If you need more data and
		3707	don't want to allocate memory separately and store a pointer to it in that
		3708	data member, you can also "subclass" the watcher type and provide your own
		3709	data:
		3710
		3711	struct my_io
		3712	{
		3713	ev_io io;
		3714	int otherfd;
		3715	void *somedata;
		3716	struct whatever *mostinteresting;
		3717	};
		3718
		3719	...
		3720	struct my_io w;
		3721	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3722
		3723	And since your callback will be called with a pointer to the watcher, you
		3724	can cast it back to your own type:
		3725
		3726	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3727	{
		3728	struct my_io *w = (struct my_io *)w_;
		3729	...
		3730	}
		3731
		3732	More interesting and less C-conformant ways of casting your callback
		3733	function type instead have been omitted.
		3734
		3735	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3736
		3737	Another common scenario is to use some data structure with multiple
		3738	embedded watchers, in effect creating your own watcher that combines
		3739	multiple libev event sources into one "super-watcher":
		3740
		3741	struct my_biggy
		3742	{
		3743	int some_data;
		3744	ev_timer t1;
		3745	ev_timer t2;
		3746	}
		3747
		3748	In this case getting the pointer to C<my_biggy> is a bit more
		3749	complicated: Either you store the address of your C<my_biggy> struct in
		3750	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3751	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3752	real programmers):
		3753
		3754	#include <stddef.h>
		3755
		3756	static void
		3757	t1_cb (EV_P_ ev_timer *w, int revents)
		3758	{
		3759	struct my_biggy big = (struct my_biggy *)
		3760	(((char *)w) - offsetof (struct my_biggy, t1));
		3761	}
		3762
		3763	static void
		3764	t2_cb (EV_P_ ev_timer *w, int revents)
		3765	{
		3766	struct my_biggy big = (struct my_biggy *)
		3767	(((char *)w) - offsetof (struct my_biggy, t2));
		3768	}
		3769
		3770	=head2 AVOIDING FINISHING BEFORE RETURNING
		3771
		3772	Often you have structures like this in event-based programs:
		3773
		3774	callback ()
2080	{	3775	{
2081	if (revents & EV_TIMEOUT)	3776	free (request);
2082	/* doh, nothing entered */;
2083	else if (revents & EV_READ)
2084	/* stdin might have data for us, joy! */;
2085	}	3777	}
2086		3778
2087	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3779	request = start_new_request (..., callback);
2088		3780
2089	=item ev_feed_event (ev_loop *, watcher *, int revents)	3781	The intent is to start some "lengthy" operation. The C<request> could be
		3782	used to cancel the operation, or do other things with it.
2090		3783
2091	Feeds the given event set into the event loop, as if the specified event	3784	It's not uncommon to have code paths in C<start_new_request> that
2092	had happened for the specified watcher (which must be a pointer to an	3785	immediately invoke the callback, for example, to report errors. Or you add
2093	initialised but not necessarily started event watcher).	3786	some caching layer that finds that it can skip the lengthy aspects of the
		3787	operation and simply invoke the callback with the result.
2094		3788
2095	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3789	The problem here is that this will happen I<before> C<start_new_request>
		3790	has returned, so C<request> is not set.
2096		3791
2097	Feed an event on the given fd, as if a file descriptor backend detected	3792	Even if you pass the request by some safer means to the callback, you
2098	the given events it.	3793	might want to do something to the request after starting it, such as
		3794	canceling it, which probably isn't working so well when the callback has
		3795	already been invoked.
2099		3796
2100	=item ev_feed_signal_event (ev_loop *loop, int signum)	3797	A common way around all these issues is to make sure that
		3798	C<start_new_request> I<always> returns before the callback is invoked. If
		3799	C<start_new_request> immediately knows the result, it can artificially
		3800	delay invoking the callback by using a C<prepare> or C<idle> watcher for
		3801	example, or more sneakily, by reusing an existing (stopped) watcher and
		3802	pushing it into the pending queue:
2101		3803
2102	Feed an event as if the given signal occured (C<loop> must be the default	3804	ev_set_cb (watcher, callback);
2103	loop!).	3805	ev_feed_event (EV_A_ watcher, 0);
2104		3806
2105	=back	3807	This way, C<start_new_request> can safely return before the callback is
		3808	invoked, while not delaying callback invocation too much.
		3809
		3810	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3811
		3812	Often (especially in GUI toolkits) there are places where you have
		3813	I<modal> interaction, which is most easily implemented by recursively
		3814	invoking C<ev_run>.
		3815
		3816	This brings the problem of exiting - a callback might want to finish the
		3817	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3818	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3819	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3820	other combination: In these cases, a simple C<ev_break> will not work.
		3821
		3822	The solution is to maintain "break this loop" variable for each C<ev_run>
		3823	invocation, and use a loop around C<ev_run> until the condition is
		3824	triggered, using C<EVRUN_ONCE>:
		3825
		3826	// main loop
		3827	int exit_main_loop = 0;
		3828
		3829	while (!exit_main_loop)
		3830	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3831
		3832	// in a modal watcher
		3833	int exit_nested_loop = 0;
		3834
		3835	while (!exit_nested_loop)
		3836	ev_run (EV_A_ EVRUN_ONCE);
		3837
		3838	To exit from any of these loops, just set the corresponding exit variable:
		3839
		3840	// exit modal loop
		3841	exit_nested_loop = 1;
		3842
		3843	// exit main program, after modal loop is finished
		3844	exit_main_loop = 1;
		3845
		3846	// exit both
		3847	exit_main_loop = exit_nested_loop = 1;
		3848
		3849	=head2 THREAD LOCKING EXAMPLE
		3850
		3851	Here is a fictitious example of how to run an event loop in a different
		3852	thread from where callbacks are being invoked and watchers are
		3853	created/added/removed.
		3854
		3855	For a real-world example, see the C<EV::Loop::Async> perl module,
		3856	which uses exactly this technique (which is suited for many high-level
		3857	languages).
		3858
		3859	The example uses a pthread mutex to protect the loop data, a condition
		3860	variable to wait for callback invocations, an async watcher to notify the
		3861	event loop thread and an unspecified mechanism to wake up the main thread.
		3862
		3863	First, you need to associate some data with the event loop:
		3864
		3865	typedef struct {
		3866	mutex_t lock; /* global loop lock */
		3867	ev_async async_w;
		3868	thread_t tid;
		3869	cond_t invoke_cv;
		3870	} userdata;
		3871
		3872	void prepare_loop (EV_P)
		3873	{
		3874	// for simplicity, we use a static userdata struct.
		3875	static userdata u;
		3876
		3877	ev_async_init (&u->async_w, async_cb);
		3878	ev_async_start (EV_A_ &u->async_w);
		3879
		3880	pthread_mutex_init (&u->lock, 0);
		3881	pthread_cond_init (&u->invoke_cv, 0);
		3882
		3883	// now associate this with the loop
		3884	ev_set_userdata (EV_A_ u);
		3885	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3886	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3887
		3888	// then create the thread running ev_run
		3889	pthread_create (&u->tid, 0, l_run, EV_A);
		3890	}
		3891
		3892	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3893	solely to wake up the event loop so it takes notice of any new watchers
		3894	that might have been added:
		3895
		3896	static void
		3897	async_cb (EV_P_ ev_async *w, int revents)
		3898	{
		3899	// just used for the side effects
		3900	}
		3901
		3902	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3903	protecting the loop data, respectively.
		3904
		3905	static void
		3906	l_release (EV_P)
		3907	{
		3908	userdata *u = ev_userdata (EV_A);
		3909	pthread_mutex_unlock (&u->lock);
		3910	}
		3911
		3912	static void
		3913	l_acquire (EV_P)
		3914	{
		3915	userdata *u = ev_userdata (EV_A);
		3916	pthread_mutex_lock (&u->lock);
		3917	}
		3918
		3919	The event loop thread first acquires the mutex, and then jumps straight
		3920	into C<ev_run>:
		3921
		3922	void *
		3923	l_run (void *thr_arg)
		3924	{
		3925	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3926
		3927	l_acquire (EV_A);
		3928	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3929	ev_run (EV_A_ 0);
		3930	l_release (EV_A);
		3931
		3932	return 0;
		3933	}
		3934
		3935	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3936	signal the main thread via some unspecified mechanism (signals? pipe
		3937	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3938	have been called (in a while loop because a) spurious wakeups are possible
		3939	and b) skipping inter-thread-communication when there are no pending
		3940	watchers is very beneficial):
		3941
		3942	static void
		3943	l_invoke (EV_P)
		3944	{
		3945	userdata *u = ev_userdata (EV_A);
		3946
		3947	while (ev_pending_count (EV_A))
		3948	{
		3949	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3950	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3951	}
		3952	}
		3953
		3954	Now, whenever the main thread gets told to invoke pending watchers, it
		3955	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3956	thread to continue:
		3957
		3958	static void
		3959	real_invoke_pending (EV_P)
		3960	{
		3961	userdata *u = ev_userdata (EV_A);
		3962
		3963	pthread_mutex_lock (&u->lock);
		3964	ev_invoke_pending (EV_A);
		3965	pthread_cond_signal (&u->invoke_cv);
		3966	pthread_mutex_unlock (&u->lock);
		3967	}
		3968
		3969	Whenever you want to start/stop a watcher or do other modifications to an
		3970	event loop, you will now have to lock:
		3971
		3972	ev_timer timeout_watcher;
		3973	userdata *u = ev_userdata (EV_A);
		3974
		3975	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3976
		3977	pthread_mutex_lock (&u->lock);
		3978	ev_timer_start (EV_A_ &timeout_watcher);
		3979	ev_async_send (EV_A_ &u->async_w);
		3980	pthread_mutex_unlock (&u->lock);
		3981
		3982	Note that sending the C<ev_async> watcher is required because otherwise
		3983	an event loop currently blocking in the kernel will have no knowledge
		3984	about the newly added timer. By waking up the loop it will pick up any new
		3985	watchers in the next event loop iteration.
		3986
		3987	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3988
		3989	While the overhead of a callback that e.g. schedules a thread is small, it
		3990	is still an overhead. If you embed libev, and your main usage is with some
		3991	kind of threads or coroutines, you might want to customise libev so that
		3992	doesn't need callbacks anymore.
		3993
		3994	Imagine you have coroutines that you can switch to using a function
		3995	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3996	and that due to some magic, the currently active coroutine is stored in a
		3997	global called C<current_coro>. Then you can build your own "wait for libev
		3998	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3999	the differing C<;> conventions):
		4000
		4001	#define EV_CB_DECLARE(type) struct my_coro *cb;
		4002	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		4003
		4004	That means instead of having a C callback function, you store the
		4005	coroutine to switch to in each watcher, and instead of having libev call
		4006	your callback, you instead have it switch to that coroutine.
		4007
		4008	A coroutine might now wait for an event with a function called
		4009	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		4010	matter when, or whether the watcher is active or not when this function is
		4011	called):
		4012
		4013	void
		4014	wait_for_event (ev_watcher *w)
		4015	{
		4016	ev_set_cb (w, current_coro);
		4017	switch_to (libev_coro);
		4018	}
		4019
		4020	That basically suspends the coroutine inside C<wait_for_event> and
		4021	continues the libev coroutine, which, when appropriate, switches back to
		4022	this or any other coroutine.
		4023
		4024	You can do similar tricks if you have, say, threads with an event queue -
		4025	instead of storing a coroutine, you store the queue object and instead of
		4026	switching to a coroutine, you push the watcher onto the queue and notify
		4027	any waiters.
