ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.204 by root, Mon Oct 27 11:08:29 2008 UTC vs.
Revision 1.388 by root, Tue Dec 20 04:08:35 2011 UTC

…		…
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
		14	#include <stdio.h> // for puts
13		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;
…		…
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_ 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	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);
…		…
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	// break 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
…		…
103	Libev is very configurable. In this manual the default (and most common)	126	Libev is very configurable. In this manual the default (and most common)
104	configuration will be described, which supports multiple event loops. For	127	configuration will be described, which supports multiple event loops. For
105	more info about various configuration options please have a look at	128	more info about various configuration options please have a look at
106	B<EMBED> section in this manual. If libev was configured without support	129	B<EMBED> section in this manual. If libev was configured without support
107	for multiple event loops, then all functions taking an initial argument of	130	for multiple event loops, then all functions taking an initial argument of
108	name C<loop> (which is always of type C<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 floating point 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.
121		145
122	=head1 ERROR HANDLING	146	=head1 ERROR HANDLING
123		147
124	Libev knows three classes of errors: operating system errors, usage errors	148	Libev knows three classes of errors: operating system errors, usage errors
125	and internal errors (bugs).	149	and internal errors (bugs).
…		…
149		173
150	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
151		175
152	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
153	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
154	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_now_update> and C<ev_now>.
155		180
156	=item ev_sleep (ev_tstamp interval)	181	=item ev_sleep (ev_tstamp interval)
157		182
158	Sleep for the given interval: The current thread will be blocked until	183	Sleep for the given interval: The current thread will be blocked
159	either it is interrupted or the given time interval has passed. Basically	184	until either it is interrupted or the given time interval has
		185	passed (approximately - it might return a bit earlier even if not
		186	interrupted). Returns immediately if C<< interval <= 0 >>.
		187
160	this is a sub-second-resolution C<sleep ()>.	188	Basically this is a sub-second-resolution C<sleep ()>.
		189
		190	The range of the C<interval> is limited - libev only guarantees to work
		191	with sleep times of up to one day (C<< interval <= 86400 >>).
161		192
162	=item int ev_version_major ()	193	=item int ev_version_major ()
163		194
164	=item int ev_version_minor ()	195	=item int ev_version_minor ()
165		196
…		…
176	as this indicates an incompatible change. Minor versions are usually	207	as this indicates an incompatible change. Minor versions are usually
177	compatible to older versions, so a larger minor version alone is usually	208	compatible to older versions, so a larger minor version alone is usually
178	not a problem.	209	not a problem.
179		210
180	Example: Make sure we haven't accidentally been linked against the wrong	211	Example: Make sure we haven't accidentally been linked against the wrong
181	version.	212	version (note, however, that this will not detect other ABI mismatches,
		213	such as LFS or reentrancy).
182		214
183	assert (("libev version mismatch",	215	assert (("libev version mismatch",
184	ev_version_major () == EV_VERSION_MAJOR	216	ev_version_major () == EV_VERSION_MAJOR
185	&& ev_version_minor () >= EV_VERSION_MINOR));	217	&& ev_version_minor () >= EV_VERSION_MINOR));
186		218
…		…
197	assert (("sorry, no epoll, no sex",	229	assert (("sorry, no epoll, no sex",
198	ev_supported_backends () & EVBACKEND_EPOLL));	230	ev_supported_backends () & EVBACKEND_EPOLL));
199		231
200	=item unsigned int ev_recommended_backends ()	232	=item unsigned int ev_recommended_backends ()
201		233
202	Return the set of all backends compiled into this binary of libev and also	234	Return the set of all backends compiled into this binary of libev and
203	recommended for this platform. This set is often smaller than the one	235	also recommended for this platform, meaning it will work for most file
		236	descriptor types. This set is often smaller than the one returned by
204	returned by C<ev_supported_backends>, as for example kqueue is broken on	237	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
205	most BSDs and will not be auto-detected unless you explicitly request it	238	and will not be auto-detected unless you explicitly request it (assuming
206	(assuming you know what you are doing). This is the set of backends that	239	you know what you are doing). This is the set of backends that libev will
207	libev will probe for if you specify no backends explicitly.	240	probe for if you specify no backends explicitly.
208		241
209	=item unsigned int ev_embeddable_backends ()	242	=item unsigned int ev_embeddable_backends ()
210		243
211	Returns the set of backends that are embeddable in other event loops. This	244	Returns the set of backends that are embeddable in other event loops. This
212	is the theoretical, all-platform, value. To find which backends	245	value is platform-specific but can include backends not available on the
213	might be supported on the current system, you would need to look at	246	current system. To find which embeddable backends might be supported on
214	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	247	the current system, you would need to look at C<ev_embeddable_backends ()
215	recommended ones.	248	& ev_supported_backends ()>, likewise for recommended ones.
216		249
217	See the description of C<ev_embed> watchers for more info.	250	See the description of C<ev_embed> watchers for more info.
218		251
219	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]	252	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
220		253
221	Sets the allocation function to use (the prototype is similar - the	254	Sets the allocation function to use (the prototype is similar - the
222	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	255	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
223	used to allocate and free memory (no surprises here). If it returns zero	256	used to allocate and free memory (no surprises here). If it returns zero
224	when memory needs to be allocated (C<size != 0>), the library might abort	257	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
250	}	283	}
251		284
252	...	285	...
253	ev_set_allocator (persistent_realloc);	286	ev_set_allocator (persistent_realloc);
254		287
255	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]	288	=item ev_set_syserr_cb (void (*cb)(const char *msg))
256		289
257	Set the callback function to call on a retryable system call error (such	290	Set the callback function to call on a retryable system call error (such
258	as failed select, poll, epoll_wait). The message is a printable string	291	as failed select, poll, epoll_wait). The message is a printable string
259	indicating the system call or subsystem causing the problem. If this	292	indicating the system call or subsystem causing the problem. If this
260	callback is set, then libev will expect it to remedy the situation, no	293	callback is set, then libev will expect it to remedy the situation, no
…		…
272	}	305	}
273		306
274	...	307	...
275	ev_set_syserr_cb (fatal_error);	308	ev_set_syserr_cb (fatal_error);
276		309
		310	=item ev_feed_signal (int signum)
		311
		312	This function can be used to "simulate" a signal receive. It is completely
		313	safe to call this function at any time, from any context, including signal
		314	handlers or random threads.
		315
		316	Its main use is to customise signal handling in your process, especially
		317	in the presence of threads. For example, you could block signals
		318	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		319	creating any loops), and in one thread, use C<sigwait> or any other
		320	mechanism to wait for signals, then "deliver" them to libev by calling
		321	C<ev_feed_signal>.
		322
277	=back	323	=back
278		324
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	325	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
280		326
281	An event loop is described by a C<struct ev_loop *> (the C<struct>	327	An event loop is described by a C<struct ev_loop *> (the C<struct> is
282	is I<not> optional in this case, as there is also an C<ev_loop>	328	I<not> optional in this case unless libev 3 compatibility is disabled, as
283	I<function>).	329	libev 3 had an C<ev_loop> function colliding with the struct name).
284		330
285	The library knows two types of such loops, the I<default> loop, which	331	The library knows two types of such loops, the I<default> loop, which
286	supports signals and child events, and dynamically created loops which do	332	supports child process events, and dynamically created event loops which
287	not.	333	do not.
288		334
289	=over 4	335	=over 4
290		336
291	=item struct ev_loop *ev_default_loop (unsigned int flags)	337	=item struct ev_loop *ev_default_loop (unsigned int flags)
292		338
293	This will initialise the default event loop if it hasn't been initialised	339	This returns the "default" event loop object, which is what you should
294	yet and return it. If the default loop could not be initialised, returns	340	normally use when you just need "the event loop". Event loop objects and
295	false. If it already was initialised it simply returns it (and ignores the	341	the C<flags> parameter are described in more detail in the entry for
296	flags. If that is troubling you, check C<ev_backend ()> afterwards).	342	C<ev_loop_new>.
		343
		344	If the default loop is already initialised then this function simply
		345	returns it (and ignores the flags. If that is troubling you, check
		346	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		347	flags, which should almost always be C<0>, unless the caller is also the
		348	one calling C<ev_run> or otherwise qualifies as "the main program".
297		349
298	If you don't know what event loop to use, use the one returned from this	350	If you don't know what event loop to use, use the one returned from this
299	function.	351	function (or via the C<EV_DEFAULT> macro).
300		352
301	Note that this function is I<not> thread-safe, so if you want to use it	353	Note that this function is I<not> thread-safe, so if you want to use it
302	from multiple threads, you have to lock (note also that this is unlikely,	354	from multiple threads, you have to employ some kind of mutex (note also
303	as loops cannot bes hared easily between threads anyway).	355	that this case is unlikely, as loops cannot be shared easily between
		356	threads anyway).
304		357
305	The default loop is the only loop that can handle C<ev_signal> and	358	The default loop is the only loop that can handle C<ev_child> watchers,
306	C<ev_child> watchers, and to do this, it always registers a handler	359	and to do this, it always registers a handler for C<SIGCHLD>. If this is
307	for C<SIGCHLD>. If this is a problem for your application you can either	360	a problem for your application you can either create a dynamic loop with
308	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	361	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
309	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	362	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
310	C<ev_default_init>.	363
		364	Example: This is the most typical usage.
		365
		366	if (!ev_default_loop (0))
		367	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		368
		369	Example: Restrict libev to the select and poll backends, and do not allow
		370	environment settings to be taken into account:
		371
		372	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		373
		374	=item struct ev_loop *ev_loop_new (unsigned int flags)
		375
		376	This will create and initialise a new event loop object. If the loop
		377	could not be initialised, returns false.
		378
		379	This function is thread-safe, and one common way to use libev with
		380	threads is indeed to create one loop per thread, and using the default
		381	loop in the "main" or "initial" thread.
311		382
312	The flags argument can be used to specify special behaviour or specific	383	The flags argument can be used to specify special behaviour or specific
313	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	384	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
314		385
315	The following flags are supported:	386	The following flags are supported:
…		…
330	useful to try out specific backends to test their performance, or to work	401	useful to try out specific backends to test their performance, or to work
331	around bugs.	402	around bugs.
332		403
333	=item C<EVFLAG_FORKCHECK>	404	=item C<EVFLAG_FORKCHECK>
334		405
335	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	406	Instead of calling C<ev_loop_fork> manually after a fork, you can also
336	a fork, you can also make libev check for a fork in each iteration by	407	make libev check for a fork in each iteration by enabling this flag.
337	enabling this flag.
338		408
339	This works by calling C<getpid ()> on every iteration of the loop,	409	This works by calling C<getpid ()> on every iteration of the loop,
340	and thus this might slow down your event loop if you do a lot of loop	410	and thus this might slow down your event loop if you do a lot of loop
341	iterations and little real work, but is usually not noticeable (on my	411	iterations and little real work, but is usually not noticeable (on my
342	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	412	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
…		…
348	flag.	418	flag.
349		419
350	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	420	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
351	environment variable.	421	environment variable.
352		422
		423	=item C<EVFLAG_NOINOTIFY>
		424
		425	When this flag is specified, then libev will not attempt to use the
		426	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
		427	testing, this flag can be useful to conserve inotify file descriptors, as
		428	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		429
		430	=item C<EVFLAG_SIGNALFD>
		431
		432	When this flag is specified, then libev will attempt to use the
		433	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
		434	delivers signals synchronously, which makes it both faster and might make
		435	it possible to get the queued signal data. It can also simplify signal
		436	handling with threads, as long as you properly block signals in your
		437	threads that are not interested in handling them.
		438
		439	Signalfd will not be used by default as this changes your signal mask, and
		440	there are a lot of shoddy libraries and programs (glib's threadpool for
		441	example) that can't properly initialise their signal masks.
		442
		443	=item C<EVFLAG_NOSIGMASK>
		444
		445	When this flag is specified, then libev will avoid to modify the signal
		446	mask. Specifically, this means you have to make sure signals are unblocked
		447	when you want to receive them.
		448
		449	This behaviour is useful when you want to do your own signal handling, or
		450	want to handle signals only in specific threads and want to avoid libev
		451	unblocking the signals.
		452
		453	It's also required by POSIX in a threaded program, as libev calls
		454	C<sigprocmask>, whose behaviour is officially unspecified.
		455
		456	This flag's behaviour will become the default in future versions of libev.
		457
353	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	458	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
354		459
355	This is your standard select(2) backend. Not I<completely> standard, as	460	This is your standard select(2) backend. Not I<completely> standard, as
356	libev tries to roll its own fd_set with no limits on the number of fds,	461	libev tries to roll its own fd_set with no limits on the number of fds,
357	but if that fails, expect a fairly low limit on the number of fds when	462	but if that fails, expect a fairly low limit on the number of fds when
…		…
381	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and	486	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
382	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.	487	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
383		488
384	=item C<EVBACKEND_EPOLL> (value 4, Linux)	489	=item C<EVBACKEND_EPOLL> (value 4, Linux)
385		490
		491	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		492	kernels).
		493
386	For few fds, this backend is a bit little slower than poll and select,	494	For few fds, this backend is a bit little slower than poll and select, but
387	but it scales phenomenally better. While poll and select usually scale	495	it scales phenomenally better. While poll and select usually scale like
388	like O(total_fds) where n is the total number of fds (or the highest fd),	496	O(total_fds) where total_fds is the total number of fds (or the highest
389	epoll scales either O(1) or O(active_fds). The epoll design has a number	497	fd), epoll scales either O(1) or O(active_fds).
390	of shortcomings, such as silently dropping events in some hard-to-detect
391	cases and requiring a system call per fd change, no fork support and bad
392	support for dup.
393		498
		499	The epoll mechanism deserves honorable mention as the most misdesigned
		500	of the more advanced event mechanisms: mere annoyances include silently
		501	dropping file descriptors, requiring a system call per change per file
		502	descriptor (and unnecessary guessing of parameters), problems with dup,
		503	returning before the timeout value, resulting in additional iterations
		504	(and only giving 5ms accuracy while select on the same platform gives
		505	0.1ms) and so on. The biggest issue is fork races, however - if a program
		506	forks then I<both> parent and child process have to recreate the epoll
		507	set, which can take considerable time (one syscall per file descriptor)
		508	and is of course hard to detect.
		509
394	Epoll is also notoriously buggy - embedding epoll fds should work, but	510	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
395	of course doesn't, and epoll just loves to report events for totally	511	but of course I<doesn't>, and epoll just loves to report events for
396	I<different> file descriptors (even already closed ones) than registered	512	totally I<different> file descriptors (even already closed ones, so
		513	one cannot even remove them from the set) than registered in the set
397	in the set (especially on SMP systems). Libev tries to counter these	514	(especially on SMP systems). Libev tries to counter these spurious
398	spurious notifications by employing an additional generation counter and	515	notifications by employing an additional generation counter and comparing
399	comparing that against the events to filter out spurious ones.	516	that against the events to filter out spurious ones, recreating the set
		517	when required. Epoll also erroneously rounds down timeouts, but gives you
		518	no way to know when and by how much, so sometimes you have to busy-wait
		519	because epoll returns immediately despite a nonzero timeout. And last
		520	not least, it also refuses to work with some file descriptors which work
		521	perfectly fine with C<select> (files, many character devices...).
		522
		523	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		524	cobbled together in a hurry, no thought to design or interaction with
		525	others. Oh, the pain, will it ever stop...
