ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.157 by root, Tue May 20 23:49:41 2008 UTC vs.
Revision 1.344 by root, Wed Nov 10 13:41:48 2010 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	// a single header file is required	11	// a single header file is required
12	#include <ev.h>	12	#include <ev.h>
13		13
		14	#include <stdio.h> // for puts
		15
14	// every watcher type has its own typedef'd struct	16	// every watcher type has its own typedef'd struct
15	// with the name ev_<type>	17	// with the name ev_TYPE
16	ev_io stdin_watcher;	18	ev_io stdin_watcher;
17	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
18		20
19	// all watcher callbacks have a similar signature	21	// all watcher callbacks have a similar signature
20	// this callback is called when data is readable on stdin	22	// this callback is called when data is readable on stdin
21	static void	23	static void
22	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
23	{	25	{
24	puts ("stdin ready");	26	puts ("stdin ready");
25	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
26	// with its corresponding stop function.	28	// with its corresponding stop function.
27	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
28		30
29	// this causes all nested ev_loop's to stop iterating	31	// this causes all nested ev_run's to stop iterating
30	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_break (EV_A_ EVBREAK_ALL);
31	}	33	}
32		34
33	// another callback, this time for a time-out	35	// another callback, this time for a time-out
34	static void	36	static void
35	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
36	{	38	{
37	puts ("timeout");	39	puts ("timeout");
38	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_run to stop iterating
39	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_break (EV_A_ EVBREAK_ONE);
40	}	42	}
41		43
42	int	44	int
43	main (void)	45	main (void)
44	{	46	{
45	// use the default event loop unless you have special needs	47	// use the default event loop unless you have special needs
46	struct ev_loop *loop = ev_default_loop (0);	48	struct ev_loop *loop = EV_DEFAULT;
47		49
48	// initialise an io watcher, then start it	50	// initialise an io watcher, then start it
49	// this one will watch for stdin to become readable	51	// this one will watch for stdin to become readable
50	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);	52	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
51	ev_io_start (loop, &stdin_watcher);	53	ev_io_start (loop, &stdin_watcher);
52		54
53	// initialise a timer watcher, then start it	55	// initialise a timer watcher, then start it
54	// simple non-repeating 5.5 second timeout	56	// simple non-repeating 5.5 second timeout
55	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);	57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
56	ev_timer_start (loop, &timeout_watcher);	58	ev_timer_start (loop, &timeout_watcher);
57		59
58	// now wait for events to arrive	60	// now wait for events to arrive
59	ev_loop (loop, 0);	61	ev_run (loop, 0);
60		62
61	// unloop was called, so exit	63	// unloop was called, so exit
62	return 0;	64	return 0;
63	}	65	}
64		66
65	=head1 DESCRIPTION	67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
66		70
67	The newest version of this document is also available as an html-formatted	71	The newest version of this document is also available as an html-formatted
68	web page you might find easier to navigate when reading it for the first	72	web page you might find easier to navigate when reading it for the first
69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.	73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familiarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 WHAT TO READ WHEN IN A HURRY
		84
		85	This manual tries to be very detailed, but unfortunately, this also makes
		86	it very long. If you just want to know the basics of libev, I suggest
		87	reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and
		88	look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and
		89	C<ev_timer> sections in L<WATCHER TYPES>.
		90
		91	=head1 ABOUT LIBEV
70		92
71	Libev is an event loop: you register interest in certain events (such as a	93	Libev is an event loop: you register interest in certain events (such as a
72	file descriptor being readable or a timeout occurring), and it will manage	94	file descriptor being readable or a timeout occurring), and it will manage
73	these event sources and provide your program with events.	95	these event sources and provide your program with events.
74		96
…		…
84	=head2 FEATURES	106	=head2 FEATURES
85		107
86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	108	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	109	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	110	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	111	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
90	with customised rescheduling (C<ev_periodic>), synchronous signals	112	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
91	(C<ev_signal>), process status change events (C<ev_child>), and event	113	timers (C<ev_timer>), absolute timers with customised rescheduling
92	watchers dealing with the event loop mechanism itself (C<ev_idle>,	114	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as	115	change events (C<ev_child>), and event watchers dealing with the event
94	file watchers (C<ev_stat>) and even limited support for fork events	116	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
95	(C<ev_fork>).	117	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		118	limited support for fork events (C<ev_fork>).
96		119
97	It also is quite fast (see this	120	It also is quite fast (see this
98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	121	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
99	for example).	122	for example).
100		123
…		…
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	131	name C<loop> (which is always of type C<struct ev_loop *>) will not have
109	this argument.	132	this argument.
110		133
111	=head2 TIME REPRESENTATION	134	=head2 TIME REPRESENTATION
112		135
113	Libev represents time as a single floating point number, representing the	136	Libev represents time as a single floating point number, representing
114	(fractional) number of seconds since the (POSIX) epoch (somewhere near	137	the (fractional) number of seconds since the (POSIX) epoch (in practice
115	the beginning of 1970, details are complicated, don't ask). This type is	138	somewhere near the beginning of 1970, details are complicated, don't
116	called C<ev_tstamp>, which is what you should use too. It usually aliases	139	ask). This type is called C<ev_tstamp>, which is what you should use
117	to the C<double> type in C, and when you need to do any calculations on	140	too. It usually aliases to the C<double> type in C. When you need to do
118	it, you should treat it as some floatingpoint value. Unlike the name	141	any calculations on it, you should treat it as some floating point value.
		142
119	component C<stamp> might indicate, it is also used for time differences	143	Unlike the name component C<stamp> might indicate, it is also used for
120	throughout libev.	144	time differences (e.g. delays) throughout libev.
		145
		146	=head1 ERROR HANDLING
		147
		148	Libev knows three classes of errors: operating system errors, usage errors
		149	and internal errors (bugs).
		150
		151	When libev catches an operating system error it cannot handle (for example
		152	a system call indicating a condition libev cannot fix), it calls the callback
		153	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		154	abort. The default is to print a diagnostic message and to call C<abort
		155	()>.
		156
		157	When libev detects a usage error such as a negative timer interval, then
		158	it will print a diagnostic message and abort (via the C<assert> mechanism,
		159	so C<NDEBUG> will disable this checking): these are programming errors in
		160	the libev caller and need to be fixed there.
		161
		162	Libev also has a few internal error-checking C<assert>ions, and also has
		163	extensive consistency checking code. These do not trigger under normal
		164	circumstances, as they indicate either a bug in libev or worse.
		165
121		166
122	=head1 GLOBAL FUNCTIONS	167	=head1 GLOBAL FUNCTIONS
123		168
124	These functions can be called anytime, even before initialising the	169	These functions can be called anytime, even before initialising the
125	library in any way.	170	library in any way.
…		…
128		173
129	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
130		175
131	Returns the current time as libev would use it. Please note that the	176	Returns the current time as libev would use it. Please note that the
132	C<ev_now> function is usually faster and also often returns the timestamp	177	C<ev_now> function is usually faster and also often returns the timestamp
133	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_update_now> and C<ev_now>.
134		180
135	=item ev_sleep (ev_tstamp interval)	181	=item ev_sleep (ev_tstamp interval)
136		182
137	Sleep for the given interval: The current thread will be blocked until	183	Sleep for the given interval: The current thread will be blocked until
138	either it is interrupted or the given time interval has passed. Basically	184	either it is interrupted or the given time interval has passed. Basically
139	this is a subsecond-resolution C<sleep ()>.	185	this is a sub-second-resolution C<sleep ()>.
140		186
141	=item int ev_version_major ()	187	=item int ev_version_major ()
142		188
143	=item int ev_version_minor ()	189	=item int ev_version_minor ()
144		190
…		…
155	as this indicates an incompatible change. Minor versions are usually	201	as this indicates an incompatible change. Minor versions are usually
156	compatible to older versions, so a larger minor version alone is usually	202	compatible to older versions, so a larger minor version alone is usually
157	not a problem.	203	not a problem.
158		204
159	Example: Make sure we haven't accidentally been linked against the wrong	205	Example: Make sure we haven't accidentally been linked against the wrong
160	version.	206	version (note, however, that this will not detect other ABI mismatches,
		207	such as LFS or reentrancy).
161		208
162	assert (("libev version mismatch",	209	assert (("libev version mismatch",
163	ev_version_major () == EV_VERSION_MAJOR	210	ev_version_major () == EV_VERSION_MAJOR
164	&& ev_version_minor () >= EV_VERSION_MINOR));	211	&& ev_version_minor () >= EV_VERSION_MINOR));
165		212
166	=item unsigned int ev_supported_backends ()	213	=item unsigned int ev_supported_backends ()
167		214
168	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	215	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
169	value) compiled into this binary of libev (independent of their	216	value) compiled into this binary of libev (independent of their
…		…
171	a description of the set values.	218	a description of the set values.
172		219
173	Example: make sure we have the epoll method, because yeah this is cool and	220	Example: make sure we have the epoll method, because yeah this is cool and
174	a must have and can we have a torrent of it please!!!11	221	a must have and can we have a torrent of it please!!!11
175		222
176	assert (("sorry, no epoll, no sex",	223	assert (("sorry, no epoll, no sex",
177	ev_supported_backends () & EVBACKEND_EPOLL));	224	ev_supported_backends () & EVBACKEND_EPOLL));
178		225
179	=item unsigned int ev_recommended_backends ()	226	=item unsigned int ev_recommended_backends ()
180		227
181	Return the set of all backends compiled into this binary of libev and also	228	Return the set of all backends compiled into this binary of libev and
182	recommended for this platform. This set is often smaller than the one	229	also recommended for this platform, meaning it will work for most file
		230	descriptor types. This set is often smaller than the one returned by
183	returned by C<ev_supported_backends>, as for example kqueue is broken on	231	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
184	most BSDs and will not be autodetected unless you explicitly request it	232	and will not be auto-detected unless you explicitly request it (assuming
185	(assuming you know what you are doing). This is the set of backends that	233	you know what you are doing). This is the set of backends that libev will
186	libev will probe for if you specify no backends explicitly.	234	probe for if you specify no backends explicitly.
187		235
188	=item unsigned int ev_embeddable_backends ()	236	=item unsigned int ev_embeddable_backends ()
189		237
190	Returns the set of backends that are embeddable in other event loops. This	238	Returns the set of backends that are embeddable in other event loops. This
191	is the theoretical, all-platform, value. To find which backends	239	value is platform-specific but can include backends not available on the
192	might be supported on the current system, you would need to look at	240	current system. To find which embeddable backends might be supported on
193	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	241	the current system, you would need to look at C<ev_embeddable_backends ()
194	recommended ones.	242	& ev_supported_backends ()>, likewise for recommended ones.
195		243
196	See the description of C<ev_embed> watchers for more info.	244	See the description of C<ev_embed> watchers for more info.
197		245
198	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	246	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
199		247
…		…
229	}	277	}
230		278
231	...	279	...
232	ev_set_allocator (persistent_realloc);	280	ev_set_allocator (persistent_realloc);
233		281
234	=item ev_set_syserr_cb (void (*cb)(const char *msg));	282	=item ev_set_syserr_cb (void (*cb)(const char *msg))
235		283
236	Set the callback function to call on a retryable syscall error (such	284	Set the callback function to call on a retryable system call error (such
237	as failed select, poll, epoll_wait). The message is a printable string	285	as failed select, poll, epoll_wait). The message is a printable string
238	indicating the system call or subsystem causing the problem. If this	286	indicating the system call or subsystem causing the problem. If this
239	callback is set, then libev will expect it to remedy the sitution, no	287	callback is set, then libev will expect it to remedy the situation, no
240	matter what, when it returns. That is, libev will generally retry the	288	matter what, when it returns. That is, libev will generally retry the
241	requested operation, or, if the condition doesn't go away, do bad stuff	289	requested operation, or, if the condition doesn't go away, do bad stuff
242	(such as abort).	290	(such as abort).
243		291
244	Example: This is basically the same thing that libev does internally, too.	292	Example: This is basically the same thing that libev does internally, too.
…		…
253	...	301	...
254	ev_set_syserr_cb (fatal_error);	302	ev_set_syserr_cb (fatal_error);
255		303
256	=back	304	=back
257		305
258	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	306	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
259		307
260	An event loop is described by a C<struct ev_loop *>. The library knows two	308	An event loop is described by a C<struct ev_loop *> (the C<struct> is
261	types of such loops, the I<default> loop, which supports signals and child	309	I<not> optional in this case unless libev 3 compatibility is disabled, as
262	events, and dynamically created loops which do not.	310	libev 3 had an C<ev_loop> function colliding with the struct name).
		311
		312	The library knows two types of such loops, the I<default> loop, which
		313	supports child process events, and dynamically created event loops which
		314	do not.
263		315
264	=over 4	316	=over 4
265		317
266	=item struct ev_loop *ev_default_loop (unsigned int flags)	318	=item struct ev_loop *ev_default_loop (unsigned int flags)
267		319
268	This will initialise the default event loop if it hasn't been initialised	320	This returns the "default" event loop object, which is what you should
269	yet and return it. If the default loop could not be initialised, returns	321	normally use when you just need "the event loop". Event loop objects and
270	false. If it already was initialised it simply returns it (and ignores the	322	the C<flags> parameter are described in more detail in the entry for
271	flags. If that is troubling you, check C<ev_backend ()> afterwards).	323	C<ev_loop_new>.
		324
		325	If the default loop is already initialised then this function simply
		326	returns it (and ignores the flags. If that is troubling you, check
		327	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		328	flags, which should almost always be C<0>, unless the caller is also the
		329	one calling C<ev_run> or otherwise qualifies as "the main program".
272		330
273	If you don't know what event loop to use, use the one returned from this	331	If you don't know what event loop to use, use the one returned from this
274	function.	332	function (or via the C<EV_DEFAULT> macro).
275		333
276	Note that this function is I<not> thread-safe, so if you want to use it	334	Note that this function is I<not> thread-safe, so if you want to use it
277	from multiple threads, you have to lock (note also that this is unlikely,	335	from multiple threads, you have to employ some kind of mutex (note also
278	as loops cannot bes hared easily between threads anyway).	336	that this case is unlikely, as loops cannot be shared easily between
		337	threads anyway).
279		338
280	The default loop is the only loop that can handle C<ev_signal> and	339	The default loop is the only loop that can handle C<ev_child> watchers,
281	C<ev_child> watchers, and to do this, it always registers a handler	340	and to do this, it always registers a handler for C<SIGCHLD>. If this is
282	for C<SIGCHLD>. If this is a problem for your app you can either	341	a problem for your application you can either create a dynamic loop with
283	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	342	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
284	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	343	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
285	C<ev_default_init>.	344
		345	Example: This is the most typical usage.
		346
		347	if (!ev_default_loop (0))
		348	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		349
		350	Example: Restrict libev to the select and poll backends, and do not allow
		351	environment settings to be taken into account:
		352
		353	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		354
		355	=item struct ev_loop *ev_loop_new (unsigned int flags)
		356
		357	This will create and initialise a new event loop object. If the loop
		358	could not be initialised, returns false.
		359
		360	This function is thread-safe, and one common way to use libev with
		361	threads is indeed to create one loop per thread, and using the default
		362	loop in the "main" or "initial" thread.
286		363
287	The flags argument can be used to specify special behaviour or specific	364	The flags argument can be used to specify special behaviour or specific
288	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	365	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
289		366
290	The following flags are supported:	367	The following flags are supported:
…		…
296	The default flags value. Use this if you have no clue (it's the right	373	The default flags value. Use this if you have no clue (it's the right
297	thing, believe me).	374	thing, believe me).
298		375
299	=item C<EVFLAG_NOENV>	376	=item C<EVFLAG_NOENV>
300		377
301	If this flag bit is ored into the flag value (or the program runs setuid	378	If this flag bit is or'ed into the flag value (or the program runs setuid
302	or setgid) then libev will I<not> look at the environment variable	379	or setgid) then libev will I<not> look at the environment variable
303	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	380	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
304	override the flags completely if it is found in the environment. This is	381	override the flags completely if it is found in the environment. This is
305	useful to try out specific backends to test their performance, or to work	382	useful to try out specific backends to test their performance, or to work
306	around bugs.	383	around bugs.
307		384
308	=item C<EVFLAG_FORKCHECK>	385	=item C<EVFLAG_FORKCHECK>
309		386
310	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	387	Instead of calling C<ev_loop_fork> manually after a fork, you can also
311	a fork, you can also make libev check for a fork in each iteration by	388	make libev check for a fork in each iteration by enabling this flag.
312	enabling this flag.
313		389
314	This works by calling C<getpid ()> on every iteration of the loop,	390	This works by calling C<getpid ()> on every iteration of the loop,
315	and thus this might slow down your event loop if you do a lot of loop	391	and thus this might slow down your event loop if you do a lot of loop
316	iterations and little real work, but is usually not noticeable (on my	392	iterations and little real work, but is usually not noticeable (on my
317	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	393	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
318	without a syscall and thus I<very> fast, but my GNU/Linux system also has	394	without a system call and thus I<very> fast, but my GNU/Linux system also has
319	C<pthread_atfork> which is even faster).	395	C<pthread_atfork> which is even faster).
320		396
321	The big advantage of this flag is that you can forget about fork (and	397	The big advantage of this flag is that you can forget about fork (and
322	forget about forgetting to tell libev about forking) when you use this	398	forget about forgetting to tell libev about forking) when you use this
323	flag.	399	flag.
324		400
325	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>	401	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
326	environment variable.	402	environment variable.
		403
		404	=item C<EVFLAG_NOINOTIFY>
		405
		406	When this flag is specified, then libev will not attempt to use the
		407	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
		408	testing, this flag can be useful to conserve inotify file descriptors, as
		409	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		410
		411	=item C<EVFLAG_SIGNALFD>
		412
		413	When this flag is specified, then libev will attempt to use the
		414	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
		415	delivers signals synchronously, which makes it both faster and might make
		416	it possible to get the queued signal data. It can also simplify signal
		417	handling with threads, as long as you properly block signals in your
		418	threads that are not interested in handling them.
		419
		420	Signalfd will not be used by default as this changes your signal mask, and
		421	there are a lot of shoddy libraries and programs (glib's threadpool for
		422	example) that can't properly initialise their signal masks.
327		423
328	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	424	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
329		425
330	This is your standard select(2) backend. Not I<completely> standard, as	426	This is your standard select(2) backend. Not I<completely> standard, as
331	libev tries to roll its own fd_set with no limits on the number of fds,	427	libev tries to roll its own fd_set with no limits on the number of fds,
332	but if that fails, expect a fairly low limit on the number of fds when	428	but if that fails, expect a fairly low limit on the number of fds when
333	using this backend. It doesn't scale too well (O(highest_fd)), but its	429	using this backend. It doesn't scale too well (O(highest_fd)), but its
334	usually the fastest backend for a low number of (low-numbered :) fds.	430	usually the fastest backend for a low number of (low-numbered :) fds.
335		431
336	To get good performance out of this backend you need a high amount of	432	To get good performance out of this backend you need a high amount of
337	parallelity (most of the file descriptors should be busy). If you are	433	parallelism (most of the file descriptors should be busy). If you are
338	writing a server, you should C<accept ()> in a loop to accept as many	434	writing a server, you should C<accept ()> in a loop to accept as many
339	connections as possible during one iteration. You might also want to have	435	connections as possible during one iteration. You might also want to have
340	a look at C<ev_set_io_collect_interval ()> to increase the amount of	436	a look at C<ev_set_io_collect_interval ()> to increase the amount of
341	readiness notifications you get per iteration.	437	readiness notifications you get per iteration.
		438
		439	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		440	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		441	C<exceptfds> set on that platform).
342		442
343	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	443	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
344		444
345	And this is your standard poll(2) backend. It's more complicated	445	And this is your standard poll(2) backend. It's more complicated
346	than select, but handles sparse fds better and has no artificial	446	than select, but handles sparse fds better and has no artificial
347	limit on the number of fds you can use (except it will slow down	447	limit on the number of fds you can use (except it will slow down
348	considerably with a lot of inactive fds). It scales similarly to select,	448	considerably with a lot of inactive fds). It scales similarly to select,
349	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	449	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
350	performance tips.	450	performance tips.
351		451
		452	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		453	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		454
352	=item C<EVBACKEND_EPOLL> (value 4, Linux)	455	=item C<EVBACKEND_EPOLL> (value 4, Linux)
		456
		457	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		458	kernels).
353		459
354	For few fds, this backend is a bit little slower than poll and select,	460	For few fds, this backend is a bit little slower than poll and select,
355	but it scales phenomenally better. While poll and select usually scale	461	but it scales phenomenally better. While poll and select usually scale
356	like O(total_fds) where n is the total number of fds (or the highest fd),	462	like O(total_fds) where n is the total number of fds (or the highest fd),
357	epoll scales either O(1) or O(active_fds). The epoll design has a number	463	epoll scales either O(1) or O(active_fds).
358	of shortcomings, such as silently dropping events in some hard-to-detect	464
359	cases and requiring a syscall per fd change, no fork support and bad	465	The epoll mechanism deserves honorable mention as the most misdesigned
360	support for dup.	466	of the more advanced event mechanisms: mere annoyances include silently
		467	dropping file descriptors, requiring a system call per change per file
		468	descriptor (and unnecessary guessing of parameters), problems with dup,
		469	returning before the timeout value, resulting in additional iterations
		470	(and only giving 5ms accuracy while select on the same platform gives
		471	0.1ms) and so on. The biggest issue is fork races, however - if a program
		472	forks then I<both> parent and child process have to recreate the epoll
		473	set, which can take considerable time (one syscall per file descriptor)
		474	and is of course hard to detect.
		475
		476	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		477	of course I<doesn't>, and epoll just loves to report events for totally
		478	I<different> file descriptors (even already closed ones, so one cannot
		479	even remove them from the set) than registered in the set (especially
		480	on SMP systems). Libev tries to counter these spurious notifications by
		481	employing an additional generation counter and comparing that against the
		482	events to filter out spurious ones, recreating the set when required. Last
		483	not least, it also refuses to work with some file descriptors which work
		484	perfectly fine with C<select> (files, many character devices...).
		485
		486	Epoll is truly the train wreck analog among event poll mechanisms.
361		487
362	While stopping, setting and starting an I/O watcher in the same iteration	488	While stopping, setting and starting an I/O watcher in the same iteration
363	will result in some caching, there is still a syscall per such incident	489	will result in some caching, there is still a system call per such
364	(because the fd could point to a different file description now), so its	490	incident (because the same I<file descriptor> could point to a different
365	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	491	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
366	very well if you register events for both fds.	492	file descriptors might not work very well if you register events for both
367		493	file descriptors.
368	Please note that epoll sometimes generates spurious notifications, so you
369	need to use non-blocking I/O or other means to avoid blocking when no data
370	(or space) is available.
371		494
372	Best performance from this backend is achieved by not unregistering all	495	Best performance from this backend is achieved by not unregistering all
373	watchers for a file descriptor until it has been closed, if possible, i.e.	496	watchers for a file descriptor until it has been closed, if possible,
374	keep at least one watcher active per fd at all times.	497	i.e. keep at least one watcher active per fd at all times. Stopping and
		498	starting a watcher (without re-setting it) also usually doesn't cause
		499	extra overhead. A fork can both result in spurious notifications as well
		500	as in libev having to destroy and recreate the epoll object, which can
		501	take considerable time and thus should be avoided.
375		502
		503	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		504	faster than epoll for maybe up to a hundred file descriptors, depending on
		505	the usage. So sad.
		506
376	While nominally embeddeble in other event loops, this feature is broken in	507	While nominally embeddable in other event loops, this feature is broken in
377	all kernel versions tested so far.	508	all kernel versions tested so far.
		509
		510	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		511	C<EVBACKEND_POLL>.
378		512
379	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	513	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
380		514
381	Kqueue deserves special mention, as at the time of this writing, it	515	Kqueue deserves special mention, as at the time of this writing, it
382	was broken on all BSDs except NetBSD (usually it doesn't work reliably	516	was broken on all BSDs except NetBSD (usually it doesn't work reliably
383	with anything but sockets and pipes, except on Darwin, where of course	517	with anything but sockets and pipes, except on Darwin, where of course
384	it's completely useless). For this reason it's not being "autodetected"	518	it's completely useless). Unlike epoll, however, whose brokenness
		519	is by design, these kqueue bugs can (and eventually will) be fixed
		520	without API changes to existing programs. For this reason it's not being
385	unless you explicitly specify it explicitly in the flags (i.e. using	521	"auto-detected" unless you explicitly specify it in the flags (i.e. using
386	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	522	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
387	system like NetBSD.	523	system like NetBSD.
388		524
389	You still can embed kqueue into a normal poll or select backend and use it	525	You still can embed kqueue into a normal poll or select backend and use it
390	only for sockets (after having made sure that sockets work with kqueue on	526	only for sockets (after having made sure that sockets work with kqueue on
391	the target platform). See C<ev_embed> watchers for more info.	527	the target platform). See C<ev_embed> watchers for more info.
392		528
393	It scales in the same way as the epoll backend, but the interface to the	529	It scales in the same way as the epoll backend, but the interface to the
394	kernel is more efficient (which says nothing about its actual speed, of	530	kernel is more efficient (which says nothing about its actual speed, of
395	course). While stopping, setting and starting an I/O watcher does never	531	course). While stopping, setting and starting an I/O watcher does never
396	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to	532	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
397	two event changes per incident, support for C<fork ()> is very bad and it	533	two event changes per incident. Support for C<fork ()> is very bad (but
398	drops fds silently in similarly hard-to-detect cases.	534	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		535	cases
399		536
400	This backend usually performs well under most conditions.	537	This backend usually performs well under most conditions.
401		538
402	While nominally embeddable in other event loops, this doesn't work	539	While nominally embeddable in other event loops, this doesn't work
403	everywhere, so you might need to test for this. And since it is broken	540	everywhere, so you might need to test for this. And since it is broken
404	almost everywhere, you should only use it when you have a lot of sockets	541	almost everywhere, you should only use it when you have a lot of sockets
405	(for which it usually works), by embedding it into another event loop	542	(for which it usually works), by embedding it into another event loop
406	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	543	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
407	sockets.	544	also broken on OS X)) and, did I mention it, using it only for sockets.
		545
		546	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		547	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		548	C<NOTE_EOF>.
408		549
409	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	550	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
410		551
411	This is not implemented yet (and might never be, unless you send me an	552	This is not implemented yet (and might never be, unless you send me an
412	implementation). According to reports, C</dev/poll> only supports sockets	553	implementation). According to reports, C</dev/poll> only supports sockets
…		…
416	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	557	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
417		558
418	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	559	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
419	it's really slow, but it still scales very well (O(active_fds)).	560	it's really slow, but it still scales very well (O(active_fds)).
420		561
421	Please note that solaris event ports can deliver a lot of spurious	562	Please note that Solaris event ports can deliver a lot of spurious
422	notifications, so you need to use non-blocking I/O or other means to avoid	563	notifications, so you need to use non-blocking I/O or other means to avoid
423	blocking when no data (or space) is available.	564	blocking when no data (or space) is available.
424		565
425	While this backend scales well, it requires one system call per active	566	While this backend scales well, it requires one system call per active
426	file descriptor per loop iteration. For small and medium numbers of file	567	file descriptor per loop iteration. For small and medium numbers of file
427	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	568	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
428	might perform better.	569	might perform better.
429		570
430	On the positive side, ignoring the spurious readiness notifications, this	571	On the positive side, with the exception of the spurious readiness
431	backend actually performed to specification in all tests and is fully	572	notifications, this backend actually performed fully to specification
432	embeddable, which is a rare feat among the OS-specific backends.	573	in all tests and is fully embeddable, which is a rare feat among the
		574	OS-specific backends (I vastly prefer correctness over speed hacks).
		575
		576	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		577	C<EVBACKEND_POLL>.
433		578
434	=item C<EVBACKEND_ALL>	579	=item C<EVBACKEND_ALL>
435		580
436	Try all backends (even potentially broken ones that wouldn't be tried	581	Try all backends (even potentially broken ones that wouldn't be tried
437	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	582	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
…		…
439		584
440	It is definitely not recommended to use this flag.	585	It is definitely not recommended to use this flag.
441		586
442	=back	587	=back
443		588
444	If one or more of these are ored into the flags value, then only these	589	If one or more of the backend flags are or'ed into the flags value,
445	backends will be tried (in the reverse order as listed here). If none are	590	then only these backends will be tried (in the reverse order as listed
446	specified, all backends in C<ev_recommended_backends ()> will be tried.	591	here). If none are specified, all backends in C<ev_recommended_backends
447		592	()> will be tried.
448	The most typical usage is like this:
449
450	if (!ev_default_loop (0))
451	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
452
453	Restrict libev to the select and poll backends, and do not allow
454	environment settings to be taken into account:
455
456	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
457
458	Use whatever libev has to offer, but make sure that kqueue is used if
459	available (warning, breaks stuff, best use only with your own private
460	event loop and only if you know the OS supports your types of fds):
461
462	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
463
464	=item struct ev_loop *ev_loop_new (unsigned int flags)
465
466	Similar to C<ev_default_loop>, but always creates a new event loop that is
467	always distinct from the default loop. Unlike the default loop, it cannot
468	handle signal and child watchers, and attempts to do so will be greeted by
469	undefined behaviour (or a failed assertion if assertions are enabled).
470
471	Note that this function I<is> thread-safe, and the recommended way to use
472	libev with threads is indeed to create one loop per thread, and using the
473	default loop in the "main" or "initial" thread.
474		593
475	Example: Try to create a event loop that uses epoll and nothing else.	594	Example: Try to create a event loop that uses epoll and nothing else.
476		595
477	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	596	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
478	if (!epoller)	597	if (!epoller)
479	fatal ("no epoll found here, maybe it hides under your chair");	598	fatal ("no epoll found here, maybe it hides under your chair");
480		599
		600	Example: Use whatever libev has to offer, but make sure that kqueue is
		601	used if available.
		602
		603	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		604
481	=item ev_default_destroy ()	605	=item ev_loop_destroy (loop)
482		606
483	Destroys the default loop again (frees all memory and kernel state	607	Destroys an event loop object (frees all memory and kernel state
484	etc.). None of the active event watchers will be stopped in the normal	608	etc.). None of the active event watchers will be stopped in the normal
485	sense, so e.g. C<ev_is_active> might still return true. It is your	609	sense, so e.g. C<ev_is_active> might still return true. It is your
486	responsibility to either stop all watchers cleanly yoursef I<before>	610	responsibility to either stop all watchers cleanly yourself I<before>
487	calling this function, or cope with the fact afterwards (which is usually	611	calling this function, or cope with the fact afterwards (which is usually
488	the easiest thing, you can just ignore the watchers and/or C<free ()> them	612	the easiest thing, you can just ignore the watchers and/or C<free ()> them
489	for example).	613	for example).