		4028
		4029	To embed libev, see L</EMBEDDING>, but in short, it's easiest to create two
		4030	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		4031
		4032	// my_ev.h
		4033	#define EV_CB_DECLARE(type) struct my_coro *cb;
		4034	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		4035	#include "../libev/ev.h"
		4036
		4037	// my_ev.c
		4038	#define EV_H "my_ev.h"
		4039	#include "../libev/ev.c"
		4040
		4041	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		4042	F<my_ev.c> into your project. When properly specifying include paths, you
		4043	can even use F<ev.h> as header file name directly.
2106		4044
2107		4045
2108	=head1 LIBEVENT EMULATION	4046	=head1 LIBEVENT EMULATION
2109		4047
2110	Libev offers a compatibility emulation layer for libevent. It cannot	4048	Libev offers a compatibility emulation layer for libevent. It cannot
2111	emulate the internals of libevent, so here are some usage hints:	4049	emulate the internals of libevent, so here are some usage hints:
2112		4050
2113	=over 4	4051	=over 4
		4052
		4053	=item * Only the libevent-1.4.1-beta API is being emulated.
		4054
		4055	This was the newest libevent version available when libev was implemented,
		4056	and is still mostly unchanged in 2010.
2114		4057
2115	=item * Use it by including <event.h>, as usual.	4058	=item * Use it by including <event.h>, as usual.
2116		4059
2117	=item * The following members are fully supported: ev_base, ev_callback,	4060	=item * The following members are fully supported: ev_base, ev_callback,
2118	ev_arg, ev_fd, ev_res, ev_events.	4061	ev_arg, ev_fd, ev_res, ev_events.
…		…
2123		4066
2124	=item * Priorities are not currently supported. Initialising priorities	4067	=item * Priorities are not currently supported. Initialising priorities
2125	will fail and all watchers will have the same priority, even though there	4068	will fail and all watchers will have the same priority, even though there
2126	is an ev_pri field.	4069	is an ev_pri field.
2127		4070
		4071	=item * In libevent, the last base created gets the signals, in libev, the
		4072	base that registered the signal gets the signals.
		4073
2128	=item * Other members are not supported.	4074	=item * Other members are not supported.
2129		4075
2130	=item * The libev emulation is I<not> ABI compatible to libevent, you need	4076	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2131	to use the libev header file and library.	4077	to use the libev header file and library.
2132		4078
2133	=back	4079	=back
2134		4080
2135	=head1 C++ SUPPORT	4081	=head1 C++ SUPPORT
2136		4082
		4083	=head2 C API
		4084
		4085	The normal C API should work fine when used from C++: both ev.h and the
		4086	libev sources can be compiled as C++. Therefore, code that uses the C API
		4087	will work fine.
		4088
		4089	Proper exception specifications might have to be added to callbacks passed
		4090	to libev: exceptions may be thrown only from watcher callbacks, all other
		4091	callbacks (allocator, syserr, loop acquire/release and periodic reschedule
		4092	callbacks) must not throw exceptions, and might need a C<noexcept>
		4093	specification. If you have code that needs to be compiled as both C and
		4094	C++ you can use the C<EV_NOEXCEPT> macro for this:
		4095
		4096	static void
		4097	fatal_error (const char *msg) EV_NOEXCEPT
		4098	{
		4099	perror (msg);
		4100	abort ();
		4101	}
		4102
		4103	...
		4104	ev_set_syserr_cb (fatal_error);
		4105
		4106	The only API functions that can currently throw exceptions are C<ev_run>,
		4107	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		4108	because it runs cleanup watchers).
		4109
		4110	Throwing exceptions in watcher callbacks is only supported if libev itself
		4111	is compiled with a C++ compiler or your C and C++ environments allow
		4112	throwing exceptions through C libraries (most do).
		4113
		4114	=head2 C++ API
		4115
2137	Libev comes with some simplistic wrapper classes for C++ that mainly allow	4116	Libev comes with some simplistic wrapper classes for C++ that mainly allow
2138	you to use some convinience methods to start/stop watchers and also change	4117	you to use some convenience methods to start/stop watchers and also change
2139	the callback model to a model using method callbacks on objects.	4118	the callback model to a model using method callbacks on objects.
2140		4119
2141	To use it,	4120	To use it,
2142		4121
2143	#include <ev++.h>	4122	#include <ev++.h>
2144		4123
2145	This automatically includes F<ev.h> and puts all of its definitions (many	4124	This automatically includes F<ev.h> and puts all of its definitions (many
2146	of them macros) into the global namespace. All C++ specific things are	4125	of them macros) into the global namespace. All C++ specific things are
2147	put into the C<ev> namespace. It should support all the same embedding	4126	put into the C<ev> namespace. It should support all the same embedding
2148	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.	4127	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
…		…
2150	Care has been taken to keep the overhead low. The only data member the C++	4129	Care has been taken to keep the overhead low. The only data member the C++
2151	classes add (compared to plain C-style watchers) is the event loop pointer	4130	classes add (compared to plain C-style watchers) is the event loop pointer
2152	that the watcher is associated with (or no additional members at all if	4131	that the watcher is associated with (or no additional members at all if
2153	you disable C<EV_MULTIPLICITY> when embedding libev).	4132	you disable C<EV_MULTIPLICITY> when embedding libev).
2154		4133
2155	Currently, functions, and static and non-static member functions can be	4134	Currently, functions, static and non-static member functions and classes
2156	used as callbacks. Other types should be easy to add as long as they only	4135	with C<operator ()> can be used as callbacks. Other types should be easy
2157	need one additional pointer for context. If you need support for other	4136	to add as long as they only need one additional pointer for context. If
2158	types of functors please contact the author (preferably after implementing	4137	you need support for other types of functors please contact the author
2159	it).	4138	(preferably after implementing it).
		4139
		4140	For all this to work, your C++ compiler either has to use the same calling
		4141	conventions as your C compiler (for static member functions), or you have
		4142	to embed libev and compile libev itself as C++.
2160		4143
2161	Here is a list of things available in the C<ev> namespace:	4144	Here is a list of things available in the C<ev> namespace:
2162		4145
2163	=over 4	4146	=over 4
2164		4147
…		…
2174	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.	4157	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
2175		4158
2176	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of	4159	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
2177	the same name in the C<ev> namespace, with the exception of C<ev_signal>	4160	the same name in the C<ev> namespace, with the exception of C<ev_signal>
2178	which is called C<ev::sig> to avoid clashes with the C<signal> macro	4161	which is called C<ev::sig> to avoid clashes with the C<signal> macro
2179	defines by many implementations.	4162	defined by many implementations.
2180		4163
2181	All of those classes have these methods:	4164	All of those classes have these methods:
2182		4165
2183	=over 4	4166	=over 4
2184		4167
2185	=item ev::TYPE::TYPE ()	4168	=item ev::TYPE::TYPE ()
2186		4169
2187	=item ev::TYPE::TYPE (struct ev_loop *)	4170	=item ev::TYPE::TYPE (loop)
2188		4171
2189	=item ev::TYPE::~TYPE	4172	=item ev::TYPE::~TYPE
2190		4173
2191	The constructor (optionally) takes an event loop to associate the watcher	4174	The constructor (optionally) takes an event loop to associate the watcher
2192	with. If it is omitted, it will use C<EV_DEFAULT>.	4175	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2215	your compiler is good :), then the method will be fully inlined into the	4198	your compiler is good :), then the method will be fully inlined into the
2216	thunking function, making it as fast as a direct C callback.	4199	thunking function, making it as fast as a direct C callback.
2217		4200
2218	Example: simple class declaration and watcher initialisation	4201	Example: simple class declaration and watcher initialisation
2219		4202
2220	struct myclass	4203	struct myclass
2221	{	4204	{
2222	void io_cb (ev::io &w, int revents) { }	4205	void io_cb (ev::io &w, int revents) { }
2223	}	4206	}
2224		4207
2225	myclass obj;	4208	myclass obj;
2226	ev::io iow;	4209	ev::io iow;
2227	iow.set <myclass, &myclass::io_cb> (&obj);	4210	iow.set <myclass, &myclass::io_cb> (&obj);
		4211
		4212	=item w->set (object *)
		4213
		4214	This is a variation of a method callback - leaving out the method to call
		4215	will default the method to C<operator ()>, which makes it possible to use
		4216	functor objects without having to manually specify the C<operator ()> all
		4217	the time. Incidentally, you can then also leave out the template argument
		4218	list.
		4219
		4220	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		4221	int revents)>.
		4222
		4223	See the method-C<set> above for more details.
		4224
		4225	Example: use a functor object as callback.
		4226
		4227	struct myfunctor
		4228	{
		4229	void operator() (ev::io &w, int revents)
		4230	{
		4231	...
		4232	}
		4233	}
		4234
		4235	myfunctor f;
		4236
		4237	ev::io w;
		4238	w.set (&f);
2228		4239
2229	=item w->set<function> (void *data = 0)	4240	=item w->set<function> (void *data = 0)
2230		4241
2231	Also sets a callback, but uses a static method or plain function as	4242	Also sets a callback, but uses a static method or plain function as
2232	callback. The optional C<data> argument will be stored in the watcher's	4243	callback. The optional C<data> argument will be stored in the watcher's
…		…
2234		4245
2235	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	4246	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2236		4247
2237	See the method-C<set> above for more details.	4248	See the method-C<set> above for more details.
2238		4249
2239	Example:	4250	Example: Use a plain function as callback.
2240		4251
2241	static void io_cb (ev::io &w, int revents) { }	4252	static void io_cb (ev::io &w, int revents) { }
2242	iow.set <io_cb> ();	4253	iow.set <io_cb> ();
2243		4254
2244	=item w->set (struct ev_loop *)	4255	=item w->set (loop)
2245		4256
2246	Associates a different C<struct ev_loop> with this watcher. You can only	4257	Associates a different C<struct ev_loop> with this watcher. You can only
2247	do this when the watcher is inactive (and not pending either).	4258	do this when the watcher is inactive (and not pending either).
2248		4259
2249	=item w->set ([args])	4260	=item w->set ([arguments])
2250		4261
2251	Basically the same as C<ev_TYPE_set>, with the same args. Must be	4262	Basically the same as C<ev_TYPE_set> (except for C<ev::embed> watchers>),
		4263	with the same arguments. Either this method or a suitable start method
2252	called at least once. Unlike the C counterpart, an active watcher gets	4264	must be called at least once. Unlike the C counterpart, an active watcher
2253	automatically stopped and restarted when reconfiguring it with this	4265	gets automatically stopped and restarted when reconfiguring it with this
2254	method.	4266	method.
		4267
		4268	For C<ev::embed> watchers this method is called C<set_embed>, to avoid
		4269	clashing with the C<set (loop)> method.
		4270
		4271	For C<ev::io> watchers there is an additional C<set> method that acepts a
		4272	new event mask only, and internally calls C<ev_io_modfify>.
2255		4273
2256	=item w->start ()	4274	=item w->start ()
2257		4275
2258	Starts the watcher. Note that there is no C<loop> argument, as the	4276	Starts the watcher. Note that there is no C<loop> argument, as the
2259	constructor already stores the event loop.	4277	constructor already stores the event loop.
2260		4278
		4279	=item w->start ([arguments])
		4280
		4281	Instead of calling C<set> and C<start> methods separately, it is often
		4282	convenient to wrap them in one call. Uses the same type of arguments as
		4283	the configure C<set> method of the watcher.
		4284
2261	=item w->stop ()	4285	=item w->stop ()
2262		4286
2263	Stops the watcher if it is active. Again, no C<loop> argument.	4287	Stops the watcher if it is active. Again, no C<loop> argument.