400		526
401	While stopping, setting and starting an I/O watcher in the same iteration	527	While stopping, setting and starting an I/O watcher in the same iteration
402	will result in some caching, there is still a system call per such incident	528	will result in some caching, there is still a system call per such
403	(because the fd could point to a different file description now), so its	529	incident (because the same I<file descriptor> could point to a different
404	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	530	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
405	very well if you register events for both fds.	531	file descriptors might not work very well if you register events for both
		532	file descriptors.
406		533
407	Best performance from this backend is achieved by not unregistering all	534	Best performance from this backend is achieved by not unregistering all
408	watchers for a file descriptor until it has been closed, if possible,	535	watchers for a file descriptor until it has been closed, if possible,
409	i.e. keep at least one watcher active per fd at all times. Stopping and	536	i.e. keep at least one watcher active per fd at all times. Stopping and
410	starting a watcher (without re-setting it) also usually doesn't cause	537	starting a watcher (without re-setting it) also usually doesn't cause
411	extra overhead.	538	extra overhead. A fork can both result in spurious notifications as well
		539	as in libev having to destroy and recreate the epoll object, which can
		540	take considerable time and thus should be avoided.
		541
		542	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		543	faster than epoll for maybe up to a hundred file descriptors, depending on
		544	the usage. So sad.
412		545
413	While nominally embeddable in other event loops, this feature is broken in	546	While nominally embeddable in other event loops, this feature is broken in
414	all kernel versions tested so far.	547	all kernel versions tested so far.
415		548
416	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	549	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
417	C<EVBACKEND_POLL>.	550	C<EVBACKEND_POLL>.
418		551
419	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	552	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
420		553
421	Kqueue deserves special mention, as at the time of this writing, it was	554	Kqueue deserves special mention, as at the time of this writing, it
422	broken on all BSDs except NetBSD (usually it doesn't work reliably with	555	was broken on all BSDs except NetBSD (usually it doesn't work reliably
423	anything but sockets and pipes, except on Darwin, where of course it's	556	with anything but sockets and pipes, except on Darwin, where of course
424	completely useless). For this reason it's not being "auto-detected" unless	557	it's completely useless). Unlike epoll, however, whose brokenness
425	you explicitly specify it in the flags (i.e. using C<EVBACKEND_KQUEUE>) or	558	is by design, these kqueue bugs can (and eventually will) be fixed
426	libev was compiled on a known-to-be-good (-enough) system like NetBSD.	559	without API changes to existing programs. For this reason it's not being
		560	"auto-detected" unless you explicitly specify it in the flags (i.e. using
		561	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		562	system like NetBSD.
427		563
428	You still can embed kqueue into a normal poll or select backend and use it	564	You still can embed kqueue into a normal poll or select backend and use it
429	only for sockets (after having made sure that sockets work with kqueue on	565	only for sockets (after having made sure that sockets work with kqueue on
430	the target platform). See C<ev_embed> watchers for more info.	566	the target platform). See C<ev_embed> watchers for more info.
431		567
432	It scales in the same way as the epoll backend, but the interface to the	568	It scales in the same way as the epoll backend, but the interface to the
433	kernel is more efficient (which says nothing about its actual speed, of	569	kernel is more efficient (which says nothing about its actual speed, of
434	course). While stopping, setting and starting an I/O watcher does never	570	course). While stopping, setting and starting an I/O watcher does never
435	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	571	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
436	two event changes per incident. Support for C<fork ()> is very bad and it	572	two event changes per incident. Support for C<fork ()> is very bad (but
437	drops fds silently in similarly hard-to-detect cases.	573	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		574	cases
438		575
439	This backend usually performs well under most conditions.	576	This backend usually performs well under most conditions.
440		577
441	While nominally embeddable in other event loops, this doesn't work	578	While nominally embeddable in other event loops, this doesn't work
442	everywhere, so you might need to test for this. And since it is broken	579	everywhere, so you might need to test for this. And since it is broken
443	almost everywhere, you should only use it when you have a lot of sockets	580	almost everywhere, you should only use it when you have a lot of sockets
444	(for which it usually works), by embedding it into another event loop	581	(for which it usually works), by embedding it into another event loop
445	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,	582	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
446	using it only for sockets.	583	also broken on OS X)) and, did I mention it, using it only for sockets.
447		584
448	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with	585	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
449	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with	586	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
450	C<NOTE_EOF>.	587	C<NOTE_EOF>.
451		588
…		…
459	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	596	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
460		597
461	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	598	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
462	it's really slow, but it still scales very well (O(active_fds)).	599	it's really slow, but it still scales very well (O(active_fds)).
463		600
464	Please note that Solaris event ports can deliver a lot of spurious
465	notifications, so you need to use non-blocking I/O or other means to avoid
466	blocking when no data (or space) is available.
467
468	While this backend scales well, it requires one system call per active	601	While this backend scales well, it requires one system call per active
469	file descriptor per loop iteration. For small and medium numbers of file	602	file descriptor per loop iteration. For small and medium numbers of file
470	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	603	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
471	might perform better.	604	might perform better.
472		605
473	On the positive side, with the exception of the spurious readiness	606	On the positive side, this backend actually performed fully to
474	notifications, this backend actually performed fully to specification
475	in all tests and is fully embeddable, which is a rare feat among the	607	specification in all tests and is fully embeddable, which is a rare feat
476	OS-specific backends.	608	among the OS-specific backends (I vastly prefer correctness over speed
		609	hacks).
		610
		611	On the negative side, the interface is I<bizarre> - so bizarre that
		612	even sun itself gets it wrong in their code examples: The event polling
		613	function sometimes returns events to the caller even though an error
		614	occurred, but with no indication whether it has done so or not (yes, it's
		615	even documented that way) - deadly for edge-triggered interfaces where you
		616	absolutely have to know whether an event occurred or not because you have
		617	to re-arm the watcher.
		618
		619	Fortunately libev seems to be able to work around these idiocies.
477		620
478	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	621	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
479	C<EVBACKEND_POLL>.	622	C<EVBACKEND_POLL>.
480		623
481	=item C<EVBACKEND_ALL>	624	=item C<EVBACKEND_ALL>
482		625
483	Try all backends (even potentially broken ones that wouldn't be tried	626	Try all backends (even potentially broken ones that wouldn't be tried
484	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	627	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
485	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	628	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
486		629
487	It is definitely not recommended to use this flag.	630	It is definitely not recommended to use this flag, use whatever
		631	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		632	at all.
		633
		634	=item C<EVBACKEND_MASK>
		635
		636	Not a backend at all, but a mask to select all backend bits from a
		637	C<flags> value, in case you want to mask out any backends from a flags
		638	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
488		639
489	=back	640	=back
490		641
491	If one or more of these are or'ed into the flags value, then only these	642	If one or more of the backend flags are or'ed into the flags value,
492	backends will be tried (in the reverse order as listed here). If none are	643	then only these backends will be tried (in the reverse order as listed
493	specified, all backends in C<ev_recommended_backends ()> will be tried.	644	here). If none are specified, all backends in C<ev_recommended_backends
494		645	()> will be tried.
495	Example: This is the most typical usage.
496
497	if (!ev_default_loop (0))
498	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
499
500	Example: Restrict libev to the select and poll backends, and do not allow
501	environment settings to be taken into account:
502
503	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
504
505	Example: Use whatever libev has to offer, but make sure that kqueue is
506	used if available (warning, breaks stuff, best use only with your own
507	private event loop and only if you know the OS supports your types of
508	fds):
509
510	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
511
512	=item struct ev_loop *ev_loop_new (unsigned int flags)
513
514	Similar to C<ev_default_loop>, but always creates a new event loop that is
515	always distinct from the default loop. Unlike the default loop, it cannot
516	handle signal and child watchers, and attempts to do so will be greeted by
517	undefined behaviour (or a failed assertion if assertions are enabled).
518
519	Note that this function I<is> thread-safe, and the recommended way to use
520	libev with threads is indeed to create one loop per thread, and using the
521	default loop in the "main" or "initial" thread.
522		646
523	Example: Try to create a event loop that uses epoll and nothing else.	647	Example: Try to create a event loop that uses epoll and nothing else.
524		648
525	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	649	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
526	if (!epoller)	650	if (!epoller)
527	fatal ("no epoll found here, maybe it hides under your chair");	651	fatal ("no epoll found here, maybe it hides under your chair");
528		652
		653	Example: Use whatever libev has to offer, but make sure that kqueue is
		654	used if available.
		655
		656	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		657
529	=item ev_default_destroy ()	658	=item ev_loop_destroy (loop)
530		659
531	Destroys the default loop again (frees all memory and kernel state	660	Destroys an event loop object (frees all memory and kernel state
532	etc.). None of the active event watchers will be stopped in the normal	661	etc.). None of the active event watchers will be stopped in the normal
533	sense, so e.g. C<ev_is_active> might still return true. It is your	662	sense, so e.g. C<ev_is_active> might still return true. It is your
534	responsibility to either stop all watchers cleanly yourself I<before>	663	responsibility to either stop all watchers cleanly yourself I<before>
535	calling this function, or cope with the fact afterwards (which is usually	664	calling this function, or cope with the fact afterwards (which is usually
536	the easiest thing, you can just ignore the watchers and/or C<free ()> them	665	the easiest thing, you can just ignore the watchers and/or C<free ()> them
…		…
538		667
539	Note that certain global state, such as signal state (and installed signal	668	Note that certain global state, such as signal state (and installed signal
540	handlers), will not be freed by this function, and related watchers (such	669	handlers), will not be freed by this function, and related watchers (such
541	as signal and child watchers) would need to be stopped manually.	670	as signal and child watchers) would need to be stopped manually.
542		671
543	In general it is not advisable to call this function except in the	672	This function is normally used on loop objects allocated by
544	rare occasion where you really need to free e.g. the signal handling	673	C<ev_loop_new>, but it can also be used on the default loop returned by
		674	C<ev_default_loop>, in which case it is not thread-safe.
		675
		676	Note that it is not advisable to call this function on the default loop
		677	except in the rare occasion where you really need to free its resources.
545	pipe fds. If you need dynamically allocated loops it is better to use	678	If you need dynamically allocated loops it is better to use C<ev_loop_new>
546	C<ev_loop_new> and C<ev_loop_destroy>).	679	and C<ev_loop_destroy>.
547		680
548	=item ev_loop_destroy (loop)	681	=item ev_loop_fork (loop)
549		682
550	Like C<ev_default_destroy>, but destroys an event loop created by an
551	earlier call to C<ev_loop_new>.
552
553	=item ev_default_fork ()
554
555	This function sets a flag that causes subsequent C<ev_loop> iterations	683	This function sets a flag that causes subsequent C<ev_run> iterations to
556	to reinitialise the kernel state for backends that have one. Despite the	684	reinitialise the kernel state for backends that have one. Despite the
557	name, you can call it anytime, but it makes most sense after forking, in	685	name, you can call it anytime, but it makes most sense after forking, in
558	the child process (or both child and parent, but that again makes little	686	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
559	sense). You I<must> call it in the child before using any of the libev	687	child before resuming or calling C<ev_run>.
560	functions, and it will only take effect at the next C<ev_loop> iteration.	688
		689	Again, you I<have> to call it on I<any> loop that you want to re-use after
		690	a fork, I<even if you do not plan to use the loop in the parent>. This is
		691	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		692	during fork.
561		693
562	On the other hand, you only need to call this function in the child	694	On the other hand, you only need to call this function in the child
563	process if and only if you want to use the event library in the child. If	695	process if and only if you want to use the event loop in the child. If
564	you just fork+exec, you don't have to call it at all.	696	you just fork+exec or create a new loop in the child, you don't have to
		697	call it at all (in fact, C<epoll> is so badly broken that it makes a
		698	difference, but libev will usually detect this case on its own and do a
		699	costly reset of the backend).
565		700
566	The function itself is quite fast and it's usually not a problem to call	701	The function itself is quite fast and it's usually not a problem to call
567	it just in case after a fork. To make this easy, the function will fit in	702	it just in case after a fork.
568	quite nicely into a call to C<pthread_atfork>:
569		703
		704	Example: Automate calling C<ev_loop_fork> on the default loop when
		705	using pthreads.
		706
		707	static void
		708	post_fork_child (void)
		709	{
		710	ev_loop_fork (EV_DEFAULT);
		711	}
		712
		713	...
570	pthread_atfork (0, 0, ev_default_fork);	714	pthread_atfork (0, 0, post_fork_child);
571
572	=item ev_loop_fork (loop)
573
574	Like C<ev_default_fork>, but acts on an event loop created by
575	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
576	after fork that you want to re-use in the child, and how you do this is
577	entirely your own problem.
578		715
579	=item int ev_is_default_loop (loop)	716	=item int ev_is_default_loop (loop)
580		717
581	Returns true when the given loop is, in fact, the default loop, and false	718	Returns true when the given loop is, in fact, the default loop, and false
582	otherwise.	719	otherwise.
583		720
584	=item unsigned int ev_loop_count (loop)	721	=item unsigned int ev_iteration (loop)
585		722
586	Returns the count of loop iterations for the loop, which is identical to	723	Returns the current iteration count for the event loop, which is identical
587	the number of times libev did poll for new events. It starts at C<0> and	724	to the number of times libev did poll for new events. It starts at C<0>
588	happily wraps around with enough iterations.	725	and happily wraps around with enough iterations.
589		726
590	This value can sometimes be useful as a generation counter of sorts (it	727	This value can sometimes be useful as a generation counter of sorts (it
591	"ticks" the number of loop iterations), as it roughly corresponds with	728	"ticks" the number of loop iterations), as it roughly corresponds with
592	C<ev_prepare> and C<ev_check> calls.	729	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		730	prepare and check phases.
		731
		732	=item unsigned int ev_depth (loop)
		733
		734	Returns the number of times C<ev_run> was entered minus the number of
		735	times C<ev_run> was exited normally, in other words, the recursion depth.
		736
		737	Outside C<ev_run>, this number is zero. In a callback, this number is
		738	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
		739	in which case it is higher.
		740
		741	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
		742	throwing an exception etc.), doesn't count as "exit" - consider this
		743	as a hint to avoid such ungentleman-like behaviour unless it's really
		744	convenient, in which case it is fully supported.
593		745
594	=item unsigned int ev_backend (loop)	746	=item unsigned int ev_backend (loop)
595		747
596	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	748	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
597	use.	749	use.
…		…
606		758
607	=item ev_now_update (loop)	759	=item ev_now_update (loop)
608		760
609	Establishes the current time by querying the kernel, updating the time	761	Establishes the current time by querying the kernel, updating the time
610	returned by C<ev_now ()> in the progress. This is a costly operation and	762	returned by C<ev_now ()> in the progress. This is a costly operation and
611	is usually done automatically within C<ev_loop ()>.	763	is usually done automatically within C<ev_run ()>.
612		764
613	This function is rarely useful, but when some event callback runs for a	765	This function is rarely useful, but when some event callback runs for a
614	very long time without entering the event loop, updating libev's idea of	766	very long time without entering the event loop, updating libev's idea of
615	the current time is a good idea.	767	the current time is a good idea.
616		768
617	See also "The special problem of time updates" in the C<ev_timer> section.	769	See also L<The special problem of time updates> in the C<ev_timer> section.
618		770
		771	=item ev_suspend (loop)
		772
		773	=item ev_resume (loop)
		774
		775	These two functions suspend and resume an event loop, for use when the
		776	loop is not used for a while and timeouts should not be processed.