490		614
491	Note that certain global state, such as signal state, will not be freed by	615	Note that certain global state, such as signal state (and installed signal
492	this function, and related watchers (such as signal and child watchers)	616	handlers), will not be freed by this function, and related watchers (such
493	would need to be stopped manually.	617	as signal and child watchers) would need to be stopped manually.
494		618
495	In general it is not advisable to call this function except in the	619	This function is normally used on loop objects allocated by
496	rare occasion where you really need to free e.g. the signal handling	620	C<ev_loop_new>, but it can also be used on the default loop returned by
		621	C<ev_default_loop>, in which case it is not thread-safe.
		622
		623	Note that it is not advisable to call this function on the default loop
		624	except in the rare occasion where you really need to free its resources.
497	pipe fds. If you need dynamically allocated loops it is better to use	625	If you need dynamically allocated loops it is better to use C<ev_loop_new>
498	C<ev_loop_new> and C<ev_loop_destroy>).	626	and C<ev_loop_destroy>.
499		627
500	=item ev_loop_destroy (loop)	628	=item ev_loop_fork (loop)
501		629
502	Like C<ev_default_destroy>, but destroys an event loop created by an
503	earlier call to C<ev_loop_new>.
504
505	=item ev_default_fork ()
506
507	This function sets a flag that causes subsequent C<ev_loop> iterations	630	This function sets a flag that causes subsequent C<ev_run> iterations to
508	to reinitialise the kernel state for backends that have one. Despite the	631	reinitialise the kernel state for backends that have one. Despite the
509	name, you can call it anytime, but it makes most sense after forking, in	632	name, you can call it anytime, but it makes most sense after forking, in
510	the child process (or both child and parent, but that again makes little	633	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
511	sense). You I<must> call it in the child before using any of the libev	634	child before resuming or calling C<ev_run>.
512	functions, and it will only take effect at the next C<ev_loop> iteration.	635
		636	Again, you I<have> to call it on I<any> loop that you want to re-use after
		637	a fork, I<even if you do not plan to use the loop in the parent>. This is
		638	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		639	during fork.
513		640
514	On the other hand, you only need to call this function in the child	641	On the other hand, you only need to call this function in the child
515	process if and only if you want to use the event library in the child. If	642	process if and only if you want to use the event loop in the child. If
516	you just fork+exec, you don't have to call it at all.	643	you just fork+exec or create a new loop in the child, you don't have to
		644	call it at all (in fact, C<epoll> is so badly broken that it makes a
		645	difference, but libev will usually detect this case on its own and do a
		646	costly reset of the backend).
517		647
518	The function itself is quite fast and it's usually not a problem to call	648	The function itself is quite fast and it's usually not a problem to call
519	it just in case after a fork. To make this easy, the function will fit in	649	it just in case after a fork.
520	quite nicely into a call to C<pthread_atfork>:
521		650
		651	Example: Automate calling C<ev_loop_fork> on the default loop when
		652	using pthreads.
		653
		654	static void
		655	post_fork_child (void)
		656	{
		657	ev_loop_fork (EV_DEFAULT);
		658	}
		659
		660	...
522	pthread_atfork (0, 0, ev_default_fork);	661	pthread_atfork (0, 0, post_fork_child);
523
524	=item ev_loop_fork (loop)
525
526	Like C<ev_default_fork>, but acts on an event loop created by
527	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
528	after fork, and how you do this is entirely your own problem.
529		662
530	=item int ev_is_default_loop (loop)	663	=item int ev_is_default_loop (loop)
531		664
532	Returns true when the given loop actually is the default loop, false otherwise.	665	Returns true when the given loop is, in fact, the default loop, and false
		666	otherwise.
533		667
534	=item unsigned int ev_loop_count (loop)	668	=item unsigned int ev_iteration (loop)
535		669
536	Returns the count of loop iterations for the loop, which is identical to	670	Returns the current iteration count for the event loop, which is identical
537	the number of times libev did poll for new events. It starts at C<0> and	671	to the number of times libev did poll for new events. It starts at C<0>
538	happily wraps around with enough iterations.	672	and happily wraps around with enough iterations.
539		673
540	This value can sometimes be useful as a generation counter of sorts (it	674	This value can sometimes be useful as a generation counter of sorts (it
541	"ticks" the number of loop iterations), as it roughly corresponds with	675	"ticks" the number of loop iterations), as it roughly corresponds with
542	C<ev_prepare> and C<ev_check> calls.	676	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		677	prepare and check phases.
		678
		679	=item unsigned int ev_depth (loop)
		680
		681	Returns the number of times C<ev_run> was entered minus the number of
		682	times C<ev_run> was exited normally, in other words, the recursion depth.
		683
		684	Outside C<ev_run>, this number is zero. In a callback, this number is
		685	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
		686	in which case it is higher.
		687
		688	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
		689	throwing an exception etc.), doesn't count as "exit" - consider this
		690	as a hint to avoid such ungentleman-like behaviour unless it's really
		691	convenient, in which case it is fully supported.
543		692
544	=item unsigned int ev_backend (loop)	693	=item unsigned int ev_backend (loop)
545		694
546	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	695	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
547	use.	696	use.
…		…
552	received events and started processing them. This timestamp does not	701	received events and started processing them. This timestamp does not
553	change as long as callbacks are being processed, and this is also the base	702	change as long as callbacks are being processed, and this is also the base
554	time used for relative timers. You can treat it as the timestamp of the	703	time used for relative timers. You can treat it as the timestamp of the
555	event occurring (or more correctly, libev finding out about it).	704	event occurring (or more correctly, libev finding out about it).
556		705
		706	=item ev_now_update (loop)
		707
		708	Establishes the current time by querying the kernel, updating the time
		709	returned by C<ev_now ()> in the progress. This is a costly operation and
		710	is usually done automatically within C<ev_run ()>.
		711
		712	This function is rarely useful, but when some event callback runs for a
		713	very long time without entering the event loop, updating libev's idea of
		714	the current time is a good idea.
		715
		716	See also L<The special problem of time updates> in the C<ev_timer> section.
		717
		718	=item ev_suspend (loop)
		719
		720	=item ev_resume (loop)
		721
		722	These two functions suspend and resume an event loop, for use when the
		723	loop is not used for a while and timeouts should not be processed.
		724
		725	A typical use case would be an interactive program such as a game: When
		726	the user presses C<^Z> to suspend the game and resumes it an hour later it
		727	would be best to handle timeouts as if no time had actually passed while
		728	the program was suspended. This can be achieved by calling C<ev_suspend>
		729	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		730	C<ev_resume> directly afterwards to resume timer processing.
		731
		732	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		733	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		734	will be rescheduled (that is, they will lose any events that would have
		735	occurred while suspended).
		736
		737	After calling C<ev_suspend> you B<must not> call I<any> function on the
		738	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		739	without a previous call to C<ev_suspend>.
		740
		741	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		742	event loop time (see C<ev_now_update>).
		743
557	=item ev_loop (loop, int flags)	744	=item ev_run (loop, int flags)
558		745
559	Finally, this is it, the event handler. This function usually is called	746	Finally, this is it, the event handler. This function usually is called
560	after you initialised all your watchers and you want to start handling	747	after you have initialised all your watchers and you want to start
561	events.	748	handling events. It will ask the operating system for any new events, call
		749	the watcher callbacks, an then repeat the whole process indefinitely: This
		750	is why event loops are called I<loops>.
562		751
563	If the flags argument is specified as C<0>, it will not return until	752	If the flags argument is specified as C<0>, it will keep handling events
564	either no event watchers are active anymore or C<ev_unloop> was called.	753	until either no event watchers are active anymore or C<ev_break> was
		754	called.
565		755
566	Please note that an explicit C<ev_unloop> is usually better than	756	Please note that an explicit C<ev_break> is usually better than
567	relying on all watchers to be stopped when deciding when a program has	757	relying on all watchers to be stopped when deciding when a program has
568	finished (especially in interactive programs), but having a program that	758	finished (especially in interactive programs), but having a program
569	automatically loops as long as it has to and no longer by virtue of	759	that automatically loops as long as it has to and no longer by virtue
570	relying on its watchers stopping correctly is a thing of beauty.	760	of relying on its watchers stopping correctly, that is truly a thing of
		761	beauty.
571		762
		763	This function is also I<mostly> exception-safe - you can break out of
		764	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		765	exception and so on. This does not decrement the C<ev_depth> value, nor
		766	will it clear any outstanding C<EVBREAK_ONE> breaks.
		767
572	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	768	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
573	those events and any outstanding ones, but will not block your process in	769	those events and any already outstanding ones, but will not wait and
574	case there are no events and will return after one iteration of the loop.	770	block your process in case there are no events and will return after one
		771	iteration of the loop. This is sometimes useful to poll and handle new
		772	events while doing lengthy calculations, to keep the program responsive.
575		773
576	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	774	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
577	neccessary) and will handle those and any outstanding ones. It will block	775	necessary) and will handle those and any already outstanding ones. It
578	your process until at least one new event arrives, and will return after	776	will block your process until at least one new event arrives (which could
579	one iteration of the loop. This is useful if you are waiting for some	777	be an event internal to libev itself, so there is no guarantee that a
580	external event in conjunction with something not expressible using other	778	user-registered callback will be called), and will return after one
		779	iteration of the loop.
		780
		781	This is useful if you are waiting for some external event in conjunction
		782	with something not expressible using other libev watchers (i.e. "roll your
581	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	783	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
582	usually a better approach for this kind of thing.	784	usually a better approach for this kind of thing.
583		785
584	Here are the gory details of what C<ev_loop> does:	786	Here are the gory details of what C<ev_run> does:
585		787
		788	- Increment loop depth.
		789	- Reset the ev_break status.
586	- Before the first iteration, call any pending watchers.	790	- Before the first iteration, call any pending watchers.
		791	LOOP:
587	* If EVFLAG_FORKCHECK was used, check for a fork.	792	- If EVFLAG_FORKCHECK was used, check for a fork.
588	- If a fork was detected, queue and call all fork watchers.	793	- If a fork was detected (by any means), queue and call all fork watchers.
589	- Queue and call all prepare watchers.	794	- Queue and call all prepare watchers.
		795	- If ev_break was called, goto FINISH.
590	- If we have been forked, recreate the kernel state.	796	- If we have been forked, detach and recreate the kernel state
		797	as to not disturb the other process.
591	- Update the kernel state with all outstanding changes.	798	- Update the kernel state with all outstanding changes.
592	- Update the "event loop time".	799	- Update the "event loop time" (ev_now ()).
593	- Calculate for how long to sleep or block, if at all	800	- Calculate for how long to sleep or block, if at all
594	(active idle watchers, EVLOOP_NONBLOCK or not having	801	(active idle watchers, EVRUN_NOWAIT or not having
595	any active watchers at all will result in not sleeping).	802	any active watchers at all will result in not sleeping).
596	- Sleep if the I/O and timer collect interval say so.	803	- Sleep if the I/O and timer collect interval say so.
		804	- Increment loop iteration counter.
597	- Block the process, waiting for any events.	805	- Block the process, waiting for any events.
598	- Queue all outstanding I/O (fd) events.	806	- Queue all outstanding I/O (fd) events.
599	- Update the "event loop time" and do time jump handling.	807	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
600	- Queue all outstanding timers.	808	- Queue all expired timers.
601	- Queue all outstanding periodics.	809	- Queue all expired periodics.
602	- If no events are pending now, queue all idle watchers.	810	- Queue all idle watchers with priority higher than that of pending events.
603	- Queue all check watchers.	811	- Queue all check watchers.
604	- Call all queued watchers in reverse order (i.e. check watchers first).	812	- Call all queued watchers in reverse order (i.e. check watchers first).
605	Signals and child watchers are implemented as I/O watchers, and will	813	Signals and child watchers are implemented as I/O watchers, and will
606	be handled here by queueing them when their watcher gets executed.	814	be handled here by queueing them when their watcher gets executed.
607	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	815	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
608	were used, or there are no active watchers, return, otherwise	816	were used, or there are no active watchers, goto FINISH, otherwise
609	continue with step *.	817	continue with step LOOP.
		818	FINISH:
		819	- Reset the ev_break status iff it was EVBREAK_ONE.
		820	- Decrement the loop depth.
		821	- Return.
610		822
611	Example: Queue some jobs and then loop until no events are outstanding	823	Example: Queue some jobs and then loop until no events are outstanding
612	anymore.	824	anymore.
613		825
614	... queue jobs here, make sure they register event watchers as long	826	... queue jobs here, make sure they register event watchers as long
615	... as they still have work to do (even an idle watcher will do..)	827	... as they still have work to do (even an idle watcher will do..)
616	ev_loop (my_loop, 0);	828	ev_run (my_loop, 0);
617	... jobs done. yeah!	829	... jobs done or somebody called unloop. yeah!
618		830
619	=item ev_unloop (loop, how)	831	=item ev_break (loop, how)
620		832
621	Can be used to make a call to C<ev_loop> return early (but only after it	833	Can be used to make a call to C<ev_run> return early (but only after it
622	has processed all outstanding events). The C<how> argument must be either	834	has processed all outstanding events). The C<how> argument must be either
623	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	835	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
624	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	836	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
625		837
626	This "unloop state" will be cleared when entering C<ev_loop> again.	838	This "break state" will be cleared on the next call to C<ev_run>.
		839
		840	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		841	which case it will have no effect.
627		842
628	=item ev_ref (loop)	843	=item ev_ref (loop)
629		844
630	=item ev_unref (loop)	845	=item ev_unref (loop)
631		846
632	Ref/unref can be used to add or remove a reference count on the event	847	Ref/unref can be used to add or remove a reference count on the event
633	loop: Every watcher keeps one reference, and as long as the reference	848	loop: Every watcher keeps one reference, and as long as the reference
634	count is nonzero, C<ev_loop> will not return on its own. If you have	849	count is nonzero, C<ev_run> will not return on its own.
635	a watcher you never unregister that should not keep C<ev_loop> from	850
636	returning, ev_unref() after starting, and ev_ref() before stopping it. For	851	This is useful when you have a watcher that you never intend to
		852	unregister, but that nevertheless should not keep C<ev_run> from
		853	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		854	before stopping it.
		855
637	example, libev itself uses this for its internal signal pipe: It is not	856	As an example, libev itself uses this for its internal signal pipe: It
638	visible to the libev user and should not keep C<ev_loop> from exiting if	857	is not visible to the libev user and should not keep C<ev_run> from
639	no event watchers registered by it are active. It is also an excellent	858	exiting if no event watchers registered by it are active. It is also an
640	way to do this for generic recurring timers or from within third-party	859	excellent way to do this for generic recurring timers or from within
641	libraries. Just remember to I<unref after start> and I<ref before stop>	860	third-party libraries. Just remember to I<unref after start> and I<ref
642	(but only if the watcher wasn't active before, or was active before,	861	before stop> (but only if the watcher wasn't active before, or was active
643	respectively).	862	before, respectively. Note also that libev might stop watchers itself
		863	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		864	in the callback).
644		865
645	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	866	Example: Create a signal watcher, but keep it from keeping C<ev_run>
646	running when nothing else is active.	867	running when nothing else is active.
647		868
648	struct ev_signal exitsig;	869	ev_signal exitsig;
649	ev_signal_init (&exitsig, sig_cb, SIGINT);	870	ev_signal_init (&exitsig, sig_cb, SIGINT);
650	ev_signal_start (loop, &exitsig);	871	ev_signal_start (loop, &exitsig);
651	evf_unref (loop);	872	evf_unref (loop);
652		873
653	Example: For some weird reason, unregister the above signal handler again.	874	Example: For some weird reason, unregister the above signal handler again.
654		875
655	ev_ref (loop);	876	ev_ref (loop);
656	ev_signal_stop (loop, &exitsig);	877	ev_signal_stop (loop, &exitsig);
657		878
658	=item ev_set_io_collect_interval (loop, ev_tstamp interval)	879	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
659		880
660	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)	881	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
661		882
662	These advanced functions influence the time that libev will spend waiting	883	These advanced functions influence the time that libev will spend waiting
663	for events. Both are by default C<0>, meaning that libev will try to	884	for events. Both time intervals are by default C<0>, meaning that libev
664	invoke timer/periodic callbacks and I/O callbacks with minimum latency.	885	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		886	latency.
665		887
666	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	888	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
667	allows libev to delay invocation of I/O and timer/periodic callbacks to	889	allows libev to delay invocation of I/O and timer/periodic callbacks
668	increase efficiency of loop iterations.	890	to increase efficiency of loop iterations (or to increase power-saving
		891	opportunities).
669		892
670	The background is that sometimes your program runs just fast enough to	893	The idea is that sometimes your program runs just fast enough to handle
671	handle one (or very few) event(s) per loop iteration. While this makes	894	one (or very few) event(s) per loop iteration. While this makes the
672	the program responsive, it also wastes a lot of CPU time to poll for new	895	program responsive, it also wastes a lot of CPU time to poll for new
673	events, especially with backends like C<select ()> which have a high	896	events, especially with backends like C<select ()> which have a high
674	overhead for the actual polling but can deliver many events at once.	897	overhead for the actual polling but can deliver many events at once.
675		898
676	By setting a higher I<io collect interval> you allow libev to spend more	899	By setting a higher I<io collect interval> you allow libev to spend more
677	time collecting I/O events, so you can handle more events per iteration,	900	time collecting I/O events, so you can handle more events per iteration,
678	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	901	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
679	C<ev_timer>) will be not affected. Setting this to a non-null value will	902	C<ev_timer>) will be not affected. Setting this to a non-null value will
680	introduce an additional C<ev_sleep ()> call into most loop iterations.	903	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		904	sleep time ensures that libev will not poll for I/O events more often then
		905	once per this interval, on average.
681		906
682	Likewise, by setting a higher I<timeout collect interval> you allow libev	907	Likewise, by setting a higher I<timeout collect interval> you allow libev
683	to spend more time collecting timeouts, at the expense of increased	908	to spend more time collecting timeouts, at the expense of increased
684	latency (the watcher callback will be called later). C<ev_io> watchers	909	latency/jitter/inexactness (the watcher callback will be called
685	will not be affected. Setting this to a non-null value will not introduce	910	later). C<ev_io> watchers will not be affected. Setting this to a non-null
686	any overhead in libev.	911	value will not introduce any overhead in libev.
687		912
688	Many (busy) programs can usually benefit by setting the io collect	913	Many (busy) programs can usually benefit by setting the I/O collect
689	interval to a value near C<0.1> or so, which is often enough for	914	interval to a value near C<0.1> or so, which is often enough for
690	interactive servers (of course not for games), likewise for timeouts. It	915	interactive servers (of course not for games), likewise for timeouts. It
691	usually doesn't make much sense to set it to a lower value than C<0.01>,	916	usually doesn't make much sense to set it to a lower value than C<0.01>,
692	as this approsaches the timing granularity of most systems.	917	as this approaches the timing granularity of most systems. Note that if
		918	you do transactions with the outside world and you can't increase the
		919	parallelity, then this setting will limit your transaction rate (if you
		920	need to poll once per transaction and the I/O collect interval is 0.01,
		921	then you can't do more than 100 transactions per second).
		922
		923	Setting the I<timeout collect interval> can improve the opportunity for
		924	saving power, as the program will "bundle" timer callback invocations that
		925	are "near" in time together, by delaying some, thus reducing the number of
		926	times the process sleeps and wakes up again. Another useful technique to
		927	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		928	they fire on, say, one-second boundaries only.
		929
		930	Example: we only need 0.1s timeout granularity, and we wish not to poll
		931	more often than 100 times per second:
		932
		933	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		934	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		935
		936	=item ev_invoke_pending (loop)
		937
		938	This call will simply invoke all pending watchers while resetting their
		939	pending state. Normally, C<ev_run> does this automatically when required,
		940	but when overriding the invoke callback this call comes handy. This
		941	function can be invoked from a watcher - this can be useful for example
		942	when you want to do some lengthy calculation and want to pass further
		943	event handling to another thread (you still have to make sure only one
		944	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
		945
		946	=item int ev_pending_count (loop)
		947
		948	Returns the number of pending watchers - zero indicates that no watchers
		949	are pending.
		950
		951	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		952
		953	This overrides the invoke pending functionality of the loop: Instead of
		954	invoking all pending watchers when there are any, C<ev_run> will call
		955	this callback instead. This is useful, for example, when you want to
		956	invoke the actual watchers inside another context (another thread etc.).
		957
		958	If you want to reset the callback, use C<ev_invoke_pending> as new
		959	callback.
		960
		961	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		962
		963	Sometimes you want to share the same loop between multiple threads. This
		964	can be done relatively simply by putting mutex_lock/unlock calls around
		965	each call to a libev function.
		966
		967	However, C<ev_run> can run an indefinite time, so it is not feasible
		968	to wait for it to return. One way around this is to wake up the event
		969	loop via C<ev_break> and C<av_async_send>, another way is to set these
		970	I<release> and I<acquire> callbacks on the loop.
		971
		972	When set, then C<release> will be called just before the thread is
		973	suspended waiting for new events, and C<acquire> is called just
		974	afterwards.
		975
		976	Ideally, C<release> will just call your mutex_unlock function, and
		977	C<acquire> will just call the mutex_lock function again.
		978
		979	While event loop modifications are allowed between invocations of
		980	C<release> and C<acquire> (that's their only purpose after all), no
		981	modifications done will affect the event loop, i.e. adding watchers will
		982	have no effect on the set of file descriptors being watched, or the time
		983	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
		984	to take note of any changes you made.
		985
		986	In theory, threads executing C<ev_run> will be async-cancel safe between
		987	invocations of C<release> and C<acquire>.
		988
		989	See also the locking example in the C<THREADS> section later in this
		990	document.
		991
		992	=item ev_set_userdata (loop, void *data)
		993
		994	=item void *ev_userdata (loop)
		995
		996	Set and retrieve a single C<void *> associated with a loop. When
		997	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		998	C<0>.
		999
		1000	These two functions can be used to associate arbitrary data with a loop,
		1001	and are intended solely for the C<invoke_pending_cb>, C<release> and
		1002	C<acquire> callbacks described above, but of course can be (ab-)used for
		1003	any other purpose as well.
		1004
		1005	=item ev_verify (loop)
		1006
		1007	This function only does something when C<EV_VERIFY> support has been
		1008	compiled in, which is the default for non-minimal builds. It tries to go
		1009	through all internal structures and checks them for validity. If anything
		1010	is found to be inconsistent, it will print an error message to standard
		1011	error and call C<abort ()>.
		1012
		1013	This can be used to catch bugs inside libev itself: under normal
		1014	circumstances, this function will never abort as of course libev keeps its
		1015	data structures consistent.
693		1016
694	=back	1017	=back
695		1018
696		1019
697	=head1 ANATOMY OF A WATCHER	1020	=head1 ANATOMY OF A WATCHER
698		1021
		1022	In the following description, uppercase C<TYPE> in names stands for the
		1023	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		1024	watchers and C<ev_io_start> for I/O watchers.
		1025
699	A watcher is a structure that you create and register to record your	1026	A watcher is an opaque structure that you allocate and register to record
700	interest in some event. For instance, if you want to wait for STDIN to	1027	your interest in some event. To make a concrete example, imagine you want
701	become readable, you would create an C<ev_io> watcher for that:	1028	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1029	for that:
702		1030
703	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1031	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
704	{	1032	{
705	ev_io_stop (w);	1033	ev_io_stop (w);
706	ev_unloop (loop, EVUNLOOP_ALL);	1034	ev_break (loop, EVBREAK_ALL);
707	}	1035	}
708		1036
709	struct ev_loop *loop = ev_default_loop (0);	1037	struct ev_loop *loop = ev_default_loop (0);
		1038
710	struct ev_io stdin_watcher;	1039	ev_io stdin_watcher;
		1040
711	ev_init (&stdin_watcher, my_cb);	1041	ev_init (&stdin_watcher, my_cb);
712	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1042	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
713	ev_io_start (loop, &stdin_watcher);	1043	ev_io_start (loop, &stdin_watcher);
		1044
714	ev_loop (loop, 0);	1045	ev_run (loop, 0);
715		1046
716	As you can see, you are responsible for allocating the memory for your	1047	As you can see, you are responsible for allocating the memory for your
717	watcher structures (and it is usually a bad idea to do this on the stack,	1048	watcher structures (and it is I<usually> a bad idea to do this on the
718	although this can sometimes be quite valid).	1049	stack).
719		1050
		1051	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1052	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
		1053
720	Each watcher structure must be initialised by a call to C<ev_init	1054	Each watcher structure must be initialised by a call to C<ev_init (watcher
721	(watcher *, callback)>, which expects a callback to be provided. This	1055	*, callback)>, which expects a callback to be provided. This callback is
722	callback gets invoked each time the event occurs (or, in the case of io	1056	invoked each time the event occurs (or, in the case of I/O watchers, each
723	watchers, each time the event loop detects that the file descriptor given	1057	time the event loop detects that the file descriptor given is readable
724	is readable and/or writable).	1058	and/or writable).
725		1059
726	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1060	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
727	with arguments specific to this watcher type. There is also a macro	1061	macro to configure it, with arguments specific to the watcher type. There
728	to combine initialisation and setting in one call: C<< ev_<type>_init	1062	is also a macro to combine initialisation and setting in one call: C<<
729	(watcher *, callback, ...) >>.	1063	ev_TYPE_init (watcher *, callback, ...) >>.
730		1064
731	To make the watcher actually watch out for events, you have to start it	1065	To make the watcher actually watch out for events, you have to start it
732	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1066	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
733	*) >>), and you can stop watching for events at any time by calling the	1067	*) >>), and you can stop watching for events at any time by calling the
734	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1068	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
735		1069
736	As long as your watcher is active (has been started but not stopped) you	1070	As long as your watcher is active (has been started but not stopped) you
737	must not touch the values stored in it. Most specifically you must never	1071	must not touch the values stored in it. Most specifically you must never
738	reinitialise it or call its C<set> macro.	1072	reinitialise it or call its C<ev_TYPE_set> macro.
739		1073
740	Each and every callback receives the event loop pointer as first, the	1074	Each and every callback receives the event loop pointer as first, the
741	registered watcher structure as second, and a bitset of received events as	1075	registered watcher structure as second, and a bitset of received events as
742	third argument.	1076	third argument.
743		1077
…		…
752	=item C<EV_WRITE>	1086	=item C<EV_WRITE>
753		1087
754	The file descriptor in the C<ev_io> watcher has become readable and/or	1088	The file descriptor in the C<ev_io> watcher has become readable and/or
755	writable.	1089	writable.
756		1090
757	=item C<EV_TIMEOUT>	1091	=item C<EV_TIMER>
758		1092
759	The C<ev_timer> watcher has timed out.	1093	The C<ev_timer> watcher has timed out.
760		1094
761	=item C<EV_PERIODIC>	1095	=item C<EV_PERIODIC>
762		1096
…		…
780		1114
781	=item C<EV_PREPARE>	1115	=item C<EV_PREPARE>
782		1116
783	=item C<EV_CHECK>	1117	=item C<EV_CHECK>
784		1118
785	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1119	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts
786	to gather new events, and all C<ev_check> watchers are invoked just after	1120	to gather new events, and all C<ev_check> watchers are invoked just after
787	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1121	C<ev_run> has gathered them, but before it invokes any callbacks for any
788	received events. Callbacks of both watcher types can start and stop as	1122	received events. Callbacks of both watcher types can start and stop as
789	many watchers as they want, and all of them will be taken into account	1123	many watchers as they want, and all of them will be taken into account
790	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1124	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
791	C<ev_loop> from blocking).	1125	C<ev_run> from blocking).
792		1126
793	=item C<EV_EMBED>	1127	=item C<EV_EMBED>
794		1128
795	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1129	The embedded event loop specified in the C<ev_embed> watcher needs attention.
796		1130
797	=item C<EV_FORK>	1131	=item C<EV_FORK>
798		1132
799	The event loop has been resumed in the child process after fork (see	1133	The event loop has been resumed in the child process after fork (see
800	C<ev_fork>).	1134	C<ev_fork>).
801		1135
		1136	=item C<EV_CLEANUP>
		1137
		1138	The event loop is about to be destroyed (see C<ev_cleanup>).
		1139
802	=item C<EV_ASYNC>	1140	=item C<EV_ASYNC>
803		1141
804	The given async watcher has been asynchronously notified (see C<ev_async>).	1142	The given async watcher has been asynchronously notified (see C<ev_async>).
805		1143
		1144	=item C<EV_CUSTOM>
		1145
		1146	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1147	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1148
806	=item C<EV_ERROR>	1149	=item C<EV_ERROR>
807		1150
808	An unspecified error has occured, the watcher has been stopped. This might	1151	An unspecified error has occurred, the watcher has been stopped. This might
809	happen because the watcher could not be properly started because libev	1152	happen because the watcher could not be properly started because libev
810	ran out of memory, a file descriptor was found to be closed or any other	1153	ran out of memory, a file descriptor was found to be closed or any other
		1154	problem. Libev considers these application bugs.
		1155
811	problem. You best act on it by reporting the problem and somehow coping	1156	You best act on it by reporting the problem and somehow coping with the
812	with the watcher being stopped.	1157	watcher being stopped. Note that well-written programs should not receive
		1158	an error ever, so when your watcher receives it, this usually indicates a
		1159	bug in your program.
813		1160
814	Libev will usually signal a few "dummy" events together with an error,	1161	Libev will usually signal a few "dummy" events together with an error, for
815	for example it might indicate that a fd is readable or writable, and if	1162	example it might indicate that a fd is readable or writable, and if your
816	your callbacks is well-written it can just attempt the operation and cope	1163	callbacks is well-written it can just attempt the operation and cope with
817	with the error from read() or write(). This will not work in multithreaded	1164	the error from read() or write(). This will not work in multi-threaded
818	programs, though, so beware.	1165	programs, though, as the fd could already be closed and reused for another
		1166	thing, so beware.
819		1167
820	=back	1168	=back
821		1169
822	=head2 GENERIC WATCHER FUNCTIONS	1170	=head2 GENERIC WATCHER FUNCTIONS
823
824	In the following description, C<TYPE> stands for the watcher type,
825	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
826		1171
827	=over 4	1172	=over 4
828		1173
829	=item C<ev_init> (ev_TYPE *watcher, callback)	1174	=item C<ev_init> (ev_TYPE *watcher, callback)
830		1175
…		…
836	which rolls both calls into one.	1181	which rolls both calls into one.
837		1182
838	You can reinitialise a watcher at any time as long as it has been stopped	1183	You can reinitialise a watcher at any time as long as it has been stopped
839	(or never started) and there are no pending events outstanding.	1184	(or never started) and there are no pending events outstanding.
840		1185
841	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1186	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
842	int revents)>.	1187	int revents)>.