2264		4288
2265	=item w->again () (C<ev::timer>, C<ev::periodic> only)	4289	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
2277		4301
2278	=back	4302	=back
2279		4303
2280	=back	4304	=back
2281		4305
2282	Example: Define a class with an IO and idle watcher, start one of them in	4306	Example: Define a class with two I/O and idle watchers, start the I/O
2283	the constructor.	4307	watchers in the constructor.
2284		4308
2285	class myclass	4309	class myclass
2286	{	4310	{
2287	ev::io io; void io_cb (ev::io &w, int revents);	4311	ev::io io ; void io_cb (ev::io &w, int revents);
		4312	ev::io io2 ; void io2_cb (ev::io &w, int revents);
2288	ev:idle idle void idle_cb (ev::idle &w, int revents);	4313	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2289		4314
2290	myclass (int fd)	4315	myclass (int fd)
2291	{	4316	{
2292	io .set <myclass, &myclass::io_cb > (this);	4317	io .set <myclass, &myclass::io_cb > (this);
		4318	io2 .set <myclass, &myclass::io2_cb > (this);
2293	idle.set <myclass, &myclass::idle_cb> (this);	4319	idle.set <myclass, &myclass::idle_cb> (this);
2294		4320
2295	io.start (fd, ev::READ);	4321	io.set (fd, ev::WRITE); // configure the watcher
		4322	io.start (); // start it whenever convenient
		4323
		4324	io2.start (fd, ev::READ); // set + start in one call
2296	}	4325	}
2297	};	4326	};
		4327
		4328
		4329	=head1 OTHER LANGUAGE BINDINGS
		4330
		4331	Libev does not offer other language bindings itself, but bindings for a
		4332	number of languages exist in the form of third-party packages. If you know
		4333	any interesting language binding in addition to the ones listed here, drop
		4334	me a note.
		4335
		4336	=over 4
		4337
		4338	=item Perl
		4339
		4340	The EV module implements the full libev API and is actually used to test
		4341	libev. EV is developed together with libev. Apart from the EV core module,
		4342	there are additional modules that implement libev-compatible interfaces
		4343	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		4344	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		4345	and C<EV::Glib>).
		4346
		4347	It can be found and installed via CPAN, its homepage is at
		4348	L<http://software.schmorp.de/pkg/EV>.
		4349
		4350	=item Python
		4351
		4352	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		4353	seems to be quite complete and well-documented.
		4354
		4355	=item Ruby
		4356
		4357	Tony Arcieri has written a ruby extension that offers access to a subset
		4358	of the libev API and adds file handle abstractions, asynchronous DNS and
		4359	more on top of it. It can be found via gem servers. Its homepage is at
		4360	L<http://rev.rubyforge.org/>.
		4361
		4362	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		4363	makes rev work even on mingw.
		4364
		4365	=item Haskell
		4366
		4367	A haskell binding to libev is available at
		4368	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		4369
		4370	=item D
		4371
		4372	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		4373	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
		4374
		4375	=item Ocaml
		4376
		4377	Erkki Seppala has written Ocaml bindings for libev, to be found at
		4378	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		4379
		4380	=item Lua
		4381
		4382	Brian Maher has written a partial interface to libev for lua (at the
		4383	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		4384	L<http://github.com/brimworks/lua-ev>.
		4385
		4386	=item Javascript
		4387
		4388	Node.js (L<http://nodejs.org>) uses libev as the underlying event library.
		4389
		4390	=item Others
		4391
		4392	There are others, and I stopped counting.
		4393
		4394	=back
2298		4395
2299		4396
2300	=head1 MACRO MAGIC	4397	=head1 MACRO MAGIC
2301		4398
2302	Libev can be compiled with a variety of options, the most fundamantal	4399	Libev can be compiled with a variety of options, the most fundamental
2303	of which is C<EV_MULTIPLICITY>. This option determines whether (most)	4400	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
2304	functions and callbacks have an initial C<struct ev_loop *> argument.	4401	functions and callbacks have an initial C<struct ev_loop *> argument.
2305		4402
2306	To make it easier to write programs that cope with either variant, the	4403	To make it easier to write programs that cope with either variant, the
2307	following macros are defined:	4404	following macros are defined:
…		…
2312		4409
2313	This provides the loop I<argument> for functions, if one is required ("ev	4410	This provides the loop I<argument> for functions, if one is required ("ev
2314	loop argument"). The C<EV_A> form is used when this is the sole argument,	4411	loop argument"). The C<EV_A> form is used when this is the sole argument,
2315	C<EV_A_> is used when other arguments are following. Example:	4412	C<EV_A_> is used when other arguments are following. Example:
2316		4413
2317	ev_unref (EV_A);	4414	ev_unref (EV_A);
2318	ev_timer_add (EV_A_ watcher);	4415	ev_timer_add (EV_A_ watcher);
2319	ev_loop (EV_A_ 0);	4416	ev_run (EV_A_ 0);
2320		4417
2321	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	4418	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
2322	which is often provided by the following macro.	4419	which is often provided by the following macro.
2323		4420
2324	=item C<EV_P>, C<EV_P_>	4421	=item C<EV_P>, C<EV_P_>
2325		4422
2326	This provides the loop I<parameter> for functions, if one is required ("ev	4423	This provides the loop I<parameter> for functions, if one is required ("ev
2327	loop parameter"). The C<EV_P> form is used when this is the sole parameter,	4424	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
2328	C<EV_P_> is used when other parameters are following. Example:	4425	C<EV_P_> is used when other parameters are following. Example:
2329		4426
2330	// this is how ev_unref is being declared	4427	// this is how ev_unref is being declared
2331	static void ev_unref (EV_P);	4428	static void ev_unref (EV_P);
2332		4429
2333	// this is how you can declare your typical callback	4430	// this is how you can declare your typical callback
2334	static void cb (EV_P_ ev_timer *w, int revents)	4431	static void cb (EV_P_ ev_timer *w, int revents)
2335		4432
2336	It declares a parameter C<loop> of type C<struct ev_loop *>, quite	4433	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
2337	suitable for use with C<EV_A>.	4434	suitable for use with C<EV_A>.
2338		4435
2339	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4436	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
2340		4437
2341	Similar to the other two macros, this gives you the value of the default	4438	Similar to the other two macros, this gives you the value of the default
2342	loop, if multiple loops are supported ("ev loop default").	4439	loop, if multiple loops are supported ("ev loop default"). The default loop
		4440	will be initialised if it isn't already initialised.
		4441
		4442	For non-multiplicity builds, these macros do nothing, so you always have
		4443	to initialise the loop somewhere.
		4444
		4445	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		4446
		4447	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		4448	default loop has been initialised (C<UC> == unchecked). Their behaviour
		4449	is undefined when the default loop has not been initialised by a previous
		4450	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		4451
		4452	It is often prudent to use C<EV_DEFAULT> when initialising the first
		4453	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
2343		4454
2344	=back	4455	=back
2345		4456
2346	Example: Declare and initialise a check watcher, utilising the above	4457	Example: Declare and initialise a check watcher, utilising the above
2347	macros so it will work regardless of whether multiple loops are supported	4458	macros so it will work regardless of whether multiple loops are supported
2348	or not.	4459	or not.
2349		4460
2350	static void	4461	static void
2351	check_cb (EV_P_ ev_timer *w, int revents)	4462	check_cb (EV_P_ ev_timer *w, int revents)
2352	{	4463	{
2353	ev_check_stop (EV_A_ w);	4464	ev_check_stop (EV_A_ w);
2354	}	4465	}
2355		4466
2356	ev_check check;	4467	ev_check check;
2357	ev_check_init (&check, check_cb);	4468	ev_check_init (&check, check_cb);
2358	ev_check_start (EV_DEFAULT_ &check);	4469	ev_check_start (EV_DEFAULT_ &check);
2359	ev_loop (EV_DEFAULT_ 0);	4470	ev_run (EV_DEFAULT_ 0);
2360		4471
2361	=head1 EMBEDDING	4472	=head1 EMBEDDING
2362		4473
2363	Libev can (and often is) directly embedded into host	4474	Libev can (and often is) directly embedded into host
2364	applications. Examples of applications that embed it include the Deliantra	4475	applications. Examples of applications that embed it include the Deliantra
…		…
2371	libev somewhere in your source tree).	4482	libev somewhere in your source tree).
2372		4483
2373	=head2 FILESETS	4484	=head2 FILESETS
2374		4485
2375	Depending on what features you need you need to include one or more sets of files	4486	Depending on what features you need you need to include one or more sets of files
2376	in your app.	4487	in your application.
2377		4488
2378	=head3 CORE EVENT LOOP	4489	=head3 CORE EVENT LOOP
2379		4490
2380	To include only the libev core (all the C<ev_*> functions), with manual	4491	To include only the libev core (all the C<ev_*> functions), with manual
2381	configuration (no autoconf):	4492	configuration (no autoconf):
2382		4493
2383	#define EV_STANDALONE 1	4494	#define EV_STANDALONE 1
2384	#include "ev.c"	4495	#include "ev.c"
2385		4496
2386	This will automatically include F<ev.h>, too, and should be done in a	4497	This will automatically include F<ev.h>, too, and should be done in a
2387	single C source file only to provide the function implementations. To use	4498	single C source file only to provide the function implementations. To use
2388	it, do the same for F<ev.h> in all files wishing to use this API (best	4499	it, do the same for F<ev.h> in all files wishing to use this API (best
2389	done by writing a wrapper around F<ev.h> that you can include instead and	4500	done by writing a wrapper around F<ev.h> that you can include instead and
2390	where you can put other configuration options):	4501	where you can put other configuration options):
2391		4502
2392	#define EV_STANDALONE 1	4503	#define EV_STANDALONE 1
2393	#include "ev.h"	4504	#include "ev.h"
2394		4505
2395	Both header files and implementation files can be compiled with a C++	4506	Both header files and implementation files can be compiled with a C++
2396	compiler (at least, thats a stated goal, and breakage will be treated	4507	compiler (at least, that's a stated goal, and breakage will be treated
2397	as a bug).	4508	as a bug).
2398		4509
2399	You need the following files in your source tree, or in a directory	4510	You need the following files in your source tree, or in a directory
2400	in your include path (e.g. in libev/ when using -Ilibev):	4511	in your include path (e.g. in libev/ when using -Ilibev):
2401		4512
2402	ev.h	4513	ev.h
2403	ev.c	4514	ev.c
2404	ev_vars.h	4515	ev_vars.h
2405	ev_wrap.h	4516	ev_wrap.h
2406		4517
2407	ev_win32.c required on win32 platforms only	4518	ev_win32.c required on win32 platforms only
2408		4519
2409	ev_select.c only when select backend is enabled (which is enabled by default)	4520	ev_select.c only when select backend is enabled
2410	ev_poll.c only when poll backend is enabled (disabled by default)	4521	ev_poll.c only when poll backend is enabled
2411	ev_epoll.c only when the epoll backend is enabled (disabled by default)	4522	ev_epoll.c only when the epoll backend is enabled
		4523	ev_linuxaio.c only when the linux aio backend is enabled
		4524	ev_iouring.c only when the linux io_uring backend is enabled
2412	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	4525	ev_kqueue.c only when the kqueue backend is enabled
2413	ev_port.c only when the solaris port backend is enabled (disabled by default)	4526	ev_port.c only when the solaris port backend is enabled
2414		4527
2415	F<ev.c> includes the backend files directly when enabled, so you only need	4528	F<ev.c> includes the backend files directly when enabled, so you only need
2416	to compile this single file.	4529	to compile this single file.