		777
		778	A typical use case would be an interactive program such as a game: When
		779	the user presses C<^Z> to suspend the game and resumes it an hour later it
		780	would be best to handle timeouts as if no time had actually passed while
		781	the program was suspended. This can be achieved by calling C<ev_suspend>
		782	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		783	C<ev_resume> directly afterwards to resume timer processing.
		784
		785	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		786	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		787	will be rescheduled (that is, they will lose any events that would have
		788	occurred while suspended).
		789
		790	After calling C<ev_suspend> you B<must not> call I<any> function on the
		791	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		792	without a previous call to C<ev_suspend>.
		793
		794	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		795	event loop time (see C<ev_now_update>).
		796
619	=item ev_loop (loop, int flags)	797	=item ev_run (loop, int flags)
620		798
621	Finally, this is it, the event handler. This function usually is called	799	Finally, this is it, the event handler. This function usually is called
622	after you initialised all your watchers and you want to start handling	800	after you have initialised all your watchers and you want to start
623	events.	801	handling events. It will ask the operating system for any new events, call
		802	the watcher callbacks, an then repeat the whole process indefinitely: This
		803	is why event loops are called I<loops>.
624		804
625	If the flags argument is specified as C<0>, it will not return until	805	If the flags argument is specified as C<0>, it will keep handling events
626	either no event watchers are active anymore or C<ev_unloop> was called.	806	until either no event watchers are active anymore or C<ev_break> was
		807	called.
627		808
628	Please note that an explicit C<ev_unloop> is usually better than	809	Please note that an explicit C<ev_break> is usually better than
629	relying on all watchers to be stopped when deciding when a program has	810	relying on all watchers to be stopped when deciding when a program has
630	finished (especially in interactive programs), but having a program	811	finished (especially in interactive programs), but having a program
631	that automatically loops as long as it has to and no longer by virtue	812	that automatically loops as long as it has to and no longer by virtue
632	of relying on its watchers stopping correctly, that is truly a thing of	813	of relying on its watchers stopping correctly, that is truly a thing of
633	beauty.	814	beauty.
634		815
		816	This function is also I<mostly> exception-safe - you can break out of
		817	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		818	exception and so on. This does not decrement the C<ev_depth> value, nor
		819	will it clear any outstanding C<EVBREAK_ONE> breaks.
		820
635	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	821	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
636	those events and any already outstanding ones, but will not block your	822	those events and any already outstanding ones, but will not wait and
637	process in case there are no events and will return after one iteration of	823	block your process in case there are no events and will return after one
638	the loop.	824	iteration of the loop. This is sometimes useful to poll and handle new
		825	events while doing lengthy calculations, to keep the program responsive.
639		826
640	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	827	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
641	necessary) and will handle those and any already outstanding ones. It	828	necessary) and will handle those and any already outstanding ones. It
642	will block your process until at least one new event arrives (which could	829	will block your process until at least one new event arrives (which could
643	be an event internal to libev itself, so there is no guarentee that a	830	be an event internal to libev itself, so there is no guarantee that a
644	user-registered callback will be called), and will return after one	831	user-registered callback will be called), and will return after one
645	iteration of the loop.	832	iteration of the loop.
646		833
647	This is useful if you are waiting for some external event in conjunction	834	This is useful if you are waiting for some external event in conjunction
648	with something not expressible using other libev watchers (i.e. "roll your	835	with something not expressible using other libev watchers (i.e. "roll your
649	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	836	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
650	usually a better approach for this kind of thing.	837	usually a better approach for this kind of thing.
651		838
652	Here are the gory details of what C<ev_loop> does:	839	Here are the gory details of what C<ev_run> does (this is for your
		840	understanding, not a guarantee that things will work exactly like this in
		841	future versions):
653		842
		843	- Increment loop depth.
		844	- Reset the ev_break status.
654	- Before the first iteration, call any pending watchers.	845	- Before the first iteration, call any pending watchers.
		846	LOOP:
655	* If EVFLAG_FORKCHECK was used, check for a fork.	847	- If EVFLAG_FORKCHECK was used, check for a fork.
656	- If a fork was detected (by any means), queue and call all fork watchers.	848	- If a fork was detected (by any means), queue and call all fork watchers.
657	- Queue and call all prepare watchers.	849	- Queue and call all prepare watchers.
		850	- If ev_break was called, goto FINISH.
658	- If we have been forked, detach and recreate the kernel state	851	- If we have been forked, detach and recreate the kernel state
659	as to not disturb the other process.	852	as to not disturb the other process.
660	- Update the kernel state with all outstanding changes.	853	- Update the kernel state with all outstanding changes.
661	- Update the "event loop time" (ev_now ()).	854	- Update the "event loop time" (ev_now ()).
662	- Calculate for how long to sleep or block, if at all	855	- Calculate for how long to sleep or block, if at all
663	(active idle watchers, EVLOOP_NONBLOCK or not having	856	(active idle watchers, EVRUN_NOWAIT or not having
664	any active watchers at all will result in not sleeping).	857	any active watchers at all will result in not sleeping).
665	- Sleep if the I/O and timer collect interval say so.	858	- Sleep if the I/O and timer collect interval say so.
		859	- Increment loop iteration counter.
666	- Block the process, waiting for any events.	860	- Block the process, waiting for any events.
667	- Queue all outstanding I/O (fd) events.	861	- Queue all outstanding I/O (fd) events.
668	- Update the "event loop time" (ev_now ()), and do time jump adjustments.	862	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
669	- Queue all expired timers.	863	- Queue all expired timers.
670	- Queue all expired periodics.	864	- Queue all expired periodics.
671	- Unless any events are pending now, queue all idle watchers.	865	- Queue all idle watchers with priority higher than that of pending events.
672	- Queue all check watchers.	866	- Queue all check watchers.
673	- Call all queued watchers in reverse order (i.e. check watchers first).	867	- Call all queued watchers in reverse order (i.e. check watchers first).
674	Signals and child watchers are implemented as I/O watchers, and will	868	Signals and child watchers are implemented as I/O watchers, and will
675	be handled here by queueing them when their watcher gets executed.	869	be handled here by queueing them when their watcher gets executed.
676	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	870	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
677	were used, or there are no active watchers, return, otherwise	871	were used, or there are no active watchers, goto FINISH, otherwise
678	continue with step *.	872	continue with step LOOP.
		873	FINISH:
		874	- Reset the ev_break status iff it was EVBREAK_ONE.
		875	- Decrement the loop depth.
		876	- Return.
679		877
680	Example: Queue some jobs and then loop until no events are outstanding	878	Example: Queue some jobs and then loop until no events are outstanding
681	anymore.	879	anymore.
682		880
683	... queue jobs here, make sure they register event watchers as long	881	... queue jobs here, make sure they register event watchers as long
684	... as they still have work to do (even an idle watcher will do..)	882	... as they still have work to do (even an idle watcher will do..)
685	ev_loop (my_loop, 0);	883	ev_run (my_loop, 0);
686	... jobs done or somebody called unloop. yeah!	884	... jobs done or somebody called break. yeah!
687		885
688	=item ev_unloop (loop, how)	886	=item ev_break (loop, how)
689		887
690	Can be used to make a call to C<ev_loop> return early (but only after it	888	Can be used to make a call to C<ev_run> return early (but only after it
691	has processed all outstanding events). The C<how> argument must be either	889	has processed all outstanding events). The C<how> argument must be either
692	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	890	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
693	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	891	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
694		892
695	This "unloop state" will be cleared when entering C<ev_loop> again.	893	This "break state" will be cleared on the next call to C<ev_run>.
696		894
697	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.	895	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		896	which case it will have no effect.
698		897
699	=item ev_ref (loop)	898	=item ev_ref (loop)
700		899
701	=item ev_unref (loop)	900	=item ev_unref (loop)
702		901
703	Ref/unref can be used to add or remove a reference count on the event	902	Ref/unref can be used to add or remove a reference count on the event
704	loop: Every watcher keeps one reference, and as long as the reference	903	loop: Every watcher keeps one reference, and as long as the reference
705	count is nonzero, C<ev_loop> will not return on its own.	904	count is nonzero, C<ev_run> will not return on its own.
706		905
707	If you have a watcher you never unregister that should not keep C<ev_loop>	906	This is useful when you have a watcher that you never intend to
708	from returning, call ev_unref() after starting, and ev_ref() before	907	unregister, but that nevertheless should not keep C<ev_run> from
		908	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
709	stopping it.	909	before stopping it.
710		910
711	As an example, libev itself uses this for its internal signal pipe: It is	911	As an example, libev itself uses this for its internal signal pipe: It
712	not visible to the libev user and should not keep C<ev_loop> from exiting	912	is not visible to the libev user and should not keep C<ev_run> from
713	if no event watchers registered by it are active. It is also an excellent	913	exiting if no event watchers registered by it are active. It is also an
714	way to do this for generic recurring timers or from within third-party	914	excellent way to do this for generic recurring timers or from within
715	libraries. Just remember to I<unref after start> and I<ref before stop>	915	third-party libraries. Just remember to I<unref after start> and I<ref
716	(but only if the watcher wasn't active before, or was active before,	916	before stop> (but only if the watcher wasn't active before, or was active
717	respectively).	917	before, respectively. Note also that libev might stop watchers itself
		918	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		919	in the callback).
718		920
719	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	921	Example: Create a signal watcher, but keep it from keeping C<ev_run>
720	running when nothing else is active.	922	running when nothing else is active.
721		923
722	ev_signal exitsig;	924	ev_signal exitsig;
723	ev_signal_init (&exitsig, sig_cb, SIGINT);	925	ev_signal_init (&exitsig, sig_cb, SIGINT);
724	ev_signal_start (loop, &exitsig);	926	ev_signal_start (loop, &exitsig);
725	evf_unref (loop);	927	ev_unref (loop);
726		928
727	Example: For some weird reason, unregister the above signal handler again.	929	Example: For some weird reason, unregister the above signal handler again.
728		930
729	ev_ref (loop);	931	ev_ref (loop);
730	ev_signal_stop (loop, &exitsig);	932	ev_signal_stop (loop, &exitsig);
…		…
750	overhead for the actual polling but can deliver many events at once.	952	overhead for the actual polling but can deliver many events at once.
751		953
752	By setting a higher I<io collect interval> you allow libev to spend more	954	By setting a higher I<io collect interval> you allow libev to spend more
753	time collecting I/O events, so you can handle more events per iteration,	955	time collecting I/O events, so you can handle more events per iteration,
754	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	956	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
755	C<ev_timer>) will be not affected. Setting this to a non-null value will	957	C<ev_timer>) will not be affected. Setting this to a non-null value will
756	introduce an additional C<ev_sleep ()> call into most loop iterations.	958	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		959	sleep time ensures that libev will not poll for I/O events more often then
		960	once per this interval, on average (as long as the host time resolution is
		961	good enough).
757		962
758	Likewise, by setting a higher I<timeout collect interval> you allow libev	963	Likewise, by setting a higher I<timeout collect interval> you allow libev
759	to spend more time collecting timeouts, at the expense of increased	964	to spend more time collecting timeouts, at the expense of increased
760	latency/jitter/inexactness (the watcher callback will be called	965	latency/jitter/inexactness (the watcher callback will be called
761	later). C<ev_io> watchers will not be affected. Setting this to a non-null	966	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
763		968
764	Many (busy) programs can usually benefit by setting the I/O collect	969	Many (busy) programs can usually benefit by setting the I/O collect
765	interval to a value near C<0.1> or so, which is often enough for	970	interval to a value near C<0.1> or so, which is often enough for
766	interactive servers (of course not for games), likewise for timeouts. It	971	interactive servers (of course not for games), likewise for timeouts. It
767	usually doesn't make much sense to set it to a lower value than C<0.01>,	972	usually doesn't make much sense to set it to a lower value than C<0.01>,
768	as this approaches the timing granularity of most systems.	973	as this approaches the timing granularity of most systems. Note that if
		974	you do transactions with the outside world and you can't increase the
		975	parallelity, then this setting will limit your transaction rate (if you
		976	need to poll once per transaction and the I/O collect interval is 0.01,
		977	then you can't do more than 100 transactions per second).
769		978
770	Setting the I<timeout collect interval> can improve the opportunity for	979	Setting the I<timeout collect interval> can improve the opportunity for
771	saving power, as the program will "bundle" timer callback invocations that	980	saving power, as the program will "bundle" timer callback invocations that
772	are "near" in time together, by delaying some, thus reducing the number of	981	are "near" in time together, by delaying some, thus reducing the number of
773	times the process sleeps and wakes up again. Another useful technique to	982	times the process sleeps and wakes up again. Another useful technique to
774	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure	983	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
775	they fire on, say, one-second boundaries only.	984	they fire on, say, one-second boundaries only.
776		985
		986	Example: we only need 0.1s timeout granularity, and we wish not to poll
		987	more often than 100 times per second:
		988
		989	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		990	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		991
		992	=item ev_invoke_pending (loop)
		993
		994	This call will simply invoke all pending watchers while resetting their
		995	pending state. Normally, C<ev_run> does this automatically when required,
		996	but when overriding the invoke callback this call comes handy. This
		997	function can be invoked from a watcher - this can be useful for example
		998	when you want to do some lengthy calculation and want to pass further
		999	event handling to another thread (you still have to make sure only one
		1000	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
		1001
		1002	=item int ev_pending_count (loop)
		1003
		1004	Returns the number of pending watchers - zero indicates that no watchers
		1005	are pending.
		1006
		1007	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		1008
		1009	This overrides the invoke pending functionality of the loop: Instead of
		1010	invoking all pending watchers when there are any, C<ev_run> will call
		1011	this callback instead. This is useful, for example, when you want to
		1012	invoke the actual watchers inside another context (another thread etc.).
		1013
		1014	If you want to reset the callback, use C<ev_invoke_pending> as new
		1015	callback.
		1016
		1017	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		1018
		1019	Sometimes you want to share the same loop between multiple threads. This
		1020	can be done relatively simply by putting mutex_lock/unlock calls around
		1021	each call to a libev function.
		1022
		1023	However, C<ev_run> can run an indefinite time, so it is not feasible
		1024	to wait for it to return. One way around this is to wake up the event
		1025	loop via C<ev_break> and C<ev_async_send>, another way is to set these
		1026	I<release> and I<acquire> callbacks on the loop.
		1027
		1028	When set, then C<release> will be called just before the thread is
		1029	suspended waiting for new events, and C<acquire> is called just
		1030	afterwards.
		1031
		1032	Ideally, C<release> will just call your mutex_unlock function, and
		1033	C<acquire> will just call the mutex_lock function again.
		1034
		1035	While event loop modifications are allowed between invocations of
		1036	C<release> and C<acquire> (that's their only purpose after all), no
		1037	modifications done will affect the event loop, i.e. adding watchers will
		1038	have no effect on the set of file descriptors being watched, or the time
		1039	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
		1040	to take note of any changes you made.
		1041
		1042	In theory, threads executing C<ev_run> will be async-cancel safe between
		1043	invocations of C<release> and C<acquire>.
		1044
		1045	See also the locking example in the C<THREADS> section later in this
		1046	document.
		1047
		1048	=item ev_set_userdata (loop, void *data)
		1049
		1050	=item void *ev_userdata (loop)
		1051
		1052	Set and retrieve a single C<void *> associated with a loop. When
		1053	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		1054	C<0>.
		1055
		1056	These two functions can be used to associate arbitrary data with a loop,
		1057	and are intended solely for the C<invoke_pending_cb>, C<release> and
		1058	C<acquire> callbacks described above, but of course can be (ab-)used for
		1059	any other purpose as well.