843		1188
		1189	Example: Initialise an C<ev_io> watcher in two steps.
		1190
		1191	ev_io w;
		1192	ev_init (&w, my_cb);
		1193	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1194
844	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1195	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
845		1196
846	This macro initialises the type-specific parts of a watcher. You need to	1197	This macro initialises the type-specific parts of a watcher. You need to
847	call C<ev_init> at least once before you call this macro, but you can	1198	call C<ev_init> at least once before you call this macro, but you can
848	call C<ev_TYPE_set> any number of times. You must not, however, call this	1199	call C<ev_TYPE_set> any number of times. You must not, however, call this
849	macro on a watcher that is active (it can be pending, however, which is a	1200	macro on a watcher that is active (it can be pending, however, which is a
850	difference to the C<ev_init> macro).	1201	difference to the C<ev_init> macro).
851		1202
852	Although some watcher types do not have type-specific arguments	1203	Although some watcher types do not have type-specific arguments
853	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1204	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
854		1205
		1206	See C<ev_init>, above, for an example.
		1207
855	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1208	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
856		1209
857	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1210	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
858	calls into a single call. This is the most convinient method to initialise	1211	calls into a single call. This is the most convenient method to initialise
859	a watcher. The same limitations apply, of course.	1212	a watcher. The same limitations apply, of course.
860		1213
		1214	Example: Initialise and set an C<ev_io> watcher in one step.
		1215
		1216	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1217
861	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1218	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
862		1219
863	Starts (activates) the given watcher. Only active watchers will receive	1220	Starts (activates) the given watcher. Only active watchers will receive
864	events. If the watcher is already active nothing will happen.	1221	events. If the watcher is already active nothing will happen.
865		1222
		1223	Example: Start the C<ev_io> watcher that is being abused as example in this
		1224	whole section.
		1225
		1226	ev_io_start (EV_DEFAULT_UC, &w);
		1227
866	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1228	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
867		1229
868	Stops the given watcher again (if active) and clears the pending	1230	Stops the given watcher if active, and clears the pending status (whether
		1231	the watcher was active or not).
		1232
869	status. It is possible that stopped watchers are pending (for example,	1233	It is possible that stopped watchers are pending - for example,
870	non-repeating timers are being stopped when they become pending), but	1234	non-repeating timers are being stopped when they become pending - but
871	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1235	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
872	you want to free or reuse the memory used by the watcher it is therefore a	1236	pending. If you want to free or reuse the memory used by the watcher it is
873	good idea to always call its C<ev_TYPE_stop> function.	1237	therefore a good idea to always call its C<ev_TYPE_stop> function.
874		1238
875	=item bool ev_is_active (ev_TYPE *watcher)	1239	=item bool ev_is_active (ev_TYPE *watcher)
876		1240
877	Returns a true value iff the watcher is active (i.e. it has been started	1241	Returns a true value iff the watcher is active (i.e. it has been started
878	and not yet been stopped). As long as a watcher is active you must not modify	1242	and not yet been stopped). As long as a watcher is active you must not modify
…		…
894	=item ev_cb_set (ev_TYPE *watcher, callback)	1258	=item ev_cb_set (ev_TYPE *watcher, callback)
895		1259
896	Change the callback. You can change the callback at virtually any time	1260	Change the callback. You can change the callback at virtually any time
897	(modulo threads).	1261	(modulo threads).
898		1262
899	=item ev_set_priority (ev_TYPE *watcher, priority)	1263	=item ev_set_priority (ev_TYPE *watcher, int priority)
900		1264
901	=item int ev_priority (ev_TYPE *watcher)	1265	=item int ev_priority (ev_TYPE *watcher)
902		1266
903	Set and query the priority of the watcher. The priority is a small	1267	Set and query the priority of the watcher. The priority is a small
904	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1268	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
905	(default: C<-2>). Pending watchers with higher priority will be invoked	1269	(default: C<-2>). Pending watchers with higher priority will be invoked
906	before watchers with lower priority, but priority will not keep watchers	1270	before watchers with lower priority, but priority will not keep watchers
907	from being executed (except for C<ev_idle> watchers).	1271	from being executed (except for C<ev_idle> watchers).
908		1272
909	This means that priorities are I<only> used for ordering callback
910	invocation after new events have been received. This is useful, for
911	example, to reduce latency after idling, or more often, to bind two
912	watchers on the same event and make sure one is called first.
913
914	If you need to suppress invocation when higher priority events are pending	1273	If you need to suppress invocation when higher priority events are pending
915	you need to look at C<ev_idle> watchers, which provide this functionality.	1274	you need to look at C<ev_idle> watchers, which provide this functionality.
916		1275
917	You I<must not> change the priority of a watcher as long as it is active or	1276	You I<must not> change the priority of a watcher as long as it is active or
918	pending.	1277	pending.
919		1278
		1279	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1280	fine, as long as you do not mind that the priority value you query might
		1281	or might not have been clamped to the valid range.
		1282
920	The default priority used by watchers when no priority has been set is	1283	The default priority used by watchers when no priority has been set is
921	always C<0>, which is supposed to not be too high and not be too low :).	1284	always C<0>, which is supposed to not be too high and not be too low :).
922		1285
923	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1286	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
924	fine, as long as you do not mind that the priority value you query might	1287	priorities.
925	or might not have been adjusted to be within valid range.
926		1288
927	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1289	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
928		1290
929	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1291	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
930	C<loop> nor C<revents> need to be valid as long as the watcher callback	1292	C<loop> nor C<revents> need to be valid as long as the watcher callback
931	can deal with that fact.	1293	can deal with that fact, as both are simply passed through to the
		1294	callback.
932		1295
933	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1296	=item int ev_clear_pending (loop, ev_TYPE *watcher)
934		1297
935	If the watcher is pending, this function returns clears its pending status	1298	If the watcher is pending, this function clears its pending status and
936	and returns its C<revents> bitset (as if its callback was invoked). If the	1299	returns its C<revents> bitset (as if its callback was invoked). If the
937	watcher isn't pending it does nothing and returns C<0>.	1300	watcher isn't pending it does nothing and returns C<0>.
938		1301
		1302	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1303	callback to be invoked, which can be accomplished with this function.
		1304
		1305	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1306
		1307	Feeds the given event set into the event loop, as if the specified event
		1308	had happened for the specified watcher (which must be a pointer to an
		1309	initialised but not necessarily started event watcher). Obviously you must
		1310	not free the watcher as long as it has pending events.
		1311
		1312	Stopping the watcher, letting libev invoke it, or calling
		1313	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1314	not started in the first place.
		1315
		1316	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1317	functions that do not need a watcher.
		1318
939	=back	1319	=back
940		1320
941
942	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1321	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
943		1322
944	Each watcher has, by default, a member C<void *data> that you can change	1323	Each watcher has, by default, a member C<void *data> that you can change
945	and read at any time, libev will completely ignore it. This can be used	1324	and read at any time: libev will completely ignore it. This can be used
946	to associate arbitrary data with your watcher. If you need more data and	1325	to associate arbitrary data with your watcher. If you need more data and
947	don't want to allocate memory and store a pointer to it in that data	1326	don't want to allocate memory and store a pointer to it in that data
948	member, you can also "subclass" the watcher type and provide your own	1327	member, you can also "subclass" the watcher type and provide your own
949	data:	1328	data:
950		1329
951	struct my_io	1330	struct my_io
952	{	1331	{
953	struct ev_io io;	1332	ev_io io;
954	int otherfd;	1333	int otherfd;
955	void *somedata;	1334	void *somedata;
956	struct whatever *mostinteresting;	1335	struct whatever *mostinteresting;
957	}	1336	};
		1337
		1338	...
		1339	struct my_io w;
		1340	ev_io_init (&w.io, my_cb, fd, EV_READ);
958		1341
959	And since your callback will be called with a pointer to the watcher, you	1342	And since your callback will be called with a pointer to the watcher, you
960	can cast it back to your own type:	1343	can cast it back to your own type:
961		1344
962	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1345	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
963	{	1346	{
964	struct my_io *w = (struct my_io *)w_;	1347	struct my_io *w = (struct my_io *)w_;
965	...	1348	...
966	}	1349	}
967		1350
968	More interesting and less C-conformant ways of casting your callback type	1351	More interesting and less C-conformant ways of casting your callback type
969	instead have been omitted.	1352	instead have been omitted.
970		1353
971	Another common scenario is having some data structure with multiple	1354	Another common scenario is to use some data structure with multiple
972	watchers:	1355	embedded watchers:
973		1356
974	struct my_biggy	1357	struct my_biggy
975	{	1358	{
976	int some_data;	1359	int some_data;
977	ev_timer t1;	1360	ev_timer t1;
978	ev_timer t2;	1361	ev_timer t2;
979	}	1362	}
980		1363
981	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1364	In this case getting the pointer to C<my_biggy> is a bit more
982	you need to use C<offsetof>:	1365	complicated: Either you store the address of your C<my_biggy> struct
		1366	in the C<data> member of the watcher (for woozies), or you need to use
		1367	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1368	programmers):
983		1369
984	#include <stddef.h>	1370	#include <stddef.h>
985		1371
986	static void	1372	static void
987	t1_cb (EV_P_ struct ev_timer *w, int revents)	1373	t1_cb (EV_P_ ev_timer *w, int revents)
988	{	1374	{
989	struct my_biggy big = (struct my_biggy *	1375	struct my_biggy big = (struct my_biggy *)
990	(((char *)w) - offsetof (struct my_biggy, t1));	1376	(((char *)w) - offsetof (struct my_biggy, t1));
991	}	1377	}
992		1378
993	static void	1379	static void
994	t2_cb (EV_P_ struct ev_timer *w, int revents)	1380	t2_cb (EV_P_ ev_timer *w, int revents)
995	{	1381	{
996	struct my_biggy big = (struct my_biggy *	1382	struct my_biggy big = (struct my_biggy *)
997	(((char *)w) - offsetof (struct my_biggy, t2));	1383	(((char *)w) - offsetof (struct my_biggy, t2));
998	}	1384	}
		1385
		1386	=head2 WATCHER STATES
		1387
		1388	There are various watcher states mentioned throughout this manual -
		1389	active, pending and so on. In this section these states and the rules to
		1390	transition between them will be described in more detail - and while these
		1391	rules might look complicated, they usually do "the right thing".
		1392
		1393	=over 4
		1394
		1395	=item initialiased
		1396
		1397	Before a watcher can be registered with the event looop it has to be
		1398	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1399	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
		1400
		1401	In this state it is simply some block of memory that is suitable for use
		1402	in an event loop. It can be moved around, freed, reused etc. at will.
		1403
		1404	=item started/running/active
		1405
		1406	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
		1407	property of the event loop, and is actively waiting for events. While in
		1408	this state it cannot be accessed (except in a few documented ways), moved,
		1409	freed or anything else - the only legal thing is to keep a pointer to it,
		1410	and call libev functions on it that are documented to work on active watchers.
		1411
		1412	=item pending
		1413
		1414	If a watcher is active and libev determines that an event it is interested
		1415	in has occurred (such as a timer expiring), it will become pending. It will
		1416	stay in this pending state until either it is stopped or its callback is
		1417	about to be invoked, so it is not normally pending inside the watcher
		1418	callback.
		1419
		1420	The watcher might or might not be active while it is pending (for example,
		1421	an expired non-repeating timer can be pending but no longer active). If it
		1422	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1423	but it is still property of the event loop at this time, so cannot be
		1424	moved, freed or reused. And if it is active the rules described in the
		1425	previous item still apply.
		1426
		1427	It is also possible to feed an event on a watcher that is not active (e.g.
		1428	via C<ev_feed_event>), in which case it becomes pending without being
		1429	active.
		1430
		1431	=item stopped
		1432
		1433	A watcher can be stopped implicitly by libev (in which case it might still
		1434	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1435	latter will clear any pending state the watcher might be in, regardless
		1436	of whether it was active or not, so stopping a watcher explicitly before
		1437	freeing it is often a good idea.
		1438
		1439	While stopped (and not pending) the watcher is essentially in the
		1440	initialised state, that is it can be reused, moved, modified in any way
		1441	you wish.
		1442
		1443	=back
		1444
		1445	=head2 WATCHER PRIORITY MODELS
		1446
		1447	Many event loops support I<watcher priorities>, which are usually small
		1448	integers that influence the ordering of event callback invocation
		1449	between watchers in some way, all else being equal.
		1450
		1451	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1452	description for the more technical details such as the actual priority
		1453	range.
		1454
		1455	There are two common ways how these these priorities are being interpreted
		1456	by event loops:
		1457
		1458	In the more common lock-out model, higher priorities "lock out" invocation
		1459	of lower priority watchers, which means as long as higher priority
		1460	watchers receive events, lower priority watchers are not being invoked.
		1461
		1462	The less common only-for-ordering model uses priorities solely to order
		1463	callback invocation within a single event loop iteration: Higher priority
		1464	watchers are invoked before lower priority ones, but they all get invoked
		1465	before polling for new events.
		1466
		1467	Libev uses the second (only-for-ordering) model for all its watchers
		1468	except for idle watchers (which use the lock-out model).
		1469
		1470	The rationale behind this is that implementing the lock-out model for
		1471	watchers is not well supported by most kernel interfaces, and most event
		1472	libraries will just poll for the same events again and again as long as
		1473	their callbacks have not been executed, which is very inefficient in the
		1474	common case of one high-priority watcher locking out a mass of lower
		1475	priority ones.
		1476
		1477	Static (ordering) priorities are most useful when you have two or more
		1478	watchers handling the same resource: a typical usage example is having an
		1479	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1480	timeouts. Under load, data might be received while the program handles
		1481	other jobs, but since timers normally get invoked first, the timeout
		1482	handler will be executed before checking for data. In that case, giving
		1483	the timer a lower priority than the I/O watcher ensures that I/O will be
		1484	handled first even under adverse conditions (which is usually, but not
		1485	always, what you want).
		1486
		1487	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1488	will only be executed when no same or higher priority watchers have
		1489	received events, they can be used to implement the "lock-out" model when
		1490	required.
		1491
		1492	For example, to emulate how many other event libraries handle priorities,
		1493	you can associate an C<ev_idle> watcher to each such watcher, and in
		1494	the normal watcher callback, you just start the idle watcher. The real
		1495	processing is done in the idle watcher callback. This causes libev to
		1496	continuously poll and process kernel event data for the watcher, but when
		1497	the lock-out case is known to be rare (which in turn is rare :), this is
		1498	workable.
		1499
		1500	Usually, however, the lock-out model implemented that way will perform
		1501	miserably under the type of load it was designed to handle. In that case,
		1502	it might be preferable to stop the real watcher before starting the
		1503	idle watcher, so the kernel will not have to process the event in case
		1504	the actual processing will be delayed for considerable time.
		1505
		1506	Here is an example of an I/O watcher that should run at a strictly lower
		1507	priority than the default, and which should only process data when no
		1508	other events are pending:
		1509
		1510	ev_idle idle; // actual processing watcher
		1511	ev_io io; // actual event watcher
		1512
		1513	static void
		1514	io_cb (EV_P_ ev_io *w, int revents)
		1515	{
		1516	// stop the I/O watcher, we received the event, but
		1517	// are not yet ready to handle it.
		1518	ev_io_stop (EV_A_ w);
		1519
		1520	// start the idle watcher to handle the actual event.
		1521	// it will not be executed as long as other watchers
		1522	// with the default priority are receiving events.
		1523	ev_idle_start (EV_A_ &idle);
		1524	}
		1525
		1526	static void
		1527	idle_cb (EV_P_ ev_idle *w, int revents)
		1528	{
		1529	// actual processing
		1530	read (STDIN_FILENO, ...);
		1531
		1532	// have to start the I/O watcher again, as
		1533	// we have handled the event
		1534	ev_io_start (EV_P_ &io);
		1535	}
		1536
		1537	// initialisation
		1538	ev_idle_init (&idle, idle_cb);
		1539	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1540	ev_io_start (EV_DEFAULT_ &io);
		1541
		1542	In the "real" world, it might also be beneficial to start a timer, so that
		1543	low-priority connections can not be locked out forever under load. This
		1544	enables your program to keep a lower latency for important connections
		1545	during short periods of high load, while not completely locking out less
		1546	important ones.
999		1547
1000		1548
1001	=head1 WATCHER TYPES	1549	=head1 WATCHER TYPES
1002		1550
1003	This section describes each watcher in detail, but will not repeat	1551	This section describes each watcher in detail, but will not repeat
…		…
1027	In general you can register as many read and/or write event watchers per	1575	In general you can register as many read and/or write event watchers per
1028	fd as you want (as long as you don't confuse yourself). Setting all file	1576	fd as you want (as long as you don't confuse yourself). Setting all file
1029	descriptors to non-blocking mode is also usually a good idea (but not	1577	descriptors to non-blocking mode is also usually a good idea (but not
1030	required if you know what you are doing).	1578	required if you know what you are doing).
1031		1579
1032	If you must do this, then force the use of a known-to-be-good backend	1580	If you cannot use non-blocking mode, then force the use of a
1033	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1581	known-to-be-good backend (at the time of this writing, this includes only
1034	C<EVBACKEND_POLL>).	1582	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1583	descriptors for which non-blocking operation makes no sense (such as
		1584	files) - libev doesn't guarantee any specific behaviour in that case.
1035		1585
1036	Another thing you have to watch out for is that it is quite easy to	1586	Another thing you have to watch out for is that it is quite easy to
1037	receive "spurious" readiness notifications, that is your callback might	1587	receive "spurious" readiness notifications, that is your callback might
1038	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1588	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1039	because there is no data. Not only are some backends known to create a	1589	because there is no data. Not only are some backends known to create a
1040	lot of those (for example solaris ports), it is very easy to get into	1590	lot of those (for example Solaris ports), it is very easy to get into
1041	this situation even with a relatively standard program structure. Thus	1591	this situation even with a relatively standard program structure. Thus
1042	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1592	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1043	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1593	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1044		1594
1045	If you cannot run the fd in non-blocking mode (for example you should not	1595	If you cannot run the fd in non-blocking mode (for example you should
1046	play around with an Xlib connection), then you have to seperately re-test	1596	not play around with an Xlib connection), then you have to separately
1047	whether a file descriptor is really ready with a known-to-be good interface	1597	re-test whether a file descriptor is really ready with a known-to-be good
1048	such as poll (fortunately in our Xlib example, Xlib already does this on	1598	interface such as poll (fortunately in our Xlib example, Xlib already
1049	its own, so its quite safe to use).	1599	does this on its own, so its quite safe to use). Some people additionally
		1600	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1601	indefinitely.
		1602
		1603	But really, best use non-blocking mode.
1050		1604
1051	=head3 The special problem of disappearing file descriptors	1605	=head3 The special problem of disappearing file descriptors
1052		1606
1053	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1607	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1054	descriptor (either by calling C<close> explicitly or by any other means,	1608	descriptor (either due to calling C<close> explicitly or any other means,
1055	such as C<dup>). The reason is that you register interest in some file	1609	such as C<dup2>). The reason is that you register interest in some file
1056	descriptor, but when it goes away, the operating system will silently drop	1610	descriptor, but when it goes away, the operating system will silently drop
1057	this interest. If another file descriptor with the same number then is	1611	this interest. If another file descriptor with the same number then is
1058	registered with libev, there is no efficient way to see that this is, in	1612	registered with libev, there is no efficient way to see that this is, in
1059	fact, a different file descriptor.	1613	fact, a different file descriptor.
1060		1614
…		…
1091	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1645	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1092	C<EVBACKEND_POLL>.	1646	C<EVBACKEND_POLL>.
1093		1647
1094	=head3 The special problem of SIGPIPE	1648	=head3 The special problem of SIGPIPE
1095		1649
1096	While not really specific to libev, it is easy to forget about SIGPIPE:	1650	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1097	when reading from a pipe whose other end has been closed, your program	1651	when writing to a pipe whose other end has been closed, your program gets
1098	gets send a SIGPIPE, which, by default, aborts your program. For most	1652	sent a SIGPIPE, which, by default, aborts your program. For most programs
1099	programs this is sensible behaviour, for daemons, this is usually	1653	this is sensible behaviour, for daemons, this is usually undesirable.
1100	undesirable.
1101		1654
1102	So when you encounter spurious, unexplained daemon exits, make sure you	1655	So when you encounter spurious, unexplained daemon exits, make sure you
1103	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1656	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1104	somewhere, as that would have given you a big clue).	1657	somewhere, as that would have given you a big clue).
1105		1658
		1659	=head3 The special problem of accept()ing when you can't
		1660
		1661	Many implementations of the POSIX C<accept> function (for example,
		1662	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1663	connection from the pending queue in all error cases.
		1664
		1665	For example, larger servers often run out of file descriptors (because
		1666	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1667	rejecting the connection, leading to libev signalling readiness on
		1668	the next iteration again (the connection still exists after all), and
		1669	typically causing the program to loop at 100% CPU usage.
		1670
		1671	Unfortunately, the set of errors that cause this issue differs between
		1672	operating systems, there is usually little the app can do to remedy the
		1673	situation, and no known thread-safe method of removing the connection to
		1674	cope with overload is known (to me).
		1675
		1676	One of the easiest ways to handle this situation is to just ignore it
		1677	- when the program encounters an overload, it will just loop until the
		1678	situation is over. While this is a form of busy waiting, no OS offers an
		1679	event-based way to handle this situation, so it's the best one can do.
		1680
		1681	A better way to handle the situation is to log any errors other than
		1682	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1683	messages, and continue as usual, which at least gives the user an idea of
		1684	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1685	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1686	usage.
		1687
		1688	If your program is single-threaded, then you could also keep a dummy file
		1689	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1690	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1691	close that fd, and create a new dummy fd. This will gracefully refuse
		1692	clients under typical overload conditions.
		1693
		1694	The last way to handle it is to simply log the error and C<exit>, as
		1695	is often done with C<malloc> failures, but this results in an easy
		1696	opportunity for a DoS attack.
1106		1697
1107	=head3 Watcher-Specific Functions	1698	=head3 Watcher-Specific Functions
1108		1699
1109	=over 4	1700	=over 4
1110		1701
1111	=item ev_io_init (ev_io *, callback, int fd, int events)	1702	=item ev_io_init (ev_io *, callback, int fd, int events)
1112		1703
1113	=item ev_io_set (ev_io *, int fd, int events)	1704	=item ev_io_set (ev_io *, int fd, int events)
1114		1705
1115	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1706	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1116	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or	1707	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1117	C<EV_READ | EV_WRITE> to receive the given events.	1708	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1118		1709
1119	=item int fd [read-only]	1710	=item int fd [read-only]
1120		1711
1121	The file descriptor being watched.	1712	The file descriptor being watched.
1122		1713
…		…
1130		1721
1131	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1722	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1132	readable, but only once. Since it is likely line-buffered, you could	1723	readable, but only once. Since it is likely line-buffered, you could
1133	attempt to read a whole line in the callback.	1724	attempt to read a whole line in the callback.
1134		1725
1135	static void	1726	static void
1136	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1727	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1137	{	1728	{
1138	ev_io_stop (loop, w);	1729	ev_io_stop (loop, w);
1139	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1730	.. read from stdin here (or from w->fd) and handle any I/O errors
1140	}	1731	}
1141		1732
1142	...	1733	...
1143	struct ev_loop *loop = ev_default_init (0);	1734	struct ev_loop *loop = ev_default_init (0);
1144	struct ev_io stdin_readable;	1735	ev_io stdin_readable;
1145	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1736	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1146	ev_io_start (loop, &stdin_readable);	1737	ev_io_start (loop, &stdin_readable);
1147	ev_loop (loop, 0);	1738	ev_run (loop, 0);
1148		1739
1149		1740
1150	=head2 C<ev_timer> - relative and optionally repeating timeouts	1741	=head2 C<ev_timer> - relative and optionally repeating timeouts
1151		1742
1152	Timer watchers are simple relative timers that generate an event after a	1743	Timer watchers are simple relative timers that generate an event after a
1153	given time, and optionally repeating in regular intervals after that.	1744	given time, and optionally repeating in regular intervals after that.
1154		1745
1155	The timers are based on real time, that is, if you register an event that	1746	The timers are based on real time, that is, if you register an event that
1156	times out after an hour and you reset your system clock to january last	1747	times out after an hour and you reset your system clock to January last
1157	year, it will still time out after (roughly) and hour. "Roughly" because	1748	year, it will still time out after (roughly) one hour. "Roughly" because
1158	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1749	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1159	monotonic clock option helps a lot here).	1750	monotonic clock option helps a lot here).
		1751
		1752	The callback is guaranteed to be invoked only I<after> its timeout has
		1753	passed (not I<at>, so on systems with very low-resolution clocks this
		1754	might introduce a small delay). If multiple timers become ready during the
		1755	same loop iteration then the ones with earlier time-out values are invoked
		1756	before ones of the same priority with later time-out values (but this is
		1757	no longer true when a callback calls C<ev_run> recursively).
		1758
		1759	=head3 Be smart about timeouts
		1760
		1761	Many real-world problems involve some kind of timeout, usually for error
		1762	recovery. A typical example is an HTTP request - if the other side hangs,
		1763	you want to raise some error after a while.
		1764
		1765	What follows are some ways to handle this problem, from obvious and
		1766	inefficient to smart and efficient.
		1767
		1768	In the following, a 60 second activity timeout is assumed - a timeout that
		1769	gets reset to 60 seconds each time there is activity (e.g. each time some
		1770	data or other life sign was received).
		1771
		1772	=over 4
		1773
		1774	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1775
		1776	This is the most obvious, but not the most simple way: In the beginning,
		1777	start the watcher:
		1778
		1779	ev_timer_init (timer, callback, 60., 0.);
		1780	ev_timer_start (loop, timer);
		1781
		1782	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1783	and start it again:
		1784
		1785	ev_timer_stop (loop, timer);
		1786	ev_timer_set (timer, 60., 0.);
		1787	ev_timer_start (loop, timer);
		1788
		1789	This is relatively simple to implement, but means that each time there is
		1790	some activity, libev will first have to remove the timer from its internal
		1791	data structure and then add it again. Libev tries to be fast, but it's
		1792	still not a constant-time operation.
		1793
		1794	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1795
		1796	This is the easiest way, and involves using C<ev_timer_again> instead of
		1797	C<ev_timer_start>.
		1798
		1799	To implement this, configure an C<ev_timer> with a C<repeat> value
		1800	of C<60> and then call C<ev_timer_again> at start and each time you
		1801	successfully read or write some data. If you go into an idle state where
		1802	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1803	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1804
		1805	That means you can ignore both the C<ev_timer_start> function and the
		1806	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1807	member and C<ev_timer_again>.
		1808
		1809	At start:
		1810
		1811	ev_init (timer, callback);
		1812	timer->repeat = 60.;
		1813	ev_timer_again (loop, timer);
		1814
		1815	Each time there is some activity:
		1816
		1817	ev_timer_again (loop, timer);
		1818
		1819	It is even possible to change the time-out on the fly, regardless of
		1820	whether the watcher is active or not:
		1821
		1822	timer->repeat = 30.;
		1823	ev_timer_again (loop, timer);
		1824
		1825	This is slightly more efficient then stopping/starting the timer each time
		1826	you want to modify its timeout value, as libev does not have to completely
		1827	remove and re-insert the timer from/into its internal data structure.
		1828
		1829	It is, however, even simpler than the "obvious" way to do it.
		1830
		1831	=item 3. Let the timer time out, but then re-arm it as required.
		1832
		1833	This method is more tricky, but usually most efficient: Most timeouts are
		1834	relatively long compared to the intervals between other activity - in
		1835	our example, within 60 seconds, there are usually many I/O events with
		1836	associated activity resets.
		1837
		1838	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1839	but remember the time of last activity, and check for a real timeout only
		1840	within the callback:
		1841
		1842	ev_tstamp last_activity; // time of last activity
		1843
		1844	static void
		1845	callback (EV_P_ ev_timer *w, int revents)
		1846	{
		1847	ev_tstamp now = ev_now (EV_A);
		1848	ev_tstamp timeout = last_activity + 60.;
		1849
		1850	// if last_activity + 60. is older than now, we did time out
		1851	if (timeout < now)
		1852	{
		1853	// timeout occurred, take action
		1854	}
		1855	else
		1856	{
		1857	// callback was invoked, but there was some activity, re-arm
		1858	// the watcher to fire in last_activity + 60, which is
		1859	// guaranteed to be in the future, so "again" is positive:
		1860	w->repeat = timeout - now;
		1861	ev_timer_again (EV_A_ w);
		1862	}
		1863	}
		1864
		1865	To summarise the callback: first calculate the real timeout (defined
		1866	as "60 seconds after the last activity"), then check if that time has
		1867	been reached, which means something I<did>, in fact, time out. Otherwise
		1868	the callback was invoked too early (C<timeout> is in the future), so
		1869	re-schedule the timer to fire at that future time, to see if maybe we have
		1870	a timeout then.
		1871
		1872	Note how C<ev_timer_again> is used, taking advantage of the
		1873	C<ev_timer_again> optimisation when the timer is already running.
		1874
		1875	This scheme causes more callback invocations (about one every 60 seconds
		1876	minus half the average time between activity), but virtually no calls to
		1877	libev to change the timeout.
		1878
		1879	To start the timer, simply initialise the watcher and set C<last_activity>
		1880	to the current time (meaning we just have some activity :), then call the
		1881	callback, which will "do the right thing" and start the timer:
		1882
		1883	ev_init (timer, callback);
		1884	last_activity = ev_now (loop);
		1885	callback (loop, timer, EV_TIMER);
		1886
		1887	And when there is some activity, simply store the current time in
		1888	C<last_activity>, no libev calls at all:
		1889
		1890	last_activity = ev_now (loop);
		1891
		1892	This technique is slightly more complex, but in most cases where the
		1893	time-out is unlikely to be triggered, much more efficient.
		1894
		1895	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1896	callback :) - just change the timeout and invoke the callback, which will
		1897	fix things for you.
		1898
		1899	=item 4. Wee, just use a double-linked list for your timeouts.
		1900
		1901	If there is not one request, but many thousands (millions...), all
		1902	employing some kind of timeout with the same timeout value, then one can
		1903	do even better:
		1904
		1905	When starting the timeout, calculate the timeout value and put the timeout
		1906	at the I<end> of the list.
		1907
		1908	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1909	the list is expected to fire (for example, using the technique #3).
		1910
		1911	When there is some activity, remove the timer from the list, recalculate
		1912	the timeout, append it to the end of the list again, and make sure to
		1913	update the C<ev_timer> if it was taken from the beginning of the list.
		1914
		1915	This way, one can manage an unlimited number of timeouts in O(1) time for
		1916	starting, stopping and updating the timers, at the expense of a major
		1917	complication, and having to use a constant timeout. The constant timeout
		1918	ensures that the list stays sorted.