2417		4530
2418	=head3 LIBEVENT COMPATIBILITY API	4531	=head3 LIBEVENT COMPATIBILITY API
2419		4532
2420	To include the libevent compatibility API, also include:	4533	To include the libevent compatibility API, also include:
2421		4534
2422	#include "event.c"	4535	#include "event.c"
2423		4536
2424	in the file including F<ev.c>, and:	4537	in the file including F<ev.c>, and:
2425		4538
2426	#include "event.h"	4539	#include "event.h"
2427		4540
2428	in the files that want to use the libevent API. This also includes F<ev.h>.	4541	in the files that want to use the libevent API. This also includes F<ev.h>.
2429		4542
2430	You need the following additional files for this:	4543	You need the following additional files for this:
2431		4544
2432	event.h	4545	event.h
2433	event.c	4546	event.c
2434		4547
2435	=head3 AUTOCONF SUPPORT	4548	=head3 AUTOCONF SUPPORT
2436		4549
2437	Instead of using C<EV_STANDALONE=1> and providing your config in	4550	Instead of using C<EV_STANDALONE=1> and providing your configuration in
2438	whatever way you want, you can also C<m4_include([libev.m4])> in your	4551	whatever way you want, you can also C<m4_include([libev.m4])> in your
2439	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then	4552	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
2440	include F<config.h> and configure itself accordingly.	4553	include F<config.h> and configure itself accordingly.
2441		4554
2442	For this of course you need the m4 file:	4555	For this of course you need the m4 file:
2443		4556
2444	libev.m4	4557	libev.m4
2445		4558
2446	=head2 PREPROCESSOR SYMBOLS/MACROS	4559	=head2 PREPROCESSOR SYMBOLS/MACROS
2447		4560
2448	Libev can be configured via a variety of preprocessor symbols you have to define	4561	Libev can be configured via a variety of preprocessor symbols you have to
2449	before including any of its files. The default is not to build for multiplicity	4562	define before including (or compiling) any of its files. The default in
2450	and only include the select backend.	4563	the absence of autoconf is documented for every option.
		4564
		4565	Symbols marked with "(h)" do not change the ABI, and can have different
		4566	values when compiling libev vs. including F<ev.h>, so it is permissible
		4567	to redefine them before including F<ev.h> without breaking compatibility
		4568	to a compiled library. All other symbols change the ABI, which means all
		4569	users of libev and the libev code itself must be compiled with compatible
		4570	settings.
2451		4571
2452	=over 4	4572	=over 4
2453		4573
		4574	=item EV_COMPAT3 (h)
		4575
		4576	Backwards compatibility is a major concern for libev. This is why this
		4577	release of libev comes with wrappers for the functions and symbols that
		4578	have been renamed between libev version 3 and 4.
		4579
		4580	You can disable these wrappers (to test compatibility with future
		4581	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4582	sources. This has the additional advantage that you can drop the C<struct>
		4583	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4584	typedef in that case.
		4585
		4586	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4587	and in some even more future version the compatibility code will be
		4588	removed completely.
		4589
2454	=item EV_STANDALONE	4590	=item EV_STANDALONE (h)
2455		4591
2456	Must always be C<1> if you do not use autoconf configuration, which	4592	Must always be C<1> if you do not use autoconf configuration, which
2457	keeps libev from including F<config.h>, and it also defines dummy	4593	keeps libev from including F<config.h>, and it also defines dummy
2458	implementations for some libevent functions (such as logging, which is not	4594	implementations for some libevent functions (such as logging, which is not
2459	supported). It will also not define any of the structs usually found in	4595	supported). It will also not define any of the structs usually found in
2460	F<event.h> that are not directly supported by the libev core alone.	4596	F<event.h> that are not directly supported by the libev core alone.
2461		4597
		4598	In standalone mode, libev will still try to automatically deduce the
		4599	configuration, but has to be more conservative.
		4600
		4601	=item EV_USE_FLOOR
		4602
		4603	If defined to be C<1>, libev will use the C<floor ()> function for its
		4604	periodic reschedule calculations, otherwise libev will fall back on a
		4605	portable (slower) implementation. If you enable this, you usually have to
		4606	link against libm or something equivalent. Enabling this when the C<floor>
		4607	function is not available will fail, so the safe default is to not enable
		4608	this.
		4609
2462	=item EV_USE_MONOTONIC	4610	=item EV_USE_MONOTONIC
2463		4611
2464	If defined to be C<1>, libev will try to detect the availability of the	4612	If defined to be C<1>, libev will try to detect the availability of the
2465	monotonic clock option at both compiletime and runtime. Otherwise no use	4613	monotonic clock option at both compile time and runtime. Otherwise no
2466	of the monotonic clock option will be attempted. If you enable this, you	4614	use of the monotonic clock option will be attempted. If you enable this,
2467	usually have to link against librt or something similar. Enabling it when	4615	you usually have to link against librt or something similar. Enabling it
2468	the functionality isn't available is safe, though, although you have	4616	when the functionality isn't available is safe, though, although you have
2469	to make sure you link against any libraries where the C<clock_gettime>	4617	to make sure you link against any libraries where the C<clock_gettime>
2470	function is hiding in (often F<-lrt>).	4618	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2471		4619
2472	=item EV_USE_REALTIME	4620	=item EV_USE_REALTIME
2473		4621
2474	If defined to be C<1>, libev will try to detect the availability of the	4622	If defined to be C<1>, libev will try to detect the availability of the
2475	realtime clock option at compiletime (and assume its availability at	4623	real-time clock option at compile time (and assume its availability
2476	runtime if successful). Otherwise no use of the realtime clock option will	4624	at runtime if successful). Otherwise no use of the real-time clock
2477	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	4625	option will be attempted. This effectively replaces C<gettimeofday>
2478	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	4626	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2479	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	4627	correctness. See the note about libraries in the description of
		4628	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		4629	C<EV_USE_CLOCK_SYSCALL>.
		4630
		4631	=item EV_USE_CLOCK_SYSCALL
		4632
		4633	If defined to be C<1>, libev will try to use a direct syscall instead
		4634	of calling the system-provided C<clock_gettime> function. This option
		4635	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		4636	unconditionally pulls in C<libpthread>, slowing down single-threaded
		4637	programs needlessly. Using a direct syscall is slightly slower (in
		4638	theory), because no optimised vdso implementation can be used, but avoids
		4639	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		4640	higher, as it simplifies linking (no need for C<-lrt>).
2480		4641
2481	=item EV_USE_NANOSLEEP	4642	=item EV_USE_NANOSLEEP
2482		4643
2483	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	4644	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2484	and will use it for delays. Otherwise it will use C<select ()>.	4645	and will use it for delays. Otherwise it will use C<select ()>.
2485		4646
		4647	=item EV_USE_EVENTFD
		4648
		4649	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		4650	available and will probe for kernel support at runtime. This will improve
		4651	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		4652	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4653	2.7 or newer, otherwise disabled.
		4654
		4655	=item EV_USE_SIGNALFD
		4656
		4657	If defined to be C<1>, then libev will assume that C<signalfd ()> is
		4658	available and will probe for kernel support at runtime. This enables
		4659	the use of EVFLAG_SIGNALFD for faster and simpler signal handling. If
		4660	undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4661	2.7 or newer, otherwise disabled.
		4662
		4663	=item EV_USE_TIMERFD
		4664
		4665	If defined to be C<1>, then libev will assume that C<timerfd ()> is
		4666	available and will probe for kernel support at runtime. This allows
		4667	libev to detect time jumps accurately. If undefined, it will be enabled
		4668	if the headers indicate GNU/Linux + Glibc 2.8 or newer and define
		4669	C<TFD_TIMER_CANCEL_ON_SET>, otherwise disabled.
		4670
		4671	=item EV_USE_EVENTFD
		4672
		4673	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		4674	available and will probe for kernel support at runtime. This will improve
		4675	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		4676	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4677	2.7 or newer, otherwise disabled.
		4678
2486	=item EV_USE_SELECT	4679	=item EV_USE_SELECT
2487		4680
2488	If undefined or defined to be C<1>, libev will compile in support for the	4681	If undefined or defined to be C<1>, libev will compile in support for the
2489	C<select>(2) backend. No attempt at autodetection will be done: if no	4682	C<select>(2) backend. No attempt at auto-detection will be done: if no
2490	other method takes over, select will be it. Otherwise the select backend	4683	other method takes over, select will be it. Otherwise the select backend
2491	will not be compiled in.	4684	will not be compiled in.
2492		4685
2493	=item EV_SELECT_USE_FD_SET	4686	=item EV_SELECT_USE_FD_SET
2494		4687
2495	If defined to C<1>, then the select backend will use the system C<fd_set>	4688	If defined to C<1>, then the select backend will use the system C<fd_set>
2496	structure. This is useful if libev doesn't compile due to a missing	4689	structure. This is useful if libev doesn't compile due to a missing
2497	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	4690	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2498	exotic systems. This usually limits the range of file descriptors to some	4691	on exotic systems. This usually limits the range of file descriptors to
2499	low limit such as 1024 or might have other limitations (winsocket only	4692	some low limit such as 1024 or might have other limitations (winsocket
2500	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	4693	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2501	influence the size of the C<fd_set> used.	4694	configures the maximum size of the C<fd_set>.
2502		4695
2503	=item EV_SELECT_IS_WINSOCKET	4696	=item EV_SELECT_IS_WINSOCKET
2504		4697
2505	When defined to C<1>, the select backend will assume that	4698	When defined to C<1>, the select backend will assume that
2506	select/socket/connect etc. don't understand file descriptors but	4699	select/socket/connect etc. don't understand file descriptors but
…		…
2508	be used is the winsock select). This means that it will call	4701	be used is the winsock select). This means that it will call
2509	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	4702	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2510	it is assumed that all these functions actually work on fds, even	4703	it is assumed that all these functions actually work on fds, even
2511	on win32. Should not be defined on non-win32 platforms.	4704	on win32. Should not be defined on non-win32 platforms.
2512		4705
2513	=item EV_FD_TO_WIN32_HANDLE	4706	=item EV_FD_TO_WIN32_HANDLE(fd)
2514		4707
2515	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	4708	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2516	file descriptors to socket handles. When not defining this symbol (the	4709	file descriptors to socket handles. When not defining this symbol (the
2517	default), then libev will call C<_get_osfhandle>, which is usually	4710	default), then libev will call C<_get_osfhandle>, which is usually
2518	correct. In some cases, programs use their own file descriptor management,	4711	correct. In some cases, programs use their own file descriptor management,
2519	in which case they can provide this function to map fds to socket handles.	4712	in which case they can provide this function to map fds to socket handles.
2520		4713
		4714	=item EV_WIN32_HANDLE_TO_FD(handle)
		4715
		4716	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		4717	using the standard C<_open_osfhandle> function. For programs implementing
		4718	their own fd to handle mapping, overwriting this function makes it easier
		4719	to do so. This can be done by defining this macro to an appropriate value.
		4720
		4721	=item EV_WIN32_CLOSE_FD(fd)
		4722
		4723	If programs implement their own fd to handle mapping on win32, then this
		4724	macro can be used to override the C<close> function, useful to unregister
		4725	file descriptors again. Note that the replacement function has to close
		4726	the underlying OS handle.
		4727
		4728	=item EV_USE_WSASOCKET
		4729
		4730	If defined to be C<1>, libev will use C<WSASocket> to create its internal
		4731	communication socket, which works better in some environments. Otherwise,
		4732	the normal C<socket> function will be used, which works better in other
		4733	environments.
		4734
2521	=item EV_USE_POLL	4735	=item EV_USE_POLL
2522		4736
2523	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4737	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2524	backend. Otherwise it will be enabled on non-win32 platforms. It	4738	backend. Otherwise it will be enabled on non-win32 platforms. It
2525	takes precedence over select.	4739	takes precedence over select.