		1060
777	=item ev_loop_verify (loop)	1061	=item ev_verify (loop)
778		1062
779	This function only does something when C<EV_VERIFY> support has been	1063	This function only does something when C<EV_VERIFY> support has been
780	compiled in, which is the default for non-minimal builds. It tries to go	1064	compiled in, which is the default for non-minimal builds. It tries to go
781	through all internal structures and checks them for validity. If anything	1065	through all internal structures and checks them for validity. If anything
782	is found to be inconsistent, it will print an error message to standard	1066	is found to be inconsistent, it will print an error message to standard
…		…
793		1077
794	In the following description, uppercase C<TYPE> in names stands for the	1078	In the following description, uppercase C<TYPE> in names stands for the
795	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer	1079	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
796	watchers and C<ev_io_start> for I/O watchers.	1080	watchers and C<ev_io_start> for I/O watchers.
797		1081
798	A watcher is a structure that you create and register to record your	1082	A watcher is an opaque structure that you allocate and register to record
799	interest in some event. For instance, if you want to wait for STDIN to	1083	your interest in some event. To make a concrete example, imagine you want
800	become readable, you would create an C<ev_io> watcher for that:	1084	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1085	for that:
801		1086
802	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)	1087	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
803	{	1088	{
804	ev_io_stop (w);	1089	ev_io_stop (w);
805	ev_unloop (loop, EVUNLOOP_ALL);	1090	ev_break (loop, EVBREAK_ALL);
806	}	1091	}
807		1092
808	struct ev_loop *loop = ev_default_loop (0);	1093	struct ev_loop *loop = ev_default_loop (0);
809		1094
810	ev_io stdin_watcher;	1095	ev_io stdin_watcher;
811		1096
812	ev_init (&stdin_watcher, my_cb);	1097	ev_init (&stdin_watcher, my_cb);
813	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1098	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
814	ev_io_start (loop, &stdin_watcher);	1099	ev_io_start (loop, &stdin_watcher);
815		1100
816	ev_loop (loop, 0);	1101	ev_run (loop, 0);
817		1102
818	As you can see, you are responsible for allocating the memory for your	1103	As you can see, you are responsible for allocating the memory for your
819	watcher structures (and it is I<usually> a bad idea to do this on the	1104	watcher structures (and it is I<usually> a bad idea to do this on the
820	stack).	1105	stack).
821		1106
822	Each watcher has an associated watcher structure (called C<struct ev_TYPE>	1107	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
823	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).	1108	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
824		1109
825	Each watcher structure must be initialised by a call to C<ev_init	1110	Each watcher structure must be initialised by a call to C<ev_init (watcher
826	(watcher *, callback)>, which expects a callback to be provided. This	1111	*, callback)>, which expects a callback to be provided. This callback is
827	callback gets invoked each time the event occurs (or, in the case of I/O	1112	invoked each time the event occurs (or, in the case of I/O watchers, each
828	watchers, each time the event loop detects that the file descriptor given	1113	time the event loop detects that the file descriptor given is readable
829	is readable and/or writable).	1114	and/or writable).
830		1115
831	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>	1116	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
832	macro to configure it, with arguments specific to the watcher type. There	1117	macro to configure it, with arguments specific to the watcher type. There
833	is also a macro to combine initialisation and setting in one call: C<<	1118	is also a macro to combine initialisation and setting in one call: C<<
834	ev_TYPE_init (watcher *, callback, ...) >>.	1119	ev_TYPE_init (watcher *, callback, ...) >>.
…		…
857	=item C<EV_WRITE>	1142	=item C<EV_WRITE>
858		1143
859	The file descriptor in the C<ev_io> watcher has become readable and/or	1144	The file descriptor in the C<ev_io> watcher has become readable and/or
860	writable.	1145	writable.
861		1146
862	=item C<EV_TIMEOUT>	1147	=item C<EV_TIMER>
863		1148
864	The C<ev_timer> watcher has timed out.	1149	The C<ev_timer> watcher has timed out.
865		1150
866	=item C<EV_PERIODIC>	1151	=item C<EV_PERIODIC>
867		1152
…		…
885		1170
886	=item C<EV_PREPARE>	1171	=item C<EV_PREPARE>
887		1172
888	=item C<EV_CHECK>	1173	=item C<EV_CHECK>
889		1174
890	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1175	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts
891	to gather new events, and all C<ev_check> watchers are invoked just after	1176	to gather new events, and all C<ev_check> watchers are invoked just after
892	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1177	C<ev_run> has gathered them, but before it invokes any callbacks for any
893	received events. Callbacks of both watcher types can start and stop as	1178	received events. Callbacks of both watcher types can start and stop as
894	many watchers as they want, and all of them will be taken into account	1179	many watchers as they want, and all of them will be taken into account
895	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1180	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
896	C<ev_loop> from blocking).	1181	C<ev_run> from blocking).
897		1182
898	=item C<EV_EMBED>	1183	=item C<EV_EMBED>
899		1184
900	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1185	The embedded event loop specified in the C<ev_embed> watcher needs attention.
901		1186
902	=item C<EV_FORK>	1187	=item C<EV_FORK>
903		1188
904	The event loop has been resumed in the child process after fork (see	1189	The event loop has been resumed in the child process after fork (see
905	C<ev_fork>).	1190	C<ev_fork>).
906		1191
		1192	=item C<EV_CLEANUP>
		1193
		1194	The event loop is about to be destroyed (see C<ev_cleanup>).
		1195
907	=item C<EV_ASYNC>	1196	=item C<EV_ASYNC>
908		1197
909	The given async watcher has been asynchronously notified (see C<ev_async>).	1198	The given async watcher has been asynchronously notified (see C<ev_async>).
		1199
		1200	=item C<EV_CUSTOM>
		1201
		1202	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1203	by libev users to signal watchers (e.g. via C<ev_feed_event>).
910		1204
911	=item C<EV_ERROR>	1205	=item C<EV_ERROR>
912		1206
913	An unspecified error has occurred, the watcher has been stopped. This might	1207	An unspecified error has occurred, the watcher has been stopped. This might
914	happen because the watcher could not be properly started because libev	1208	happen because the watcher could not be properly started because libev
…		…
952		1246
953	ev_io w;	1247	ev_io w;
954	ev_init (&w, my_cb);	1248	ev_init (&w, my_cb);
955	ev_io_set (&w, STDIN_FILENO, EV_READ);	1249	ev_io_set (&w, STDIN_FILENO, EV_READ);
956		1250
957	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1251	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
958		1252
959	This macro initialises the type-specific parts of a watcher. You need to	1253	This macro initialises the type-specific parts of a watcher. You need to
960	call C<ev_init> at least once before you call this macro, but you can	1254	call C<ev_init> at least once before you call this macro, but you can
961	call C<ev_TYPE_set> any number of times. You must not, however, call this	1255	call C<ev_TYPE_set> any number of times. You must not, however, call this
962	macro on a watcher that is active (it can be pending, however, which is a	1256	macro on a watcher that is active (it can be pending, however, which is a
…		…
975		1269
976	Example: Initialise and set an C<ev_io> watcher in one step.	1270	Example: Initialise and set an C<ev_io> watcher in one step.
977		1271
978	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);	1272	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
979		1273
980	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1274	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
981		1275
982	Starts (activates) the given watcher. Only active watchers will receive	1276	Starts (activates) the given watcher. Only active watchers will receive
983	events. If the watcher is already active nothing will happen.	1277	events. If the watcher is already active nothing will happen.
984		1278
985	Example: Start the C<ev_io> watcher that is being abused as example in this	1279	Example: Start the C<ev_io> watcher that is being abused as example in this
986	whole section.	1280	whole section.
987		1281
988	ev_io_start (EV_DEFAULT_UC, &w);	1282	ev_io_start (EV_DEFAULT_UC, &w);
989		1283
990	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1284	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
991		1285
992	Stops the given watcher if active, and clears the pending status (whether	1286	Stops the given watcher if active, and clears the pending status (whether
993	the watcher was active or not).	1287	the watcher was active or not).
994		1288
995	It is possible that stopped watchers are pending - for example,	1289	It is possible that stopped watchers are pending - for example,
…		…
1020	=item ev_cb_set (ev_TYPE *watcher, callback)	1314	=item ev_cb_set (ev_TYPE *watcher, callback)
1021		1315
1022	Change the callback. You can change the callback at virtually any time	1316	Change the callback. You can change the callback at virtually any time
1023	(modulo threads).	1317	(modulo threads).
1024		1318
1025	=item ev_set_priority (ev_TYPE *watcher, priority)	1319	=item ev_set_priority (ev_TYPE *watcher, int priority)
1026		1320
1027	=item int ev_priority (ev_TYPE *watcher)	1321	=item int ev_priority (ev_TYPE *watcher)
1028		1322
1029	Set and query the priority of the watcher. The priority is a small	1323	Set and query the priority of the watcher. The priority is a small
1030	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1324	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1031	(default: C<-2>). Pending watchers with higher priority will be invoked	1325	(default: C<-2>). Pending watchers with higher priority will be invoked
1032	before watchers with lower priority, but priority will not keep watchers	1326	before watchers with lower priority, but priority will not keep watchers
1033	from being executed (except for C<ev_idle> watchers).	1327	from being executed (except for C<ev_idle> watchers).
1034		1328
1035	This means that priorities are I<only> used for ordering callback
1036	invocation after new events have been received. This is useful, for
1037	example, to reduce latency after idling, or more often, to bind two
1038	watchers on the same event and make sure one is called first.
1039
1040	If you need to suppress invocation when higher priority events are pending	1329	If you need to suppress invocation when higher priority events are pending
1041	you need to look at C<ev_idle> watchers, which provide this functionality.	1330	you need to look at C<ev_idle> watchers, which provide this functionality.
1042		1331
1043	You I<must not> change the priority of a watcher as long as it is active or	1332	You I<must not> change the priority of a watcher as long as it is active or
1044	pending.	1333	pending.
1045
1046	The default priority used by watchers when no priority has been set is
1047	always C<0>, which is supposed to not be too high and not be too low :).
1048		1334
1049	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1335	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
1050	fine, as long as you do not mind that the priority value you query might	1336	fine, as long as you do not mind that the priority value you query might
1051	or might not have been clamped to the valid range.	1337	or might not have been clamped to the valid range.
		1338
		1339	The default priority used by watchers when no priority has been set is
		1340	always C<0>, which is supposed to not be too high and not be too low :).
		1341
		1342	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
		1343	priorities.
1052		1344
1053	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1345	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1054		1346
1055	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1347	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1056	C<loop> nor C<revents> need to be valid as long as the watcher callback	1348	C<loop> nor C<revents> need to be valid as long as the watcher callback
…		…
1064	watcher isn't pending it does nothing and returns C<0>.	1356	watcher isn't pending it does nothing and returns C<0>.
1065		1357
1066	Sometimes it can be useful to "poll" a watcher instead of waiting for its	1358	Sometimes it can be useful to "poll" a watcher instead of waiting for its
1067	callback to be invoked, which can be accomplished with this function.	1359	callback to be invoked, which can be accomplished with this function.
1068		1360
		1361	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1362
		1363	Feeds the given event set into the event loop, as if the specified event
		1364	had happened for the specified watcher (which must be a pointer to an
		1365	initialised but not necessarily started event watcher). Obviously you must
		1366	not free the watcher as long as it has pending events.
		1367
		1368	Stopping the watcher, letting libev invoke it, or calling
		1369	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1370	not started in the first place.
		1371
		1372	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1373	functions that do not need a watcher.
		1374
1069	=back	1375	=back
1070		1376
		1377	See also the L<ASSOCIATING CUSTOM DATA WITH A WATCHER> and L<BUILDING YOUR
		1378	OWN COMPOSITE WATCHERS> idioms.
1071		1379
1072	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1380	=head2 WATCHER STATES
1073		1381
1074	Each watcher has, by default, a member C<void *data> that you can change	1382	There are various watcher states mentioned throughout this manual -
1075	and read at any time: libev will completely ignore it. This can be used	1383	active, pending and so on. In this section these states and the rules to
1076	to associate arbitrary data with your watcher. If you need more data and	1384	transition between them will be described in more detail - and while these
1077	don't want to allocate memory and store a pointer to it in that data	1385	rules might look complicated, they usually do "the right thing".
1078	member, you can also "subclass" the watcher type and provide your own
1079	data:
1080		1386
1081	struct my_io	1387	=over 4
		1388
		1389	=item initialiased
		1390
		1391	Before a watcher can be registered with the event loop it has to be
		1392	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1393	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
		1394
		1395	In this state it is simply some block of memory that is suitable for
		1396	use in an event loop. It can be moved around, freed, reused etc. at
		1397	will - as long as you either keep the memory contents intact, or call
		1398	C<ev_TYPE_init> again.
		1399
		1400	=item started/running/active
		1401
		1402	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
		1403	property of the event loop, and is actively waiting for events. While in
		1404	this state it cannot be accessed (except in a few documented ways), moved,
		1405	freed or anything else - the only legal thing is to keep a pointer to it,
		1406	and call libev functions on it that are documented to work on active watchers.
		1407
		1408	=item pending
		1409
		1410	If a watcher is active and libev determines that an event it is interested
		1411	in has occurred (such as a timer expiring), it will become pending. It will
		1412	stay in this pending state until either it is stopped or its callback is
		1413	about to be invoked, so it is not normally pending inside the watcher
		1414	callback.
		1415
		1416	The watcher might or might not be active while it is pending (for example,
		1417	an expired non-repeating timer can be pending but no longer active). If it
		1418	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1419	but it is still property of the event loop at this time, so cannot be
		1420	moved, freed or reused. And if it is active the rules described in the
		1421	previous item still apply.
		1422
		1423	It is also possible to feed an event on a watcher that is not active (e.g.
		1424	via C<ev_feed_event>), in which case it becomes pending without being
		1425	active.
		1426
		1427	=item stopped
		1428
		1429	A watcher can be stopped implicitly by libev (in which case it might still
		1430	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1431	latter will clear any pending state the watcher might be in, regardless
		1432	of whether it was active or not, so stopping a watcher explicitly before
		1433	freeing it is often a good idea.
		1434
		1435	While stopped (and not pending) the watcher is essentially in the
		1436	initialised state, that is, it can be reused, moved, modified in any way
		1437	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1438	it again).
		1439
		1440	=back
		1441
		1442	=head2 WATCHER PRIORITY MODELS
		1443
		1444	Many event loops support I<watcher priorities>, which are usually small
		1445	integers that influence the ordering of event callback invocation
		1446	between watchers in some way, all else being equal.
		1447
		1448	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1449	description for the more technical details such as the actual priority
		1450	range.
		1451
		1452	There are two common ways how these these priorities are being interpreted
		1453	by event loops:
		1454
		1455	In the more common lock-out model, higher priorities "lock out" invocation
		1456	of lower priority watchers, which means as long as higher priority
		1457	watchers receive events, lower priority watchers are not being invoked.
		1458
		1459	The less common only-for-ordering model uses priorities solely to order
		1460	callback invocation within a single event loop iteration: Higher priority
		1461	watchers are invoked before lower priority ones, but they all get invoked
		1462	before polling for new events.
		1463
		1464	Libev uses the second (only-for-ordering) model for all its watchers
		1465	except for idle watchers (which use the lock-out model).
		1466
		1467	The rationale behind this is that implementing the lock-out model for
		1468	watchers is not well supported by most kernel interfaces, and most event
		1469	libraries will just poll for the same events again and again as long as
		1470	their callbacks have not been executed, which is very inefficient in the
		1471	common case of one high-priority watcher locking out a mass of lower
		1472	priority ones.