		1919
		1920	=back
		1921
		1922	So which method the best?
		1923
		1924	Method #2 is a simple no-brain-required solution that is adequate in most
		1925	situations. Method #3 requires a bit more thinking, but handles many cases
		1926	better, and isn't very complicated either. In most case, choosing either
		1927	one is fine, with #3 being better in typical situations.
		1928
		1929	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1930	rather complicated, but extremely efficient, something that really pays
		1931	off after the first million or so of active timers, i.e. it's usually
		1932	overkill :)
		1933
		1934	=head3 The special problem of time updates
		1935
		1936	Establishing the current time is a costly operation (it usually takes at
		1937	least two system calls): EV therefore updates its idea of the current
		1938	time only before and after C<ev_run> collects new events, which causes a
		1939	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1940	lots of events in one iteration.
1160		1941
1161	The relative timeouts are calculated relative to the C<ev_now ()>	1942	The relative timeouts are calculated relative to the C<ev_now ()>
1162	time. This is usually the right thing as this timestamp refers to the time	1943	time. This is usually the right thing as this timestamp refers to the time
1163	of the event triggering whatever timeout you are modifying/starting. If	1944	of the event triggering whatever timeout you are modifying/starting. If
1164	you suspect event processing to be delayed and you I<need> to base the timeout	1945	you suspect event processing to be delayed and you I<need> to base the
1165	on the current time, use something like this to adjust for this:	1946	timeout on the current time, use something like this to adjust for this:
1166		1947
1167	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1948	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1168		1949
1169	The callback is guarenteed to be invoked only after its timeout has passed,	1950	If the event loop is suspended for a long time, you can also force an
1170	but if multiple timers become ready during the same loop iteration then	1951	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1171	order of execution is undefined.	1952	()>.
		1953
		1954	=head3 The special problems of suspended animation
		1955
		1956	When you leave the server world it is quite customary to hit machines that
		1957	can suspend/hibernate - what happens to the clocks during such a suspend?
		1958
		1959	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1960	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1961	to run until the system is suspended, but they will not advance while the
		1962	system is suspended. That means, on resume, it will be as if the program
		1963	was frozen for a few seconds, but the suspend time will not be counted
		1964	towards C<ev_timer> when a monotonic clock source is used. The real time
		1965	clock advanced as expected, but if it is used as sole clocksource, then a
		1966	long suspend would be detected as a time jump by libev, and timers would
		1967	be adjusted accordingly.
		1968
		1969	I would not be surprised to see different behaviour in different between
		1970	operating systems, OS versions or even different hardware.
		1971
		1972	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1973	time jump in the monotonic clocks and the realtime clock. If the program
		1974	is suspended for a very long time, and monotonic clock sources are in use,
		1975	then you can expect C<ev_timer>s to expire as the full suspension time
		1976	will be counted towards the timers. When no monotonic clock source is in
		1977	use, then libev will again assume a timejump and adjust accordingly.
		1978
		1979	It might be beneficial for this latter case to call C<ev_suspend>
		1980	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1981	deterministic behaviour in this case (you can do nothing against
		1982	C<SIGSTOP>).
1172		1983
1173	=head3 Watcher-Specific Functions and Data Members	1984	=head3 Watcher-Specific Functions and Data Members
1174		1985
1175	=over 4	1986	=over 4
1176		1987
…		…
1195	This will act as if the timer timed out and restart it again if it is	2006	This will act as if the timer timed out and restart it again if it is
1196	repeating. The exact semantics are:	2007	repeating. The exact semantics are:
1197		2008
1198	If the timer is pending, its pending status is cleared.	2009	If the timer is pending, its pending status is cleared.
1199		2010
1200	If the timer is started but nonrepeating, stop it (as if it timed out).	2011	If the timer is started but non-repeating, stop it (as if it timed out).
1201		2012
1202	If the timer is repeating, either start it if necessary (with the	2013	If the timer is repeating, either start it if necessary (with the
1203	C<repeat> value), or reset the running timer to the C<repeat> value.	2014	C<repeat> value), or reset the running timer to the C<repeat> value.
1204		2015
1205	This sounds a bit complicated, but here is a useful and typical	2016	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1206	example: Imagine you have a tcp connection and you want a so-called idle	2017	usage example.
1207	timeout, that is, you want to be called when there have been, say, 60
1208	seconds of inactivity on the socket. The easiest way to do this is to
1209	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1210	C<ev_timer_again> each time you successfully read or write some data. If
1211	you go into an idle state where you do not expect data to travel on the
1212	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1213	automatically restart it if need be.
1214		2018
1215	That means you can ignore the C<after> value and C<ev_timer_start>	2019	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1216	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1217		2020
1218	ev_timer_init (timer, callback, 0., 5.);	2021	Returns the remaining time until a timer fires. If the timer is active,
1219	ev_timer_again (loop, timer);	2022	then this time is relative to the current event loop time, otherwise it's
1220	...	2023	the timeout value currently configured.
1221	timer->again = 17.;
1222	ev_timer_again (loop, timer);
1223	...
1224	timer->again = 10.;
1225	ev_timer_again (loop, timer);
1226		2024
1227	This is more slightly efficient then stopping/starting the timer each time	2025	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1228	you want to modify its timeout value.	2026	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		2027	will return C<4>. When the timer expires and is restarted, it will return
		2028	roughly C<7> (likely slightly less as callback invocation takes some time,
		2029	too), and so on.
1229		2030
1230	=item ev_tstamp repeat [read-write]	2031	=item ev_tstamp repeat [read-write]
1231		2032
1232	The current C<repeat> value. Will be used each time the watcher times out	2033	The current C<repeat> value. Will be used each time the watcher times out
1233	or C<ev_timer_again> is called and determines the next timeout (if any),	2034	or C<ev_timer_again> is called, and determines the next timeout (if any),
1234	which is also when any modifications are taken into account.	2035	which is also when any modifications are taken into account.
1235		2036
1236	=back	2037	=back
1237		2038
1238	=head3 Examples	2039	=head3 Examples
1239		2040
1240	Example: Create a timer that fires after 60 seconds.	2041	Example: Create a timer that fires after 60 seconds.
1241		2042
1242	static void	2043	static void
1243	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2044	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1244	{	2045	{
1245	.. one minute over, w is actually stopped right here	2046	.. one minute over, w is actually stopped right here
1246	}	2047	}
1247		2048
1248	struct ev_timer mytimer;	2049	ev_timer mytimer;
1249	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	2050	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1250	ev_timer_start (loop, &mytimer);	2051	ev_timer_start (loop, &mytimer);
1251		2052
1252	Example: Create a timeout timer that times out after 10 seconds of	2053	Example: Create a timeout timer that times out after 10 seconds of
1253	inactivity.	2054	inactivity.
1254		2055
1255	static void	2056	static void
1256	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2057	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1257	{	2058	{
1258	.. ten seconds without any activity	2059	.. ten seconds without any activity
1259	}	2060	}
1260		2061
1261	struct ev_timer mytimer;	2062	ev_timer mytimer;
1262	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2063	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1263	ev_timer_again (&mytimer); /* start timer */	2064	ev_timer_again (&mytimer); /* start timer */
1264	ev_loop (loop, 0);	2065	ev_run (loop, 0);
1265		2066
1266	// and in some piece of code that gets executed on any "activity":	2067	// and in some piece of code that gets executed on any "activity":
1267	// reset the timeout to start ticking again at 10 seconds	2068	// reset the timeout to start ticking again at 10 seconds
1268	ev_timer_again (&mytimer);	2069	ev_timer_again (&mytimer);
1269		2070
1270		2071
1271	=head2 C<ev_periodic> - to cron or not to cron?	2072	=head2 C<ev_periodic> - to cron or not to cron?
1272		2073
1273	Periodic watchers are also timers of a kind, but they are very versatile	2074	Periodic watchers are also timers of a kind, but they are very versatile
1274	(and unfortunately a bit complex).	2075	(and unfortunately a bit complex).
1275		2076
1276	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	2077	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1277	but on wallclock time (absolute time). You can tell a periodic watcher	2078	relative time, the physical time that passes) but on wall clock time
1278	to trigger after some specific point in time. For example, if you tell a	2079	(absolute time, the thing you can read on your calender or clock). The
1279	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	2080	difference is that wall clock time can run faster or slower than real
1280	+ 10.>, that is, an absolute time not a delay) and then reset your system	2081	time, and time jumps are not uncommon (e.g. when you adjust your
1281	clock to january of the previous year, then it will take more than year	2082	wrist-watch).
1282	to trigger the event (unlike an C<ev_timer>, which would still trigger
1283	roughly 10 seconds later as it uses a relative timeout).
1284		2083
		2084	You can tell a periodic watcher to trigger after some specific point
		2085	in time: for example, if you tell a periodic watcher to trigger "in 10
		2086	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		2087	not a delay) and then reset your system clock to January of the previous
		2088	year, then it will take a year or more to trigger the event (unlike an
		2089	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		2090	it, as it uses a relative timeout).
		2091
1285	C<ev_periodic>s can also be used to implement vastly more complex timers,	2092	C<ev_periodic> watchers can also be used to implement vastly more complex
1286	such as triggering an event on each "midnight, local time", or other	2093	timers, such as triggering an event on each "midnight, local time", or
1287	complicated, rules.	2094	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		2095	those cannot react to time jumps.
1288		2096
1289	As with timers, the callback is guarenteed to be invoked only when the	2097	As with timers, the callback is guaranteed to be invoked only when the
1290	time (C<at>) has passed, but if multiple periodic timers become ready	2098	point in time where it is supposed to trigger has passed. If multiple
1291	during the same loop iteration then order of execution is undefined.	2099	timers become ready during the same loop iteration then the ones with
		2100	earlier time-out values are invoked before ones with later time-out values
		2101	(but this is no longer true when a callback calls C<ev_run> recursively).
1292		2102
1293	=head3 Watcher-Specific Functions and Data Members	2103	=head3 Watcher-Specific Functions and Data Members
1294		2104
1295	=over 4	2105	=over 4
1296		2106
1297	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	2107	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1298		2108
1299	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2109	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1300		2110
1301	Lots of arguments, lets sort it out... There are basically three modes of	2111	Lots of arguments, let's sort it out... There are basically three modes of
1302	operation, and we will explain them from simplest to complex:	2112	operation, and we will explain them from simplest to most complex:
1303		2113
1304	=over 4	2114	=over 4
1305		2115
1306	=item * absolute timer (at = time, interval = reschedule_cb = 0)	2116	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1307		2117
1308	In this configuration the watcher triggers an event after the wallclock	2118	In this configuration the watcher triggers an event after the wall clock
1309	time C<at> has passed and doesn't repeat. It will not adjust when a time	2119	time C<offset> has passed. It will not repeat and will not adjust when a
1310	jump occurs, that is, if it is to be run at January 1st 2011 then it will	2120	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1311	run when the system time reaches or surpasses this time.	2121	will be stopped and invoked when the system clock reaches or surpasses
		2122	this point in time.
1312		2123
1313	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	2124	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1314		2125
1315	In this mode the watcher will always be scheduled to time out at the next	2126	In this mode the watcher will always be scheduled to time out at the next
1316	C<at + N * interval> time (for some integer N, which can also be negative)	2127	C<offset + N * interval> time (for some integer N, which can also be
1317	and then repeat, regardless of any time jumps.	2128	negative) and then repeat, regardless of any time jumps. The C<offset>
		2129	argument is merely an offset into the C<interval> periods.
1318		2130
1319	This can be used to create timers that do not drift with respect to system	2131	This can be used to create timers that do not drift with respect to the
1320	time, for example, here is a C<ev_periodic> that triggers each hour, on	2132	system clock, for example, here is an C<ev_periodic> that triggers each
1321	the hour:	2133	hour, on the hour (with respect to UTC):
1322		2134
1323	ev_periodic_set (&periodic, 0., 3600., 0);	2135	ev_periodic_set (&periodic, 0., 3600., 0);
1324		2136
1325	This doesn't mean there will always be 3600 seconds in between triggers,	2137	This doesn't mean there will always be 3600 seconds in between triggers,
1326	but only that the the callback will be called when the system time shows a	2138	but only that the callback will be called when the system time shows a
1327	full hour (UTC), or more correctly, when the system time is evenly divisible	2139	full hour (UTC), or more correctly, when the system time is evenly divisible
1328	by 3600.	2140	by 3600.
1329		2141
1330	Another way to think about it (for the mathematically inclined) is that	2142	Another way to think about it (for the mathematically inclined) is that
1331	C<ev_periodic> will try to run the callback in this mode at the next possible	2143	C<ev_periodic> will try to run the callback in this mode at the next possible
1332	time where C<time = at (mod interval)>, regardless of any time jumps.	2144	time where C<time = offset (mod interval)>, regardless of any time jumps.
1333		2145
1334	For numerical stability it is preferable that the C<at> value is near	2146	For numerical stability it is preferable that the C<offset> value is near
1335	C<ev_now ()> (the current time), but there is no range requirement for	2147	C<ev_now ()> (the current time), but there is no range requirement for
1336	this value, and in fact is often specified as zero.	2148	this value, and in fact is often specified as zero.
1337		2149
		2150	Note also that there is an upper limit to how often a timer can fire (CPU
		2151	speed for example), so if C<interval> is very small then timing stability
		2152	will of course deteriorate. Libev itself tries to be exact to be about one
		2153	millisecond (if the OS supports it and the machine is fast enough).
		2154
1338	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2155	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1339		2156
1340	In this mode the values for C<interval> and C<at> are both being	2157	In this mode the values for C<interval> and C<offset> are both being
1341	ignored. Instead, each time the periodic watcher gets scheduled, the	2158	ignored. Instead, each time the periodic watcher gets scheduled, the
1342	reschedule callback will be called with the watcher as first, and the	2159	reschedule callback will be called with the watcher as first, and the
1343	current time as second argument.	2160	current time as second argument.
1344		2161
1345	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2162	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1346	ever, or make ANY event loop modifications whatsoever>.	2163	or make ANY other event loop modifications whatsoever, unless explicitly
		2164	allowed by documentation here>.
1347		2165
1348	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2166	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1349	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2167	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1350	only event loop modification you are allowed to do).	2168	only event loop modification you are allowed to do).
1351		2169
1352	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	2170	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1353	*w, ev_tstamp now)>, e.g.:	2171	*w, ev_tstamp now)>, e.g.:
1354		2172
		2173	static ev_tstamp
1355	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2174	my_rescheduler (ev_periodic *w, ev_tstamp now)
1356	{	2175	{
1357	return now + 60.;	2176	return now + 60.;
1358	}	2177	}
1359		2178
1360	It must return the next time to trigger, based on the passed time value	2179	It must return the next time to trigger, based on the passed time value
…		…
1380	a different time than the last time it was called (e.g. in a crond like	2199	a different time than the last time it was called (e.g. in a crond like
1381	program when the crontabs have changed).	2200	program when the crontabs have changed).
1382		2201
1383	=item ev_tstamp ev_periodic_at (ev_periodic *)	2202	=item ev_tstamp ev_periodic_at (ev_periodic *)
1384		2203
1385	When active, returns the absolute time that the watcher is supposed to	2204	When active, returns the absolute time that the watcher is supposed
1386	trigger next.	2205	to trigger next. This is not the same as the C<offset> argument to
		2206	C<ev_periodic_set>, but indeed works even in interval and manual
		2207	rescheduling modes.
1387		2208
1388	=item ev_tstamp offset [read-write]	2209	=item ev_tstamp offset [read-write]
1389		2210
1390	When repeating, this contains the offset value, otherwise this is the	2211	When repeating, this contains the offset value, otherwise this is the
1391	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2212	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2213	although libev might modify this value for better numerical stability).
1392		2214
1393	Can be modified any time, but changes only take effect when the periodic	2215	Can be modified any time, but changes only take effect when the periodic
1394	timer fires or C<ev_periodic_again> is being called.	2216	timer fires or C<ev_periodic_again> is being called.
1395		2217
1396	=item ev_tstamp interval [read-write]	2218	=item ev_tstamp interval [read-write]
1397		2219
1398	The current interval value. Can be modified any time, but changes only	2220	The current interval value. Can be modified any time, but changes only
1399	take effect when the periodic timer fires or C<ev_periodic_again> is being	2221	take effect when the periodic timer fires or C<ev_periodic_again> is being
1400	called.	2222	called.
1401		2223
1402	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2224	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1403		2225
1404	The current reschedule callback, or C<0>, if this functionality is	2226	The current reschedule callback, or C<0>, if this functionality is
1405	switched off. Can be changed any time, but changes only take effect when	2227	switched off. Can be changed any time, but changes only take effect when
1406	the periodic timer fires or C<ev_periodic_again> is being called.	2228	the periodic timer fires or C<ev_periodic_again> is being called.
1407		2229
1408	=back	2230	=back
1409		2231
1410	=head3 Examples	2232	=head3 Examples
1411		2233
1412	Example: Call a callback every hour, or, more precisely, whenever the	2234	Example: Call a callback every hour, or, more precisely, whenever the
1413	system clock is divisible by 3600. The callback invocation times have	2235	system time is divisible by 3600. The callback invocation times have
1414	potentially a lot of jittering, but good long-term stability.	2236	potentially a lot of jitter, but good long-term stability.
1415		2237
1416	static void	2238	static void
1417	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2239	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
1418	{	2240	{
1419	... its now a full hour (UTC, or TAI or whatever your clock follows)	2241	... its now a full hour (UTC, or TAI or whatever your clock follows)
1420	}	2242	}
1421		2243
1422	struct ev_periodic hourly_tick;	2244	ev_periodic hourly_tick;
1423	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2245	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1424	ev_periodic_start (loop, &hourly_tick);	2246	ev_periodic_start (loop, &hourly_tick);
1425		2247
1426	Example: The same as above, but use a reschedule callback to do it:	2248	Example: The same as above, but use a reschedule callback to do it:
1427		2249
1428	#include <math.h>	2250	#include <math.h>
1429		2251
1430	static ev_tstamp	2252	static ev_tstamp
1431	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2253	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1432	{	2254	{
1433	return fmod (now, 3600.) + 3600.;	2255	return now + (3600. - fmod (now, 3600.));
1434	}	2256	}
1435		2257
1436	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2258	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1437		2259
1438	Example: Call a callback every hour, starting now:	2260	Example: Call a callback every hour, starting now:
1439		2261
1440	struct ev_periodic hourly_tick;	2262	ev_periodic hourly_tick;
1441	ev_periodic_init (&hourly_tick, clock_cb,	2263	ev_periodic_init (&hourly_tick, clock_cb,
1442	fmod (ev_now (loop), 3600.), 3600., 0);	2264	fmod (ev_now (loop), 3600.), 3600., 0);
1443	ev_periodic_start (loop, &hourly_tick);	2265	ev_periodic_start (loop, &hourly_tick);
1444		2266
1445		2267
1446	=head2 C<ev_signal> - signal me when a signal gets signalled!	2268	=head2 C<ev_signal> - signal me when a signal gets signalled!
1447		2269
1448	Signal watchers will trigger an event when the process receives a specific	2270	Signal watchers will trigger an event when the process receives a specific
1449	signal one or more times. Even though signals are very asynchronous, libev	2271	signal one or more times. Even though signals are very asynchronous, libev
1450	will try it's best to deliver signals synchronously, i.e. as part of the	2272	will try its best to deliver signals synchronously, i.e. as part of the
1451	normal event processing, like any other event.	2273	normal event processing, like any other event.
1452		2274
		2275	If you want signals to be delivered truly asynchronously, just use
		2276	C<sigaction> as you would do without libev and forget about sharing
		2277	the signal. You can even use C<ev_async> from a signal handler to
		2278	synchronously wake up an event loop.
		2279
1453	You can configure as many watchers as you like per signal. Only when the	2280	You can configure as many watchers as you like for the same signal, but
		2281	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2282	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2283	C<SIGINT> in both the default loop and another loop at the same time. At
		2284	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2285
1454	first watcher gets started will libev actually register a signal watcher	2286	When the first watcher gets started will libev actually register something
1455	with the kernel (thus it coexists with your own signal handlers as long	2287	with the kernel (thus it coexists with your own signal handlers as long as
1456	as you don't register any with libev). Similarly, when the last signal	2288	you don't register any with libev for the same signal).
1457	watcher for a signal is stopped libev will reset the signal handler to
1458	SIG_DFL (regardless of what it was set to before).
1459		2289
1460	If possible and supported, libev will install its handlers with	2290	If possible and supported, libev will install its handlers with
1461	C<SA_RESTART> behaviour enabled, so syscalls should not be unduly	2291	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1462	interrupted. If you have a problem with syscalls getting interrupted by	2292	not be unduly interrupted. If you have a problem with system calls getting
1463	signals you can block all signals in an C<ev_check> watcher and unblock	2293	interrupted by signals you can block all signals in an C<ev_check> watcher
1464	them in an C<ev_prepare> watcher.	2294	and unblock them in an C<ev_prepare> watcher.
		2295
		2296	=head3 The special problem of inheritance over fork/execve/pthread_create
		2297
		2298	Both the signal mask (C<sigprocmask>) and the signal disposition
		2299	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2300	stopping it again), that is, libev might or might not block the signal,
		2301	and might or might not set or restore the installed signal handler.
		2302
		2303	While this does not matter for the signal disposition (libev never
		2304	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2305	C<execve>), this matters for the signal mask: many programs do not expect
		2306	certain signals to be blocked.
		2307
		2308	This means that before calling C<exec> (from the child) you should reset
		2309	the signal mask to whatever "default" you expect (all clear is a good
		2310	choice usually).
		2311
		2312	The simplest way to ensure that the signal mask is reset in the child is
		2313	to install a fork handler with C<pthread_atfork> that resets it. That will
		2314	catch fork calls done by libraries (such as the libc) as well.
		2315
		2316	In current versions of libev, the signal will not be blocked indefinitely
		2317	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2318	the window of opportunity for problems, it will not go away, as libev
		2319	I<has> to modify the signal mask, at least temporarily.
		2320
		2321	So I can't stress this enough: I<If you do not reset your signal mask when
		2322	you expect it to be empty, you have a race condition in your code>. This
		2323	is not a libev-specific thing, this is true for most event libraries.
1465		2324
1466	=head3 Watcher-Specific Functions and Data Members	2325	=head3 Watcher-Specific Functions and Data Members
1467		2326
1468	=over 4	2327	=over 4
1469		2328
…		…
1480		2339
1481	=back	2340	=back
1482		2341
1483	=head3 Examples	2342	=head3 Examples
1484		2343
1485	Example: Try to exit cleanly on SIGINT and SIGTERM.	2344	Example: Try to exit cleanly on SIGINT.
1486		2345
1487	static void	2346	static void
1488	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2347	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1489	{	2348	{
1490	ev_unloop (loop, EVUNLOOP_ALL);	2349	ev_break (loop, EVBREAK_ALL);
1491	}	2350	}
1492		2351
1493	struct ev_signal signal_watcher;	2352	ev_signal signal_watcher;
1494	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2353	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1495	ev_signal_start (loop, &sigint_cb);	2354	ev_signal_start (loop, &signal_watcher);
1496		2355
1497		2356
1498	=head2 C<ev_child> - watch out for process status changes	2357	=head2 C<ev_child> - watch out for process status changes
1499		2358
1500	Child watchers trigger when your process receives a SIGCHLD in response to	2359	Child watchers trigger when your process receives a SIGCHLD in response to
1501	some child status changes (most typically when a child of yours dies). It	2360	some child status changes (most typically when a child of yours dies or
1502	is permissible to install a child watcher I<after> the child has been	2361	exits). It is permissible to install a child watcher I<after> the child
1503	forked (which implies it might have already exited), as long as the event	2362	has been forked (which implies it might have already exited), as long
1504	loop isn't entered (or is continued from a watcher).	2363	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2364	forking and then immediately registering a watcher for the child is fine,
		2365	but forking and registering a watcher a few event loop iterations later or
		2366	in the next callback invocation is not.
1505		2367
1506	Only the default event loop is capable of handling signals, and therefore	2368	Only the default event loop is capable of handling signals, and therefore
1507	you can only rgeister child watchers in the default event loop.	2369	you can only register child watchers in the default event loop.
		2370
		2371	Due to some design glitches inside libev, child watchers will always be
		2372	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2373	libev)
1508		2374
1509	=head3 Process Interaction	2375	=head3 Process Interaction
1510		2376
1511	Libev grabs C<SIGCHLD> as soon as the default event loop is	2377	Libev grabs C<SIGCHLD> as soon as the default event loop is
1512	initialised. This is necessary to guarantee proper behaviour even if	2378	initialised. This is necessary to guarantee proper behaviour even if the
1513	the first child watcher is started after the child exits. The occurance	2379	first child watcher is started after the child exits. The occurrence
1514	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2380	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1515	synchronously as part of the event loop processing. Libev always reaps all	2381	synchronously as part of the event loop processing. Libev always reaps all
1516	children, even ones not watched.	2382	children, even ones not watched.
1517		2383
1518	=head3 Overriding the Built-In Processing	2384	=head3 Overriding the Built-In Processing
…		…
1522	handler, you can override it easily by installing your own handler for	2388	handler, you can override it easily by installing your own handler for
1523	C<SIGCHLD> after initialising the default loop, and making sure the	2389	C<SIGCHLD> after initialising the default loop, and making sure the
1524	default loop never gets destroyed. You are encouraged, however, to use an	2390	default loop never gets destroyed. You are encouraged, however, to use an
1525	event-based approach to child reaping and thus use libev's support for	2391	event-based approach to child reaping and thus use libev's support for
1526	that, so other libev users can use C<ev_child> watchers freely.	2392	that, so other libev users can use C<ev_child> watchers freely.
		2393
		2394	=head3 Stopping the Child Watcher
		2395
		2396	Currently, the child watcher never gets stopped, even when the
		2397	child terminates, so normally one needs to stop the watcher in the
		2398	callback. Future versions of libev might stop the watcher automatically
		2399	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2400	problem).
1527		2401
1528	=head3 Watcher-Specific Functions and Data Members	2402	=head3 Watcher-Specific Functions and Data Members
1529		2403
1530	=over 4	2404	=over 4
1531		2405
…		…
1560	=head3 Examples	2434	=head3 Examples
1561		2435
1562	Example: C<fork()> a new process and install a child handler to wait for	2436	Example: C<fork()> a new process and install a child handler to wait for
1563	its completion.	2437	its completion.
1564		2438
1565	ev_child cw;	2439	ev_child cw;
1566		2440
1567	static void	2441	static void
1568	child_cb (EV_P_ struct ev_child *w, int revents)	2442	child_cb (EV_P_ ev_child *w, int revents)
1569	{	2443	{
1570	ev_child_stop (EV_A_ w);	2444	ev_child_stop (EV_A_ w);
1571	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2445	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1572	}	2446	}
1573		2447
1574	pid_t pid = fork ();	2448	pid_t pid = fork ();
1575		2449
1576	if (pid < 0)	2450	if (pid < 0)
1577	// error	2451	// error
1578	else if (pid == 0)	2452	else if (pid == 0)
1579	{	2453	{
1580	// the forked child executes here	2454	// the forked child executes here
1581	exit (1);	2455	exit (1);
1582	}	2456	}
1583	else	2457	else
1584	{	2458	{
1585	ev_child_init (&cw, child_cb, pid, 0);	2459	ev_child_init (&cw, child_cb, pid, 0);
1586	ev_child_start (EV_DEFAULT_ &cw);	2460	ev_child_start (EV_DEFAULT_ &cw);
1587	}	2461	}
1588		2462
1589		2463
1590	=head2 C<ev_stat> - did the file attributes just change?	2464	=head2 C<ev_stat> - did the file attributes just change?
1591		2465
1592	This watches a filesystem path for attribute changes. That is, it calls	2466	This watches a file system path for attribute changes. That is, it calls
1593	C<stat> regularly (or when the OS says it changed) and sees if it changed	2467	C<stat> on that path in regular intervals (or when the OS says it changed)
1594	compared to the last time, invoking the callback if it did.	2468	and sees if it changed compared to the last time, invoking the callback if
		2469	it did.
1595		2470
1596	The path does not need to exist: changing from "path exists" to "path does	2471	The path does not need to exist: changing from "path exists" to "path does
1597	not exist" is a status change like any other. The condition "path does	2472	not exist" is a status change like any other. The condition "path does not
1598	not exist" is signified by the C<st_nlink> field being zero (which is	2473	exist" (or more correctly "path cannot be stat'ed") is signified by the
1599	otherwise always forced to be at least one) and all the other fields of	2474	C<st_nlink> field being zero (which is otherwise always forced to be at
1600	the stat buffer having unspecified contents.	2475	least one) and all the other fields of the stat buffer having unspecified
		2476	contents.
1601		2477
1602	The path I<should> be absolute and I<must not> end in a slash. If it is	2478	The path I<must not> end in a slash or contain special components such as
		2479	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1603	relative and your working directory changes, the behaviour is undefined.	2480	your working directory changes, then the behaviour is undefined.
1604		2481
1605	Since there is no standard to do this, the portable implementation simply	2482	Since there is no portable change notification interface available, the
1606	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2483	portable implementation simply calls C<stat(2)> regularly on the path
1607	can specify a recommended polling interval for this case. If you specify	2484	to see if it changed somehow. You can specify a recommended polling
1608	a polling interval of C<0> (highly recommended!) then a I<suitable,	2485	interval for this case. If you specify a polling interval of C<0> (highly
1609	unspecified default> value will be used (which you can expect to be around	2486	recommended!) then a I<suitable, unspecified default> value will be used
1610	five seconds, although this might change dynamically). Libev will also	2487	(which you can expect to be around five seconds, although this might
1611	impose a minimum interval which is currently around C<0.1>, but thats	2488	change dynamically). Libev will also impose a minimum interval which is
1612	usually overkill.	2489	currently around C<0.1>, but that's usually overkill.
1613		2490
1614	This watcher type is not meant for massive numbers of stat watchers,	2491	This watcher type is not meant for massive numbers of stat watchers,
1615	as even with OS-supported change notifications, this can be	2492	as even with OS-supported change notifications, this can be
1616	resource-intensive.	2493	resource-intensive.
1617		2494
1618	At the time of this writing, only the Linux inotify interface is	2495	At the time of this writing, the only OS-specific interface implemented
1619	implemented (implementing kqueue support is left as an exercise for the	2496	is the Linux inotify interface (implementing kqueue support is left as an
1620	reader, note, however, that the author sees no way of implementing ev_stat	2497	exercise for the reader. Note, however, that the author sees no way of
1621	semantics with kqueue). Inotify will be used to give hints only and should	2498	implementing C<ev_stat> semantics with kqueue, except as a hint).
1622	not change the semantics of C<ev_stat> watchers, which means that libev
1623	sometimes needs to fall back to regular polling again even with inotify,
1624	but changes are usually detected immediately, and if the file exists there
1625	will be no polling.