2526		4740
2527	=item EV_USE_EPOLL	4741	=item EV_USE_EPOLL
2528		4742
2529	If defined to be C<1>, libev will compile in support for the Linux	4743	If defined to be C<1>, libev will compile in support for the Linux
2530	C<epoll>(7) backend. Its availability will be detected at runtime,	4744	C<epoll>(7) backend. Its availability will be detected at runtime,
2531	otherwise another method will be used as fallback. This is the	4745	otherwise another method will be used as fallback. This is the preferred
2532	preferred backend for GNU/Linux systems.	4746	backend for GNU/Linux systems. If undefined, it will be enabled if the
		4747	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4748
		4749	=item EV_USE_LINUXAIO
		4750
		4751	If defined to be C<1>, libev will compile in support for the Linux aio
		4752	backend (C<EV_USE_EPOLL> must also be enabled). If undefined, it will be
		4753	enabled on linux, otherwise disabled.
		4754
		4755	=item EV_USE_IOURING
		4756
		4757	If defined to be C<1>, libev will compile in support for the Linux
		4758	io_uring backend (C<EV_USE_EPOLL> must also be enabled). Due to it's
		4759	current limitations it has to be requested explicitly. If undefined, it
		4760	will be enabled on linux, otherwise disabled.
2533		4761
2534	=item EV_USE_KQUEUE	4762	=item EV_USE_KQUEUE
2535		4763
2536	If defined to be C<1>, libev will compile in support for the BSD style	4764	If defined to be C<1>, libev will compile in support for the BSD style
2537	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	4765	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
2550	otherwise another method will be used as fallback. This is the preferred	4778	otherwise another method will be used as fallback. This is the preferred
2551	backend for Solaris 10 systems.	4779	backend for Solaris 10 systems.
2552		4780
2553	=item EV_USE_DEVPOLL	4781	=item EV_USE_DEVPOLL
2554		4782
2555	reserved for future expansion, works like the USE symbols above.	4783	Reserved for future expansion, works like the USE symbols above.
2556		4784
2557	=item EV_USE_INOTIFY	4785	=item EV_USE_INOTIFY
2558		4786
2559	If defined to be C<1>, libev will compile in support for the Linux inotify	4787	If defined to be C<1>, libev will compile in support for the Linux inotify
2560	interface to speed up C<ev_stat> watchers. Its actual availability will	4788	interface to speed up C<ev_stat> watchers. Its actual availability will
2561	be detected at runtime.	4789	be detected at runtime. If undefined, it will be enabled if the headers
		4790	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2562		4791
		4792	=item EV_NO_SMP
		4793
		4794	If defined to be C<1>, libev will assume that memory is always coherent
		4795	between threads, that is, threads can be used, but threads never run on
		4796	different cpus (or different cpu cores). This reduces dependencies
		4797	and makes libev faster.
		4798
		4799	=item EV_NO_THREADS
		4800
		4801	If defined to be C<1>, libev will assume that it will never be called from
		4802	different threads (that includes signal handlers), which is a stronger
		4803	assumption than C<EV_NO_SMP>, above. This reduces dependencies and makes
		4804	libev faster.
		4805
		4806	=item EV_ATOMIC_T
		4807
		4808	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		4809	access is atomic with respect to other threads or signal contexts. No
		4810	such type is easily found in the C language, so you can provide your own
		4811	type that you know is safe for your purposes. It is used both for signal
		4812	handler "locking" as well as for signal and thread safety in C<ev_async>
		4813	watchers.
		4814
		4815	In the absence of this define, libev will use C<sig_atomic_t volatile>
		4816	(from F<signal.h>), which is usually good enough on most platforms.
		4817
2563	=item EV_H	4818	=item EV_H (h)
2564		4819
2565	The name of the F<ev.h> header file used to include it. The default if	4820	The name of the F<ev.h> header file used to include it. The default if
2566	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	4821	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
2567	used to virtually rename the F<ev.h> header file in case of conflicts.	4822	used to virtually rename the F<ev.h> header file in case of conflicts.
2568		4823
2569	=item EV_CONFIG_H	4824	=item EV_CONFIG_H (h)
2570		4825
2571	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	4826	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2572	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	4827	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2573	C<EV_H>, above.	4828	C<EV_H>, above.
2574		4829
2575	=item EV_EVENT_H	4830	=item EV_EVENT_H (h)
2576		4831
2577	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	4832	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2578	of how the F<event.h> header can be found, the default is C<"event.h">.	4833	of how the F<event.h> header can be found, the default is C<"event.h">.
2579		4834
2580	=item EV_PROTOTYPES	4835	=item EV_PROTOTYPES (h)
2581		4836
2582	If defined to be C<0>, then F<ev.h> will not define any function	4837	If defined to be C<0>, then F<ev.h> will not define any function
2583	prototypes, but still define all the structs and other symbols. This is	4838	prototypes, but still define all the structs and other symbols. This is
2584	occasionally useful if you want to provide your own wrapper functions	4839	occasionally useful if you want to provide your own wrapper functions
2585	around libev functions.	4840	around libev functions.
…		…
2590	will have the C<struct ev_loop *> as first argument, and you can create	4845	will have the C<struct ev_loop *> as first argument, and you can create
2591	additional independent event loops. Otherwise there will be no support	4846	additional independent event loops. Otherwise there will be no support
2592	for multiple event loops and there is no first event loop pointer	4847	for multiple event loops and there is no first event loop pointer
2593	argument. Instead, all functions act on the single default loop.	4848	argument. Instead, all functions act on the single default loop.
2594		4849
		4850	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4851	default loop when multiplicity is switched off - you always have to
		4852	initialise the loop manually in this case.
		4853
2595	=item EV_MINPRI	4854	=item EV_MINPRI
2596		4855
2597	=item EV_MAXPRI	4856	=item EV_MAXPRI
2598		4857
2599	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4858	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
2604	When doing priority-based operations, libev usually has to linearly search	4863	When doing priority-based operations, libev usually has to linearly search
2605	all the priorities, so having many of them (hundreds) uses a lot of space	4864	all the priorities, so having many of them (hundreds) uses a lot of space
2606	and time, so using the defaults of five priorities (-2 .. +2) is usually	4865	and time, so using the defaults of five priorities (-2 .. +2) is usually
2607	fine.	4866	fine.
2608		4867
2609	If your embedding app does not need any priorities, defining these both to	4868	If your embedding application does not need any priorities, defining these
2610	C<0> will save some memory and cpu.	4869	both to C<0> will save some memory and CPU.
2611		4870
2612	=item EV_PERIODIC_ENABLE	4871	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4872	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4873	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
2613		4874
2614	If undefined or defined to be C<1>, then periodic timers are supported. If	4875	If undefined or defined to be C<1> (and the platform supports it), then
2615	defined to be C<0>, then they are not. Disabling them saves a few kB of	4876	the respective watcher type is supported. If defined to be C<0>, then it
2616	code.	4877	is not. Disabling watcher types mainly saves code size.
2617		4878
2618	=item EV_IDLE_ENABLE	4879	=item EV_FEATURES
2619
2620	If undefined or defined to be C<1>, then idle watchers are supported. If
2621	defined to be C<0>, then they are not. Disabling them saves a few kB of
2622	code.
2623
2624	=item EV_EMBED_ENABLE
2625
2626	If undefined or defined to be C<1>, then embed watchers are supported. If
2627	defined to be C<0>, then they are not.
2628
2629	=item EV_STAT_ENABLE
2630
2631	If undefined or defined to be C<1>, then stat watchers are supported. If
2632	defined to be C<0>, then they are not.
2633
2634	=item EV_FORK_ENABLE
2635
2636	If undefined or defined to be C<1>, then fork watchers are supported. If
2637	defined to be C<0>, then they are not.
2638
2639	=item EV_MINIMAL
2640		4880
2641	If you need to shave off some kilobytes of code at the expense of some	4881	If you need to shave off some kilobytes of code at the expense of some
2642	speed, define this symbol to C<1>. Currently only used for gcc to override	4882	speed (but with the full API), you can define this symbol to request
2643	some inlining decisions, saves roughly 30% codesize of amd64.	4883	certain subsets of functionality. The default is to enable all features
		4884	that can be enabled on the platform.
		4885
		4886	A typical way to use this symbol is to define it to C<0> (or to a bitset
		4887	with some broad features you want) and then selectively re-enable
		4888	additional parts you want, for example if you want everything minimal,
		4889	but multiple event loop support, async and child watchers and the poll
		4890	backend, use this:
		4891
		4892	#define EV_FEATURES 0
		4893	#define EV_MULTIPLICITY 1
		4894	#define EV_USE_POLL 1
		4895	#define EV_CHILD_ENABLE 1
		4896	#define EV_ASYNC_ENABLE 1
		4897
		4898	The actual value is a bitset, it can be a combination of the following
		4899	values (by default, all of these are enabled):
		4900
		4901	=over 4
		4902
		4903	=item C<1> - faster/larger code
		4904
		4905	Use larger code to speed up some operations.
		4906
		4907	Currently this is used to override some inlining decisions (enlarging the
		4908	code size by roughly 30% on amd64).
		4909
		4910	When optimising for size, use of compiler flags such as C<-Os> with
		4911	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4912	assertions.
		4913
		4914	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4915	(e.g. gcc with C<-Os>).
		4916
		4917	=item C<2> - faster/larger data structures
		4918
		4919	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4920	hash table sizes and so on. This will usually further increase code size
		4921	and can additionally have an effect on the size of data structures at
		4922	runtime.
		4923
		4924	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4925	(e.g. gcc with C<-Os>).
		4926
		4927	=item C<4> - full API configuration
		4928
		4929	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4930	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4931
		4932	=item C<8> - full API
		4933
		4934	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4935	details on which parts of the API are still available without this
		4936	feature, and do not complain if this subset changes over time.
		4937
		4938	=item C<16> - enable all optional watcher types
		4939
		4940	Enables all optional watcher types. If you want to selectively enable
		4941	only some watcher types other than I/O and timers (e.g. prepare,
		4942	embed, async, child...) you can enable them manually by defining
		4943	C<EV_watchertype_ENABLE> to C<1> instead.
		4944
		4945	=item C<32> - enable all backends
		4946
		4947	This enables all backends - without this feature, you need to enable at
		4948	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4949
		4950	=item C<64> - enable OS-specific "helper" APIs
		4951
		4952	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4953	default.
		4954
		4955	=back
		4956
		4957	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4958	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4959	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4960	watchers, timers and monotonic clock support.
		4961
		4962	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4963	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4964	your program might be left out as well - a binary starting a timer and an
		4965	I/O watcher then might come out at only 5Kb.
		4966
		4967	=item EV_API_STATIC
		4968
		4969	If this symbol is defined (by default it is not), then all identifiers
		4970	will have static linkage. This means that libev will not export any
		4971	identifiers, and you cannot link against libev anymore. This can be useful
		4972	when you embed libev, only want to use libev functions in a single file,
		4973	and do not want its identifiers to be visible.
		4974
		4975	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4976	wants to use libev.
		4977
		4978	This option only works when libev is compiled with a C compiler, as C++
		4979	doesn't support the required declaration syntax.
		4980
		4981	=item EV_AVOID_STDIO
		4982
		4983	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4984	functions (printf, scanf, perror etc.). This will increase the code size
		4985	somewhat, but if your program doesn't otherwise depend on stdio and your
		4986	libc allows it, this avoids linking in the stdio library which is quite
		4987	big.
		4988
		4989	Note that error messages might become less precise when this option is
		4990	enabled.