		1473
		1474	Static (ordering) priorities are most useful when you have two or more
		1475	watchers handling the same resource: a typical usage example is having an
		1476	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1477	timeouts. Under load, data might be received while the program handles
		1478	other jobs, but since timers normally get invoked first, the timeout
		1479	handler will be executed before checking for data. In that case, giving
		1480	the timer a lower priority than the I/O watcher ensures that I/O will be
		1481	handled first even under adverse conditions (which is usually, but not
		1482	always, what you want).
		1483
		1484	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1485	will only be executed when no same or higher priority watchers have
		1486	received events, they can be used to implement the "lock-out" model when
		1487	required.
		1488
		1489	For example, to emulate how many other event libraries handle priorities,
		1490	you can associate an C<ev_idle> watcher to each such watcher, and in
		1491	the normal watcher callback, you just start the idle watcher. The real
		1492	processing is done in the idle watcher callback. This causes libev to
		1493	continuously poll and process kernel event data for the watcher, but when
		1494	the lock-out case is known to be rare (which in turn is rare :), this is
		1495	workable.
		1496
		1497	Usually, however, the lock-out model implemented that way will perform
		1498	miserably under the type of load it was designed to handle. In that case,
		1499	it might be preferable to stop the real watcher before starting the
		1500	idle watcher, so the kernel will not have to process the event in case
		1501	the actual processing will be delayed for considerable time.
		1502
		1503	Here is an example of an I/O watcher that should run at a strictly lower
		1504	priority than the default, and which should only process data when no
		1505	other events are pending:
		1506
		1507	ev_idle idle; // actual processing watcher
		1508	ev_io io; // actual event watcher
		1509
		1510	static void
		1511	io_cb (EV_P_ ev_io *w, int revents)
1082	{	1512	{
1083	ev_io io;	1513	// stop the I/O watcher, we received the event, but
1084	int otherfd;	1514	// are not yet ready to handle it.
1085	void *somedata;	1515	ev_io_stop (EV_A_ w);
1086	struct whatever *mostinteresting;	1516
		1517	// start the idle watcher to handle the actual event.
		1518	// it will not be executed as long as other watchers
		1519	// with the default priority are receiving events.
		1520	ev_idle_start (EV_A_ &idle);
1087	};	1521	}
1088		1522
1089	...	1523	static void
1090	struct my_io w;	1524	idle_cb (EV_P_ ev_idle *w, int revents)
1091	ev_io_init (&w.io, my_cb, fd, EV_READ);
1092
1093	And since your callback will be called with a pointer to the watcher, you
1094	can cast it back to your own type:
1095
1096	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1097	{	1525	{
1098	struct my_io *w = (struct my_io *)w_;	1526	// actual processing
1099	...	1527	read (STDIN_FILENO, ...);
		1528
		1529	// have to start the I/O watcher again, as
		1530	// we have handled the event
		1531	ev_io_start (EV_P_ &io);
1100	}	1532	}
1101		1533
1102	More interesting and less C-conformant ways of casting your callback type	1534	// initialisation
1103	instead have been omitted.	1535	ev_idle_init (&idle, idle_cb);
		1536	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1537	ev_io_start (EV_DEFAULT_ &io);
1104		1538
1105	Another common scenario is to use some data structure with multiple	1539	In the "real" world, it might also be beneficial to start a timer, so that
1106	embedded watchers:	1540	low-priority connections can not be locked out forever under load. This
1107		1541	enables your program to keep a lower latency for important connections
1108	struct my_biggy	1542	during short periods of high load, while not completely locking out less
1109	{	1543	important ones.
1110	int some_data;
1111	ev_timer t1;
1112	ev_timer t2;
1113	}
1114
1115	In this case getting the pointer to C<my_biggy> is a bit more
1116	complicated: Either you store the address of your C<my_biggy> struct
1117	in the C<data> member of the watcher (for woozies), or you need to use
1118	some pointer arithmetic using C<offsetof> inside your watchers (for real
1119	programmers):
1120
1121	#include <stddef.h>
1122
1123	static void
1124	t1_cb (EV_P_ ev_timer *w, int revents)
1125	{
1126	struct my_biggy big = (struct my_biggy *
1127	(((char *)w) - offsetof (struct my_biggy, t1));
1128	}
1129
1130	static void
1131	t2_cb (EV_P_ ev_timer *w, int revents)
1132	{
1133	struct my_biggy big = (struct my_biggy *
1134	(((char *)w) - offsetof (struct my_biggy, t2));
1135	}
1136		1544
1137		1545
1138	=head1 WATCHER TYPES	1546	=head1 WATCHER TYPES
1139		1547
1140	This section describes each watcher in detail, but will not repeat	1548	This section describes each watcher in detail, but will not repeat
…		…
1164	In general you can register as many read and/or write event watchers per	1572	In general you can register as many read and/or write event watchers per
1165	fd as you want (as long as you don't confuse yourself). Setting all file	1573	fd as you want (as long as you don't confuse yourself). Setting all file
1166	descriptors to non-blocking mode is also usually a good idea (but not	1574	descriptors to non-blocking mode is also usually a good idea (but not
1167	required if you know what you are doing).	1575	required if you know what you are doing).
1168		1576
1169	If you cannot use non-blocking mode, then force the use of a
1170	known-to-be-good backend (at the time of this writing, this includes only
1171	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>).
1172
1173	Another thing you have to watch out for is that it is quite easy to	1577	Another thing you have to watch out for is that it is quite easy to
1174	receive "spurious" readiness notifications, that is your callback might	1578	receive "spurious" readiness notifications, that is, your callback might
1175	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1579	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1176	because there is no data. Not only are some backends known to create a	1580	because there is no data. It is very easy to get into this situation even
1177	lot of those (for example Solaris ports), it is very easy to get into	1581	with a relatively standard program structure. Thus it is best to always
1178	this situation even with a relatively standard program structure. Thus	1582	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1179	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1180	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1583	preferable to a program hanging until some data arrives.
1181		1584
1182	If you cannot run the fd in non-blocking mode (for example you should	1585	If you cannot run the fd in non-blocking mode (for example you should
1183	not play around with an Xlib connection), then you have to separately	1586	not play around with an Xlib connection), then you have to separately
1184	re-test whether a file descriptor is really ready with a known-to-be good	1587	re-test whether a file descriptor is really ready with a known-to-be good
1185	interface such as poll (fortunately in our Xlib example, Xlib already	1588	interface such as poll (fortunately in the case of Xlib, it already does
1186	does this on its own, so its quite safe to use). Some people additionally	1589	this on its own, so its quite safe to use). Some people additionally
1187	use C<SIGALRM> and an interval timer, just to be sure you won't block	1590	use C<SIGALRM> and an interval timer, just to be sure you won't block
1188	indefinitely.	1591	indefinitely.
1189		1592
1190	But really, best use non-blocking mode.	1593	But really, best use non-blocking mode.
1191		1594
…		…
1219		1622
1220	There is no workaround possible except not registering events	1623	There is no workaround possible except not registering events
1221	for potentially C<dup ()>'ed file descriptors, or to resort to	1624	for potentially C<dup ()>'ed file descriptors, or to resort to
1222	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1625	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1223		1626
		1627	=head3 The special problem of files
		1628
		1629	Many people try to use C<select> (or libev) on file descriptors
		1630	representing files, and expect it to become ready when their program
		1631	doesn't block on disk accesses (which can take a long time on their own).
		1632
		1633	However, this cannot ever work in the "expected" way - you get a readiness
		1634	notification as soon as the kernel knows whether and how much data is
		1635	there, and in the case of open files, that's always the case, so you
		1636	always get a readiness notification instantly, and your read (or possibly
		1637	write) will still block on the disk I/O.
		1638
		1639	Another way to view it is that in the case of sockets, pipes, character
		1640	devices and so on, there is another party (the sender) that delivers data
		1641	on its own, but in the case of files, there is no such thing: the disk
		1642	will not send data on its own, simply because it doesn't know what you
		1643	wish to read - you would first have to request some data.
		1644
		1645	Since files are typically not-so-well supported by advanced notification
		1646	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1647	to files, even though you should not use it. The reason for this is
		1648	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1649	usually a tty, often a pipe, but also sometimes files or special devices
		1650	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1651	F</dev/urandom>), and even though the file might better be served with
		1652	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1653	it "just works" instead of freezing.
		1654
		1655	So avoid file descriptors pointing to files when you know it (e.g. use
		1656	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1657	when you rarely read from a file instead of from a socket, and want to
		1658	reuse the same code path.
		1659
1224	=head3 The special problem of fork	1660	=head3 The special problem of fork
1225		1661
1226	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1662	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1227	useless behaviour. Libev fully supports fork, but needs to be told about	1663	useless behaviour. Libev fully supports fork, but needs to be told about
1228	it in the child.	1664	it in the child if you want to continue to use it in the child.
1229		1665
1230	To support fork in your programs, you either have to call	1666	To support fork in your child processes, you have to call C<ev_loop_fork
1231	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1667	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1232	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1668	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1233	C<EVBACKEND_POLL>.
1234		1669
1235	=head3 The special problem of SIGPIPE	1670	=head3 The special problem of SIGPIPE
1236		1671
1237	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1672	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1238	when writing to a pipe whose other end has been closed, your program gets	1673	when writing to a pipe whose other end has been closed, your program gets
…		…
1241		1676
1242	So when you encounter spurious, unexplained daemon exits, make sure you	1677	So when you encounter spurious, unexplained daemon exits, make sure you
1243	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1678	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1244	somewhere, as that would have given you a big clue).	1679	somewhere, as that would have given you a big clue).
1245		1680
		1681	=head3 The special problem of accept()ing when you can't
		1682
		1683	Many implementations of the POSIX C<accept> function (for example,
		1684	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1685	connection from the pending queue in all error cases.
		1686
		1687	For example, larger servers often run out of file descriptors (because
		1688	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1689	rejecting the connection, leading to libev signalling readiness on
		1690	the next iteration again (the connection still exists after all), and
		1691	typically causing the program to loop at 100% CPU usage.
		1692
		1693	Unfortunately, the set of errors that cause this issue differs between
		1694	operating systems, there is usually little the app can do to remedy the
		1695	situation, and no known thread-safe method of removing the connection to
		1696	cope with overload is known (to me).
		1697
		1698	One of the easiest ways to handle this situation is to just ignore it
		1699	- when the program encounters an overload, it will just loop until the
		1700	situation is over. While this is a form of busy waiting, no OS offers an
		1701	event-based way to handle this situation, so it's the best one can do.
		1702
		1703	A better way to handle the situation is to log any errors other than
		1704	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1705	messages, and continue as usual, which at least gives the user an idea of
		1706	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1707	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1708	usage.
		1709
		1710	If your program is single-threaded, then you could also keep a dummy file
		1711	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1712	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1713	close that fd, and create a new dummy fd. This will gracefully refuse
		1714	clients under typical overload conditions.
		1715
		1716	The last way to handle it is to simply log the error and C<exit>, as
		1717	is often done with C<malloc> failures, but this results in an easy
		1718	opportunity for a DoS attack.
1246		1719
1247	=head3 Watcher-Specific Functions	1720	=head3 Watcher-Specific Functions
1248		1721
1249	=over 4	1722	=over 4
1250		1723
…		…
1282	...	1755	...
1283	struct ev_loop *loop = ev_default_init (0);	1756	struct ev_loop *loop = ev_default_init (0);
1284	ev_io stdin_readable;	1757	ev_io stdin_readable;
1285	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1758	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1286	ev_io_start (loop, &stdin_readable);	1759	ev_io_start (loop, &stdin_readable);
1287	ev_loop (loop, 0);	1760	ev_run (loop, 0);
1288		1761
1289		1762
1290	=head2 C<ev_timer> - relative and optionally repeating timeouts	1763	=head2 C<ev_timer> - relative and optionally repeating timeouts
1291		1764
1292	Timer watchers are simple relative timers that generate an event after a	1765	Timer watchers are simple relative timers that generate an event after a
…		…
1297	year, it will still time out after (roughly) one hour. "Roughly" because	1770	year, it will still time out after (roughly) one hour. "Roughly" because
1298	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1771	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1299	monotonic clock option helps a lot here).	1772	monotonic clock option helps a lot here).
1300		1773
1301	The callback is guaranteed to be invoked only I<after> its timeout has	1774	The callback is guaranteed to be invoked only I<after> its timeout has
1302	passed, but if multiple timers become ready during the same loop iteration	1775	passed (not I<at>, so on systems with very low-resolution clocks this
1303	then order of execution is undefined.	1776	might introduce a small delay, see "the special problem of being too
		1777	early", below). If multiple timers become ready during the same loop
		1778	iteration then the ones with earlier time-out values are invoked before
		1779	ones of the same priority with later time-out values (but this is no
		1780	longer true when a callback calls C<ev_run> recursively).
1304		1781
1305	=head3 Be smart about timeouts	1782	=head3 Be smart about timeouts
1306		1783
1307	Many real-world problems involve some kind of timeout, usually for error	1784	Many real-world problems involve some kind of timeout, usually for error
1308	recovery. A typical example is an HTTP request - if the other side hangs,	1785	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1352	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>	1829	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
1353	member and C<ev_timer_again>.	1830	member and C<ev_timer_again>.
1354		1831
1355	At start:	1832	At start:
1356		1833
1357	ev_timer_init (timer, callback);	1834	ev_init (timer, callback);
1358	timer->repeat = 60.;	1835	timer->repeat = 60.;
1359	ev_timer_again (loop, timer);	1836	ev_timer_again (loop, timer);
1360		1837
1361	Each time there is some activity:	1838	Each time there is some activity:
1362		1839
…		…
1383		1860
1384	In this case, it would be more efficient to leave the C<ev_timer> alone,	1861	In this case, it would be more efficient to leave the C<ev_timer> alone,
1385	but remember the time of last activity, and check for a real timeout only	1862	but remember the time of last activity, and check for a real timeout only
1386	within the callback:	1863	within the callback:
1387		1864
		1865	ev_tstamp timeout = 60.;
1388	ev_tstamp last_activity; // time of last activity	1866	ev_tstamp last_activity; // time of last activity
		1867	ev_timer timer;
1389		1868
1390	static void	1869	static void
1391	callback (EV_P_ ev_timer *w, int revents)	1870	callback (EV_P_ ev_timer *w, int revents)
1392	{	1871	{
1393	ev_tstamp now = ev_now (EV_A);	1872	// calculate when the timeout would happen
1394	ev_tstamp timeout = last_activity + 60.;	1873	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1395		1874
1396	// if last_activity + 60. is older than now, we did time out	1875	// if negative, it means we the timeout already occured
1397	if (timeout < now)	1876	if (after < 0.)
1398	{	1877	{
1399	// timeout occured, take action	1878	// timeout occurred, take action
1400	}	1879	}
1401	else	1880	else
1402	{	1881	{
1403	// callback was invoked, but there was some activity, re-arm	1882	// callback was invoked, but there was some recent
1404	// the watcher to fire in last_activity + 60, which is	1883	// activity. simply restart the timer to time out
1405	// guaranteed to be in the future, so "again" is positive:	1884	// after "after" seconds, which is the earliest time
1406	w->again = timeout - now;	1885	// the timeout can occur.