1626		2499
1627	=head3 ABI Issues (Largefile Support)	2500	=head3 ABI Issues (Largefile Support)
1628		2501
1629	Libev by default (unless the user overrides this) uses the default	2502	Libev by default (unless the user overrides this) uses the default
1630	compilation environment, which means that on systems with optionally	2503	compilation environment, which means that on systems with large file
1631	disabled large file support, you get the 32 bit version of the stat	2504	support disabled by default, you get the 32 bit version of the stat
1632	structure. When using the library from programs that change the ABI to	2505	structure. When using the library from programs that change the ABI to
1633	use 64 bit file offsets the programs will fail. In that case you have to	2506	use 64 bit file offsets the programs will fail. In that case you have to
1634	compile libev with the same flags to get binary compatibility. This is	2507	compile libev with the same flags to get binary compatibility. This is
1635	obviously the case with any flags that change the ABI, but the problem is	2508	obviously the case with any flags that change the ABI, but the problem is
1636	most noticably with ev_stat and largefile support.	2509	most noticeably displayed with ev_stat and large file support.
1637		2510
1638	=head3 Inotify	2511	The solution for this is to lobby your distribution maker to make large
		2512	file interfaces available by default (as e.g. FreeBSD does) and not
		2513	optional. Libev cannot simply switch on large file support because it has
		2514	to exchange stat structures with application programs compiled using the
		2515	default compilation environment.
1639		2516
		2517	=head3 Inotify and Kqueue
		2518
1640	When C<inotify (7)> support has been compiled into libev (generally only	2519	When C<inotify (7)> support has been compiled into libev and present at
1641	available on Linux) and present at runtime, it will be used to speed up	2520	runtime, it will be used to speed up change detection where possible. The
1642	change detection where possible. The inotify descriptor will be created lazily	2521	inotify descriptor will be created lazily when the first C<ev_stat>
1643	when the first C<ev_stat> watcher is being started.	2522	watcher is being started.
1644		2523
1645	Inotify presence does not change the semantics of C<ev_stat> watchers	2524	Inotify presence does not change the semantics of C<ev_stat> watchers
1646	except that changes might be detected earlier, and in some cases, to avoid	2525	except that changes might be detected earlier, and in some cases, to avoid
1647	making regular C<stat> calls. Even in the presence of inotify support	2526	making regular C<stat> calls. Even in the presence of inotify support
1648	there are many cases where libev has to resort to regular C<stat> polling.	2527	there are many cases where libev has to resort to regular C<stat> polling,
		2528	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2529	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2530	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2531	xfs are fully working) libev usually gets away without polling.
1649		2532
1650	(There is no support for kqueue, as apparently it cannot be used to	2533	There is no support for kqueue, as apparently it cannot be used to
1651	implement this functionality, due to the requirement of having a file	2534	implement this functionality, due to the requirement of having a file
1652	descriptor open on the object at all times).	2535	descriptor open on the object at all times, and detecting renames, unlinks
		2536	etc. is difficult.
		2537
		2538	=head3 C<stat ()> is a synchronous operation
		2539
		2540	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2541	the process. The exception are C<ev_stat> watchers - those call C<stat
		2542	()>, which is a synchronous operation.
		2543
		2544	For local paths, this usually doesn't matter: unless the system is very
		2545	busy or the intervals between stat's are large, a stat call will be fast,
		2546	as the path data is usually in memory already (except when starting the
		2547	watcher).
		2548
		2549	For networked file systems, calling C<stat ()> can block an indefinite
		2550	time due to network issues, and even under good conditions, a stat call
		2551	often takes multiple milliseconds.
		2552
		2553	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2554	paths, although this is fully supported by libev.
1653		2555
1654	=head3 The special problem of stat time resolution	2556	=head3 The special problem of stat time resolution
1655		2557
1656	The C<stat ()> syscall only supports full-second resolution portably, and	2558	The C<stat ()> system call only supports full-second resolution portably,
1657	even on systems where the resolution is higher, many filesystems still	2559	and even on systems where the resolution is higher, most file systems
1658	only support whole seconds.	2560	still only support whole seconds.
1659		2561
1660	That means that, if the time is the only thing that changes, you can	2562	That means that, if the time is the only thing that changes, you can
1661	easily miss updates: on the first update, C<ev_stat> detects a change and	2563	easily miss updates: on the first update, C<ev_stat> detects a change and
1662	calls your callback, which does something. When there is another update	2564	calls your callback, which does something. When there is another update
1663	within the same second, C<ev_stat> will be unable to detect it as the stat	2565	within the same second, C<ev_stat> will be unable to detect unless the
1664	data does not change.	2566	stat data does change in other ways (e.g. file size).
1665		2567
1666	The solution to this is to delay acting on a change for slightly more	2568	The solution to this is to delay acting on a change for slightly more
1667	than a second (or till slightly after the next full second boundary), using	2569	than a second (or till slightly after the next full second boundary), using
1668	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);	2570	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1669	ev_timer_again (loop, w)>).	2571	ev_timer_again (loop, w)>).
…		…
1689	C<path>. The C<interval> is a hint on how quickly a change is expected to	2591	C<path>. The C<interval> is a hint on how quickly a change is expected to
1690	be detected and should normally be specified as C<0> to let libev choose	2592	be detected and should normally be specified as C<0> to let libev choose
1691	a suitable value. The memory pointed to by C<path> must point to the same	2593	a suitable value. The memory pointed to by C<path> must point to the same
1692	path for as long as the watcher is active.	2594	path for as long as the watcher is active.
1693		2595
1694	The callback will receive C<EV_STAT> when a change was detected, relative	2596	The callback will receive an C<EV_STAT> event when a change was detected,
1695	to the attributes at the time the watcher was started (or the last change	2597	relative to the attributes at the time the watcher was started (or the
1696	was detected).	2598	last change was detected).
1697		2599
1698	=item ev_stat_stat (loop, ev_stat *)	2600	=item ev_stat_stat (loop, ev_stat *)
1699		2601
1700	Updates the stat buffer immediately with new values. If you change the	2602	Updates the stat buffer immediately with new values. If you change the
1701	watched path in your callback, you could call this function to avoid	2603	watched path in your callback, you could call this function to avoid
…		…
1722		2624
1723	The specified interval.	2625	The specified interval.
1724		2626
1725	=item const char *path [read-only]	2627	=item const char *path [read-only]
1726		2628
1727	The filesystem path that is being watched.	2629	The file system path that is being watched.
1728		2630
1729	=back	2631	=back
1730		2632
1731	=head3 Examples	2633	=head3 Examples
1732		2634
1733	Example: Watch C</etc/passwd> for attribute changes.	2635	Example: Watch C</etc/passwd> for attribute changes.
1734		2636
1735	static void	2637	static void
1736	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	2638	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1737	{	2639	{
1738	/* /etc/passwd changed in some way */	2640	/* /etc/passwd changed in some way */
1739	if (w->attr.st_nlink)	2641	if (w->attr.st_nlink)
1740	{	2642	{
1741	printf ("passwd current size %ld\n", (long)w->attr.st_size);	2643	printf ("passwd current size %ld\n", (long)w->attr.st_size);
1742	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);	2644	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
1743	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);	2645	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
1744	}	2646	}
1745	else	2647	else
1746	/* you shalt not abuse printf for puts */	2648	/* you shalt not abuse printf for puts */
1747	puts ("wow, /etc/passwd is not there, expect problems. "	2649	puts ("wow, /etc/passwd is not there, expect problems. "
1748	"if this is windows, they already arrived\n");	2650	"if this is windows, they already arrived\n");
1749	}	2651	}
1750		2652
1751	...	2653	...
1752	ev_stat passwd;	2654	ev_stat passwd;
1753		2655
1754	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);	2656	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1755	ev_stat_start (loop, &passwd);	2657	ev_stat_start (loop, &passwd);
1756		2658
1757	Example: Like above, but additionally use a one-second delay so we do not	2659	Example: Like above, but additionally use a one-second delay so we do not
1758	miss updates (however, frequent updates will delay processing, too, so	2660	miss updates (however, frequent updates will delay processing, too, so
1759	one might do the work both on C<ev_stat> callback invocation I<and> on	2661	one might do the work both on C<ev_stat> callback invocation I<and> on
1760	C<ev_timer> callback invocation).	2662	C<ev_timer> callback invocation).
1761		2663
1762	static ev_stat passwd;	2664	static ev_stat passwd;
1763	static ev_timer timer;	2665	static ev_timer timer;
1764		2666
1765	static void	2667	static void
1766	timer_cb (EV_P_ ev_timer *w, int revents)	2668	timer_cb (EV_P_ ev_timer *w, int revents)
1767	{	2669	{
1768	ev_timer_stop (EV_A_ w);	2670	ev_timer_stop (EV_A_ w);
1769		2671
1770	/* now it's one second after the most recent passwd change */	2672	/* now it's one second after the most recent passwd change */
1771	}	2673	}
1772		2674
1773	static void	2675	static void
1774	stat_cb (EV_P_ ev_stat *w, int revents)	2676	stat_cb (EV_P_ ev_stat *w, int revents)
1775	{	2677	{
1776	/* reset the one-second timer */	2678	/* reset the one-second timer */
1777	ev_timer_again (EV_A_ &timer);	2679	ev_timer_again (EV_A_ &timer);
1778	}	2680	}
1779		2681
1780	...	2682	...
1781	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	2683	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1782	ev_stat_start (loop, &passwd);	2684	ev_stat_start (loop, &passwd);
1783	ev_timer_init (&timer, timer_cb, 0., 1.02);	2685	ev_timer_init (&timer, timer_cb, 0., 1.02);
1784		2686
1785		2687
1786	=head2 C<ev_idle> - when you've got nothing better to do...	2688	=head2 C<ev_idle> - when you've got nothing better to do...
1787		2689
1788	Idle watchers trigger events when no other events of the same or higher	2690	Idle watchers trigger events when no other events of the same or higher
1789	priority are pending (prepare, check and other idle watchers do not	2691	priority are pending (prepare, check and other idle watchers do not count
1790	count).	2692	as receiving "events").
1791		2693
1792	That is, as long as your process is busy handling sockets or timeouts	2694	That is, as long as your process is busy handling sockets or timeouts
1793	(or even signals, imagine) of the same or higher priority it will not be	2695	(or even signals, imagine) of the same or higher priority it will not be
1794	triggered. But when your process is idle (or only lower-priority watchers	2696	triggered. But when your process is idle (or only lower-priority watchers
1795	are pending), the idle watchers are being called once per event loop	2697	are pending), the idle watchers are being called once per event loop
…		…
1806		2708
1807	=head3 Watcher-Specific Functions and Data Members	2709	=head3 Watcher-Specific Functions and Data Members
1808		2710
1809	=over 4	2711	=over 4
1810		2712
1811	=item ev_idle_init (ev_signal *, callback)	2713	=item ev_idle_init (ev_idle *, callback)
1812		2714
1813	Initialises and configures the idle watcher - it has no parameters of any	2715	Initialises and configures the idle watcher - it has no parameters of any
1814	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2716	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1815	believe me.	2717	believe me.
1816		2718
…		…
1819	=head3 Examples	2721	=head3 Examples
1820		2722
1821	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2723	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1822	callback, free it. Also, use no error checking, as usual.	2724	callback, free it. Also, use no error checking, as usual.
1823		2725
1824	static void	2726	static void
1825	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2727	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1826	{	2728	{
1827	free (w);	2729	free (w);
1828	// now do something you wanted to do when the program has	2730	// now do something you wanted to do when the program has
1829	// no longer anything immediate to do.	2731	// no longer anything immediate to do.
1830	}	2732	}
1831		2733
1832	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2734	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1833	ev_idle_init (idle_watcher, idle_cb);	2735	ev_idle_init (idle_watcher, idle_cb);
1834	ev_idle_start (loop, idle_cb);	2736	ev_idle_start (loop, idle_watcher);
1835		2737
1836		2738
1837	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2739	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1838		2740
1839	Prepare and check watchers are usually (but not always) used in tandem:	2741	Prepare and check watchers are usually (but not always) used in pairs:
1840	prepare watchers get invoked before the process blocks and check watchers	2742	prepare watchers get invoked before the process blocks and check watchers
1841	afterwards.	2743	afterwards.
1842		2744
1843	You I<must not> call C<ev_loop> or similar functions that enter	2745	You I<must not> call C<ev_run> or similar functions that enter
1844	the current event loop from either C<ev_prepare> or C<ev_check>	2746	the current event loop from either C<ev_prepare> or C<ev_check>
1845	watchers. Other loops than the current one are fine, however. The	2747	watchers. Other loops than the current one are fine, however. The
1846	rationale behind this is that you do not need to check for recursion in	2748	rationale behind this is that you do not need to check for recursion in
1847	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2749	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1848	C<ev_check> so if you have one watcher of each kind they will always be	2750	C<ev_check> so if you have one watcher of each kind they will always be
1849	called in pairs bracketing the blocking call.	2751	called in pairs bracketing the blocking call.
1850		2752
1851	Their main purpose is to integrate other event mechanisms into libev and	2753	Their main purpose is to integrate other event mechanisms into libev and
1852	their use is somewhat advanced. This could be used, for example, to track	2754	their use is somewhat advanced. They could be used, for example, to track
1853	variable changes, implement your own watchers, integrate net-snmp or a	2755	variable changes, implement your own watchers, integrate net-snmp or a
1854	coroutine library and lots more. They are also occasionally useful if	2756	coroutine library and lots more. They are also occasionally useful if
1855	you cache some data and want to flush it before blocking (for example,	2757	you cache some data and want to flush it before blocking (for example,
1856	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2758	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1857	watcher).	2759	watcher).
1858		2760
1859	This is done by examining in each prepare call which file descriptors need	2761	This is done by examining in each prepare call which file descriptors
1860	to be watched by the other library, registering C<ev_io> watchers for	2762	need to be watched by the other library, registering C<ev_io> watchers
1861	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2763	for them and starting an C<ev_timer> watcher for any timeouts (many
1862	provide just this functionality). Then, in the check watcher you check for	2764	libraries provide exactly this functionality). Then, in the check watcher,
1863	any events that occured (by checking the pending status of all watchers	2765	you check for any events that occurred (by checking the pending status
1864	and stopping them) and call back into the library. The I/O and timer	2766	of all watchers and stopping them) and call back into the library. The
1865	callbacks will never actually be called (but must be valid nevertheless,	2767	I/O and timer callbacks will never actually be called (but must be valid
1866	because you never know, you know?).	2768	nevertheless, because you never know, you know?).
1867		2769
1868	As another example, the Perl Coro module uses these hooks to integrate	2770	As another example, the Perl Coro module uses these hooks to integrate
1869	coroutines into libev programs, by yielding to other active coroutines	2771	coroutines into libev programs, by yielding to other active coroutines
1870	during each prepare and only letting the process block if no coroutines	2772	during each prepare and only letting the process block if no coroutines
1871	are ready to run (it's actually more complicated: it only runs coroutines	2773	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1874	loop from blocking if lower-priority coroutines are active, thus mapping	2776	loop from blocking if lower-priority coroutines are active, thus mapping
1875	low-priority coroutines to idle/background tasks).	2777	low-priority coroutines to idle/background tasks).
1876		2778
1877	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2779	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1878	priority, to ensure that they are being run before any other watchers	2780	priority, to ensure that they are being run before any other watchers
		2781	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2782
1879	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2783	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1880	too) should not activate ("feed") events into libev. While libev fully	2784	activate ("feed") events into libev. While libev fully supports this, they
1881	supports this, they might get executed before other C<ev_check> watchers	2785	might get executed before other C<ev_check> watchers did their job. As
1882	did their job. As C<ev_check> watchers are often used to embed other	2786	C<ev_check> watchers are often used to embed other (non-libev) event
1883	(non-libev) event loops those other event loops might be in an unusable	2787	loops those other event loops might be in an unusable state until their
1884	state until their C<ev_check> watcher ran (always remind yourself to	2788	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1885	coexist peacefully with others).	2789	others).
1886		2790
1887	=head3 Watcher-Specific Functions and Data Members	2791	=head3 Watcher-Specific Functions and Data Members
1888		2792
1889	=over 4	2793	=over 4
1890		2794
…		…
1892		2796
1893	=item ev_check_init (ev_check *, callback)	2797	=item ev_check_init (ev_check *, callback)
1894		2798
1895	Initialises and configures the prepare or check watcher - they have no	2799	Initialises and configures the prepare or check watcher - they have no
1896	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2800	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1897	macros, but using them is utterly, utterly and completely pointless.	2801	macros, but using them is utterly, utterly, utterly and completely
		2802	pointless.
1898		2803
1899	=back	2804	=back
1900		2805
1901	=head3 Examples	2806	=head3 Examples
1902		2807
…		…
1911	and in a check watcher, destroy them and call into libadns. What follows	2816	and in a check watcher, destroy them and call into libadns. What follows
1912	is pseudo-code only of course. This requires you to either use a low	2817	is pseudo-code only of course. This requires you to either use a low
1913	priority for the check watcher or use C<ev_clear_pending> explicitly, as	2818	priority for the check watcher or use C<ev_clear_pending> explicitly, as
1914	the callbacks for the IO/timeout watchers might not have been called yet.	2819	the callbacks for the IO/timeout watchers might not have been called yet.
1915		2820
1916	static ev_io iow [nfd];	2821	static ev_io iow [nfd];
1917	static ev_timer tw;	2822	static ev_timer tw;
1918		2823
1919	static void	2824	static void
1920	io_cb (ev_loop *loop, ev_io *w, int revents)	2825	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1921	{	2826	{
1922	}	2827	}
1923		2828
1924	// create io watchers for each fd and a timer before blocking	2829	// create io watchers for each fd and a timer before blocking
1925	static void	2830	static void
1926	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2831	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1927	{	2832	{
1928	int timeout = 3600000;	2833	int timeout = 3600000;
1929	struct pollfd fds [nfd];	2834	struct pollfd fds [nfd];
1930	// actual code will need to loop here and realloc etc.	2835	// actual code will need to loop here and realloc etc.
1931	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2836	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1932		2837
1933	/* the callback is illegal, but won't be called as we stop during check */	2838	/* the callback is illegal, but won't be called as we stop during check */
1934	ev_timer_init (&tw, 0, timeout * 1e-3);	2839	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
1935	ev_timer_start (loop, &tw);	2840	ev_timer_start (loop, &tw);
1936		2841
1937	// create one ev_io per pollfd	2842	// create one ev_io per pollfd
1938	for (int i = 0; i < nfd; ++i)	2843	for (int i = 0; i < nfd; ++i)
1939	{	2844	{
1940	ev_io_init (iow + i, io_cb, fds [i].fd,	2845	ev_io_init (iow + i, io_cb, fds [i].fd,
1941	((fds [i].events & POLLIN ? EV_READ : 0)	2846	((fds [i].events & POLLIN ? EV_READ : 0)
1942	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	2847	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1943		2848
1944	fds [i].revents = 0;	2849	fds [i].revents = 0;
1945	ev_io_start (loop, iow + i);	2850	ev_io_start (loop, iow + i);
1946	}	2851	}
1947	}	2852	}
1948		2853
1949	// stop all watchers after blocking	2854	// stop all watchers after blocking
1950	static void	2855	static void
1951	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2856	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
1952	{	2857	{
1953	ev_timer_stop (loop, &tw);	2858	ev_timer_stop (loop, &tw);
1954		2859
1955	for (int i = 0; i < nfd; ++i)	2860	for (int i = 0; i < nfd; ++i)
1956	{	2861	{
1957	// set the relevant poll flags	2862	// set the relevant poll flags
1958	// could also call adns_processreadable etc. here	2863	// could also call adns_processreadable etc. here
1959	struct pollfd *fd = fds + i;	2864	struct pollfd *fd = fds + i;
1960	int revents = ev_clear_pending (iow + i);	2865	int revents = ev_clear_pending (iow + i);
1961	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;	2866	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1962	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;	2867	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1963		2868
1964	// now stop the watcher	2869	// now stop the watcher
1965	ev_io_stop (loop, iow + i);	2870	ev_io_stop (loop, iow + i);
1966	}	2871	}
1967		2872
1968	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	2873	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1969	}	2874	}
1970		2875
1971	Method 2: This would be just like method 1, but you run C<adns_afterpoll>	2876	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
1972	in the prepare watcher and would dispose of the check watcher.	2877	in the prepare watcher and would dispose of the check watcher.
1973		2878
1974	Method 3: If the module to be embedded supports explicit event	2879	Method 3: If the module to be embedded supports explicit event
1975	notification (adns does), you can also make use of the actual watcher	2880	notification (libadns does), you can also make use of the actual watcher
1976	callbacks, and only destroy/create the watchers in the prepare watcher.	2881	callbacks, and only destroy/create the watchers in the prepare watcher.
1977		2882
1978	static void	2883	static void
1979	timer_cb (EV_P_ ev_timer *w, int revents)	2884	timer_cb (EV_P_ ev_timer *w, int revents)
1980	{	2885	{
1981	adns_state ads = (adns_state)w->data;	2886	adns_state ads = (adns_state)w->data;
1982	update_now (EV_A);	2887	update_now (EV_A);
1983		2888
1984	adns_processtimeouts (ads, &tv_now);	2889	adns_processtimeouts (ads, &tv_now);
1985	}	2890	}
1986		2891
1987	static void	2892	static void
1988	io_cb (EV_P_ ev_io *w, int revents)	2893	io_cb (EV_P_ ev_io *w, int revents)
1989	{	2894	{
1990	adns_state ads = (adns_state)w->data;	2895	adns_state ads = (adns_state)w->data;
1991	update_now (EV_A);	2896	update_now (EV_A);
1992		2897
1993	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);	2898	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
1994	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);	2899	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
1995	}	2900	}
1996		2901
1997	// do not ever call adns_afterpoll	2902	// do not ever call adns_afterpoll
1998		2903
1999	Method 4: Do not use a prepare or check watcher because the module you	2904	Method 4: Do not use a prepare or check watcher because the module you
2000	want to embed is too inflexible to support it. Instead, youc na override	2905	want to embed is not flexible enough to support it. Instead, you can
2001	their poll function. The drawback with this solution is that the main	2906	override their poll function. The drawback with this solution is that the
2002	loop is now no longer controllable by EV. The C<Glib::EV> module does	2907	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
2003	this.	2908	this approach, effectively embedding EV as a client into the horrible
		2909	libglib event loop.
2004		2910
2005	static gint	2911	static gint
2006	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2912	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
2007	{	2913	{
2008	int got_events = 0;	2914	int got_events = 0;
2009		2915
2010	for (n = 0; n < nfds; ++n)	2916	for (n = 0; n < nfds; ++n)
2011	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events	2917	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
2012		2918
2013	if (timeout >= 0)	2919	if (timeout >= 0)
2014	// create/start timer	2920	// create/start timer
2015		2921
2016	// poll	2922	// poll
2017	ev_loop (EV_A_ 0);	2923	ev_run (EV_A_ 0);
2018		2924
2019	// stop timer again	2925	// stop timer again
2020	if (timeout >= 0)	2926	if (timeout >= 0)
2021	ev_timer_stop (EV_A_ &to);	2927	ev_timer_stop (EV_A_ &to);
2022		2928
2023	// stop io watchers again - their callbacks should have set	2929	// stop io watchers again - their callbacks should have set
2024	for (n = 0; n < nfds; ++n)	2930	for (n = 0; n < nfds; ++n)
2025	ev_io_stop (EV_A_ iow [n]);	2931	ev_io_stop (EV_A_ iow [n]);
2026		2932
2027	return got_events;	2933	return got_events;
2028	}	2934	}
2029		2935
2030		2936
2031	=head2 C<ev_embed> - when one backend isn't enough...	2937	=head2 C<ev_embed> - when one backend isn't enough...
2032		2938
2033	This is a rather advanced watcher type that lets you embed one event loop	2939	This is a rather advanced watcher type that lets you embed one event loop
…		…
2039	prioritise I/O.	2945	prioritise I/O.
2040		2946
2041	As an example for a bug workaround, the kqueue backend might only support	2947	As an example for a bug workaround, the kqueue backend might only support
2042	sockets on some platform, so it is unusable as generic backend, but you	2948	sockets on some platform, so it is unusable as generic backend, but you
2043	still want to make use of it because you have many sockets and it scales	2949	still want to make use of it because you have many sockets and it scales
2044	so nicely. In this case, you would create a kqueue-based loop and embed it	2950	so nicely. In this case, you would create a kqueue-based loop and embed
2045	into your default loop (which might use e.g. poll). Overall operation will	2951	it into your default loop (which might use e.g. poll). Overall operation
2046	be a bit slower because first libev has to poll and then call kevent, but	2952	will be a bit slower because first libev has to call C<poll> and then
2047	at least you can use both at what they are best.	2953	C<kevent>, but at least you can use both mechanisms for what they are
		2954	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
2048		2955
2049	As for prioritising I/O: rarely you have the case where some fds have	2956	As for prioritising I/O: under rare circumstances you have the case where
2050	to be watched and handled very quickly (with low latency), and even	2957	some fds have to be watched and handled very quickly (with low latency),
2051	priorities and idle watchers might have too much overhead. In this case	2958	and even priorities and idle watchers might have too much overhead. In
2052	you would put all the high priority stuff in one loop and all the rest in	2959	this case you would put all the high priority stuff in one loop and all
2053	a second one, and embed the second one in the first.	2960	the rest in a second one, and embed the second one in the first.
2054		2961
2055	As long as the watcher is active, the callback will be invoked every time	2962	As long as the watcher is active, the callback will be invoked every
2056	there might be events pending in the embedded loop. The callback must then	2963	time there might be events pending in the embedded loop. The callback
2057	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2964	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2058	their callbacks (you could also start an idle watcher to give the embedded	2965	sweep and invoke their callbacks (the callback doesn't need to invoke the
2059	loop strictly lower priority for example). You can also set the callback	2966	C<ev_embed_sweep> function directly, it could also start an idle watcher
2060	to C<0>, in which case the embed watcher will automatically execute the	2967	to give the embedded loop strictly lower priority for example).
2061	embedded loop sweep.
2062		2968
2063	As long as the watcher is started it will automatically handle events. The	2969	You can also set the callback to C<0>, in which case the embed watcher
2064	callback will be invoked whenever some events have been handled. You can	2970	will automatically execute the embedded loop sweep whenever necessary.
2065	set the callback to C<0> to avoid having to specify one if you are not
2066	interested in that.
2067		2971
2068	Also, there have not currently been made special provisions for forking:	2972	Fork detection will be handled transparently while the C<ev_embed> watcher
2069	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2973	is active, i.e., the embedded loop will automatically be forked when the
2070	but you will also have to stop and restart any C<ev_embed> watchers	2974	embedding loop forks. In other cases, the user is responsible for calling
2071	yourself.	2975	C<ev_loop_fork> on the embedded loop.
2072		2976
2073	Unfortunately, not all backends are embeddable, only the ones returned by	2977	Unfortunately, not all backends are embeddable: only the ones returned by
2074	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2978	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2075	portable one.	2979	portable one.
2076		2980
2077	So when you want to use this feature you will always have to be prepared	2981	So when you want to use this feature you will always have to be prepared
2078	that you cannot get an embeddable loop. The recommended way to get around	2982	that you cannot get an embeddable loop. The recommended way to get around
2079	this is to have a separate variables for your embeddable loop, try to	2983	this is to have a separate variables for your embeddable loop, try to
2080	create it, and if that fails, use the normal loop for everything.	2984	create it, and if that fails, use the normal loop for everything.
2081		2985
		2986	=head3 C<ev_embed> and fork
		2987
		2988	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2989	automatically be applied to the embedded loop as well, so no special
		2990	fork handling is required in that case. When the watcher is not running,
		2991	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2992	as applicable.
		2993
2082	=head3 Watcher-Specific Functions and Data Members	2994	=head3 Watcher-Specific Functions and Data Members
2083		2995
2084	=over 4	2996	=over 4
2085		2997
2086	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2998	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
…		…
2089		3001
2090	Configures the watcher to embed the given loop, which must be	3002	Configures the watcher to embed the given loop, which must be
2091	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3003	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
2092	invoked automatically, otherwise it is the responsibility of the callback	3004	invoked automatically, otherwise it is the responsibility of the callback
2093	to invoke it (it will continue to be called until the sweep has been done,	3005	to invoke it (it will continue to be called until the sweep has been done,
2094	if you do not want thta, you need to temporarily stop the embed watcher).	3006	if you do not want that, you need to temporarily stop the embed watcher).
2095		3007
2096	=item ev_embed_sweep (loop, ev_embed *)	3008	=item ev_embed_sweep (loop, ev_embed *)
2097		3009
2098	Make a single, non-blocking sweep over the embedded loop. This works	3010	Make a single, non-blocking sweep over the embedded loop. This works
2099	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3011	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
2100	apropriate way for embedded loops.	3012	appropriate way for embedded loops.
2101		3013
2102	=item struct ev_loop *other [read-only]	3014	=item struct ev_loop *other [read-only]
2103		3015
2104	The embedded event loop.	3016	The embedded event loop.
2105		3017
…		…
2107		3019
2108	=head3 Examples	3020	=head3 Examples
2109		3021
2110	Example: Try to get an embeddable event loop and embed it into the default	3022	Example: Try to get an embeddable event loop and embed it into the default
2111	event loop. If that is not possible, use the default loop. The default	3023	event loop. If that is not possible, use the default loop. The default
2112	loop is stored in C<loop_hi>, while the mebeddable loop is stored in	3024	loop is stored in C<loop_hi>, while the embeddable loop is stored in
2113	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be	3025	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2114	used).	3026	used).