		4991
		4992	=item EV_NSIG
		4993
		4994	The highest supported signal number, +1 (or, the number of
		4995	signals): Normally, libev tries to deduce the maximum number of signals
		4996	automatically, but sometimes this fails, in which case it can be
		4997	specified. Also, using a lower number than detected (C<32> should be
		4998	good for about any system in existence) can save some memory, as libev
		4999	statically allocates some 12-24 bytes per signal number.
2644		5000
2645	=item EV_PID_HASHSIZE	5001	=item EV_PID_HASHSIZE
2646		5002
2647	C<ev_child> watchers use a small hash table to distribute workload by	5003	C<ev_child> watchers use a small hash table to distribute workload by
2648	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	5004	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
2649	than enough. If you need to manage thousands of children you might want to	5005	usually more than enough. If you need to manage thousands of children you
2650	increase this value (I<must> be a power of two).	5006	might want to increase this value (I<must> be a power of two).
2651		5007
2652	=item EV_INOTIFY_HASHSIZE	5008	=item EV_INOTIFY_HASHSIZE
2653		5009
2654	C<ev_stat> watchers use a small hash table to distribute workload by	5010	C<ev_stat> watchers use a small hash table to distribute workload by
2655	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	5011	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
2656	usually more than enough. If you need to manage thousands of C<ev_stat>	5012	disabled), usually more than enough. If you need to manage thousands of
2657	watchers you might want to increase this value (I<must> be a power of	5013	C<ev_stat> watchers you might want to increase this value (I<must> be a
2658	two).	5014	power of two).
		5015
		5016	=item EV_USE_4HEAP
		5017
		5018	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		5019	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		5020	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		5021	faster performance with many (thousands) of watchers.
		5022
		5023	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		5024	will be C<0>.
		5025
		5026	=item EV_HEAP_CACHE_AT
		5027
		5028	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		5029	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		5030	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		5031	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		5032	but avoids random read accesses on heap changes. This improves performance
		5033	noticeably with many (hundreds) of watchers.
		5034
		5035	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		5036	will be C<0>.
		5037
		5038	=item EV_VERIFY
		5039
		5040	Controls how much internal verification (see C<ev_verify ()>) will
		5041	be done: If set to C<0>, no internal verification code will be compiled
		5042	in. If set to C<1>, then verification code will be compiled in, but not
		5043	called. If set to C<2>, then the internal verification code will be
		5044	called once per loop, which can slow down libev. If set to C<3>, then the
		5045	verification code will be called very frequently, which will slow down
		5046	libev considerably.
		5047
		5048	Verification errors are reported via C's C<assert> mechanism, so if you
		5049	disable that (e.g. by defining C<NDEBUG>) then no errors will be reported.
		5050
		5051	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		5052	will be C<0>.
2659		5053
2660	=item EV_COMMON	5054	=item EV_COMMON
2661		5055
2662	By default, all watchers have a C<void *data> member. By redefining	5056	By default, all watchers have a C<void *data> member. By redefining
2663	this macro to a something else you can include more and other types of	5057	this macro to something else you can include more and other types of
2664	members. You have to define it each time you include one of the files,	5058	members. You have to define it each time you include one of the files,
2665	though, and it must be identical each time.	5059	though, and it must be identical each time.
2666		5060
2667	For example, the perl EV module uses something like this:	5061	For example, the perl EV module uses something like this:
2668		5062
2669	#define EV_COMMON \	5063	#define EV_COMMON \
2670	SV *self; /* contains this struct */ \	5064	SV *self; /* contains this struct */ \
2671	SV *cb_sv, *fh /* note no trailing ";" */	5065	SV *cb_sv, *fh /* note no trailing ";" */
2672		5066
2673	=item EV_CB_DECLARE (type)	5067	=item EV_CB_DECLARE (type)
2674		5068
2675	=item EV_CB_INVOKE (watcher, revents)	5069	=item EV_CB_INVOKE (watcher, revents)
2676		5070
…		…
2681	definition and a statement, respectively. See the F<ev.h> header file for	5075	definition and a statement, respectively. See the F<ev.h> header file for
2682	their default definitions. One possible use for overriding these is to	5076	their default definitions. One possible use for overriding these is to
2683	avoid the C<struct ev_loop *> as first argument in all cases, or to use	5077	avoid the C<struct ev_loop *> as first argument in all cases, or to use
2684	method calls instead of plain function calls in C++.	5078	method calls instead of plain function calls in C++.
2685		5079
		5080	=back
		5081
2686	=head2 EXPORTED API SYMBOLS	5082	=head2 EXPORTED API SYMBOLS
2687		5083
2688	If you need to re-export the API (e.g. via a dll) and you need a list of	5084	If you need to re-export the API (e.g. via a DLL) and you need a list of
2689	exported symbols, you can use the provided F<Symbol.*> files which list	5085	exported symbols, you can use the provided F<Symbol.*> files which list
2690	all public symbols, one per line:	5086	all public symbols, one per line:
2691		5087
2692	Symbols.ev for libev proper	5088	Symbols.ev for libev proper
2693	Symbols.event for the libevent emulation	5089	Symbols.event for the libevent emulation
2694		5090
2695	This can also be used to rename all public symbols to avoid clashes with	5091	This can also be used to rename all public symbols to avoid clashes with
2696	multiple versions of libev linked together (which is obviously bad in	5092	multiple versions of libev linked together (which is obviously bad in
2697	itself, but sometimes it is inconvinient to avoid this).	5093	itself, but sometimes it is inconvenient to avoid this).
2698		5094
2699	A sed command like this will create wrapper C<#define>'s that you need to	5095	A sed command like this will create wrapper C<#define>'s that you need to
2700	include before including F<ev.h>:	5096	include before including F<ev.h>:
2701		5097
2702	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h	5098	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
…		…
2719	file.	5115	file.
2720		5116
2721	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	5117	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2722	that everybody includes and which overrides some configure choices:	5118	that everybody includes and which overrides some configure choices:
2723		5119
2724	#define EV_MINIMAL 1	5120	#define EV_FEATURES 8
2725	#define EV_USE_POLL 0	5121	#define EV_USE_SELECT 1
2726	#define EV_MULTIPLICITY 0
2727	#define EV_PERIODIC_ENABLE 0	5122	#define EV_PREPARE_ENABLE 1
		5123	#define EV_IDLE_ENABLE 1
2728	#define EV_STAT_ENABLE 0	5124	#define EV_SIGNAL_ENABLE 1
2729	#define EV_FORK_ENABLE 0	5125	#define EV_CHILD_ENABLE 1
		5126	#define EV_USE_STDEXCEPT 0
2730	#define EV_CONFIG_H <config.h>	5127	#define EV_CONFIG_H <config.h>
2731	#define EV_MINPRI 0
2732	#define EV_MAXPRI 0
2733		5128
2734	#include "ev++.h"	5129	#include "ev++.h"
2735		5130
2736	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	5131	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2737		5132
2738	#include "ev_cpp.h"	5133	#include "ev_cpp.h"
2739	#include "ev.c"	5134	#include "ev.c"
2740		5135
		5136	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
2741		5137
2742	=head1 COMPLEXITIES	5138	=head2 THREADS AND COROUTINES
2743		5139
2744	In this section the complexities of (many of) the algorithms used inside	5140	=head3 THREADS
2745	libev will be explained. For complexity discussions about backends see the
2746	documentation for C<ev_default_init>.
2747		5141
2748	All of the following are about amortised time: If an array needs to be	5142	All libev functions are reentrant and thread-safe unless explicitly
2749	extended, libev needs to realloc and move the whole array, but this	5143	documented otherwise, but libev implements no locking itself. This means
2750	happens asymptotically never with higher number of elements, so O(1) might	5144	that you can use as many loops as you want in parallel, as long as there
2751	mean it might do a lengthy realloc operation in rare cases, but on average	5145	are no concurrent calls into any libev function with the same loop
2752	it is much faster and asymptotically approaches constant time.	5146	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		5147	of course): libev guarantees that different event loops share no data
		5148	structures that need any locking.
		5149
		5150	Or to put it differently: calls with different loop parameters can be done
		5151	concurrently from multiple threads, calls with the same loop parameter
		5152	must be done serially (but can be done from different threads, as long as
		5153	only one thread ever is inside a call at any point in time, e.g. by using
		5154	a mutex per loop).
		5155
		5156	Specifically to support threads (and signal handlers), libev implements
		5157	so-called C<ev_async> watchers, which allow some limited form of
		5158	concurrency on the same event loop, namely waking it up "from the
		5159	outside".
		5160
		5161	If you want to know which design (one loop, locking, or multiple loops
		5162	without or something else still) is best for your problem, then I cannot
		5163	help you, but here is some generic advice:
2753		5164
2754	=over 4	5165	=over 4
2755		5166
2756	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	5167	=item * most applications have a main thread: use the default libev loop
		5168	in that thread, or create a separate thread running only the default loop.
2757		5169
2758	This means that, when you have a watcher that triggers in one hour and	5170	This helps integrating other libraries or software modules that use libev
2759	there are 100 watchers that would trigger before that then inserting will	5171	themselves and don't care/know about threading.
2760	have to skip roughly seven (C<ld 100>) of these watchers.
2761		5172
2762	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	5173	=item * one loop per thread is usually a good model.
2763		5174
2764	That means that changing a timer costs less than removing/adding them	5175	Doing this is almost never wrong, sometimes a better-performance model
2765	as only the relative motion in the event queue has to be paid for.	5176	exists, but it is always a good start.
2766		5177
2767	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	5178	=item * other models exist, such as the leader/follower pattern, where one
		5179	loop is handed through multiple threads in a kind of round-robin fashion.
2768		5180
2769	These just add the watcher into an array or at the head of a list.	5181	Choosing a model is hard - look around, learn, know that usually you can do
		5182	better than you currently do :-)
2770		5183
2771	=item Stopping check/prepare/idle watchers: O(1)	5184	=item * often you need to talk to some other thread which blocks in the
		5185	event loop.
2772		5186
2773	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	5187	C<ev_async> watchers can be used to wake them up from other threads safely
		5188	(or from signal contexts...).
2774		5189
2775	These watchers are stored in lists then need to be walked to find the	5190	An example use would be to communicate signals or other events that only
2776	correct watcher to remove. The lists are usually short (you don't usually	5191	work in the default loop by registering the signal watcher with the
2777	have many watchers waiting for the same fd or signal).	5192	default loop and triggering an C<ev_async> watcher from the default loop
2778		5193	watcher callback into the event loop interested in the signal.
2779	=item Finding the next timer in each loop iteration: O(1)
2780
2781	By virtue of using a binary heap, the next timer is always found at the
2782	beginning of the storage array.
2783
2784	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2785
2786	A change means an I/O watcher gets started or stopped, which requires
2787	libev to recalculate its status (and possibly tell the kernel, depending
2788	on backend and wether C<ev_io_set> was used).
2789
2790	=item Activating one watcher (putting it into the pending state): O(1)
2791
2792	=item Priority handling: O(number_of_priorities)
2793
2794	Priorities are implemented by allocating some space for each
2795	priority. When doing priority-based operations, libev usually has to
2796	linearly search all the priorities, but starting/stopping and activating
2797	watchers becomes O(1) w.r.t. prioritiy handling.
2798		5194
2799	=back	5195	=back
2800		5196
		5197	See also L</THREAD LOCKING EXAMPLE>.
2801		5198
2802	=head1 Win32 platform limitations and workarounds	5199	=head3 COROUTINES
		5200
		5201	Libev is very accommodating to coroutines ("cooperative threads"):
		5202	libev fully supports nesting calls to its functions from different
		5203	coroutines (e.g. you can call C<ev_run> on the same loop from two
		5204	different coroutines, and switch freely between both coroutines running
		5205	the loop, as long as you don't confuse yourself). The only exception is
		5206	that you must not do this from C<ev_periodic> reschedule callbacks.