		1886	ev_timer_set (w, after, 0.);
1407	ev_timer_again (EV_A_ w);	1887	ev_timer_start (EV_A_ w);
1408	}	1888	}
1409	}	1889	}
1410		1890
1411	To summarise the callback: first calculate the real timeout (defined	1891	To summarise the callback: first calculate in how many seconds the
1412	as "60 seconds after the last activity"), then check if that time has	1892	timeout will occur (by calculating the absolute time when it would occur,
1413	been reached, which means something I<did>, in fact, time out. Otherwise	1893	C<last_activity + timeout>, and subtracting the current time, C<ev_now
1414	the callback was invoked too early (C<timeout> is in the future), so	1894	(EV_A)> from that).
1415	re-schedule the timer to fire at that future time, to see if maybe we have
1416	a timeout then.
1417		1895
1418	Note how C<ev_timer_again> is used, taking advantage of the	1896	If this value is negative, then we are already past the timeout, i.e. we
1419	C<ev_timer_again> optimisation when the timer is already running.	1897	timed out, and need to do whatever is needed in this case.
		1898
		1899	Otherwise, we now the earliest time at which the timeout would trigger,
		1900	and simply start the timer with this timeout value.
		1901
		1902	In other words, each time the callback is invoked it will check whether
		1903	the timeout cocured. If not, it will simply reschedule itself to check
		1904	again at the earliest time it could time out. Rinse. Repeat.
1420		1905
1421	This scheme causes more callback invocations (about one every 60 seconds	1906	This scheme causes more callback invocations (about one every 60 seconds
1422	minus half the average time between activity), but virtually no calls to	1907	minus half the average time between activity), but virtually no calls to
1423	libev to change the timeout.	1908	libev to change the timeout.
1424		1909
1425	To start the timer, simply initialise the watcher and set C<last_activity>	1910	To start the machinery, simply initialise the watcher and set
1426	to the current time (meaning we just have some activity :), then call the	1911	C<last_activity> to the current time (meaning there was some activity just
1427	callback, which will "do the right thing" and start the timer:	1912	now), then call the callback, which will "do the right thing" and start
		1913	the timer:
1428		1914
		1915	last_activity = ev_now (EV_A);
1429	ev_timer_init (timer, callback);	1916	ev_init (&timer, callback);
1430	last_activity = ev_now (loop);	1917	callback (EV_A_ &timer, 0);
1431	callback (loop, timer, EV_TIMEOUT);
1432		1918
1433	And when there is some activity, simply store the current time in	1919	When there is some activity, simply store the current time in
1434	C<last_activity>, no libev calls at all:	1920	C<last_activity>, no libev calls at all:
1435		1921
		1922	if (activity detected)
1436	last_actiivty = ev_now (loop);	1923	last_activity = ev_now (EV_A);
		1924
		1925	When your timeout value changes, then the timeout can be changed by simply
		1926	providing a new value, stopping the timer and calling the callback, which
		1927	will agaion do the right thing (for example, time out immediately :).
		1928
		1929	timeout = new_value;
		1930	ev_timer_stop (EV_A_ &timer);
		1931	callback (EV_A_ &timer, 0);
1437		1932
1438	This technique is slightly more complex, but in most cases where the	1933	This technique is slightly more complex, but in most cases where the
1439	time-out is unlikely to be triggered, much more efficient.	1934	time-out is unlikely to be triggered, much more efficient.
1440
1441	Changing the timeout is trivial as well (if it isn't hard-coded in the
1442	callback :) - just change the timeout and invoke the callback, which will
1443	fix things for you.
1444		1935
1445	=item 4. Wee, just use a double-linked list for your timeouts.	1936	=item 4. Wee, just use a double-linked list for your timeouts.
1446		1937
1447	If there is not one request, but many thousands (millions...), all	1938	If there is not one request, but many thousands (millions...), all
1448	employing some kind of timeout with the same timeout value, then one can	1939	employing some kind of timeout with the same timeout value, then one can
…		…
1475	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is	1966	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1476	rather complicated, but extremely efficient, something that really pays	1967	rather complicated, but extremely efficient, something that really pays
1477	off after the first million or so of active timers, i.e. it's usually	1968	off after the first million or so of active timers, i.e. it's usually
1478	overkill :)	1969	overkill :)
1479		1970
		1971	=head3 The special problem of being too early
		1972
		1973	If you ask a timer to call your callback after three seconds, then
		1974	you expect it to be invoked after three seconds - but of course, this
		1975	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		1976	guaranteed to any precision by libev - imagine somebody suspending the
		1977	process with a STOP signal for a few hours for example.
		1978
		1979	So, libev tries to invoke your callback as soon as possible I<after> the
		1980	delay has occurred, but cannot guarantee this.
		1981
		1982	A less obvious failure mode is calling your callback too early: many event
		1983	loops compare timestamps with a "elapsed delay >= requested delay", but
		1984	this can cause your callback to be invoked much earlier than you would
		1985	expect.
		1986
		1987	To see why, imagine a system with a clock that only offers full second
		1988	resolution (think windows if you can't come up with a broken enough OS
		1989	yourself). If you schedule a one-second timer at the time 500.9, then the
		1990	event loop will schedule your timeout to elapse at a system time of 500
		1991	(500.9 truncated to the resolution) + 1, or 501.
		1992
		1993	If an event library looks at the timeout 0.1s later, it will see "501 >=
		1994	501" and invoke the callback 0.1s after it was started, even though a
		1995	one-second delay was requested - this is being "too early", despite best
		1996	intentions.
		1997
		1998	This is the reason why libev will never invoke the callback if the elapsed
		1999	delay equals the requested delay, but only when the elapsed delay is
		2000	larger than the requested delay. In the example above, libev would only invoke
		2001	the callback at system time 502, or 1.1s after the timer was started.
		2002
		2003	So, while libev cannot guarantee that your callback will be invoked
		2004	exactly when requested, it I<can> and I<does> guarantee that the requested
		2005	delay has actually elapsed, or in other words, it always errs on the "too
		2006	late" side of things.
		2007
1480	=head3 The special problem of time updates	2008	=head3 The special problem of time updates
1481		2009
1482	Establishing the current time is a costly operation (it usually takes at	2010	Establishing the current time is a costly operation (it usually takes
1483	least two system calls): EV therefore updates its idea of the current	2011	at least one system call): EV therefore updates its idea of the current
1484	time only before and after C<ev_loop> collects new events, which causes a	2012	time only before and after C<ev_run> collects new events, which causes a
1485	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2013	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1486	lots of events in one iteration.	2014	lots of events in one iteration.
1487		2015
1488	The relative timeouts are calculated relative to the C<ev_now ()>	2016	The relative timeouts are calculated relative to the C<ev_now ()>
1489	time. This is usually the right thing as this timestamp refers to the time	2017	time. This is usually the right thing as this timestamp refers to the time
…		…
1495		2023
1496	If the event loop is suspended for a long time, you can also force an	2024	If the event loop is suspended for a long time, you can also force an
1497	update of the time returned by C<ev_now ()> by calling C<ev_now_update	2025	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1498	()>.	2026	()>.
1499		2027
		2028	=head3 The special problem of unsynchronised clocks
		2029
		2030	Modern systems have a variety of clocks - libev itself uses the normal
		2031	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2032	jumps).
		2033
		2034	Neither of these clocks is synchronised with each other or any other clock
		2035	on the system, so C<ev_time ()> might return a considerably different time
		2036	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2037	a call to C<gettimeofday> might return a second count that is one higher
		2038	than a directly following call to C<time>.
		2039
		2040	The moral of this is to only compare libev-related timestamps with
		2041	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2042	a second or so.
		2043
		2044	One more problem arises due to this lack of synchronisation: if libev uses
		2045	the system monotonic clock and you compare timestamps from C<ev_time>
		2046	or C<ev_now> from when you started your timer and when your callback is
		2047	invoked, you will find that sometimes the callback is a bit "early".
		2048
		2049	This is because C<ev_timer>s work in real time, not wall clock time, so
		2050	libev makes sure your callback is not invoked before the delay happened,
		2051	I<measured according to the real time>, not the system clock.
		2052
		2053	If your timeouts are based on a physical timescale (e.g. "time out this
		2054	connection after 100 seconds") then this shouldn't bother you as it is
		2055	exactly the right behaviour.
		2056
		2057	If you want to compare wall clock/system timestamps to your timers, then
		2058	you need to use C<ev_periodic>s, as these are based on the wall clock
		2059	time, where your comparisons will always generate correct results.
		2060
		2061	=head3 The special problems of suspended animation
		2062
		2063	When you leave the server world it is quite customary to hit machines that
		2064	can suspend/hibernate - what happens to the clocks during such a suspend?
		2065
		2066	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		2067	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		2068	to run until the system is suspended, but they will not advance while the
		2069	system is suspended. That means, on resume, it will be as if the program
		2070	was frozen for a few seconds, but the suspend time will not be counted
		2071	towards C<ev_timer> when a monotonic clock source is used. The real time
		2072	clock advanced as expected, but if it is used as sole clocksource, then a
		2073	long suspend would be detected as a time jump by libev, and timers would
		2074	be adjusted accordingly.
		2075
		2076	I would not be surprised to see different behaviour in different between
		2077	operating systems, OS versions or even different hardware.
		2078
		2079	The other form of suspend (job control, or sending a SIGSTOP) will see a
		2080	time jump in the monotonic clocks and the realtime clock. If the program
		2081	is suspended for a very long time, and monotonic clock sources are in use,
		2082	then you can expect C<ev_timer>s to expire as the full suspension time
		2083	will be counted towards the timers. When no monotonic clock source is in
		2084	use, then libev will again assume a timejump and adjust accordingly.
		2085
		2086	It might be beneficial for this latter case to call C<ev_suspend>
		2087	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		2088	deterministic behaviour in this case (you can do nothing against
		2089	C<SIGSTOP>).
		2090
1500	=head3 Watcher-Specific Functions and Data Members	2091	=head3 Watcher-Specific Functions and Data Members
1501		2092
1502	=over 4	2093	=over 4
1503		2094
1504	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2095	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
…		…
1517	keep up with the timer (because it takes longer than those 10 seconds to	2108	keep up with the timer (because it takes longer than those 10 seconds to
1518	do stuff) the timer will not fire more than once per event loop iteration.	2109	do stuff) the timer will not fire more than once per event loop iteration.
1519		2110
1520	=item ev_timer_again (loop, ev_timer *)	2111	=item ev_timer_again (loop, ev_timer *)
1521		2112
1522	This will act as if the timer timed out and restart it again if it is	2113	This will act as if the timer timed out and restarts it again if it is
1523	repeating. The exact semantics are:	2114	repeating. The exact semantics are:
1524		2115
1525	If the timer is pending, its pending status is cleared.	2116	If the timer is pending, its pending status is cleared.
1526		2117
1527	If the timer is started but non-repeating, stop it (as if it timed out).	2118	If the timer is started but non-repeating, stop it (as if it timed out).
1528		2119
1529	If the timer is repeating, either start it if necessary (with the	2120	If the timer is repeating, either start it if necessary (with the
1530	C<repeat> value), or reset the running timer to the C<repeat> value.	2121	C<repeat> value), or reset the running timer to the C<repeat> value.
1531		2122
1532	This sounds a bit complicated, see "Be smart about timeouts", above, for a	2123	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1533	usage example.	2124	usage example.
		2125
		2126	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
		2127
		2128	Returns the remaining time until a timer fires. If the timer is active,
		2129	then this time is relative to the current event loop time, otherwise it's
		2130	the timeout value currently configured.
		2131
		2132	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
		2133	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		2134	will return C<4>. When the timer expires and is restarted, it will return
		2135	roughly C<7> (likely slightly less as callback invocation takes some time,
		2136	too), and so on.
1534		2137
1535	=item ev_tstamp repeat [read-write]	2138	=item ev_tstamp repeat [read-write]
1536		2139
1537	The current C<repeat> value. Will be used each time the watcher times out	2140	The current C<repeat> value. Will be used each time the watcher times out
1538	or C<ev_timer_again> is called, and determines the next timeout (if any),	2141	or C<ev_timer_again> is called, and determines the next timeout (if any),
…		…
1564	}	2167	}
1565		2168
1566	ev_timer mytimer;	2169	ev_timer mytimer;
1567	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2170	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1568	ev_timer_again (&mytimer); /* start timer */	2171	ev_timer_again (&mytimer); /* start timer */
1569	ev_loop (loop, 0);	2172	ev_run (loop, 0);
1570		2173
1571	// and in some piece of code that gets executed on any "activity":	2174	// and in some piece of code that gets executed on any "activity":
1572	// reset the timeout to start ticking again at 10 seconds	2175	// reset the timeout to start ticking again at 10 seconds
1573	ev_timer_again (&mytimer);	2176	ev_timer_again (&mytimer);
1574		2177
…		…
1576	=head2 C<ev_periodic> - to cron or not to cron?	2179	=head2 C<ev_periodic> - to cron or not to cron?
1577		2180
1578	Periodic watchers are also timers of a kind, but they are very versatile	2181	Periodic watchers are also timers of a kind, but they are very versatile
1579	(and unfortunately a bit complex).	2182	(and unfortunately a bit complex).
1580		2183
1581	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	2184	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1582	but on wall clock time (absolute time). You can tell a periodic watcher	2185	relative time, the physical time that passes) but on wall clock time
1583	to trigger after some specific point in time. For example, if you tell a	2186	(absolute time, the thing you can read on your calender or clock). The
1584	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	2187	difference is that wall clock time can run faster or slower than real
1585	+ 10.>, that is, an absolute time not a delay) and then reset your system	2188	time, and time jumps are not uncommon (e.g. when you adjust your
1586	clock to January of the previous year, then it will take more than year	2189	wrist-watch).
1587	to trigger the event (unlike an C<ev_timer>, which would still trigger
1588	roughly 10 seconds later as it uses a relative timeout).
1589		2190
		2191	You can tell a periodic watcher to trigger after some specific point
		2192	in time: for example, if you tell a periodic watcher to trigger "in 10
		2193	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		2194	not a delay) and then reset your system clock to January of the previous
		2195	year, then it will take a year or more to trigger the event (unlike an
		2196	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		2197	it, as it uses a relative timeout).
		2198
1590	C<ev_periodic>s can also be used to implement vastly more complex timers,	2199	C<ev_periodic> watchers can also be used to implement vastly more complex
1591	such as triggering an event on each "midnight, local time", or other	2200	timers, such as triggering an event on each "midnight, local time", or
1592	complicated rules.	2201	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		2202	those cannot react to time jumps.
1593		2203
1594	As with timers, the callback is guaranteed to be invoked only when the	2204	As with timers, the callback is guaranteed to be invoked only when the
1595	time (C<at>) has passed, but if multiple periodic timers become ready	2205	point in time where it is supposed to trigger has passed. If multiple
1596	during the same loop iteration, then order of execution is undefined.	2206	timers become ready during the same loop iteration then the ones with
		2207	earlier time-out values are invoked before ones with later time-out values
		2208	(but this is no longer true when a callback calls C<ev_run> recursively).