2115		3027
2116	struct ev_loop *loop_hi = ev_default_init (0);	3028	struct ev_loop *loop_hi = ev_default_init (0);
2117	struct ev_loop *loop_lo = 0;	3029	struct ev_loop *loop_lo = 0;
2118	struct ev_embed embed;	3030	ev_embed embed;
2119		3031
2120	// see if there is a chance of getting one that works	3032	// see if there is a chance of getting one that works
2121	// (remember that a flags value of 0 means autodetection)	3033	// (remember that a flags value of 0 means autodetection)
2122	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	3034	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2123	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	3035	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
2124	: 0;	3036	: 0;
2125		3037
2126	// if we got one, then embed it, otherwise default to loop_hi	3038	// if we got one, then embed it, otherwise default to loop_hi
2127	if (loop_lo)	3039	if (loop_lo)
2128	{	3040	{
2129	ev_embed_init (&embed, 0, loop_lo);	3041	ev_embed_init (&embed, 0, loop_lo);
2130	ev_embed_start (loop_hi, &embed);	3042	ev_embed_start (loop_hi, &embed);
2131	}	3043	}
2132	else	3044	else
2133	loop_lo = loop_hi;	3045	loop_lo = loop_hi;
2134		3046
2135	Example: Check if kqueue is available but not recommended and create	3047	Example: Check if kqueue is available but not recommended and create
2136	a kqueue backend for use with sockets (which usually work with any	3048	a kqueue backend for use with sockets (which usually work with any
2137	kqueue implementation). Store the kqueue/socket-only event loop in	3049	kqueue implementation). Store the kqueue/socket-only event loop in
2138	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	3050	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2139		3051
2140	struct ev_loop *loop = ev_default_init (0);	3052	struct ev_loop *loop = ev_default_init (0);
2141	struct ev_loop *loop_socket = 0;	3053	struct ev_loop *loop_socket = 0;
2142	struct ev_embed embed;	3054	ev_embed embed;
2143		3055
2144	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	3056	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2145	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	3057	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2146	{	3058	{
2147	ev_embed_init (&embed, 0, loop_socket);	3059	ev_embed_init (&embed, 0, loop_socket);
2148	ev_embed_start (loop, &embed);	3060	ev_embed_start (loop, &embed);
2149	}	3061	}
2150		3062
2151	if (!loop_socket)	3063	if (!loop_socket)
2152	loop_socket = loop;	3064	loop_socket = loop;
2153		3065
2154	// now use loop_socket for all sockets, and loop for everything else	3066	// now use loop_socket for all sockets, and loop for everything else
2155		3067
2156		3068
2157	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	3069	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
2158		3070
2159	Fork watchers are called when a C<fork ()> was detected (usually because	3071	Fork watchers are called when a C<fork ()> was detected (usually because
…		…
2162	event loop blocks next and before C<ev_check> watchers are being called,	3074	event loop blocks next and before C<ev_check> watchers are being called,
2163	and only in the child after the fork. If whoever good citizen calling	3075	and only in the child after the fork. If whoever good citizen calling
2164	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3076	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2165	handlers will be invoked, too, of course.	3077	handlers will be invoked, too, of course.
2166		3078
		3079	=head3 The special problem of life after fork - how is it possible?
		3080
		3081	Most uses of C<fork()> consist of forking, then some simple calls to set
		3082	up/change the process environment, followed by a call to C<exec()>. This
		3083	sequence should be handled by libev without any problems.
		3084
		3085	This changes when the application actually wants to do event handling
		3086	in the child, or both parent in child, in effect "continuing" after the
		3087	fork.
		3088
		3089	The default mode of operation (for libev, with application help to detect
		3090	forks) is to duplicate all the state in the child, as would be expected
		3091	when I<either> the parent I<or> the child process continues.
		3092
		3093	When both processes want to continue using libev, then this is usually the
		3094	wrong result. In that case, usually one process (typically the parent) is
		3095	supposed to continue with all watchers in place as before, while the other
		3096	process typically wants to start fresh, i.e. without any active watchers.
		3097
		3098	The cleanest and most efficient way to achieve that with libev is to
		3099	simply create a new event loop, which of course will be "empty", and
		3100	use that for new watchers. This has the advantage of not touching more
		3101	memory than necessary, and thus avoiding the copy-on-write, and the
		3102	disadvantage of having to use multiple event loops (which do not support
		3103	signal watchers).
		3104
		3105	When this is not possible, or you want to use the default loop for
		3106	other reasons, then in the process that wants to start "fresh", call
		3107	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
		3108	Destroying the default loop will "orphan" (not stop) all registered
		3109	watchers, so you have to be careful not to execute code that modifies
		3110	those watchers. Note also that in that case, you have to re-register any
		3111	signal watchers.
		3112
2167	=head3 Watcher-Specific Functions and Data Members	3113	=head3 Watcher-Specific Functions and Data Members
2168		3114
2169	=over 4	3115	=over 4
2170		3116
2171	=item ev_fork_init (ev_signal *, callback)	3117	=item ev_fork_init (ev_fork *, callback)
2172		3118
2173	Initialises and configures the fork watcher - it has no parameters of any	3119	Initialises and configures the fork watcher - it has no parameters of any
2174	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3120	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
2175	believe me.	3121	really.
2176		3122
2177	=back	3123	=back
2178		3124
2179		3125
		3126	=head2 C<ev_cleanup> - even the best things end
		3127
		3128	Cleanup watchers are called just before the event loop is being destroyed
		3129	by a call to C<ev_loop_destroy>.
		3130
		3131	While there is no guarantee that the event loop gets destroyed, cleanup
		3132	watchers provide a convenient method to install cleanup hooks for your
		3133	program, worker threads and so on - you just to make sure to destroy the
		3134	loop when you want them to be invoked.
		3135
		3136	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3137	all other watchers, they do not keep a reference to the event loop (which
		3138	makes a lot of sense if you think about it). Like all other watchers, you
		3139	can call libev functions in the callback, except C<ev_cleanup_start>.
		3140
		3141	=head3 Watcher-Specific Functions and Data Members
		3142
		3143	=over 4
		3144
		3145	=item ev_cleanup_init (ev_cleanup *, callback)
		3146
		3147	Initialises and configures the cleanup watcher - it has no parameters of
		3148	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3149	pointless, I assure you.
		3150
		3151	=back
		3152
		3153	Example: Register an atexit handler to destroy the default loop, so any
		3154	cleanup functions are called.
		3155
		3156	static void
		3157	program_exits (void)
		3158	{
		3159	ev_loop_destroy (EV_DEFAULT_UC);
		3160	}
		3161
		3162	...
		3163	atexit (program_exits);
		3164
		3165
2180	=head2 C<ev_async> - how to wake up another event loop	3166	=head2 C<ev_async> - how to wake up an event loop
2181		3167
2182	In general, you cannot use an C<ev_loop> from multiple threads or other	3168	In general, you cannot use an C<ev_run> from multiple threads or other
2183	asynchronous sources such as signal handlers (as opposed to multiple event	3169	asynchronous sources such as signal handlers (as opposed to multiple event
2184	loops - those are of course safe to use in different threads).	3170	loops - those are of course safe to use in different threads).
2185		3171
2186	Sometimes, however, you need to wake up another event loop you do not	3172	Sometimes, however, you need to wake up an event loop you do not control,
2187	control, for example because it belongs to another thread. This is what	3173	for example because it belongs to another thread. This is what C<ev_async>
2188	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3174	watchers do: as long as the C<ev_async> watcher is active, you can signal
2189	can signal it by calling C<ev_async_send>, which is thread- and signal	3175	it by calling C<ev_async_send>, which is thread- and signal safe.
2190	safe.
2191		3176
2192	This functionality is very similar to C<ev_signal> watchers, as signals,	3177	This functionality is very similar to C<ev_signal> watchers, as signals,
2193	too, are asynchronous in nature, and signals, too, will be compressed	3178	too, are asynchronous in nature, and signals, too, will be compressed
2194	(i.e. the number of callback invocations may be less than the number of	3179	(i.e. the number of callback invocations may be less than the number of
2195	C<ev_async_sent> calls).	3180	C<ev_async_sent> calls).
…		…
2200	=head3 Queueing	3185	=head3 Queueing
2201		3186
2202	C<ev_async> does not support queueing of data in any way. The reason	3187	C<ev_async> does not support queueing of data in any way. The reason
2203	is that the author does not know of a simple (or any) algorithm for a	3188	is that the author does not know of a simple (or any) algorithm for a
2204	multiple-writer-single-reader queue that works in all cases and doesn't	3189	multiple-writer-single-reader queue that works in all cases and doesn't
2205	need elaborate support such as pthreads.	3190	need elaborate support such as pthreads or unportable memory access
		3191	semantics.
2206		3192
2207	That means that if you want to queue data, you have to provide your own	3193	That means that if you want to queue data, you have to provide your own
2208	queue. But at least I can tell you would implement locking around your	3194	queue. But at least I can tell you how to implement locking around your
2209	queue:	3195	queue:
2210		3196
2211	=over 4	3197	=over 4
2212		3198
2213	=item queueing from a signal handler context	3199	=item queueing from a signal handler context
2214		3200
2215	To implement race-free queueing, you simply add to the queue in the signal	3201	To implement race-free queueing, you simply add to the queue in the signal
2216	handler but you block the signal handler in the watcher callback. Here is an example that does that for	3202	handler but you block the signal handler in the watcher callback. Here is
2217	some fictitiuous SIGUSR1 handler:	3203	an example that does that for some fictitious SIGUSR1 handler:
2218		3204
2219	static ev_async mysig;	3205	static ev_async mysig;
2220		3206
2221	static void	3207	static void
2222	sigusr1_handler (void)	3208	sigusr1_handler (void)
…		…
2288	=over 4	3274	=over 4
2289		3275
2290	=item ev_async_init (ev_async *, callback)	3276	=item ev_async_init (ev_async *, callback)
2291		3277
2292	Initialises and configures the async watcher - it has no parameters of any	3278	Initialises and configures the async watcher - it has no parameters of any
2293	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3279	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2294	believe me.	3280	trust me.
2295		3281
2296	=item ev_async_send (loop, ev_async *)	3282	=item ev_async_send (loop, ev_async *)
2297		3283
2298	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3284	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2299	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3285	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2300	C<ev_feed_event>, this call is safe to do in other threads, signal or	3286	C<ev_feed_event>, this call is safe to do from other threads, signal or
2301	similar contexts (see the dicusssion of C<EV_ATOMIC_T> in the embedding	3287	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2302	section below on what exactly this means).	3288	section below on what exactly this means).
2303		3289
		3290	Note that, as with other watchers in libev, multiple events might get
		3291	compressed into a single callback invocation (another way to look at this
		3292	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3293	reset when the event loop detects that).
		3294
2304	This call incurs the overhead of a syscall only once per loop iteration,	3295	This call incurs the overhead of a system call only once per event loop
2305	so while the overhead might be noticable, it doesn't apply to repeated	3296	iteration, so while the overhead might be noticeable, it doesn't apply to
2306	calls to C<ev_async_send>.	3297	repeated calls to C<ev_async_send> for the same event loop.
2307		3298
2308	=item bool = ev_async_pending (ev_async *)	3299	=item bool = ev_async_pending (ev_async *)
2309		3300
2310	Returns a non-zero value when C<ev_async_send> has been called on the	3301	Returns a non-zero value when C<ev_async_send> has been called on the
2311	watcher but the event has not yet been processed (or even noted) by the	3302	watcher but the event has not yet been processed (or even noted) by the
2312	event loop.	3303	event loop.
2313		3304
2314	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3305	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2315	the loop iterates next and checks for the watcher to have become active,	3306	the loop iterates next and checks for the watcher to have become active,
2316	it will reset the flag again. C<ev_async_pending> can be used to very	3307	it will reset the flag again. C<ev_async_pending> can be used to very
2317	quickly check wether invoking the loop might be a good idea.	3308	quickly check whether invoking the loop might be a good idea.
2318		3309
2319	Not that this does I<not> check wether the watcher itself is pending, only	3310	Not that this does I<not> check whether the watcher itself is pending,
2320	wether it has been requested to make this watcher pending.	3311	only whether it has been requested to make this watcher pending: there
		3312	is a time window between the event loop checking and resetting the async
		3313	notification, and the callback being invoked.
2321		3314
2322	=back	3315	=back
2323		3316
2324		3317
2325	=head1 OTHER FUNCTIONS	3318	=head1 OTHER FUNCTIONS
…		…
2329	=over 4	3322	=over 4
2330		3323
2331	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3324	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2332		3325
2333	This function combines a simple timer and an I/O watcher, calls your	3326	This function combines a simple timer and an I/O watcher, calls your
2334	callback on whichever event happens first and automatically stop both	3327	callback on whichever event happens first and automatically stops both
2335	watchers. This is useful if you want to wait for a single event on an fd	3328	watchers. This is useful if you want to wait for a single event on an fd
2336	or timeout without having to allocate/configure/start/stop/free one or	3329	or timeout without having to allocate/configure/start/stop/free one or
2337	more watchers yourself.	3330	more watchers yourself.
2338		3331
2339	If C<fd> is less than 0, then no I/O watcher will be started and events	3332	If C<fd> is less than 0, then no I/O watcher will be started and the
2340	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3333	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2341	C<events> set will be craeted and started.	3334	the given C<fd> and C<events> set will be created and started.
2342		3335
2343	If C<timeout> is less than 0, then no timeout watcher will be	3336	If C<timeout> is less than 0, then no timeout watcher will be
2344	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3337	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2345	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3338	repeat = 0) will be started. C<0> is a valid timeout.
2346	dubious value.
2347		3339
2348	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3340	The callback has the type C<void (*cb)(int revents, void *arg)> and is
2349	passed an C<revents> set like normal event callbacks (a combination of	3341	passed an C<revents> set like normal event callbacks (a combination of
2350	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3342	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
2351	value passed to C<ev_once>:	3343	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3344	a timeout and an io event at the same time - you probably should give io
		3345	events precedence.
2352		3346
		3347	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3348
2353	static void stdin_ready (int revents, void *arg)	3349	static void stdin_ready (int revents, void *arg)
2354	{	3350	{
2355	if (revents & EV_TIMEOUT)
2356	/* doh, nothing entered */;
2357	else if (revents & EV_READ)	3351	if (revents & EV_READ)
2358	/* stdin might have data for us, joy! */;	3352	/* stdin might have data for us, joy! */;
		3353	else if (revents & EV_TIMER)
		3354	/* doh, nothing entered */;
2359	}	3355	}
2360		3356
2361	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3357	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2362		3358
2363	=item ev_feed_event (ev_loop *, watcher *, int revents)
2364
2365	Feeds the given event set into the event loop, as if the specified event
2366	had happened for the specified watcher (which must be a pointer to an
2367	initialised but not necessarily started event watcher).
2368
2369	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3359	=item ev_feed_fd_event (loop, int fd, int revents)
2370		3360
2371	Feed an event on the given fd, as if a file descriptor backend detected	3361	Feed an event on the given fd, as if a file descriptor backend detected
2372	the given events it.	3362	the given events it.
2373		3363
2374	=item ev_feed_signal_event (ev_loop *loop, int signum)	3364	=item ev_feed_signal_event (loop, int signum)
2375		3365
2376	Feed an event as if the given signal occured (C<loop> must be the default	3366	Feed an event as if the given signal occurred (C<loop> must be the default
2377	loop!).	3367	loop!).
2378		3368
2379	=back	3369	=back
2380		3370
2381		3371
…		…
2398	=item * Priorities are not currently supported. Initialising priorities	3388	=item * Priorities are not currently supported. Initialising priorities
2399	will fail and all watchers will have the same priority, even though there	3389	will fail and all watchers will have the same priority, even though there
2400	is an ev_pri field.	3390	is an ev_pri field.
2401		3391
2402	=item * In libevent, the last base created gets the signals, in libev, the	3392	=item * In libevent, the last base created gets the signals, in libev, the
2403	first base created (== the default loop) gets the signals.	3393	base that registered the signal gets the signals.
2404		3394
2405	=item * Other members are not supported.	3395	=item * Other members are not supported.
2406		3396
2407	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3397	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2408	to use the libev header file and library.	3398	to use the libev header file and library.
…		…
2410	=back	3400	=back
2411		3401
2412	=head1 C++ SUPPORT	3402	=head1 C++ SUPPORT
2413		3403
2414	Libev comes with some simplistic wrapper classes for C++ that mainly allow	3404	Libev comes with some simplistic wrapper classes for C++ that mainly allow
2415	you to use some convinience methods to start/stop watchers and also change	3405	you to use some convenience methods to start/stop watchers and also change
2416	the callback model to a model using method callbacks on objects.	3406	the callback model to a model using method callbacks on objects.
2417		3407
2418	To use it,	3408	To use it,
2419		3409
2420	#include <ev++.h>	3410	#include <ev++.h>
2421		3411
2422	This automatically includes F<ev.h> and puts all of its definitions (many	3412	This automatically includes F<ev.h> and puts all of its definitions (many
2423	of them macros) into the global namespace. All C++ specific things are	3413	of them macros) into the global namespace. All C++ specific things are
2424	put into the C<ev> namespace. It should support all the same embedding	3414	put into the C<ev> namespace. It should support all the same embedding
2425	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.	3415	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
…		…
2459		3449
2460	=over 4	3450	=over 4
2461		3451
2462	=item ev::TYPE::TYPE ()	3452	=item ev::TYPE::TYPE ()
2463		3453
2464	=item ev::TYPE::TYPE (struct ev_loop *)	3454	=item ev::TYPE::TYPE (loop)
2465		3455
2466	=item ev::TYPE::~TYPE	3456	=item ev::TYPE::~TYPE
2467		3457
2468	The constructor (optionally) takes an event loop to associate the watcher	3458	The constructor (optionally) takes an event loop to associate the watcher
2469	with. If it is omitted, it will use C<EV_DEFAULT>.	3459	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2492	your compiler is good :), then the method will be fully inlined into the	3482	your compiler is good :), then the method will be fully inlined into the
2493	thunking function, making it as fast as a direct C callback.	3483	thunking function, making it as fast as a direct C callback.
2494		3484
2495	Example: simple class declaration and watcher initialisation	3485	Example: simple class declaration and watcher initialisation
2496		3486
2497	struct myclass	3487	struct myclass
2498	{	3488	{
2499	void io_cb (ev::io &w, int revents) { }	3489	void io_cb (ev::io &w, int revents) { }
2500	}	3490	}
2501		3491
2502	myclass obj;	3492	myclass obj;
2503	ev::io iow;	3493	ev::io iow;
2504	iow.set <myclass, &myclass::io_cb> (&obj);	3494	iow.set <myclass, &myclass::io_cb> (&obj);
		3495
		3496	=item w->set (object *)
		3497
		3498	This is a variation of a method callback - leaving out the method to call
		3499	will default the method to C<operator ()>, which makes it possible to use
		3500	functor objects without having to manually specify the C<operator ()> all
		3501	the time. Incidentally, you can then also leave out the template argument
		3502	list.
		3503
		3504	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3505	int revents)>.
		3506
		3507	See the method-C<set> above for more details.
		3508
		3509	Example: use a functor object as callback.
		3510
		3511	struct myfunctor
		3512	{
		3513	void operator() (ev::io &w, int revents)
		3514	{
		3515	...
		3516	}
		3517	}
		3518
		3519	myfunctor f;
		3520
		3521	ev::io w;
		3522	w.set (&f);
2505		3523
2506	=item w->set<function> (void *data = 0)	3524	=item w->set<function> (void *data = 0)
2507		3525
2508	Also sets a callback, but uses a static method or plain function as	3526	Also sets a callback, but uses a static method or plain function as
2509	callback. The optional C<data> argument will be stored in the watcher's	3527	callback. The optional C<data> argument will be stored in the watcher's
…		…
2511		3529
2512	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	3530	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2513		3531
2514	See the method-C<set> above for more details.	3532	See the method-C<set> above for more details.
2515		3533
2516	Example:	3534	Example: Use a plain function as callback.
2517		3535
2518	static void io_cb (ev::io &w, int revents) { }	3536	static void io_cb (ev::io &w, int revents) { }
2519	iow.set <io_cb> ();	3537	iow.set <io_cb> ();
2520		3538
2521	=item w->set (struct ev_loop *)	3539	=item w->set (loop)
2522		3540
2523	Associates a different C<struct ev_loop> with this watcher. You can only	3541	Associates a different C<struct ev_loop> with this watcher. You can only
2524	do this when the watcher is inactive (and not pending either).	3542	do this when the watcher is inactive (and not pending either).
2525		3543
2526	=item w->set ([args])	3544	=item w->set ([arguments])
2527		3545
2528	Basically the same as C<ev_TYPE_set>, with the same args. Must be	3546	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
2529	called at least once. Unlike the C counterpart, an active watcher gets	3547	method or a suitable start method must be called at least once. Unlike the
2530	automatically stopped and restarted when reconfiguring it with this	3548	C counterpart, an active watcher gets automatically stopped and restarted
2531	method.	3549	when reconfiguring it with this method.
2532		3550
2533	=item w->start ()	3551	=item w->start ()
2534		3552
2535	Starts the watcher. Note that there is no C<loop> argument, as the	3553	Starts the watcher. Note that there is no C<loop> argument, as the
2536	constructor already stores the event loop.	3554	constructor already stores the event loop.
2537		3555
		3556	=item w->start ([arguments])
		3557
		3558	Instead of calling C<set> and C<start> methods separately, it is often
		3559	convenient to wrap them in one call. Uses the same type of arguments as
		3560	the configure C<set> method of the watcher.
		3561
2538	=item w->stop ()	3562	=item w->stop ()
2539		3563
2540	Stops the watcher if it is active. Again, no C<loop> argument.	3564	Stops the watcher if it is active. Again, no C<loop> argument.
2541		3565
2542	=item w->again () (C<ev::timer>, C<ev::periodic> only)	3566	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
2554		3578
2555	=back	3579	=back
2556		3580
2557	=back	3581	=back
2558		3582
2559	Example: Define a class with an IO and idle watcher, start one of them in	3583	Example: Define a class with two I/O and idle watchers, start the I/O
2560	the constructor.	3584	watchers in the constructor.
2561		3585
2562	class myclass	3586	class myclass
2563	{	3587	{
2564	ev::io io; void io_cb (ev::io &w, int revents);	3588	ev::io io ; void io_cb (ev::io &w, int revents);
		3589	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);
2565	ev:idle idle void idle_cb (ev::idle &w, int revents);	3590	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2566		3591
2567	myclass (int fd)	3592	myclass (int fd)
2568	{	3593	{
2569	io .set <myclass, &myclass::io_cb > (this);	3594	io .set <myclass, &myclass::io_cb > (this);
		3595	io2 .set <myclass, &myclass::io2_cb > (this);
2570	idle.set <myclass, &myclass::idle_cb> (this);	3596	idle.set <myclass, &myclass::idle_cb> (this);
2571		3597
2572	io.start (fd, ev::READ);	3598	io.set (fd, ev::WRITE); // configure the watcher
		3599	io.start (); // start it whenever convenient
		3600
		3601	io2.start (fd, ev::READ); // set + start in one call
2573	}	3602	}
2574	};	3603	};
2575		3604
2576		3605
2577	=head1 OTHER LANGUAGE BINDINGS	3606	=head1 OTHER LANGUAGE BINDINGS
2578		3607
2579	Libev does not offer other language bindings itself, but bindings for a	3608	Libev does not offer other language bindings itself, but bindings for a
2580	numbe rof languages exist in the form of third-party packages. If you know	3609	number of languages exist in the form of third-party packages. If you know
2581	any interesting language binding in addition to the ones listed here, drop	3610	any interesting language binding in addition to the ones listed here, drop
2582	me a note.	3611	me a note.
2583		3612
2584	=over 4	3613	=over 4
2585		3614
2586	=item Perl	3615	=item Perl
2587		3616
2588	The EV module implements the full libev API and is actually used to test	3617	The EV module implements the full libev API and is actually used to test
2589	libev. EV is developed together with libev. Apart from the EV core module,	3618	libev. EV is developed together with libev. Apart from the EV core module,
2590	there are additional modules that implement libev-compatible interfaces	3619	there are additional modules that implement libev-compatible interfaces
2591	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the	3620	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
2592	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).	3621	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3622	and C<EV::Glib>).
2593		3623
2594	It can be found and installed via CPAN, its homepage is found at	3624	It can be found and installed via CPAN, its homepage is at
2595	L<http://software.schmorp.de/pkg/EV>.	3625	L<http://software.schmorp.de/pkg/EV>.
2596		3626
		3627	=item Python
		3628
		3629	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		3630	seems to be quite complete and well-documented.
		3631
2597	=item Ruby	3632	=item Ruby
2598		3633
2599	Tony Arcieri has written a ruby extension that offers access to a subset	3634	Tony Arcieri has written a ruby extension that offers access to a subset
2600	of the libev API and adds filehandle abstractions, asynchronous DNS and	3635	of the libev API and adds file handle abstractions, asynchronous DNS and
2601	more on top of it. It can be found via gem servers. Its homepage is at	3636	more on top of it. It can be found via gem servers. Its homepage is at
2602	L<http://rev.rubyforge.org/>.	3637	L<http://rev.rubyforge.org/>.
2603		3638
		3639	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3640	makes rev work even on mingw.
		3641
		3642	=item Haskell
		3643
		3644	A haskell binding to libev is available at
		3645	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3646
2604	=item D	3647	=item D
2605		3648
2606	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3649	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2607	be found at L<http://git.llucax.com.ar/?p=software/ev.d.git;a=summary>.	3650	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3651
		3652	=item Ocaml
		3653
		3654	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3655	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3656
		3657	=item Lua
		3658
		3659	Brian Maher has written a partial interface to libev for lua (at the
		3660	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		3661	L<http://github.com/brimworks/lua-ev>.
2608		3662
2609	=back	3663	=back
2610		3664
2611		3665
2612	=head1 MACRO MAGIC	3666	=head1 MACRO MAGIC
2613		3667
2614	Libev can be compiled with a variety of options, the most fundamantal	3668	Libev can be compiled with a variety of options, the most fundamental
2615	of which is C<EV_MULTIPLICITY>. This option determines whether (most)	3669	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
2616	functions and callbacks have an initial C<struct ev_loop *> argument.	3670	functions and callbacks have an initial C<struct ev_loop *> argument.
2617		3671
2618	To make it easier to write programs that cope with either variant, the	3672	To make it easier to write programs that cope with either variant, the
2619	following macros are defined:	3673	following macros are defined:
…		…
2624		3678
2625	This provides the loop I<argument> for functions, if one is required ("ev	3679	This provides the loop I<argument> for functions, if one is required ("ev
2626	loop argument"). The C<EV_A> form is used when this is the sole argument,	3680	loop argument"). The C<EV_A> form is used when this is the sole argument,
2627	C<EV_A_> is used when other arguments are following. Example:	3681	C<EV_A_> is used when other arguments are following. Example:
2628		3682
2629	ev_unref (EV_A);	3683	ev_unref (EV_A);
2630	ev_timer_add (EV_A_ watcher);	3684	ev_timer_add (EV_A_ watcher);
2631	ev_loop (EV_A_ 0);	3685	ev_run (EV_A_ 0);
2632		3686
2633	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	3687	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
2634	which is often provided by the following macro.	3688	which is often provided by the following macro.
2635		3689
2636	=item C<EV_P>, C<EV_P_>	3690	=item C<EV_P>, C<EV_P_>
2637		3691
2638	This provides the loop I<parameter> for functions, if one is required ("ev	3692	This provides the loop I<parameter> for functions, if one is required ("ev
2639	loop parameter"). The C<EV_P> form is used when this is the sole parameter,	3693	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
2640	C<EV_P_> is used when other parameters are following. Example:	3694	C<EV_P_> is used when other parameters are following. Example:
2641		3695
2642	// this is how ev_unref is being declared	3696	// this is how ev_unref is being declared
2643	static void ev_unref (EV_P);	3697	static void ev_unref (EV_P);
2644		3698
2645	// this is how you can declare your typical callback	3699	// this is how you can declare your typical callback
2646	static void cb (EV_P_ ev_timer *w, int revents)	3700	static void cb (EV_P_ ev_timer *w, int revents)
2647		3701
2648	It declares a parameter C<loop> of type C<struct ev_loop *>, quite	3702	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
2649	suitable for use with C<EV_A>.	3703	suitable for use with C<EV_A>.
2650		3704
2651	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	3705	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
…		…
2667		3721
2668	Example: Declare and initialise a check watcher, utilising the above	3722	Example: Declare and initialise a check watcher, utilising the above
2669	macros so it will work regardless of whether multiple loops are supported	3723	macros so it will work regardless of whether multiple loops are supported
2670	or not.	3724	or not.
2671		3725
2672	static void	3726	static void
2673	check_cb (EV_P_ ev_timer *w, int revents)	3727	check_cb (EV_P_ ev_timer *w, int revents)
2674	{	3728	{
2675	ev_check_stop (EV_A_ w);	3729	ev_check_stop (EV_A_ w);
2676	}	3730	}
2677		3731
2678	ev_check check;	3732	ev_check check;
2679	ev_check_init (&check, check_cb);	3733	ev_check_init (&check, check_cb);
2680	ev_check_start (EV_DEFAULT_ &check);	3734	ev_check_start (EV_DEFAULT_ &check);
2681	ev_loop (EV_DEFAULT_ 0);	3735	ev_run (EV_DEFAULT_ 0);
2682		3736
2683	=head1 EMBEDDING	3737	=head1 EMBEDDING
2684		3738
2685	Libev can (and often is) directly embedded into host	3739	Libev can (and often is) directly embedded into host
2686	applications. Examples of applications that embed it include the Deliantra	3740	applications. Examples of applications that embed it include the Deliantra
…		…
2693	libev somewhere in your source tree).	3747	libev somewhere in your source tree).
2694		3748
2695	=head2 FILESETS	3749	=head2 FILESETS
2696		3750
2697	Depending on what features you need you need to include one or more sets of files	3751	Depending on what features you need you need to include one or more sets of files
2698	in your app.	3752	in your application.
2699		3753
2700	=head3 CORE EVENT LOOP	3754	=head3 CORE EVENT LOOP
2701		3755
2702	To include only the libev core (all the C<ev_*> functions), with manual	3756	To include only the libev core (all the C<ev_*> functions), with manual
2703	configuration (no autoconf):	3757	configuration (no autoconf):
2704		3758
2705	#define EV_STANDALONE 1	3759	#define EV_STANDALONE 1
2706	#include "ev.c"	3760	#include "ev.c"
2707		3761
2708	This will automatically include F<ev.h>, too, and should be done in a	3762	This will automatically include F<ev.h>, too, and should be done in a
2709	single C source file only to provide the function implementations. To use	3763	single C source file only to provide the function implementations. To use
2710	it, do the same for F<ev.h> in all files wishing to use this API (best	3764	it, do the same for F<ev.h> in all files wishing to use this API (best
2711	done by writing a wrapper around F<ev.h> that you can include instead and	3765	done by writing a wrapper around F<ev.h> that you can include instead and
2712	where you can put other configuration options):	3766	where you can put other configuration options):
2713		3767
2714	#define EV_STANDALONE 1	3768	#define EV_STANDALONE 1
2715	#include "ev.h"	3769	#include "ev.h"
2716		3770
2717	Both header files and implementation files can be compiled with a C++	3771	Both header files and implementation files can be compiled with a C++
2718	compiler (at least, thats a stated goal, and breakage will be treated	3772	compiler (at least, that's a stated goal, and breakage will be treated
2719	as a bug).	3773	as a bug).