		5207
		5208	Care has been taken to ensure that libev does not keep local state inside
		5209	C<ev_run>, and other calls do not usually allow for coroutine switches as
		5210	they do not call any callbacks.
		5211
		5212	=head2 COMPILER WARNINGS
		5213
		5214	Depending on your compiler and compiler settings, you might get no or a
		5215	lot of warnings when compiling libev code. Some people are apparently
		5216	scared by this.
		5217
		5218	However, these are unavoidable for many reasons. For one, each compiler
		5219	has different warnings, and each user has different tastes regarding
		5220	warning options. "Warn-free" code therefore cannot be a goal except when
		5221	targeting a specific compiler and compiler-version.
		5222
		5223	Another reason is that some compiler warnings require elaborate
		5224	workarounds, or other changes to the code that make it less clear and less
		5225	maintainable.
		5226
		5227	And of course, some compiler warnings are just plain stupid, or simply
		5228	wrong (because they don't actually warn about the condition their message
		5229	seems to warn about). For example, certain older gcc versions had some
		5230	warnings that resulted in an extreme number of false positives. These have
		5231	been fixed, but some people still insist on making code warn-free with
		5232	such buggy versions.
		5233
		5234	While libev is written to generate as few warnings as possible,
		5235	"warn-free" code is not a goal, and it is recommended not to build libev
		5236	with any compiler warnings enabled unless you are prepared to cope with
		5237	them (e.g. by ignoring them). Remember that warnings are just that:
		5238	warnings, not errors, or proof of bugs.
		5239
		5240
		5241	=head2 VALGRIND
		5242
		5243	Valgrind has a special section here because it is a popular tool that is
		5244	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		5245
		5246	If you think you found a bug (memory leak, uninitialised data access etc.)
		5247	in libev, then check twice: If valgrind reports something like:
		5248
		5249	==2274== definitely lost: 0 bytes in 0 blocks.
		5250	==2274== possibly lost: 0 bytes in 0 blocks.
		5251	==2274== still reachable: 256 bytes in 1 blocks.
		5252
		5253	Then there is no memory leak, just as memory accounted to global variables
		5254	is not a memleak - the memory is still being referenced, and didn't leak.
		5255
		5256	Similarly, under some circumstances, valgrind might report kernel bugs
		5257	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		5258	although an acceptable workaround has been found here), or it might be
		5259	confused.
		5260
		5261	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		5262	make it into some kind of religion.
		5263
		5264	If you are unsure about something, feel free to contact the mailing list
		5265	with the full valgrind report and an explanation on why you think this
		5266	is a bug in libev (best check the archives, too :). However, don't be
		5267	annoyed when you get a brisk "this is no bug" answer and take the chance
		5268	of learning how to interpret valgrind properly.
		5269
		5270	If you need, for some reason, empty reports from valgrind for your project
		5271	I suggest using suppression lists.
		5272
		5273
		5274	=head1 PORTABILITY NOTES
		5275
		5276	=head2 GNU/LINUX 32 BIT LIMITATIONS
		5277
		5278	GNU/Linux is the only common platform that supports 64 bit file/large file
		5279	interfaces but I<disables> them by default.
		5280
		5281	That means that libev compiled in the default environment doesn't support
		5282	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		5283
		5284	Unfortunately, many programs try to work around this GNU/Linux issue
		5285	by enabling the large file API, which makes them incompatible with the
		5286	standard libev compiled for their system.
		5287
		5288	Likewise, libev cannot enable the large file API itself as this would
		5289	suddenly make it incompatible to the default compile time environment,
		5290	i.e. all programs not using special compile switches.
		5291
		5292	=head2 OS/X AND DARWIN BUGS
		5293
		5294	The whole thing is a bug if you ask me - basically any system interface
		5295	you touch is broken, whether it is locales, poll, kqueue or even the
		5296	OpenGL drivers.
		5297
		5298	=head3 C<kqueue> is buggy
		5299
		5300	The kqueue syscall is broken in all known versions - most versions support
		5301	only sockets, many support pipes.
		5302
		5303	Libev tries to work around this by not using C<kqueue> by default on this
		5304	rotten platform, but of course you can still ask for it when creating a
		5305	loop - embedding a socket-only kqueue loop into a select-based one is
		5306	probably going to work well.
		5307
		5308	=head3 C<poll> is buggy
		5309
		5310	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		5311	implementation by something calling C<kqueue> internally around the 10.5.6
		5312	release, so now C<kqueue> I<and> C<poll> are broken.
		5313
		5314	Libev tries to work around this by not using C<poll> by default on
		5315	this rotten platform, but of course you can still ask for it when creating
		5316	a loop.
		5317
		5318	=head3 C<select> is buggy
		5319
		5320	All that's left is C<select>, and of course Apple found a way to fuck this
		5321	one up as well: On OS/X, C<select> actively limits the number of file
		5322	descriptors you can pass in to 1024 - your program suddenly crashes when
		5323	you use more.
		5324
		5325	There is an undocumented "workaround" for this - defining
		5326	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		5327	work on OS/X.
		5328
		5329	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		5330
		5331	=head3 C<errno> reentrancy
		5332
		5333	The default compile environment on Solaris is unfortunately so
		5334	thread-unsafe that you can't even use components/libraries compiled
		5335	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		5336	defined by default. A valid, if stupid, implementation choice.
		5337
		5338	If you want to use libev in threaded environments you have to make sure
		5339	it's compiled with C<_REENTRANT> defined.
		5340
		5341	=head3 Event port backend
		5342
		5343	The scalable event interface for Solaris is called "event
		5344	ports". Unfortunately, this mechanism is very buggy in all major
		5345	releases. If you run into high CPU usage, your program freezes or you get
		5346	a large number of spurious wakeups, make sure you have all the relevant
		5347	and latest kernel patches applied. No, I don't know which ones, but there
		5348	are multiple ones to apply, and afterwards, event ports actually work
		5349	great.
		5350
		5351	If you can't get it to work, you can try running the program by setting
		5352	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		5353	C<select> backends.
		5354
		5355	=head2 AIX POLL BUG
		5356
		5357	AIX unfortunately has a broken C<poll.h> header. Libev works around
		5358	this by trying to avoid the poll backend altogether (i.e. it's not even
		5359	compiled in), which normally isn't a big problem as C<select> works fine
		5360	with large bitsets on AIX, and AIX is dead anyway.
		5361
		5362	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		5363
		5364	=head3 General issues
2803		5365
2804	Win32 doesn't support any of the standards (e.g. POSIX) that libev	5366	Win32 doesn't support any of the standards (e.g. POSIX) that libev
2805	requires, and its I/O model is fundamentally incompatible with the POSIX	5367	requires, and its I/O model is fundamentally incompatible with the POSIX
2806	model. Libev still offers limited functionality on this platform in	5368	model. Libev still offers limited functionality on this platform in
2807	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5369	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
2808	descriptors. This only applies when using Win32 natively, not when using	5370	descriptors. This only applies when using Win32 natively, not when using
2809	e.g. cygwin.	5371	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		5372	as every compiler comes with a slightly differently broken/incompatible
		5373	environment.
		5374
		5375	Lifting these limitations would basically require the full
		5376	re-implementation of the I/O system. If you are into this kind of thing,
		5377	then note that glib does exactly that for you in a very portable way (note
		5378	also that glib is the slowest event library known to man).
2810		5379
2811	There is no supported compilation method available on windows except	5380	There is no supported compilation method available on windows except
2812	embedding it into other applications.	5381	embedding it into other applications.
2813		5382
		5383	Sensible signal handling is officially unsupported by Microsoft - libev
		5384	tries its best, but under most conditions, signals will simply not work.
		5385
		5386	Not a libev limitation but worth mentioning: windows apparently doesn't
		5387	accept large writes: instead of resulting in a partial write, windows will
		5388	either accept everything or return C<ENOBUFS> if the buffer is too large,
		5389	so make sure you only write small amounts into your sockets (less than a
		5390	megabyte seems safe, but this apparently depends on the amount of memory
		5391	available).
		5392
2814	Due to the many, low, and arbitrary limits on the win32 platform and the	5393	Due to the many, low, and arbitrary limits on the win32 platform and
2815	abysmal performance of winsockets, using a large number of sockets is not	5394	the abysmal performance of winsockets, using a large number of sockets
2816	recommended (and not reasonable). If your program needs to use more than	5395	is not recommended (and not reasonable). If your program needs to use
2817	a hundred or so sockets, then likely it needs to use a totally different	5396	more than a hundred or so sockets, then likely it needs to use a totally
2818	implementation for windows, as libev offers the POSIX model, which cannot	5397	different implementation for windows, as libev offers the POSIX readiness
2819	be implemented efficiently on windows (microsoft monopoly games).	5398	notification model, which cannot be implemented efficiently on windows
		5399	(due to Microsoft monopoly games).
2820		5400
2821	=over 4	5401	A typical way to use libev under windows is to embed it (see the embedding
		5402	section for details) and use the following F<evwrap.h> header file instead
		5403	of F<ev.h>:
2822		5404
		5405	#define EV_STANDALONE /* keeps ev from requiring config.h */
		5406	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		5407
		5408	#include "ev.h"
		5409
		5410	And compile the following F<evwrap.c> file into your project (make sure
		5411	you do I<not> compile the F<ev.c> or any other embedded source files!):
		5412
		5413	#include "evwrap.h"
		5414	#include "ev.c"
		5415
2823	=item The winsocket select function	5416	=head3 The winsocket C<select> function
2824		5417
2825	The winsocket C<select> function doesn't follow POSIX in that it requires	5418	The winsocket C<select> function doesn't follow POSIX in that it
2826	socket I<handles> and not socket I<file descriptors>. This makes select	5419	requires socket I<handles> and not socket I<file descriptors> (it is
2827	very inefficient, and also requires a mapping from file descriptors	5420	also extremely buggy). This makes select very inefficient, and also
2828	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,	5421	requires a mapping from file descriptors to socket handles (the Microsoft
2829	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor	5422	C runtime provides the function C<_open_osfhandle> for this). See the
2830	symbols for more info.	5423	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		5424	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
2831		5425
2832	The configuration for a "naked" win32 using the microsoft runtime	5426	The configuration for a "naked" win32 using the Microsoft runtime
2833	libraries and raw winsocket select is:	5427	libraries and raw winsocket select is:
2834		5428
2835	#define EV_USE_SELECT 1	5429	#define EV_USE_SELECT 1
2836	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	5430	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
2837		5431
2838	Note that winsockets handling of fd sets is O(n), so you can easily get a	5432	Note that winsockets handling of fd sets is O(n), so you can easily get a
2839	complexity in the O(n²) range when using win32.	5433	complexity in the O(n²) range when using win32.
2840		5434
2841	=item Limited number of file descriptors	5435	=head3 Limited number of file descriptors
2842		5436
2843	Windows has numerous arbitrary (and low) limits on things. Early versions	5437	Windows has numerous arbitrary (and low) limits on things.
2844	of winsocket's select only supported waiting for a max. of C<64> handles	5438
		5439	Early versions of winsocket's select only supported waiting for a maximum
2845	(probably owning to the fact that all windows kernels can only wait for	5440	of C<64> handles (probably owning to the fact that all windows kernels
2846	C<64> things at the same time internally; microsoft recommends spawning a	5441	can only wait for C<64> things at the same time internally; Microsoft
2847	chain of threads and wait for 63 handles and the previous thread in each).	5442	recommends spawning a chain of threads and wait for 63 handles and the
		5443	previous thread in each. Sounds great!).