1597		2209
1598	=head3 Watcher-Specific Functions and Data Members	2210	=head3 Watcher-Specific Functions and Data Members
1599		2211
1600	=over 4	2212	=over 4
1601		2213
1602	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	2214	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1603		2215
1604	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2216	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1605		2217
1606	Lots of arguments, lets sort it out... There are basically three modes of	2218	Lots of arguments, let's sort it out... There are basically three modes of
1607	operation, and we will explain them from simplest to most complex:	2219	operation, and we will explain them from simplest to most complex:
1608		2220
1609	=over 4	2221	=over 4
1610		2222
1611	=item * absolute timer (at = time, interval = reschedule_cb = 0)	2223	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1612		2224
1613	In this configuration the watcher triggers an event after the wall clock	2225	In this configuration the watcher triggers an event after the wall clock
1614	time C<at> has passed. It will not repeat and will not adjust when a time	2226	time C<offset> has passed. It will not repeat and will not adjust when a
1615	jump occurs, that is, if it is to be run at January 1st 2011 then it will	2227	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1616	only run when the system clock reaches or surpasses this time.	2228	will be stopped and invoked when the system clock reaches or surpasses
		2229	this point in time.
1617		2230
1618	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	2231	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1619		2232
1620	In this mode the watcher will always be scheduled to time out at the next	2233	In this mode the watcher will always be scheduled to time out at the next
1621	C<at + N * interval> time (for some integer N, which can also be negative)	2234	C<offset + N * interval> time (for some integer N, which can also be
1622	and then repeat, regardless of any time jumps.	2235	negative) and then repeat, regardless of any time jumps. The C<offset>
		2236	argument is merely an offset into the C<interval> periods.
1623		2237
1624	This can be used to create timers that do not drift with respect to the	2238	This can be used to create timers that do not drift with respect to the
1625	system clock, for example, here is a C<ev_periodic> that triggers each	2239	system clock, for example, here is an C<ev_periodic> that triggers each
1626	hour, on the hour:	2240	hour, on the hour (with respect to UTC):
1627		2241
1628	ev_periodic_set (&periodic, 0., 3600., 0);	2242	ev_periodic_set (&periodic, 0., 3600., 0);
1629		2243
1630	This doesn't mean there will always be 3600 seconds in between triggers,	2244	This doesn't mean there will always be 3600 seconds in between triggers,
1631	but only that the callback will be called when the system time shows a	2245	but only that the callback will be called when the system time shows a
1632	full hour (UTC), or more correctly, when the system time is evenly divisible	2246	full hour (UTC), or more correctly, when the system time is evenly divisible
1633	by 3600.	2247	by 3600.
1634		2248
1635	Another way to think about it (for the mathematically inclined) is that	2249	Another way to think about it (for the mathematically inclined) is that
1636	C<ev_periodic> will try to run the callback in this mode at the next possible	2250	C<ev_periodic> will try to run the callback in this mode at the next possible
1637	time where C<time = at (mod interval)>, regardless of any time jumps.	2251	time where C<time = offset (mod interval)>, regardless of any time jumps.
1638		2252
1639	For numerical stability it is preferable that the C<at> value is near	2253	The C<interval> I<MUST> be positive, and for numerical stability, the
1640	C<ev_now ()> (the current time), but there is no range requirement for	2254	interval value should be higher than C<1/8192> (which is around 100
1641	this value, and in fact is often specified as zero.	2255	microseconds) and C<offset> should be higher than C<0> and should have
		2256	at most a similar magnitude as the current time (say, within a factor of
		2257	ten). Typical values for offset are, in fact, C<0> or something between
		2258	C<0> and C<interval>, which is also the recommended range.
1642		2259
1643	Note also that there is an upper limit to how often a timer can fire (CPU	2260	Note also that there is an upper limit to how often a timer can fire (CPU
1644	speed for example), so if C<interval> is very small then timing stability	2261	speed for example), so if C<interval> is very small then timing stability
1645	will of course deteriorate. Libev itself tries to be exact to be about one	2262	will of course deteriorate. Libev itself tries to be exact to be about one
1646	millisecond (if the OS supports it and the machine is fast enough).	2263	millisecond (if the OS supports it and the machine is fast enough).
1647		2264
1648	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2265	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1649		2266
1650	In this mode the values for C<interval> and C<at> are both being	2267	In this mode the values for C<interval> and C<offset> are both being
1651	ignored. Instead, each time the periodic watcher gets scheduled, the	2268	ignored. Instead, each time the periodic watcher gets scheduled, the
1652	reschedule callback will be called with the watcher as first, and the	2269	reschedule callback will be called with the watcher as first, and the
1653	current time as second argument.	2270	current time as second argument.
1654		2271
1655	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2272	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1656	ever, or make ANY event loop modifications whatsoever>.	2273	or make ANY other event loop modifications whatsoever, unless explicitly
		2274	allowed by documentation here>.
1657		2275
1658	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2276	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1659	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2277	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1660	only event loop modification you are allowed to do).	2278	only event loop modification you are allowed to do).
1661		2279
…		…
1691	a different time than the last time it was called (e.g. in a crond like	2309	a different time than the last time it was called (e.g. in a crond like
1692	program when the crontabs have changed).	2310	program when the crontabs have changed).
1693		2311
1694	=item ev_tstamp ev_periodic_at (ev_periodic *)	2312	=item ev_tstamp ev_periodic_at (ev_periodic *)
1695		2313
1696	When active, returns the absolute time that the watcher is supposed to	2314	When active, returns the absolute time that the watcher is supposed
1697	trigger next.	2315	to trigger next. This is not the same as the C<offset> argument to
		2316	C<ev_periodic_set>, but indeed works even in interval and manual
		2317	rescheduling modes.
1698		2318
1699	=item ev_tstamp offset [read-write]	2319	=item ev_tstamp offset [read-write]
1700		2320
1701	When repeating, this contains the offset value, otherwise this is the	2321	When repeating, this contains the offset value, otherwise this is the
1702	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2322	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2323	although libev might modify this value for better numerical stability).
1703		2324
1704	Can be modified any time, but changes only take effect when the periodic	2325	Can be modified any time, but changes only take effect when the periodic
1705	timer fires or C<ev_periodic_again> is being called.	2326	timer fires or C<ev_periodic_again> is being called.
1706		2327
1707	=item ev_tstamp interval [read-write]	2328	=item ev_tstamp interval [read-write]
…		…
1723	Example: Call a callback every hour, or, more precisely, whenever the	2344	Example: Call a callback every hour, or, more precisely, whenever the
1724	system time is divisible by 3600. The callback invocation times have	2345	system time is divisible by 3600. The callback invocation times have
1725	potentially a lot of jitter, but good long-term stability.	2346	potentially a lot of jitter, but good long-term stability.
1726		2347
1727	static void	2348	static void
1728	clock_cb (struct ev_loop *loop, ev_io *w, int revents)	2349	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
1729	{	2350	{
1730	... its now a full hour (UTC, or TAI or whatever your clock follows)	2351	... its now a full hour (UTC, or TAI or whatever your clock follows)
1731	}	2352	}
1732		2353
1733	ev_periodic hourly_tick;	2354	ev_periodic hourly_tick;
…		…
1756		2377
1757	=head2 C<ev_signal> - signal me when a signal gets signalled!	2378	=head2 C<ev_signal> - signal me when a signal gets signalled!
1758		2379
1759	Signal watchers will trigger an event when the process receives a specific	2380	Signal watchers will trigger an event when the process receives a specific
1760	signal one or more times. Even though signals are very asynchronous, libev	2381	signal one or more times. Even though signals are very asynchronous, libev
1761	will try it's best to deliver signals synchronously, i.e. as part of the	2382	will try its best to deliver signals synchronously, i.e. as part of the
1762	normal event processing, like any other event.	2383	normal event processing, like any other event.
1763		2384
1764	If you want signals asynchronously, just use C<sigaction> as you would	2385	If you want signals to be delivered truly asynchronously, just use
1765	do without libev and forget about sharing the signal. You can even use	2386	C<sigaction> as you would do without libev and forget about sharing
1766	C<ev_async> from a signal handler to synchronously wake up an event loop.	2387	the signal. You can even use C<ev_async> from a signal handler to
		2388	synchronously wake up an event loop.
1767		2389
1768	You can configure as many watchers as you like per signal. Only when the	2390	You can configure as many watchers as you like for the same signal, but
		2391	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2392	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2393	C<SIGINT> in both the default loop and another loop at the same time. At
		2394	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2395
1769	first watcher gets started will libev actually register a signal handler	2396	When the first watcher gets started will libev actually register something
1770	with the kernel (thus it coexists with your own signal handlers as long as	2397	with the kernel (thus it coexists with your own signal handlers as long as
1771	you don't register any with libev for the same signal). Similarly, when	2398	you don't register any with libev for the same signal).
1772	the last signal watcher for a signal is stopped, libev will reset the
1773	signal handler to SIG_DFL (regardless of what it was set to before).
1774		2399
1775	If possible and supported, libev will install its handlers with	2400	If possible and supported, libev will install its handlers with
1776	C<SA_RESTART> behaviour enabled, so system calls should not be unduly	2401	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1777	interrupted. If you have a problem with system calls getting interrupted by	2402	not be unduly interrupted. If you have a problem with system calls getting
1778	signals you can block all signals in an C<ev_check> watcher and unblock	2403	interrupted by signals you can block all signals in an C<ev_check> watcher
1779	them in an C<ev_prepare> watcher.	2404	and unblock them in an C<ev_prepare> watcher.
		2405
		2406	=head3 The special problem of inheritance over fork/execve/pthread_create
		2407
		2408	Both the signal mask (C<sigprocmask>) and the signal disposition
		2409	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2410	stopping it again), that is, libev might or might not block the signal,
		2411	and might or might not set or restore the installed signal handler (but
		2412	see C<EVFLAG_NOSIGMASK>).
		2413
		2414	While this does not matter for the signal disposition (libev never
		2415	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2416	C<execve>), this matters for the signal mask: many programs do not expect
		2417	certain signals to be blocked.
		2418
		2419	This means that before calling C<exec> (from the child) you should reset
		2420	the signal mask to whatever "default" you expect (all clear is a good
		2421	choice usually).
		2422
		2423	The simplest way to ensure that the signal mask is reset in the child is
		2424	to install a fork handler with C<pthread_atfork> that resets it. That will
		2425	catch fork calls done by libraries (such as the libc) as well.
		2426
		2427	In current versions of libev, the signal will not be blocked indefinitely
		2428	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2429	the window of opportunity for problems, it will not go away, as libev
		2430	I<has> to modify the signal mask, at least temporarily.
		2431
		2432	So I can't stress this enough: I<If you do not reset your signal mask when
		2433	you expect it to be empty, you have a race condition in your code>. This
		2434	is not a libev-specific thing, this is true for most event libraries.
		2435
		2436	=head3 The special problem of threads signal handling
		2437
		2438	POSIX threads has problematic signal handling semantics, specifically,
		2439	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2440	threads in a process block signals, which is hard to achieve.
		2441
		2442	When you want to use sigwait (or mix libev signal handling with your own
		2443	for the same signals), you can tackle this problem by globally blocking
		2444	all signals before creating any threads (or creating them with a fully set
		2445	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2446	loops. Then designate one thread as "signal receiver thread" which handles
		2447	these signals. You can pass on any signals that libev might be interested
		2448	in by calling C<ev_feed_signal>.
1780		2449
1781	=head3 Watcher-Specific Functions and Data Members	2450	=head3 Watcher-Specific Functions and Data Members
1782		2451
1783	=over 4	2452	=over 4
1784		2453
…		…
1800	Example: Try to exit cleanly on SIGINT.	2469	Example: Try to exit cleanly on SIGINT.
1801		2470
1802	static void	2471	static void
1803	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)	2472	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1804	{	2473	{
1805	ev_unloop (loop, EVUNLOOP_ALL);	2474	ev_break (loop, EVBREAK_ALL);
1806	}	2475	}
1807		2476
1808	ev_signal signal_watcher;	2477	ev_signal signal_watcher;
1809	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2478	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1810	ev_signal_start (loop, &signal_watcher);	2479	ev_signal_start (loop, &signal_watcher);
…		…
1816	some child status changes (most typically when a child of yours dies or	2485	some child status changes (most typically when a child of yours dies or
1817	exits). It is permissible to install a child watcher I<after> the child	2486	exits). It is permissible to install a child watcher I<after> the child
1818	has been forked (which implies it might have already exited), as long	2487	has been forked (which implies it might have already exited), as long
1819	as the event loop isn't entered (or is continued from a watcher), i.e.,	2488	as the event loop isn't entered (or is continued from a watcher), i.e.,
1820	forking and then immediately registering a watcher for the child is fine,	2489	forking and then immediately registering a watcher for the child is fine,
1821	but forking and registering a watcher a few event loop iterations later is	2490	but forking and registering a watcher a few event loop iterations later or
1822	not.	2491	in the next callback invocation is not.
1823		2492
1824	Only the default event loop is capable of handling signals, and therefore	2493	Only the default event loop is capable of handling signals, and therefore
1825	you can only register child watchers in the default event loop.	2494	you can only register child watchers in the default event loop.
1826		2495
		2496	Due to some design glitches inside libev, child watchers will always be
		2497	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2498	libev)
		2499
1827	=head3 Process Interaction	2500	=head3 Process Interaction
1828		2501
1829	Libev grabs C<SIGCHLD> as soon as the default event loop is	2502	Libev grabs C<SIGCHLD> as soon as the default event loop is
1830	initialised. This is necessary to guarantee proper behaviour even if	2503	initialised. This is necessary to guarantee proper behaviour even if the
1831	the first child watcher is started after the child exits. The occurrence	2504	first child watcher is started after the child exits. The occurrence
1832	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2505	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1833	synchronously as part of the event loop processing. Libev always reaps all	2506	synchronously as part of the event loop processing. Libev always reaps all
1834	children, even ones not watched.	2507	children, even ones not watched.
1835		2508
1836	=head3 Overriding the Built-In Processing	2509	=head3 Overriding the Built-In Processing
…		…
1846	=head3 Stopping the Child Watcher	2519	=head3 Stopping the Child Watcher
1847		2520
1848	Currently, the child watcher never gets stopped, even when the	2521	Currently, the child watcher never gets stopped, even when the
1849	child terminates, so normally one needs to stop the watcher in the	2522	child terminates, so normally one needs to stop the watcher in the
1850	callback. Future versions of libev might stop the watcher automatically	2523	callback. Future versions of libev might stop the watcher automatically
1851	when a child exit is detected.	2524	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2525	problem).
1852		2526
1853	=head3 Watcher-Specific Functions and Data Members	2527	=head3 Watcher-Specific Functions and Data Members
1854		2528
1855	=over 4	2529	=over 4
1856		2530
…		…
1913		2587
1914		2588
1915	=head2 C<ev_stat> - did the file attributes just change?	2589	=head2 C<ev_stat> - did the file attributes just change?
1916		2590
1917	This watches a file system path for attribute changes. That is, it calls	2591	This watches a file system path for attribute changes. That is, it calls
1918	C<stat> regularly (or when the OS says it changed) and sees if it changed	2592	C<stat> on that path in regular intervals (or when the OS says it changed)
1919	compared to the last time, invoking the callback if it did.	2593	and sees if it changed compared to the last time, invoking the callback if
		2594	it did.
1920		2595
1921	The path does not need to exist: changing from "path exists" to "path does	2596	The path does not need to exist: changing from "path exists" to "path does
1922	not exist" is a status change like any other. The condition "path does	2597	not exist" is a status change like any other. The condition "path does not
1923	not exist" is signified by the C<st_nlink> field being zero (which is	2598	exist" (or more correctly "path cannot be stat'ed") is signified by the
1924	otherwise always forced to be at least one) and all the other fields of	2599	C<st_nlink> field being zero (which is otherwise always forced to be at
1925	the stat buffer having unspecified contents.	2600	least one) and all the other fields of the stat buffer having unspecified
		2601	contents.