2720		3774
2721	You need the following files in your source tree, or in a directory	3775	You need the following files in your source tree, or in a directory
2722	in your include path (e.g. in libev/ when using -Ilibev):	3776	in your include path (e.g. in libev/ when using -Ilibev):
2723		3777
2724	ev.h	3778	ev.h
2725	ev.c	3779	ev.c
2726	ev_vars.h	3780	ev_vars.h
2727	ev_wrap.h	3781	ev_wrap.h
2728		3782
2729	ev_win32.c required on win32 platforms only	3783	ev_win32.c required on win32 platforms only
2730		3784
2731	ev_select.c only when select backend is enabled (which is enabled by default)	3785	ev_select.c only when select backend is enabled (which is enabled by default)
2732	ev_poll.c only when poll backend is enabled (disabled by default)	3786	ev_poll.c only when poll backend is enabled (disabled by default)
2733	ev_epoll.c only when the epoll backend is enabled (disabled by default)	3787	ev_epoll.c only when the epoll backend is enabled (disabled by default)
2734	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	3788	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
2735	ev_port.c only when the solaris port backend is enabled (disabled by default)	3789	ev_port.c only when the solaris port backend is enabled (disabled by default)
2736		3790
2737	F<ev.c> includes the backend files directly when enabled, so you only need	3791	F<ev.c> includes the backend files directly when enabled, so you only need
2738	to compile this single file.	3792	to compile this single file.
2739		3793
2740	=head3 LIBEVENT COMPATIBILITY API	3794	=head3 LIBEVENT COMPATIBILITY API
2741		3795
2742	To include the libevent compatibility API, also include:	3796	To include the libevent compatibility API, also include:
2743		3797
2744	#include "event.c"	3798	#include "event.c"
2745		3799
2746	in the file including F<ev.c>, and:	3800	in the file including F<ev.c>, and:
2747		3801
2748	#include "event.h"	3802	#include "event.h"
2749		3803
2750	in the files that want to use the libevent API. This also includes F<ev.h>.	3804	in the files that want to use the libevent API. This also includes F<ev.h>.
2751		3805
2752	You need the following additional files for this:	3806	You need the following additional files for this:
2753		3807
2754	event.h	3808	event.h
2755	event.c	3809	event.c
2756		3810
2757	=head3 AUTOCONF SUPPORT	3811	=head3 AUTOCONF SUPPORT
2758		3812
2759	Instead of using C<EV_STANDALONE=1> and providing your config in	3813	Instead of using C<EV_STANDALONE=1> and providing your configuration in
2760	whatever way you want, you can also C<m4_include([libev.m4])> in your	3814	whatever way you want, you can also C<m4_include([libev.m4])> in your
2761	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then	3815	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
2762	include F<config.h> and configure itself accordingly.	3816	include F<config.h> and configure itself accordingly.
2763		3817
2764	For this of course you need the m4 file:	3818	For this of course you need the m4 file:
2765		3819
2766	libev.m4	3820	libev.m4
2767		3821
2768	=head2 PREPROCESSOR SYMBOLS/MACROS	3822	=head2 PREPROCESSOR SYMBOLS/MACROS
2769		3823
2770	Libev can be configured via a variety of preprocessor symbols you have to	3824	Libev can be configured via a variety of preprocessor symbols you have to
2771	define before including any of its files. The default in the absense of	3825	define before including (or compiling) any of its files. The default in
2772	autoconf is noted for every option.	3826	the absence of autoconf is documented for every option.
		3827
		3828	Symbols marked with "(h)" do not change the ABI, and can have different
		3829	values when compiling libev vs. including F<ev.h>, so it is permissible
		3830	to redefine them before including F<ev.h> without breaking compatibility
		3831	to a compiled library. All other symbols change the ABI, which means all
		3832	users of libev and the libev code itself must be compiled with compatible
		3833	settings.
2773		3834
2774	=over 4	3835	=over 4
2775		3836
		3837	=item EV_COMPAT3 (h)
		3838
		3839	Backwards compatibility is a major concern for libev. This is why this
		3840	release of libev comes with wrappers for the functions and symbols that
		3841	have been renamed between libev version 3 and 4.
		3842
		3843	You can disable these wrappers (to test compatibility with future
		3844	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		3845	sources. This has the additional advantage that you can drop the C<struct>
		3846	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		3847	typedef in that case.
		3848
		3849	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		3850	and in some even more future version the compatibility code will be
		3851	removed completely.
		3852
2776	=item EV_STANDALONE	3853	=item EV_STANDALONE (h)
2777		3854
2778	Must always be C<1> if you do not use autoconf configuration, which	3855	Must always be C<1> if you do not use autoconf configuration, which
2779	keeps libev from including F<config.h>, and it also defines dummy	3856	keeps libev from including F<config.h>, and it also defines dummy
2780	implementations for some libevent functions (such as logging, which is not	3857	implementations for some libevent functions (such as logging, which is not
2781	supported). It will also not define any of the structs usually found in	3858	supported). It will also not define any of the structs usually found in
2782	F<event.h> that are not directly supported by the libev core alone.	3859	F<event.h> that are not directly supported by the libev core alone.
2783		3860
		3861	In standalone mode, libev will still try to automatically deduce the
		3862	configuration, but has to be more conservative.
		3863
2784	=item EV_USE_MONOTONIC	3864	=item EV_USE_MONOTONIC
2785		3865
2786	If defined to be C<1>, libev will try to detect the availability of the	3866	If defined to be C<1>, libev will try to detect the availability of the
2787	monotonic clock option at both compiletime and runtime. Otherwise no use	3867	monotonic clock option at both compile time and runtime. Otherwise no
2788	of the monotonic clock option will be attempted. If you enable this, you	3868	use of the monotonic clock option will be attempted. If you enable this,
2789	usually have to link against librt or something similar. Enabling it when	3869	you usually have to link against librt or something similar. Enabling it
2790	the functionality isn't available is safe, though, although you have	3870	when the functionality isn't available is safe, though, although you have
2791	to make sure you link against any libraries where the C<clock_gettime>	3871	to make sure you link against any libraries where the C<clock_gettime>
2792	function is hiding in (often F<-lrt>).	3872	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2793		3873
2794	=item EV_USE_REALTIME	3874	=item EV_USE_REALTIME
2795		3875
2796	If defined to be C<1>, libev will try to detect the availability of the	3876	If defined to be C<1>, libev will try to detect the availability of the
2797	realtime clock option at compiletime (and assume its availability at	3877	real-time clock option at compile time (and assume its availability
2798	runtime if successful). Otherwise no use of the realtime clock option will	3878	at runtime if successful). Otherwise no use of the real-time clock
2799	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3879	option will be attempted. This effectively replaces C<gettimeofday>
2800	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3880	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2801	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3881	correctness. See the note about libraries in the description of
		3882	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3883	C<EV_USE_CLOCK_SYSCALL>.
		3884
		3885	=item EV_USE_CLOCK_SYSCALL
		3886
		3887	If defined to be C<1>, libev will try to use a direct syscall instead
		3888	of calling the system-provided C<clock_gettime> function. This option
		3889	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3890	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3891	programs needlessly. Using a direct syscall is slightly slower (in
		3892	theory), because no optimised vdso implementation can be used, but avoids
		3893	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3894	higher, as it simplifies linking (no need for C<-lrt>).
2802		3895
2803	=item EV_USE_NANOSLEEP	3896	=item EV_USE_NANOSLEEP
2804		3897
2805	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3898	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2806	and will use it for delays. Otherwise it will use C<select ()>.	3899	and will use it for delays. Otherwise it will use C<select ()>.
…		…
2814	2.7 or newer, otherwise disabled.	3907	2.7 or newer, otherwise disabled.
2815		3908
2816	=item EV_USE_SELECT	3909	=item EV_USE_SELECT
2817		3910
2818	If undefined or defined to be C<1>, libev will compile in support for the	3911	If undefined or defined to be C<1>, libev will compile in support for the
2819	C<select>(2) backend. No attempt at autodetection will be done: if no	3912	C<select>(2) backend. No attempt at auto-detection will be done: if no
2820	other method takes over, select will be it. Otherwise the select backend	3913	other method takes over, select will be it. Otherwise the select backend
2821	will not be compiled in.	3914	will not be compiled in.
2822		3915
2823	=item EV_SELECT_USE_FD_SET	3916	=item EV_SELECT_USE_FD_SET
2824		3917
2825	If defined to C<1>, then the select backend will use the system C<fd_set>	3918	If defined to C<1>, then the select backend will use the system C<fd_set>
2826	structure. This is useful if libev doesn't compile due to a missing	3919	structure. This is useful if libev doesn't compile due to a missing
2827	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	3920	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2828	exotic systems. This usually limits the range of file descriptors to some	3921	on exotic systems. This usually limits the range of file descriptors to
2829	low limit such as 1024 or might have other limitations (winsocket only	3922	some low limit such as 1024 or might have other limitations (winsocket
2830	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3923	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2831	influence the size of the C<fd_set> used.	3924	configures the maximum size of the C<fd_set>.
2832		3925
2833	=item EV_SELECT_IS_WINSOCKET	3926	=item EV_SELECT_IS_WINSOCKET
2834		3927
2835	When defined to C<1>, the select backend will assume that	3928	When defined to C<1>, the select backend will assume that
2836	select/socket/connect etc. don't understand file descriptors but	3929	select/socket/connect etc. don't understand file descriptors but
…		…
2838	be used is the winsock select). This means that it will call	3931	be used is the winsock select). This means that it will call
2839	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3932	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2840	it is assumed that all these functions actually work on fds, even	3933	it is assumed that all these functions actually work on fds, even
2841	on win32. Should not be defined on non-win32 platforms.	3934	on win32. Should not be defined on non-win32 platforms.
2842		3935
2843	=item EV_FD_TO_WIN32_HANDLE	3936	=item EV_FD_TO_WIN32_HANDLE(fd)
2844		3937
2845	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	3938	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2846	file descriptors to socket handles. When not defining this symbol (the	3939	file descriptors to socket handles. When not defining this symbol (the
2847	default), then libev will call C<_get_osfhandle>, which is usually	3940	default), then libev will call C<_get_osfhandle>, which is usually
2848	correct. In some cases, programs use their own file descriptor management,	3941	correct. In some cases, programs use their own file descriptor management,
2849	in which case they can provide this function to map fds to socket handles.	3942	in which case they can provide this function to map fds to socket handles.
		3943
		3944	=item EV_WIN32_HANDLE_TO_FD(handle)
		3945
		3946	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3947	using the standard C<_open_osfhandle> function. For programs implementing
		3948	their own fd to handle mapping, overwriting this function makes it easier
		3949	to do so. This can be done by defining this macro to an appropriate value.
		3950
		3951	=item EV_WIN32_CLOSE_FD(fd)
		3952
		3953	If programs implement their own fd to handle mapping on win32, then this
		3954	macro can be used to override the C<close> function, useful to unregister
		3955	file descriptors again. Note that the replacement function has to close
		3956	the underlying OS handle.
2850		3957
2851	=item EV_USE_POLL	3958	=item EV_USE_POLL
2852		3959
2853	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3960	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2854	backend. Otherwise it will be enabled on non-win32 platforms. It	3961	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
2881	otherwise another method will be used as fallback. This is the preferred	3988	otherwise another method will be used as fallback. This is the preferred
2882	backend for Solaris 10 systems.	3989	backend for Solaris 10 systems.
2883		3990
2884	=item EV_USE_DEVPOLL	3991	=item EV_USE_DEVPOLL
2885		3992
2886	reserved for future expansion, works like the USE symbols above.	3993	Reserved for future expansion, works like the USE symbols above.
2887		3994
2888	=item EV_USE_INOTIFY	3995	=item EV_USE_INOTIFY
2889		3996
2890	If defined to be C<1>, libev will compile in support for the Linux inotify	3997	If defined to be C<1>, libev will compile in support for the Linux inotify
2891	interface to speed up C<ev_stat> watchers. Its actual availability will	3998	interface to speed up C<ev_stat> watchers. Its actual availability will
…		…
2898	access is atomic with respect to other threads or signal contexts. No such	4005	access is atomic with respect to other threads or signal contexts. No such
2899	type is easily found in the C language, so you can provide your own type	4006	type is easily found in the C language, so you can provide your own type
2900	that you know is safe for your purposes. It is used both for signal handler "locking"	4007	that you know is safe for your purposes. It is used both for signal handler "locking"
2901	as well as for signal and thread safety in C<ev_async> watchers.	4008	as well as for signal and thread safety in C<ev_async> watchers.
2902		4009
2903	In the absense of this define, libev will use C<sig_atomic_t volatile>	4010	In the absence of this define, libev will use C<sig_atomic_t volatile>
2904	(from F<signal.h>), which is usually good enough on most platforms.	4011	(from F<signal.h>), which is usually good enough on most platforms.
2905		4012
2906	=item EV_H	4013	=item EV_H (h)
2907		4014
2908	The name of the F<ev.h> header file used to include it. The default if	4015	The name of the F<ev.h> header file used to include it. The default if
2909	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	4016	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
2910	used to virtually rename the F<ev.h> header file in case of conflicts.	4017	used to virtually rename the F<ev.h> header file in case of conflicts.
2911		4018
2912	=item EV_CONFIG_H	4019	=item EV_CONFIG_H (h)
2913		4020
2914	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	4021	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2915	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	4022	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2916	C<EV_H>, above.	4023	C<EV_H>, above.
2917		4024
2918	=item EV_EVENT_H	4025	=item EV_EVENT_H (h)
2919		4026
2920	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	4027	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2921	of how the F<event.h> header can be found, the default is C<"event.h">.	4028	of how the F<event.h> header can be found, the default is C<"event.h">.
2922		4029
2923	=item EV_PROTOTYPES	4030	=item EV_PROTOTYPES (h)
2924		4031
2925	If defined to be C<0>, then F<ev.h> will not define any function	4032	If defined to be C<0>, then F<ev.h> will not define any function
2926	prototypes, but still define all the structs and other symbols. This is	4033	prototypes, but still define all the structs and other symbols. This is
2927	occasionally useful if you want to provide your own wrapper functions	4034	occasionally useful if you want to provide your own wrapper functions
2928	around libev functions.	4035	around libev functions.
…		…
2947	When doing priority-based operations, libev usually has to linearly search	4054	When doing priority-based operations, libev usually has to linearly search
2948	all the priorities, so having many of them (hundreds) uses a lot of space	4055	all the priorities, so having many of them (hundreds) uses a lot of space
2949	and time, so using the defaults of five priorities (-2 .. +2) is usually	4056	and time, so using the defaults of five priorities (-2 .. +2) is usually
2950	fine.	4057	fine.
2951		4058
2952	If your embedding app does not need any priorities, defining these both to	4059	If your embedding application does not need any priorities, defining these
2953	C<0> will save some memory and cpu.	4060	both to C<0> will save some memory and CPU.
2954		4061
2955	=item EV_PERIODIC_ENABLE	4062	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4063	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4064	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
2956		4065
2957	If undefined or defined to be C<1>, then periodic timers are supported. If	4066	If undefined or defined to be C<1> (and the platform supports it), then
2958	defined to be C<0>, then they are not. Disabling them saves a few kB of	4067	the respective watcher type is supported. If defined to be C<0>, then it
2959	code.	4068	is not. Disabling watcher types mainly saves code size.
2960		4069
2961	=item EV_IDLE_ENABLE	4070	=item EV_FEATURES
2962
2963	If undefined or defined to be C<1>, then idle watchers are supported. If
2964	defined to be C<0>, then they are not. Disabling them saves a few kB of
2965	code.
2966
2967	=item EV_EMBED_ENABLE
2968
2969	If undefined or defined to be C<1>, then embed watchers are supported. If
2970	defined to be C<0>, then they are not.
2971
2972	=item EV_STAT_ENABLE
2973
2974	If undefined or defined to be C<1>, then stat watchers are supported. If
2975	defined to be C<0>, then they are not.
2976
2977	=item EV_FORK_ENABLE
2978
2979	If undefined or defined to be C<1>, then fork watchers are supported. If
2980	defined to be C<0>, then they are not.
2981
2982	=item EV_ASYNC_ENABLE
2983
2984	If undefined or defined to be C<1>, then async watchers are supported. If
2985	defined to be C<0>, then they are not.
2986
2987	=item EV_MINIMAL
2988		4071
2989	If you need to shave off some kilobytes of code at the expense of some	4072	If you need to shave off some kilobytes of code at the expense of some
2990	speed, define this symbol to C<1>. Currently this is used to override some	4073	speed (but with the full API), you can define this symbol to request
2991	inlining decisions, saves roughly 30% codesize of amd64. It also selects a	4074	certain subsets of functionality. The default is to enable all features
2992	much smaller 2-heap for timer management over the default 4-heap.	4075	that can be enabled on the platform.
		4076
		4077	A typical way to use this symbol is to define it to C<0> (or to a bitset
		4078	with some broad features you want) and then selectively re-enable
		4079	additional parts you want, for example if you want everything minimal,
		4080	but multiple event loop support, async and child watchers and the poll
		4081	backend, use this:
		4082
		4083	#define EV_FEATURES 0
		4084	#define EV_MULTIPLICITY 1
		4085	#define EV_USE_POLL 1
		4086	#define EV_CHILD_ENABLE 1
		4087	#define EV_ASYNC_ENABLE 1
		4088
		4089	The actual value is a bitset, it can be a combination of the following
		4090	values:
		4091
		4092	=over 4
		4093
		4094	=item C<1> - faster/larger code
		4095
		4096	Use larger code to speed up some operations.
		4097
		4098	Currently this is used to override some inlining decisions (enlarging the
		4099	code size by roughly 30% on amd64).
		4100
		4101	When optimising for size, use of compiler flags such as C<-Os> with
		4102	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4103	assertions.
		4104
		4105	=item C<2> - faster/larger data structures
		4106
		4107	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4108	hash table sizes and so on. This will usually further increase code size
		4109	and can additionally have an effect on the size of data structures at
		4110	runtime.
		4111
		4112	=item C<4> - full API configuration
		4113
		4114	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4115	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4116
		4117	=item C<8> - full API
		4118
		4119	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4120	details on which parts of the API are still available without this
		4121	feature, and do not complain if this subset changes over time.
		4122
		4123	=item C<16> - enable all optional watcher types
		4124
		4125	Enables all optional watcher types. If you want to selectively enable
		4126	only some watcher types other than I/O and timers (e.g. prepare,
		4127	embed, async, child...) you can enable them manually by defining
		4128	C<EV_watchertype_ENABLE> to C<1> instead.
		4129
		4130	=item C<32> - enable all backends
		4131
		4132	This enables all backends - without this feature, you need to enable at
		4133	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4134
		4135	=item C<64> - enable OS-specific "helper" APIs
		4136
		4137	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4138	default.
		4139
		4140	=back
		4141
		4142	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4143	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4144	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4145	watchers, timers and monotonic clock support.
		4146
		4147	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4148	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4149	your program might be left out as well - a binary starting a timer and an
		4150	I/O watcher then might come out at only 5Kb.
		4151
		4152	=item EV_AVOID_STDIO
		4153
		4154	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4155	functions (printf, scanf, perror etc.). This will increase the code size
		4156	somewhat, but if your program doesn't otherwise depend on stdio and your
		4157	libc allows it, this avoids linking in the stdio library which is quite
		4158	big.
		4159
		4160	Note that error messages might become less precise when this option is
		4161	enabled.
		4162
		4163	=item EV_NSIG
		4164
		4165	The highest supported signal number, +1 (or, the number of
		4166	signals): Normally, libev tries to deduce the maximum number of signals
		4167	automatically, but sometimes this fails, in which case it can be
		4168	specified. Also, using a lower number than detected (C<32> should be
		4169	good for about any system in existence) can save some memory, as libev
		4170	statically allocates some 12-24 bytes per signal number.
2993		4171
2994	=item EV_PID_HASHSIZE	4172	=item EV_PID_HASHSIZE
2995		4173
2996	C<ev_child> watchers use a small hash table to distribute workload by	4174	C<ev_child> watchers use a small hash table to distribute workload by
2997	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	4175	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
2998	than enough. If you need to manage thousands of children you might want to	4176	usually more than enough. If you need to manage thousands of children you
2999	increase this value (I<must> be a power of two).	4177	might want to increase this value (I<must> be a power of two).
3000		4178
3001	=item EV_INOTIFY_HASHSIZE	4179	=item EV_INOTIFY_HASHSIZE
3002		4180
3003	C<ev_stat> watchers use a small hash table to distribute workload by	4181	C<ev_stat> watchers use a small hash table to distribute workload by
3004	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4182	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3005	usually more than enough. If you need to manage thousands of C<ev_stat>	4183	disabled), usually more than enough. If you need to manage thousands of
3006	watchers you might want to increase this value (I<must> be a power of	4184	C<ev_stat> watchers you might want to increase this value (I<must> be a
3007	two).	4185	power of two).
3008		4186
3009	=item EV_USE_4HEAP	4187	=item EV_USE_4HEAP
3010		4188
3011	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4189	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3012	timer and periodics heap, libev uses a 4-heap when this symbol is defined	4190	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3013	to C<1>. The 4-heap uses more complicated (longer) code but has	4191	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3014	noticably faster performance with many (thousands) of watchers.	4192	faster performance with many (thousands) of watchers.
3015		4193
3016	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4194	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3017	(disabled).	4195	will be C<0>.
3018		4196
3019	=item EV_HEAP_CACHE_AT	4197	=item EV_HEAP_CACHE_AT
3020		4198
3021	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4199	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3022	timer and periodics heap, libev can cache the timestamp (I<at>) within	4200	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3023	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4201	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3024	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4202	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3025	but avoids random read accesses on heap changes. This improves performance	4203	but avoids random read accesses on heap changes. This improves performance
3026	noticably with with many (hundreds) of watchers.	4204	noticeably with many (hundreds) of watchers.
3027		4205
3028	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4206	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3029	(disabled).	4207	will be C<0>.
		4208
		4209	=item EV_VERIFY
		4210
		4211	Controls how much internal verification (see C<ev_verify ()>) will
		4212	be done: If set to C<0>, no internal verification code will be compiled
		4213	in. If set to C<1>, then verification code will be compiled in, but not
		4214	called. If set to C<2>, then the internal verification code will be
		4215	called once per loop, which can slow down libev. If set to C<3>, then the
		4216	verification code will be called very frequently, which will slow down
		4217	libev considerably.
		4218
		4219	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4220	will be C<0>.
3030		4221
3031	=item EV_COMMON	4222	=item EV_COMMON
3032		4223
3033	By default, all watchers have a C<void *data> member. By redefining	4224	By default, all watchers have a C<void *data> member. By redefining
3034	this macro to a something else you can include more and other types of	4225	this macro to something else you can include more and other types of
3035	members. You have to define it each time you include one of the files,	4226	members. You have to define it each time you include one of the files,
3036	though, and it must be identical each time.	4227	though, and it must be identical each time.
3037		4228
3038	For example, the perl EV module uses something like this:	4229	For example, the perl EV module uses something like this:
3039		4230
3040	#define EV_COMMON \	4231	#define EV_COMMON \
3041	SV *self; /* contains this struct */ \	4232	SV *self; /* contains this struct */ \
3042	SV *cb_sv, *fh /* note no trailing ";" */	4233	SV *cb_sv, *fh /* note no trailing ";" */
3043		4234
3044	=item EV_CB_DECLARE (type)	4235	=item EV_CB_DECLARE (type)
3045		4236
3046	=item EV_CB_INVOKE (watcher, revents)	4237	=item EV_CB_INVOKE (watcher, revents)
3047		4238
…		…
3052	definition and a statement, respectively. See the F<ev.h> header file for	4243	definition and a statement, respectively. See the F<ev.h> header file for
3053	their default definitions. One possible use for overriding these is to	4244	their default definitions. One possible use for overriding these is to
3054	avoid the C<struct ev_loop *> as first argument in all cases, or to use	4245	avoid the C<struct ev_loop *> as first argument in all cases, or to use
3055	method calls instead of plain function calls in C++.	4246	method calls instead of plain function calls in C++.
3056		4247
		4248	=back
		4249
3057	=head2 EXPORTED API SYMBOLS	4250	=head2 EXPORTED API SYMBOLS
3058		4251
3059	If you need to re-export the API (e.g. via a dll) and you need a list of	4252	If you need to re-export the API (e.g. via a DLL) and you need a list of
3060	exported symbols, you can use the provided F<Symbol.*> files which list	4253	exported symbols, you can use the provided F<Symbol.*> files which list
3061	all public symbols, one per line:	4254	all public symbols, one per line:
3062		4255
3063	Symbols.ev for libev proper	4256	Symbols.ev for libev proper
3064	Symbols.event for the libevent emulation	4257	Symbols.event for the libevent emulation
3065		4258
3066	This can also be used to rename all public symbols to avoid clashes with	4259	This can also be used to rename all public symbols to avoid clashes with
3067	multiple versions of libev linked together (which is obviously bad in	4260	multiple versions of libev linked together (which is obviously bad in
3068	itself, but sometimes it is inconvinient to avoid this).	4261	itself, but sometimes it is inconvenient to avoid this).
3069		4262
3070	A sed command like this will create wrapper C<#define>'s that you need to	4263	A sed command like this will create wrapper C<#define>'s that you need to
3071	include before including F<ev.h>:	4264	include before including F<ev.h>:
3072		4265
3073	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h	4266	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
…		…
3090	file.	4283	file.
3091		4284
3092	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4285	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3093	that everybody includes and which overrides some configure choices:	4286	that everybody includes and which overrides some configure choices:
3094		4287
3095	#define EV_MINIMAL 1	4288	#define EV_FEATURES 8
3096	#define EV_USE_POLL 0	4289	#define EV_USE_SELECT 1
3097	#define EV_MULTIPLICITY 0
3098	#define EV_PERIODIC_ENABLE 0	4290	#define EV_PREPARE_ENABLE 1
		4291	#define EV_IDLE_ENABLE 1
3099	#define EV_STAT_ENABLE 0	4292	#define EV_SIGNAL_ENABLE 1
3100	#define EV_FORK_ENABLE 0	4293	#define EV_CHILD_ENABLE 1
		4294	#define EV_USE_STDEXCEPT 0
3101	#define EV_CONFIG_H <config.h>	4295	#define EV_CONFIG_H <config.h>
3102	#define EV_MINPRI 0
3103	#define EV_MAXPRI 0
3104		4296
3105	#include "ev++.h"	4297	#include "ev++.h"
3106		4298
3107	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4299	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3108		4300
3109	#include "ev_cpp.h"	4301	#include "ev_cpp.h"
3110	#include "ev.c"	4302	#include "ev.c"
3111		4303
		4304	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3112		4305
3113	=head1 THREADS AND COROUTINES	4306	=head2 THREADS AND COROUTINES
3114		4307
3115	=head2 THREADS	4308	=head3 THREADS
3116		4309
3117	Libev itself is completely threadsafe, but it uses no locking. This	4310	All libev functions are reentrant and thread-safe unless explicitly
		4311	documented otherwise, but libev implements no locking itself. This means
3118	means that you can use as many loops as you want in parallel, as long as	4312	that you can use as many loops as you want in parallel, as long as there
3119	only one thread ever calls into one libev function with the same loop	4313	are no concurrent calls into any libev function with the same loop
3120	parameter.	4314	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4315	of course): libev guarantees that different event loops share no data
		4316	structures that need any locking.
3121		4317
3122	Or put differently: calls with different loop parameters can be done in	4318	Or to put it differently: calls with different loop parameters can be done
3123	parallel from multiple threads, calls with the same loop parameter must be	4319	concurrently from multiple threads, calls with the same loop parameter
3124	done serially (but can be done from different threads, as long as only one	4320	must be done serially (but can be done from different threads, as long as
3125	thread ever is inside a call at any point in time, e.g. by using a mutex	4321	only one thread ever is inside a call at any point in time, e.g. by using
3126	per loop).	4322	a mutex per loop).
3127		4323
3128	If you want to know which design is best for your problem, then I cannot	4324	Specifically to support threads (and signal handlers), libev implements
		4325	so-called C<ev_async> watchers, which allow some limited form of
		4326	concurrency on the same event loop, namely waking it up "from the
		4327	outside".
		4328
		4329	If you want to know which design (one loop, locking, or multiple loops
		4330	without or something else still) is best for your problem, then I cannot
3129	help you but by giving some generic advice:	4331	help you, but here is some generic advice:
3130		4332
3131	=over 4	4333	=over 4
3132		4334
3133	=item * most applications have a main thread: use the default libev loop	4335	=item * most applications have a main thread: use the default libev loop
3134	in that thread, or create a seperate thread running only the default loop.	4336	in that thread, or create a separate thread running only the default loop.
3135		4337
3136	This helps integrating other libraries or software modules that use libev	4338	This helps integrating other libraries or software modules that use libev
3137	themselves and don't care/know about threading.	4339	themselves and don't care/know about threading.
3138		4340
3139	=item * one loop per thread is usually a good model.	4341	=item * one loop per thread is usually a good model.
3140		4342
3141	Doing this is almost never wrong, sometimes a better-performance model	4343	Doing this is almost never wrong, sometimes a better-performance model
3142	exists, but it is always a good start.	4344	exists, but it is always a good start.
3143		4345
3144	=item * other models exist, such as the leader/follower pattern, where one	4346	=item * other models exist, such as the leader/follower pattern, where one
3145	loop is handed through multiple threads in a kind of round-robbin fashion.	4347	loop is handed through multiple threads in a kind of round-robin fashion.
3146		4348
3147	Chosing a model is hard - look around, learn, know that usually you cna do	4349	Choosing a model is hard - look around, learn, know that usually you can do
3148	better than you currently do :-)	4350	better than you currently do :-)
3149		4351
3150	=item * often you need to talk to some other thread which blocks in the	4352	=item * often you need to talk to some other thread which blocks in the
		4353	event loop.
		4354
3151	event loop - C<ev_async> watchers can be used to wake them up from other	4355	C<ev_async> watchers can be used to wake them up from other threads safely
3152	threads safely (or from signal contexts...).	4356	(or from signal contexts...).
		4357
		4358	An example use would be to communicate signals or other events that only
		4359	work in the default loop by registering the signal watcher with the
		4360	default loop and triggering an C<ev_async> watcher from the default loop
		4361	watcher callback into the event loop interested in the signal.
3153		4362
3154	=back	4363	=back
3155		4364
		4365	=head4 THREAD LOCKING EXAMPLE
		4366
		4367	Here is a fictitious example of how to run an event loop in a different
		4368	thread than where callbacks are being invoked and watchers are
		4369	created/added/removed.
		4370
		4371	For a real-world example, see the C<EV::Loop::Async> perl module,
		4372	which uses exactly this technique (which is suited for many high-level
		4373	languages).
		4374
		4375	The example uses a pthread mutex to protect the loop data, a condition
		4376	variable to wait for callback invocations, an async watcher to notify the
		4377	event loop thread and an unspecified mechanism to wake up the main thread.
		4378
		4379	First, you need to associate some data with the event loop:
		4380
		4381	typedef struct {
		4382	mutex_t lock; /* global loop lock */
		4383	ev_async async_w;
		4384	thread_t tid;
		4385	cond_t invoke_cv;
		4386	} userdata;
		4387
		4388	void prepare_loop (EV_P)
		4389	{
		4390	// for simplicity, we use a static userdata struct.