2848		5444
2849	Newer versions support more handles, but you need to define C<FD_SETSIZE>	5445	Newer versions support more handles, but you need to define C<FD_SETSIZE>
2850	to some high number (e.g. C<2048>) before compiling the winsocket select	5446	to some high number (e.g. C<2048>) before compiling the winsocket select
2851	call (which might be in libev or elsewhere, for example, perl does its own	5447	call (which might be in libev or elsewhere, for example, perl and many
2852	select emulation on windows).	5448	other interpreters do their own select emulation on windows).
2853		5449
2854	Another limit is the number of file descriptors in the microsoft runtime	5450	Another limit is the number of file descriptors in the Microsoft runtime
2855	libraries, which by default is C<64> (there must be a hidden I<64> fetish	5451	libraries, which by default is C<64> (there must be a hidden I<64>
2856	or something like this inside microsoft). You can increase this by calling	5452	fetish or something like this inside Microsoft). You can increase this
2857	C<_setmaxstdio>, which can increase this limit to C<2048> (another	5453	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
2858	arbitrary limit), but is broken in many versions of the microsoft runtime	5454	(another arbitrary limit), but is broken in many versions of the Microsoft
2859	libraries.
2860
2861	This might get you to about C<512> or C<2048> sockets (depending on	5455	runtime libraries. This might get you to about C<512> or C<2048> sockets
2862	windows version and/or the phase of the moon). To get more, you need to	5456	(depending on windows version and/or the phase of the moon). To get more,
2863	wrap all I/O functions and provide your own fd management, but the cost of	5457	you need to wrap all I/O functions and provide your own fd management, but
2864	calling select (O(n²)) will likely make this unworkable.	5458	the cost of calling select (O(n²)) will likely make this unworkable.
		5459
		5460	=head2 PORTABILITY REQUIREMENTS
		5461
		5462	In addition to a working ISO-C implementation and of course the
		5463	backend-specific APIs, libev relies on a few additional extensions:
		5464
		5465	=over 4
		5466
		5467	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		5468	calling conventions regardless of C<ev_watcher_type *>.
		5469
		5470	Libev assumes not only that all watcher pointers have the same internal
		5471	structure (guaranteed by POSIX but not by ISO C for example), but it also
		5472	assumes that the same (machine) code can be used to call any watcher
		5473	callback: The watcher callbacks have different type signatures, but libev
		5474	calls them using an C<ev_watcher *> internally.
		5475
		5476	=item null pointers and integer zero are represented by 0 bytes
		5477
		5478	Libev uses C<memset> to initialise structs and arrays to C<0> bytes, and
		5479	relies on this setting pointers and integers to null.
		5480
		5481	=item pointer accesses must be thread-atomic
		5482
		5483	Accessing a pointer value must be atomic, it must both be readable and
		5484	writable in one piece - this is the case on all current architectures.
		5485
		5486	=item C<sig_atomic_t volatile> must be thread-atomic as well
		5487
		5488	The type C<sig_atomic_t volatile> (or whatever is defined as
		5489	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		5490	threads. This is not part of the specification for C<sig_atomic_t>, but is
		5491	believed to be sufficiently portable.
		5492
		5493	=item C<sigprocmask> must work in a threaded environment
		5494
		5495	Libev uses C<sigprocmask> to temporarily block signals. This is not
		5496	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		5497	pthread implementations will either allow C<sigprocmask> in the "main
		5498	thread" or will block signals process-wide, both behaviours would
		5499	be compatible with libev. Interaction between C<sigprocmask> and
		5500	C<pthread_sigmask> could complicate things, however.
		5501
		5502	The most portable way to handle signals is to block signals in all threads
		5503	except the initial one, and run the signal handling loop in the initial
		5504	thread as well.
		5505
		5506	=item C<long> must be large enough for common memory allocation sizes
		5507
		5508	To improve portability and simplify its API, libev uses C<long> internally
		5509	instead of C<size_t> when allocating its data structures. On non-POSIX
		5510	systems (Microsoft...) this might be unexpectedly low, but is still at
		5511	least 31 bits everywhere, which is enough for hundreds of millions of
		5512	watchers.
		5513
		5514	=item C<double> must hold a time value in seconds with enough accuracy
		5515
		5516	The type C<double> is used to represent timestamps. It is required to
		5517	have at least 51 bits of mantissa (and 9 bits of exponent), which is
		5518	good enough for at least into the year 4000 with millisecond accuracy
		5519	(the design goal for libev). This requirement is overfulfilled by
		5520	implementations using IEEE 754, which is basically all existing ones.
		5521
		5522	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5523	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5524	is either obsolete or somebody patched it to use C<long double> or
		5525	something like that, just kidding).
2865		5526
2866	=back	5527	=back
2867		5528
		5529	If you know of other additional requirements drop me a note.
		5530
		5531
		5532	=head1 ALGORITHMIC COMPLEXITIES
		5533
		5534	In this section the complexities of (many of) the algorithms used inside
		5535	libev will be documented. For complexity discussions about backends see
		5536	the documentation for C<ev_default_init>.
		5537
		5538	All of the following are about amortised time: If an array needs to be
		5539	extended, libev needs to realloc and move the whole array, but this
		5540	happens asymptotically rarer with higher number of elements, so O(1) might
		5541	mean that libev does a lengthy realloc operation in rare cases, but on
		5542	average it is much faster and asymptotically approaches constant time.
		5543
		5544	=over 4
		5545
		5546	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		5547
		5548	This means that, when you have a watcher that triggers in one hour and
		5549	there are 100 watchers that would trigger before that, then inserting will
		5550	have to skip roughly seven (C<ld 100>) of these watchers.
		5551
		5552	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		5553
		5554	That means that changing a timer costs less than removing/adding them,
		5555	as only the relative motion in the event queue has to be paid for.
		5556
		5557	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		5558
		5559	These just add the watcher into an array or at the head of a list.
		5560
		5561	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		5562
		5563	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		5564
		5565	These watchers are stored in lists, so they need to be walked to find the
		5566	correct watcher to remove. The lists are usually short (you don't usually
		5567	have many watchers waiting for the same fd or signal: one is typical, two
		5568	is rare).
		5569
		5570	=item Finding the next timer in each loop iteration: O(1)
		5571
		5572	By virtue of using a binary or 4-heap, the next timer is always found at a
		5573	fixed position in the storage array.
		5574
		5575	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		5576
		5577	A change means an I/O watcher gets started or stopped, which requires
		5578	libev to recalculate its status (and possibly tell the kernel, depending
		5579	on backend and whether C<ev_io_set> was used).
		5580
		5581	=item Activating one watcher (putting it into the pending state): O(1)
		5582
		5583	=item Priority handling: O(number_of_priorities)
		5584
		5585	Priorities are implemented by allocating some space for each
		5586	priority. When doing priority-based operations, libev usually has to
		5587	linearly search all the priorities, but starting/stopping and activating
		5588	watchers becomes O(1) with respect to priority handling.
		5589
		5590	=item Sending an ev_async: O(1)
		5591
		5592	=item Processing ev_async_send: O(number_of_async_watchers)
		5593
		5594	=item Processing signals: O(max_signal_number)
		5595
		5596	Sending involves a system call I<iff> there were no other C<ev_async_send>
		5597	calls in the current loop iteration and the loop is currently
		5598	blocked. Checking for async and signal events involves iterating over all
		5599	running async watchers or all signal numbers.
		5600
		5601	=back
		5602
		5603
		5604	=head1 PORTING FROM LIBEV 3.X TO 4.X
		5605
		5606	The major version 4 introduced some incompatible changes to the API.
		5607
		5608	At the moment, the C<ev.h> header file provides compatibility definitions
		5609	for all changes, so most programs should still compile. The compatibility
		5610	layer might be removed in later versions of libev, so better update to the
		5611	new API early than late.
		5612
		5613	=over 4
		5614
		5615	=item C<EV_COMPAT3> backwards compatibility mechanism
		5616
		5617	The backward compatibility mechanism can be controlled by
		5618	C<EV_COMPAT3>. See L</"PREPROCESSOR SYMBOLS/MACROS"> in the L</EMBEDDING>
		5619	section.
		5620
		5621	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5622
		5623	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5624
		5625	ev_loop_destroy (EV_DEFAULT_UC);
		5626	ev_loop_fork (EV_DEFAULT);
		5627
		5628	=item function/symbol renames
		5629
		5630	A number of functions and symbols have been renamed:
		5631
		5632	ev_loop => ev_run
		5633	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5634	EVLOOP_ONESHOT => EVRUN_ONCE
		5635
		5636	ev_unloop => ev_break
		5637	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5638	EVUNLOOP_ONE => EVBREAK_ONE
		5639	EVUNLOOP_ALL => EVBREAK_ALL
		5640
		5641	EV_TIMEOUT => EV_TIMER
		5642
		5643	ev_loop_count => ev_iteration
		5644	ev_loop_depth => ev_depth
		5645	ev_loop_verify => ev_verify
		5646
		5647	Most functions working on C<struct ev_loop> objects don't have an
		5648	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5649	associated constants have been renamed to not collide with the C<struct
		5650	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5651	as all other watcher types. Note that C<ev_loop_fork> is still called
		5652	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		5653	typedef.
		5654
		5655	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		5656
		5657	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		5658	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		5659	and work, but the library code will of course be larger.
		5660
		5661	=back
		5662
		5663
		5664	=head1 GLOSSARY
		5665
		5666	=over 4
		5667
		5668	=item active
		5669
		5670	A watcher is active as long as it has been started and not yet stopped.
		5671	See L</WATCHER STATES> for details.
		5672
		5673	=item application
		5674
		5675	In this document, an application is whatever is using libev.
		5676
		5677	=item backend
		5678
		5679	The part of the code dealing with the operating system interfaces.
		5680
		5681	=item callback
		5682
		5683	The address of a function that is called when some event has been
		5684	detected. Callbacks are being passed the event loop, the watcher that
		5685	received the event, and the actual event bitset.
		5686
		5687	=item callback/watcher invocation
		5688
		5689	The act of calling the callback associated with a watcher.
		5690
		5691	=item event
		5692
		5693	A change of state of some external event, such as data now being available
		5694	for reading on a file descriptor, time having passed or simply not having
		5695	any other events happening anymore.
		5696
		5697	In libev, events are represented as single bits (such as C<EV_READ> or
		5698	C<EV_TIMER>).
		5699
		5700	=item event library
		5701
		5702	A software package implementing an event model and loop.
		5703
		5704	=item event loop
		5705
		5706	An entity that handles and processes external events and converts them
		5707	into callback invocations.
		5708
		5709	=item event model
		5710
		5711	The model used to describe how an event loop handles and processes
		5712	watchers and events.
		5713
		5714	=item pending
		5715
		5716	A watcher is pending as soon as the corresponding event has been
		5717	detected. See L</WATCHER STATES> for details.
		5718
		5719	=item real time
		5720
		5721	The physical time that is observed. It is apparently strictly monotonic :)
		5722
		5723	=item wall-clock time
		5724
		5725	The time and date as shown on clocks. Unlike real time, it can actually
		5726	be wrong and jump forwards and backwards, e.g. when you adjust your
		5727	clock.
		5728
		5729	=item watcher
		5730
		5731	A data structure that describes interest in certain events. Watchers need
		5732	to be started (attached to an event loop) before they can receive events.
		5733
		5734	=back
2868		5735
2869	=head1 AUTHOR	5736	=head1 AUTHOR
2870		5737
2871	Marc Lehmann <libev@schmorp.de>.	5738	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5739	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
2872		5740

Diff Legend

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