		4391	static userdata u;
		4392
		4393	ev_async_init (&u->async_w, async_cb);
		4394	ev_async_start (EV_A_ &u->async_w);
		4395
		4396	pthread_mutex_init (&u->lock, 0);
		4397	pthread_cond_init (&u->invoke_cv, 0);
		4398
		4399	// now associate this with the loop
		4400	ev_set_userdata (EV_A_ u);
		4401	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4402	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4403
		4404	// then create the thread running ev_loop
		4405	pthread_create (&u->tid, 0, l_run, EV_A);
		4406	}
		4407
		4408	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4409	solely to wake up the event loop so it takes notice of any new watchers
		4410	that might have been added:
		4411
		4412	static void
		4413	async_cb (EV_P_ ev_async *w, int revents)
		4414	{
		4415	// just used for the side effects
		4416	}
		4417
		4418	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4419	protecting the loop data, respectively.
		4420
		4421	static void
		4422	l_release (EV_P)
		4423	{
		4424	userdata *u = ev_userdata (EV_A);
		4425	pthread_mutex_unlock (&u->lock);
		4426	}
		4427
		4428	static void
		4429	l_acquire (EV_P)
		4430	{
		4431	userdata *u = ev_userdata (EV_A);
		4432	pthread_mutex_lock (&u->lock);
		4433	}
		4434
		4435	The event loop thread first acquires the mutex, and then jumps straight
		4436	into C<ev_run>:
		4437
		4438	void *
		4439	l_run (void *thr_arg)
		4440	{
		4441	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4442
		4443	l_acquire (EV_A);
		4444	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4445	ev_run (EV_A_ 0);
		4446	l_release (EV_A);
		4447
		4448	return 0;
		4449	}
		4450
		4451	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4452	signal the main thread via some unspecified mechanism (signals? pipe
		4453	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4454	have been called (in a while loop because a) spurious wakeups are possible
		4455	and b) skipping inter-thread-communication when there are no pending
		4456	watchers is very beneficial):
		4457
		4458	static void
		4459	l_invoke (EV_P)
		4460	{
		4461	userdata *u = ev_userdata (EV_A);
		4462
		4463	while (ev_pending_count (EV_A))
		4464	{
		4465	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4466	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4467	}
		4468	}
		4469
		4470	Now, whenever the main thread gets told to invoke pending watchers, it
		4471	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4472	thread to continue:
		4473
		4474	static void
		4475	real_invoke_pending (EV_P)
		4476	{
		4477	userdata *u = ev_userdata (EV_A);
		4478
		4479	pthread_mutex_lock (&u->lock);
		4480	ev_invoke_pending (EV_A);
		4481	pthread_cond_signal (&u->invoke_cv);
		4482	pthread_mutex_unlock (&u->lock);
		4483	}
		4484
		4485	Whenever you want to start/stop a watcher or do other modifications to an
		4486	event loop, you will now have to lock:
		4487
		4488	ev_timer timeout_watcher;
		4489	userdata *u = ev_userdata (EV_A);
		4490
		4491	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4492
		4493	pthread_mutex_lock (&u->lock);
		4494	ev_timer_start (EV_A_ &timeout_watcher);
		4495	ev_async_send (EV_A_ &u->async_w);
		4496	pthread_mutex_unlock (&u->lock);
		4497
		4498	Note that sending the C<ev_async> watcher is required because otherwise
		4499	an event loop currently blocking in the kernel will have no knowledge
		4500	about the newly added timer. By waking up the loop it will pick up any new
		4501	watchers in the next event loop iteration.
		4502
3156	=head2 COROUTINES	4503	=head3 COROUTINES
3157		4504
3158	Libev is much more accomodating to coroutines ("cooperative threads"):	4505	Libev is very accommodating to coroutines ("cooperative threads"):
3159	libev fully supports nesting calls to it's functions from different	4506	libev fully supports nesting calls to its functions from different
3160	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4507	coroutines (e.g. you can call C<ev_run> on the same loop from two
3161	different coroutines and switch freely between both coroutines running the	4508	different coroutines, and switch freely between both coroutines running
3162	loop, as long as you don't confuse yourself). The only exception is that	4509	the loop, as long as you don't confuse yourself). The only exception is
3163	you must not do this from C<ev_periodic> reschedule callbacks.	4510	that you must not do this from C<ev_periodic> reschedule callbacks.
3164		4511
3165	Care has been invested into making sure that libev does not keep local	4512	Care has been taken to ensure that libev does not keep local state inside
3166	state inside C<ev_loop>, and other calls do not usually allow coroutine	4513	C<ev_run>, and other calls do not usually allow for coroutine switches as
3167	switches.	4514	they do not call any callbacks.
3168		4515
		4516	=head2 COMPILER WARNINGS
3169		4517
3170	=head1 COMPLEXITIES	4518	Depending on your compiler and compiler settings, you might get no or a
		4519	lot of warnings when compiling libev code. Some people are apparently
		4520	scared by this.
3171		4521
3172	In this section the complexities of (many of) the algorithms used inside	4522	However, these are unavoidable for many reasons. For one, each compiler
3173	libev will be explained. For complexity discussions about backends see the	4523	has different warnings, and each user has different tastes regarding
3174	documentation for C<ev_default_init>.	4524	warning options. "Warn-free" code therefore cannot be a goal except when
		4525	targeting a specific compiler and compiler-version.
3175		4526
3176	All of the following are about amortised time: If an array needs to be	4527	Another reason is that some compiler warnings require elaborate
3177	extended, libev needs to realloc and move the whole array, but this	4528	workarounds, or other changes to the code that make it less clear and less
3178	happens asymptotically never with higher number of elements, so O(1) might	4529	maintainable.
3179	mean it might do a lengthy realloc operation in rare cases, but on average
3180	it is much faster and asymptotically approaches constant time.
3181		4530
3182	=over 4	4531	And of course, some compiler warnings are just plain stupid, or simply
		4532	wrong (because they don't actually warn about the condition their message
		4533	seems to warn about). For example, certain older gcc versions had some
		4534	warnings that resulted in an extreme number of false positives. These have
		4535	been fixed, but some people still insist on making code warn-free with
		4536	such buggy versions.
3183		4537
3184	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4538	While libev is written to generate as few warnings as possible,
		4539	"warn-free" code is not a goal, and it is recommended not to build libev
		4540	with any compiler warnings enabled unless you are prepared to cope with
		4541	them (e.g. by ignoring them). Remember that warnings are just that:
		4542	warnings, not errors, or proof of bugs.
3185		4543
3186	This means that, when you have a watcher that triggers in one hour and
3187	there are 100 watchers that would trigger before that then inserting will
3188	have to skip roughly seven (C<ld 100>) of these watchers.
3189		4544
3190	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4545	=head2 VALGRIND
3191		4546
3192	That means that changing a timer costs less than removing/adding them	4547	Valgrind has a special section here because it is a popular tool that is
3193	as only the relative motion in the event queue has to be paid for.	4548	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3194		4549
3195	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4550	If you think you found a bug (memory leak, uninitialised data access etc.)
		4551	in libev, then check twice: If valgrind reports something like:
3196		4552
3197	These just add the watcher into an array or at the head of a list.	4553	==2274== definitely lost: 0 bytes in 0 blocks.
		4554	==2274== possibly lost: 0 bytes in 0 blocks.
		4555	==2274== still reachable: 256 bytes in 1 blocks.
3198		4556
3199	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4557	Then there is no memory leak, just as memory accounted to global variables
		4558	is not a memleak - the memory is still being referenced, and didn't leak.
3200		4559
3201	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4560	Similarly, under some circumstances, valgrind might report kernel bugs
		4561	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4562	although an acceptable workaround has been found here), or it might be
		4563	confused.
3202		4564
3203	These watchers are stored in lists then need to be walked to find the	4565	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3204	correct watcher to remove. The lists are usually short (you don't usually	4566	make it into some kind of religion.
3205	have many watchers waiting for the same fd or signal).
3206		4567
3207	=item Finding the next timer in each loop iteration: O(1)	4568	If you are unsure about something, feel free to contact the mailing list
		4569	with the full valgrind report and an explanation on why you think this
		4570	is a bug in libev (best check the archives, too :). However, don't be
		4571	annoyed when you get a brisk "this is no bug" answer and take the chance
		4572	of learning how to interpret valgrind properly.
3208		4573
3209	By virtue of using a binary or 4-heap, the next timer is always found at a	4574	If you need, for some reason, empty reports from valgrind for your project
3210	fixed position in the storage array.	4575	I suggest using suppression lists.
3211		4576
3212	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3213		4577
3214	A change means an I/O watcher gets started or stopped, which requires	4578	=head1 PORTABILITY NOTES
3215	libev to recalculate its status (and possibly tell the kernel, depending
3216	on backend and wether C<ev_io_set> was used).
3217		4579
3218	=item Activating one watcher (putting it into the pending state): O(1)	4580	=head2 GNU/LINUX 32 BIT LIMITATIONS
3219		4581
3220	=item Priority handling: O(number_of_priorities)	4582	GNU/Linux is the only common platform that supports 64 bit file/large file
		4583	interfaces but I<disables> them by default.
3221		4584
3222	Priorities are implemented by allocating some space for each	4585	That means that libev compiled in the default environment doesn't support
3223	priority. When doing priority-based operations, libev usually has to	4586	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
3224	linearly search all the priorities, but starting/stopping and activating
3225	watchers becomes O(1) w.r.t. priority handling.
3226		4587
3227	=item Sending an ev_async: O(1)	4588	Unfortunately, many programs try to work around this GNU/Linux issue
		4589	by enabling the large file API, which makes them incompatible with the
		4590	standard libev compiled for their system.
3228		4591
3229	=item Processing ev_async_send: O(number_of_async_watchers)	4592	Likewise, libev cannot enable the large file API itself as this would
		4593	suddenly make it incompatible to the default compile time environment,
		4594	i.e. all programs not using special compile switches.
3230		4595
3231	=item Processing signals: O(max_signal_number)	4596	=head2 OS/X AND DARWIN BUGS
3232		4597
3233	Sending involves a syscall I<iff> there were no other C<ev_async_send>	4598	The whole thing is a bug if you ask me - basically any system interface
3234	calls in the current loop iteration. Checking for async and signal events	4599	you touch is broken, whether it is locales, poll, kqueue or even the
3235	involves iterating over all running async watchers or all signal numbers.	4600	OpenGL drivers.
3236		4601
3237	=back	4602	=head3 C<kqueue> is buggy
3238		4603
		4604	The kqueue syscall is broken in all known versions - most versions support
		4605	only sockets, many support pipes.
3239		4606
3240	=head1 Win32 platform limitations and workarounds	4607	Libev tries to work around this by not using C<kqueue> by default on this
		4608	rotten platform, but of course you can still ask for it when creating a
		4609	loop - embedding a socket-only kqueue loop into a select-based one is
		4610	probably going to work well.
		4611
		4612	=head3 C<poll> is buggy
		4613
		4614	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4615	implementation by something calling C<kqueue> internally around the 10.5.6
		4616	release, so now C<kqueue> I<and> C<poll> are broken.
		4617
		4618	Libev tries to work around this by not using C<poll> by default on
		4619	this rotten platform, but of course you can still ask for it when creating
		4620	a loop.
		4621
		4622	=head3 C<select> is buggy
		4623
		4624	All that's left is C<select>, and of course Apple found a way to fuck this
		4625	one up as well: On OS/X, C<select> actively limits the number of file
		4626	descriptors you can pass in to 1024 - your program suddenly crashes when
		4627	you use more.
		4628
		4629	There is an undocumented "workaround" for this - defining
		4630	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		4631	work on OS/X.
		4632
		4633	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		4634
		4635	=head3 C<errno> reentrancy
		4636
		4637	The default compile environment on Solaris is unfortunately so
		4638	thread-unsafe that you can't even use components/libraries compiled
		4639	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		4640	defined by default. A valid, if stupid, implementation choice.
		4641
		4642	If you want to use libev in threaded environments you have to make sure
		4643	it's compiled with C<_REENTRANT> defined.
		4644
		4645	=head3 Event port backend
		4646
		4647	The scalable event interface for Solaris is called "event
		4648	ports". Unfortunately, this mechanism is very buggy in all major
		4649	releases. If you run into high CPU usage, your program freezes or you get
		4650	a large number of spurious wakeups, make sure you have all the relevant
		4651	and latest kernel patches applied. No, I don't know which ones, but there
		4652	are multiple ones to apply, and afterwards, event ports actually work
		4653	great.
		4654
		4655	If you can't get it to work, you can try running the program by setting
		4656	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		4657	C<select> backends.
		4658
		4659	=head2 AIX POLL BUG
		4660
		4661	AIX unfortunately has a broken C<poll.h> header. Libev works around
		4662	this by trying to avoid the poll backend altogether (i.e. it's not even
		4663	compiled in), which normally isn't a big problem as C<select> works fine
		4664	with large bitsets on AIX, and AIX is dead anyway.
		4665
		4666	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4667
		4668	=head3 General issues
3241		4669
3242	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4670	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3243	requires, and its I/O model is fundamentally incompatible with the POSIX	4671	requires, and its I/O model is fundamentally incompatible with the POSIX
3244	model. Libev still offers limited functionality on this platform in	4672	model. Libev still offers limited functionality on this platform in
3245	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4673	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
3246	descriptors. This only applies when using Win32 natively, not when using	4674	descriptors. This only applies when using Win32 natively, not when using
3247	e.g. cygwin.	4675	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		4676	as every compielr comes with a slightly differently broken/incompatible
		4677	environment.
3248		4678
3249	Lifting these limitations would basically require the full	4679	Lifting these limitations would basically require the full
3250	re-implementation of the I/O system. If you are into these kinds of	4680	re-implementation of the I/O system. If you are into this kind of thing,
3251	things, then note that glib does exactly that for you in a very portable	4681	then note that glib does exactly that for you in a very portable way (note
3252	way (note also that glib is the slowest event library known to man).	4682	also that glib is the slowest event library known to man).
3253		4683
3254	There is no supported compilation method available on windows except	4684	There is no supported compilation method available on windows except
3255	embedding it into other applications.	4685	embedding it into other applications.
		4686
		4687	Sensible signal handling is officially unsupported by Microsoft - libev
		4688	tries its best, but under most conditions, signals will simply not work.
		4689
		4690	Not a libev limitation but worth mentioning: windows apparently doesn't
		4691	accept large writes: instead of resulting in a partial write, windows will
		4692	either accept everything or return C<ENOBUFS> if the buffer is too large,
		4693	so make sure you only write small amounts into your sockets (less than a
		4694	megabyte seems safe, but this apparently depends on the amount of memory
		4695	available).
3256		4696
3257	Due to the many, low, and arbitrary limits on the win32 platform and	4697	Due to the many, low, and arbitrary limits on the win32 platform and
3258	the abysmal performance of winsockets, using a large number of sockets	4698	the abysmal performance of winsockets, using a large number of sockets
3259	is not recommended (and not reasonable). If your program needs to use	4699	is not recommended (and not reasonable). If your program needs to use
3260	more than a hundred or so sockets, then likely it needs to use a totally	4700	more than a hundred or so sockets, then likely it needs to use a totally
3261	different implementation for windows, as libev offers the POSIX readiness	4701	different implementation for windows, as libev offers the POSIX readiness
3262	notification model, which cannot be implemented efficiently on windows	4702	notification model, which cannot be implemented efficiently on windows
3263	(microsoft monopoly games).	4703	(due to Microsoft monopoly games).
3264		4704
3265	=over 4	4705	A typical way to use libev under windows is to embed it (see the embedding
		4706	section for details) and use the following F<evwrap.h> header file instead
		4707	of F<ev.h>:
3266		4708
		4709	#define EV_STANDALONE /* keeps ev from requiring config.h */
		4710	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		4711
		4712	#include "ev.h"
		4713
		4714	And compile the following F<evwrap.c> file into your project (make sure
		4715	you do I<not> compile the F<ev.c> or any other embedded source files!):
		4716
		4717	#include "evwrap.h"
		4718	#include "ev.c"
		4719
3267	=item The winsocket select function	4720	=head3 The winsocket C<select> function
3268		4721
3269	The winsocket C<select> function doesn't follow POSIX in that it requires	4722	The winsocket C<select> function doesn't follow POSIX in that it
3270	socket I<handles> and not socket I<file descriptors>. This makes select	4723	requires socket I<handles> and not socket I<file descriptors> (it is
3271	very inefficient, and also requires a mapping from file descriptors	4724	also extremely buggy). This makes select very inefficient, and also
3272	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,	4725	requires a mapping from file descriptors to socket handles (the Microsoft
3273	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor	4726	C runtime provides the function C<_open_osfhandle> for this). See the
3274	symbols for more info.	4727	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		4728	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
3275		4729
3276	The configuration for a "naked" win32 using the microsoft runtime	4730	The configuration for a "naked" win32 using the Microsoft runtime
3277	libraries and raw winsocket select is:	4731	libraries and raw winsocket select is:
3278		4732
3279	#define EV_USE_SELECT 1	4733	#define EV_USE_SELECT 1
3280	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	4734	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
3281		4735
3282	Note that winsockets handling of fd sets is O(n), so you can easily get a	4736	Note that winsockets handling of fd sets is O(n), so you can easily get a
3283	complexity in the O(n²) range when using win32.	4737	complexity in the O(n²) range when using win32.
3284		4738
3285	=item Limited number of file descriptors	4739	=head3 Limited number of file descriptors
3286		4740
3287	Windows has numerous arbitrary (and low) limits on things.	4741	Windows has numerous arbitrary (and low) limits on things.
3288		4742
3289	Early versions of winsocket's select only supported waiting for a maximum	4743	Early versions of winsocket's select only supported waiting for a maximum
3290	of C<64> handles (probably owning to the fact that all windows kernels	4744	of C<64> handles (probably owning to the fact that all windows kernels
3291	can only wait for C<64> things at the same time internally; microsoft	4745	can only wait for C<64> things at the same time internally; Microsoft
3292	recommends spawning a chain of threads and wait for 63 handles and the	4746	recommends spawning a chain of threads and wait for 63 handles and the
3293	previous thread in each. Great).	4747	previous thread in each. Sounds great!).
3294		4748
3295	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4749	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3296	to some high number (e.g. C<2048>) before compiling the winsocket select	4750	to some high number (e.g. C<2048>) before compiling the winsocket select
3297	call (which might be in libev or elsewhere, for example, perl does its own	4751	call (which might be in libev or elsewhere, for example, perl and many
3298	select emulation on windows).	4752	other interpreters do their own select emulation on windows).
3299		4753
3300	Another limit is the number of file descriptors in the microsoft runtime	4754	Another limit is the number of file descriptors in the Microsoft runtime
3301	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4755	libraries, which by default is C<64> (there must be a hidden I<64>
3302	or something like this inside microsoft). You can increase this by calling	4756	fetish or something like this inside Microsoft). You can increase this
3303	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4757	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3304	arbitrary limit), but is broken in many versions of the microsoft runtime	4758	(another arbitrary limit), but is broken in many versions of the Microsoft
3305	libraries.
3306
3307	This might get you to about C<512> or C<2048> sockets (depending on	4759	runtime libraries. This might get you to about C<512> or C<2048> sockets
3308	windows version and/or the phase of the moon). To get more, you need to	4760	(depending on windows version and/or the phase of the moon). To get more,
3309	wrap all I/O functions and provide your own fd management, but the cost of	4761	you need to wrap all I/O functions and provide your own fd management, but
3310	calling select (O(n²)) will likely make this unworkable.	4762	the cost of calling select (O(n²)) will likely make this unworkable.
3311		4763
3312	=back
3313
3314
3315	=head1 PORTABILITY REQUIREMENTS	4764	=head2 PORTABILITY REQUIREMENTS
3316		4765
3317	In addition to a working ISO-C implementation, libev relies on a few	4766	In addition to a working ISO-C implementation and of course the
3318	additional extensions:	4767	backend-specific APIs, libev relies on a few additional extensions:
3319		4768
3320	=over 4	4769	=over 4
3321		4770
		4771	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		4772	calling conventions regardless of C<ev_watcher_type *>.
		4773
		4774	Libev assumes not only that all watcher pointers have the same internal
		4775	structure (guaranteed by POSIX but not by ISO C for example), but it also
		4776	assumes that the same (machine) code can be used to call any watcher
		4777	callback: The watcher callbacks have different type signatures, but libev
		4778	calls them using an C<ev_watcher *> internally.
		4779
		4780	=item pointer accesses must be thread-atomic
		4781
		4782	Accessing a pointer value must be atomic, it must both be readable and
		4783	writable in one piece - this is the case on all current architectures.
		4784
3322	=item C<sig_atomic_t volatile> must be thread-atomic as well	4785	=item C<sig_atomic_t volatile> must be thread-atomic as well
3323		4786
3324	The type C<sig_atomic_t volatile> (or whatever is defined as	4787	The type C<sig_atomic_t volatile> (or whatever is defined as
3325	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different	4788	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
3326	threads. This is not part of the specification for C<sig_atomic_t>, but is	4789	threads. This is not part of the specification for C<sig_atomic_t>, but is
3327	believed to be sufficiently portable.	4790	believed to be sufficiently portable.
3328		4791
3329	=item C<sigprocmask> must work in a threaded environment	4792	=item C<sigprocmask> must work in a threaded environment
3330		4793
…		…
3339	except the initial one, and run the default loop in the initial thread as	4802	except the initial one, and run the default loop in the initial thread as
3340	well.	4803	well.
3341		4804
3342	=item C<long> must be large enough for common memory allocation sizes	4805	=item C<long> must be large enough for common memory allocation sizes
3343		4806
3344	To improve portability and simplify using libev, libev uses C<long>	4807	To improve portability and simplify its API, libev uses C<long> internally
3345	internally instead of C<size_t> when allocating its data structures. On	4808	instead of C<size_t> when allocating its data structures. On non-POSIX
3346	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4809	systems (Microsoft...) this might be unexpectedly low, but is still at
3347	is still at least 31 bits everywhere, which is enough for hundreds of	4810	least 31 bits everywhere, which is enough for hundreds of millions of
3348	millions of watchers.	4811	watchers.
3349		4812
3350	=item C<double> must hold a time value in seconds with enough accuracy	4813	=item C<double> must hold a time value in seconds with enough accuracy
3351		4814
3352	The type C<double> is used to represent timestamps. It is required to	4815	The type C<double> is used to represent timestamps. It is required to
3353	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4816	have at least 51 bits of mantissa (and 9 bits of exponent), which is
3354	enough for at least into the year 4000. This requirement is fulfilled by	4817	good enough for at least into the year 4000 with millisecond accuracy
		4818	(the design goal for libev). This requirement is overfulfilled by
3355	implementations implementing IEEE 754 (basically all existing ones).	4819	implementations using IEEE 754, which is basically all existing ones. With
		4820	IEEE 754 doubles, you get microsecond accuracy until at least 2200.
3356		4821
3357	=back	4822	=back
3358		4823
3359	If you know of other additional requirements drop me a note.	4824	If you know of other additional requirements drop me a note.
3360		4825
3361		4826
3362	=head1 VALGRIND	4827	=head1 ALGORITHMIC COMPLEXITIES
3363		4828
3364	Valgrind has a special section here because it is a popular tool that is	4829	In this section the complexities of (many of) the algorithms used inside
3365	highly useful, but valgrind reports are very hard to interpret.	4830	libev will be documented. For complexity discussions about backends see
		4831	the documentation for C<ev_default_init>.
3366		4832
3367	If you think you found a bug (memory leak, uninitialised data access etc.)	4833	All of the following are about amortised time: If an array needs to be
3368	in libev, then check twice: If valgrind reports something like:	4834	extended, libev needs to realloc and move the whole array, but this
		4835	happens asymptotically rarer with higher number of elements, so O(1) might
		4836	mean that libev does a lengthy realloc operation in rare cases, but on
		4837	average it is much faster and asymptotically approaches constant time.
3369		4838
3370	==2274== definitely lost: 0 bytes in 0 blocks.	4839	=over 4
3371	==2274== possibly lost: 0 bytes in 0 blocks.
3372	==2274== still reachable: 256 bytes in 1 blocks.
3373		4840
3374	then there is no memory leak. Similarly, under some circumstances,	4841	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3375	valgrind might report kernel bugs as if it were a bug in libev, or it
3376	might be confused (it is a very good tool, but only a tool).
3377		4842
3378	If you are unsure about something, feel free to contact the mailing list	4843	This means that, when you have a watcher that triggers in one hour and
3379	with the full valgrind report and an explanation on why you think this is	4844	there are 100 watchers that would trigger before that, then inserting will
3380	a bug in libev. However, don't be annoyed when you get a brisk "this is	4845	have to skip roughly seven (C<ld 100>) of these watchers.
3381	no bug" answer and take the chance of learning how to interpret valgrind
3382	properly.
3383		4846
3384	If you need, for some reason, empty reports from valgrind for your project	4847	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3385	I suggest using suppression lists.
3386		4848
		4849	That means that changing a timer costs less than removing/adding them,
		4850	as only the relative motion in the event queue has to be paid for.
		4851
		4852	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		4853
		4854	These just add the watcher into an array or at the head of a list.
		4855
		4856	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		4857
		4858	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		4859
		4860	These watchers are stored in lists, so they need to be walked to find the
		4861	correct watcher to remove. The lists are usually short (you don't usually
		4862	have many watchers waiting for the same fd or signal: one is typical, two
		4863	is rare).
		4864
		4865	=item Finding the next timer in each loop iteration: O(1)
		4866
		4867	By virtue of using a binary or 4-heap, the next timer is always found at a
		4868	fixed position in the storage array.
		4869
		4870	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4871
		4872	A change means an I/O watcher gets started or stopped, which requires
		4873	libev to recalculate its status (and possibly tell the kernel, depending
		4874	on backend and whether C<ev_io_set> was used).
		4875
		4876	=item Activating one watcher (putting it into the pending state): O(1)
		4877
		4878	=item Priority handling: O(number_of_priorities)
		4879
		4880	Priorities are implemented by allocating some space for each
		4881	priority. When doing priority-based operations, libev usually has to
		4882	linearly search all the priorities, but starting/stopping and activating
		4883	watchers becomes O(1) with respect to priority handling.
		4884
		4885	=item Sending an ev_async: O(1)
		4886
		4887	=item Processing ev_async_send: O(number_of_async_watchers)
		4888
		4889	=item Processing signals: O(max_signal_number)
		4890
		4891	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4892	calls in the current loop iteration. Checking for async and signal events
		4893	involves iterating over all running async watchers or all signal numbers.
		4894
		4895	=back
		4896
		4897
		4898	=head1 PORTING FROM LIBEV 3.X TO 4.X
		4899
		4900	The major version 4 introduced some incompatible changes to the API.
		4901
		4902	At the moment, the C<ev.h> header file provides compatibility definitions
		4903	for all changes, so most programs should still compile. The compatibility
		4904	layer might be removed in later versions of libev, so better update to the
		4905	new API early than late.
		4906
		4907	=over 4
		4908
		4909	=item C<EV_COMPAT3> backwards compatibility mechanism
		4910
		4911	The backward compatibility mechanism can be controlled by
		4912	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		4913	section.
		4914
		4915	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		4916
		4917	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		4918
		4919	ev_loop_destroy (EV_DEFAULT_UC);
		4920	ev_loop_fork (EV_DEFAULT);
		4921
		4922	=item function/symbol renames
		4923
		4924	A number of functions and symbols have been renamed:
		4925
		4926	ev_loop => ev_run
		4927	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		4928	EVLOOP_ONESHOT => EVRUN_ONCE
		4929
		4930	ev_unloop => ev_break
		4931	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		4932	EVUNLOOP_ONE => EVBREAK_ONE
		4933	EVUNLOOP_ALL => EVBREAK_ALL
		4934
		4935	EV_TIMEOUT => EV_TIMER
		4936
		4937	ev_loop_count => ev_iteration
		4938	ev_loop_depth => ev_depth
		4939	ev_loop_verify => ev_verify
		4940
		4941	Most functions working on C<struct ev_loop> objects don't have an
		4942	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		4943	associated constants have been renamed to not collide with the C<struct
		4944	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		4945	as all other watcher types. Note that C<ev_loop_fork> is still called
		4946	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		4947	typedef.
		4948
		4949	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		4950
		4951	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		4952	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		4953	and work, but the library code will of course be larger.
		4954
		4955	=back
		4956
		4957
		4958	=head1 GLOSSARY
		4959
		4960	=over 4
		4961
		4962	=item active
		4963
		4964	A watcher is active as long as it has been started and not yet stopped.
		4965	See L<WATCHER STATES> for details.
		4966
		4967	=item application
		4968
		4969	In this document, an application is whatever is using libev.
		4970
		4971	=item backend
		4972
		4973	The part of the code dealing with the operating system interfaces.
		4974
		4975	=item callback
		4976
		4977	The address of a function that is called when some event has been
		4978	detected. Callbacks are being passed the event loop, the watcher that
		4979	received the event, and the actual event bitset.
		4980
		4981	=item callback/watcher invocation
		4982
		4983	The act of calling the callback associated with a watcher.
		4984
		4985	=item event
		4986
		4987	A change of state of some external event, such as data now being available
		4988	for reading on a file descriptor, time having passed or simply not having
		4989	any other events happening anymore.
		4990
		4991	In libev, events are represented as single bits (such as C<EV_READ> or
		4992	C<EV_TIMER>).
		4993
		4994	=item event library
		4995
		4996	A software package implementing an event model and loop.
		4997
		4998	=item event loop
		4999
		5000	An entity that handles and processes external events and converts them
		5001	into callback invocations.
		5002
		5003	=item event model
		5004
		5005	The model used to describe how an event loop handles and processes
		5006	watchers and events.
		5007
		5008	=item pending
		5009
		5010	A watcher is pending as soon as the corresponding event has been
		5011	detected. See L<WATCHER STATES> for details.
		5012
		5013	=item real time
		5014
		5015	The physical time that is observed. It is apparently strictly monotonic :)
		5016
		5017	=item wall-clock time
		5018
		5019	The time and date as shown on clocks. Unlike real time, it can actually
		5020	be wrong and jump forwards and backwards, e.g. when the you adjust your
		5021	clock.
		5022
		5023	=item watcher
		5024
		5025	A data structure that describes interest in certain events. Watchers need
		5026	to be started (attached to an event loop) before they can receive events.
		5027
		5028	=back
3387		5029
3388	=head1 AUTHOR	5030	=head1 AUTHOR
3389		5031
3390	Marc Lehmann <libev@schmorp.de>.	5032	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5033	Magnusson and Emanuele Giaquinta.
3391		5034

Diff Legend

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