1926		2602
1927	The path I<should> be absolute and I<must not> end in a slash. If it is	2603	The path I<must not> end in a slash or contain special components such as
		2604	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1928	relative and your working directory changes, the behaviour is undefined.	2605	your working directory changes, then the behaviour is undefined.
1929		2606
1930	Since there is no standard kernel interface to do this, the portable	2607	Since there is no portable change notification interface available, the
1931	implementation simply calls C<stat (2)> regularly on the path to see if	2608	portable implementation simply calls C<stat(2)> regularly on the path
1932	it changed somehow. You can specify a recommended polling interval for	2609	to see if it changed somehow. You can specify a recommended polling
1933	this case. If you specify a polling interval of C<0> (highly recommended!)	2610	interval for this case. If you specify a polling interval of C<0> (highly
1934	then a I<suitable, unspecified default> value will be used (which	2611	recommended!) then a I<suitable, unspecified default> value will be used
1935	you can expect to be around five seconds, although this might change	2612	(which you can expect to be around five seconds, although this might
1936	dynamically). Libev will also impose a minimum interval which is currently	2613	change dynamically). Libev will also impose a minimum interval which is
1937	around C<0.1>, but thats usually overkill.	2614	currently around C<0.1>, but that's usually overkill.
1938		2615
1939	This watcher type is not meant for massive numbers of stat watchers,	2616	This watcher type is not meant for massive numbers of stat watchers,
1940	as even with OS-supported change notifications, this can be	2617	as even with OS-supported change notifications, this can be
1941	resource-intensive.	2618	resource-intensive.
1942		2619
1943	At the time of this writing, the only OS-specific interface implemented	2620	At the time of this writing, the only OS-specific interface implemented
1944	is the Linux inotify interface (implementing kqueue support is left as	2621	is the Linux inotify interface (implementing kqueue support is left as an
1945	an exercise for the reader. Note, however, that the author sees no way	2622	exercise for the reader. Note, however, that the author sees no way of
1946	of implementing C<ev_stat> semantics with kqueue).	2623	implementing C<ev_stat> semantics with kqueue, except as a hint).
1947		2624
1948	=head3 ABI Issues (Largefile Support)	2625	=head3 ABI Issues (Largefile Support)
1949		2626
1950	Libev by default (unless the user overrides this) uses the default	2627	Libev by default (unless the user overrides this) uses the default
1951	compilation environment, which means that on systems with large file	2628	compilation environment, which means that on systems with large file
1952	support disabled by default, you get the 32 bit version of the stat	2629	support disabled by default, you get the 32 bit version of the stat
1953	structure. When using the library from programs that change the ABI to	2630	structure. When using the library from programs that change the ABI to
1954	use 64 bit file offsets the programs will fail. In that case you have to	2631	use 64 bit file offsets the programs will fail. In that case you have to
1955	compile libev with the same flags to get binary compatibility. This is	2632	compile libev with the same flags to get binary compatibility. This is
1956	obviously the case with any flags that change the ABI, but the problem is	2633	obviously the case with any flags that change the ABI, but the problem is
1957	most noticeably disabled with ev_stat and large file support.	2634	most noticeably displayed with ev_stat and large file support.
1958		2635
1959	The solution for this is to lobby your distribution maker to make large	2636	The solution for this is to lobby your distribution maker to make large
1960	file interfaces available by default (as e.g. FreeBSD does) and not	2637	file interfaces available by default (as e.g. FreeBSD does) and not
1961	optional. Libev cannot simply switch on large file support because it has	2638	optional. Libev cannot simply switch on large file support because it has
1962	to exchange stat structures with application programs compiled using the	2639	to exchange stat structures with application programs compiled using the
1963	default compilation environment.	2640	default compilation environment.
1964		2641
1965	=head3 Inotify and Kqueue	2642	=head3 Inotify and Kqueue
1966		2643
1967	When C<inotify (7)> support has been compiled into libev (generally	2644	When C<inotify (7)> support has been compiled into libev and present at
1968	only available with Linux 2.6.25 or above due to bugs in earlier	2645	runtime, it will be used to speed up change detection where possible. The
1969	implementations) and present at runtime, it will be used to speed up	2646	inotify descriptor will be created lazily when the first C<ev_stat>
1970	change detection where possible. The inotify descriptor will be created	2647	watcher is being started.
1971	lazily when the first C<ev_stat> watcher is being started.
1972		2648
1973	Inotify presence does not change the semantics of C<ev_stat> watchers	2649	Inotify presence does not change the semantics of C<ev_stat> watchers
1974	except that changes might be detected earlier, and in some cases, to avoid	2650	except that changes might be detected earlier, and in some cases, to avoid
1975	making regular C<stat> calls. Even in the presence of inotify support	2651	making regular C<stat> calls. Even in the presence of inotify support
1976	there are many cases where libev has to resort to regular C<stat> polling,	2652	there are many cases where libev has to resort to regular C<stat> polling,
1977	but as long as the path exists, libev usually gets away without polling.	2653	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2654	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2655	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2656	xfs are fully working) libev usually gets away without polling.
1978		2657
1979	There is no support for kqueue, as apparently it cannot be used to	2658	There is no support for kqueue, as apparently it cannot be used to
1980	implement this functionality, due to the requirement of having a file	2659	implement this functionality, due to the requirement of having a file
1981	descriptor open on the object at all times, and detecting renames, unlinks	2660	descriptor open on the object at all times, and detecting renames, unlinks
1982	etc. is difficult.	2661	etc. is difficult.
1983		2662
		2663	=head3 C<stat ()> is a synchronous operation
		2664
		2665	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2666	the process. The exception are C<ev_stat> watchers - those call C<stat
		2667	()>, which is a synchronous operation.
		2668
		2669	For local paths, this usually doesn't matter: unless the system is very
		2670	busy or the intervals between stat's are large, a stat call will be fast,
		2671	as the path data is usually in memory already (except when starting the
		2672	watcher).
		2673
		2674	For networked file systems, calling C<stat ()> can block an indefinite
		2675	time due to network issues, and even under good conditions, a stat call
		2676	often takes multiple milliseconds.
		2677
		2678	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2679	paths, although this is fully supported by libev.
		2680
1984	=head3 The special problem of stat time resolution	2681	=head3 The special problem of stat time resolution
1985		2682
1986	The C<stat ()> system call only supports full-second resolution portably, and	2683	The C<stat ()> system call only supports full-second resolution portably,
1987	even on systems where the resolution is higher, most file systems still	2684	and even on systems where the resolution is higher, most file systems
1988	only support whole seconds.	2685	still only support whole seconds.
1989		2686
1990	That means that, if the time is the only thing that changes, you can	2687	That means that, if the time is the only thing that changes, you can
1991	easily miss updates: on the first update, C<ev_stat> detects a change and	2688	easily miss updates: on the first update, C<ev_stat> detects a change and
1992	calls your callback, which does something. When there is another update	2689	calls your callback, which does something. When there is another update
1993	within the same second, C<ev_stat> will be unable to detect unless the	2690	within the same second, C<ev_stat> will be unable to detect unless the
…		…
2136		2833
2137	=head3 Watcher-Specific Functions and Data Members	2834	=head3 Watcher-Specific Functions and Data Members
2138		2835
2139	=over 4	2836	=over 4
2140		2837
2141	=item ev_idle_init (ev_signal *, callback)	2838	=item ev_idle_init (ev_idle *, callback)
2142		2839
2143	Initialises and configures the idle watcher - it has no parameters of any	2840	Initialises and configures the idle watcher - it has no parameters of any
2144	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2841	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
2145	believe me.	2842	believe me.
2146		2843
…		…
2159	// no longer anything immediate to do.	2856	// no longer anything immediate to do.
2160	}	2857	}
2161		2858
2162	ev_idle *idle_watcher = malloc (sizeof (ev_idle));	2859	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
2163	ev_idle_init (idle_watcher, idle_cb);	2860	ev_idle_init (idle_watcher, idle_cb);
2164	ev_idle_start (loop, idle_cb);	2861	ev_idle_start (loop, idle_watcher);
2165		2862
2166		2863
2167	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2864	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
2168		2865
2169	Prepare and check watchers are usually (but not always) used in pairs:	2866	Prepare and check watchers are usually (but not always) used in pairs:
2170	prepare watchers get invoked before the process blocks and check watchers	2867	prepare watchers get invoked before the process blocks and check watchers
2171	afterwards.	2868	afterwards.
2172		2869
2173	You I<must not> call C<ev_loop> or similar functions that enter	2870	You I<must not> call C<ev_run> or similar functions that enter
2174	the current event loop from either C<ev_prepare> or C<ev_check>	2871	the current event loop from either C<ev_prepare> or C<ev_check>
2175	watchers. Other loops than the current one are fine, however. The	2872	watchers. Other loops than the current one are fine, however. The
2176	rationale behind this is that you do not need to check for recursion in	2873	rationale behind this is that you do not need to check for recursion in
2177	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2874	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
2178	C<ev_check> so if you have one watcher of each kind they will always be	2875	C<ev_check> so if you have one watcher of each kind they will always be
…		…
2262	struct pollfd fds [nfd];	2959	struct pollfd fds [nfd];
2263	// actual code will need to loop here and realloc etc.	2960	// actual code will need to loop here and realloc etc.
2264	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2961	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
2265		2962
2266	/* the callback is illegal, but won't be called as we stop during check */	2963	/* the callback is illegal, but won't be called as we stop during check */
2267	ev_timer_init (&tw, 0, timeout * 1e-3);	2964	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
2268	ev_timer_start (loop, &tw);	2965	ev_timer_start (loop, &tw);
2269		2966
2270	// create one ev_io per pollfd	2967	// create one ev_io per pollfd
2271	for (int i = 0; i < nfd; ++i)	2968	for (int i = 0; i < nfd; ++i)
2272	{	2969	{
…		…
2346		3043
2347	if (timeout >= 0)	3044	if (timeout >= 0)
2348	// create/start timer	3045	// create/start timer
2349		3046
2350	// poll	3047	// poll
2351	ev_loop (EV_A_ 0);	3048	ev_run (EV_A_ 0);
2352		3049
2353	// stop timer again	3050	// stop timer again
2354	if (timeout >= 0)	3051	if (timeout >= 0)
2355	ev_timer_stop (EV_A_ &to);	3052	ev_timer_stop (EV_A_ &to);
2356		3053
…		…
2385	some fds have to be watched and handled very quickly (with low latency),	3082	some fds have to be watched and handled very quickly (with low latency),
2386	and even priorities and idle watchers might have too much overhead. In	3083	and even priorities and idle watchers might have too much overhead. In
2387	this case you would put all the high priority stuff in one loop and all	3084	this case you would put all the high priority stuff in one loop and all
2388	the rest in a second one, and embed the second one in the first.	3085	the rest in a second one, and embed the second one in the first.
2389		3086
2390	As long as the watcher is active, the callback will be invoked every time	3087	As long as the watcher is active, the callback will be invoked every
2391	there might be events pending in the embedded loop. The callback must then	3088	time there might be events pending in the embedded loop. The callback
2392	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	3089	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2393	their callbacks (you could also start an idle watcher to give the embedded	3090	sweep and invoke their callbacks (the callback doesn't need to invoke the
2394	loop strictly lower priority for example). You can also set the callback	3091	C<ev_embed_sweep> function directly, it could also start an idle watcher
2395	to C<0>, in which case the embed watcher will automatically execute the	3092	to give the embedded loop strictly lower priority for example).
2396	embedded loop sweep.
2397		3093
2398	As long as the watcher is started it will automatically handle events. The	3094	You can also set the callback to C<0>, in which case the embed watcher
2399	callback will be invoked whenever some events have been handled. You can	3095	will automatically execute the embedded loop sweep whenever necessary.
2400	set the callback to C<0> to avoid having to specify one if you are not
2401	interested in that.
2402		3096
2403	Also, there have not currently been made special provisions for forking:	3097	Fork detection will be handled transparently while the C<ev_embed> watcher
2404	when you fork, you not only have to call C<ev_loop_fork> on both loops,	3098	is active, i.e., the embedded loop will automatically be forked when the
2405	but you will also have to stop and restart any C<ev_embed> watchers	3099	embedding loop forks. In other cases, the user is responsible for calling
2406	yourself - but you can use a fork watcher to handle this automatically,	3100	C<ev_loop_fork> on the embedded loop.
2407	and future versions of libev might do just that.
2408		3101
2409	Unfortunately, not all backends are embeddable: only the ones returned by	3102	Unfortunately, not all backends are embeddable: only the ones returned by
2410	C<ev_embeddable_backends> are, which, unfortunately, does not include any	3103	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2411	portable one.	3104	portable one.
2412		3105
…		…
2438	if you do not want that, you need to temporarily stop the embed watcher).	3131	if you do not want that, you need to temporarily stop the embed watcher).
2439		3132
2440	=item ev_embed_sweep (loop, ev_embed *)	3133	=item ev_embed_sweep (loop, ev_embed *)
2441		3134
2442	Make a single, non-blocking sweep over the embedded loop. This works	3135	Make a single, non-blocking sweep over the embedded loop. This works
2443	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3136	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
2444	appropriate way for embedded loops.	3137	appropriate way for embedded loops.
2445		3138
2446	=item struct ev_loop *other [read-only]	3139	=item struct ev_loop *other [read-only]
2447		3140
2448	The embedded event loop.	3141	The embedded event loop.
…		…
2506	event loop blocks next and before C<ev_check> watchers are being called,	3199	event loop blocks next and before C<ev_check> watchers are being called,
2507	and only in the child after the fork. If whoever good citizen calling	3200	and only in the child after the fork. If whoever good citizen calling
2508	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3201	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2509	handlers will be invoked, too, of course.	3202	handlers will be invoked, too, of course.
2510		3203
		3204	=head3 The special problem of life after fork - how is it possible?
		3205
		3206	Most uses of C<fork()> consist of forking, then some simple calls to set
		3207	up/change the process environment, followed by a call to C<exec()>. This
		3208	sequence should be handled by libev without any problems.
		3209
		3210	This changes when the application actually wants to do event handling
		3211	in the child, or both parent in child, in effect "continuing" after the
		3212	fork.
		3213
		3214	The default mode of operation (for libev, with application help to detect
		3215	forks) is to duplicate all the state in the child, as would be expected
		3216	when I<either> the parent I<or> the child process continues.
		3217
		3218	When both processes want to continue using libev, then this is usually the
		3219	wrong result. In that case, usually one process (typically the parent) is
		3220	supposed to continue with all watchers in place as before, while the other
		3221	process typically wants to start fresh, i.e. without any active watchers.
		3222
		3223	The cleanest and most efficient way to achieve that with libev is to
		3224	simply create a new event loop, which of course will be "empty", and
		3225	use that for new watchers. This has the advantage of not touching more
		3226	memory than necessary, and thus avoiding the copy-on-write, and the
		3227	disadvantage of having to use multiple event loops (which do not support
		3228	signal watchers).
		3229
		3230	When this is not possible, or you want to use the default loop for
		3231	other reasons, then in the process that wants to start "fresh", call
		3232	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
		3233	Destroying the default loop will "orphan" (not stop) all registered
		3234	watchers, so you have to be careful not to execute code that modifies
		3235	those watchers. Note also that in that case, you have to re-register any
		3236	signal watchers.
		3237
2511	=head3 Watcher-Specific Functions and Data Members	3238	=head3 Watcher-Specific Functions and Data Members
2512		3239
2513	=over 4	3240	=over 4
2514		3241
2515	=item ev_fork_init (ev_signal *, callback)	3242	=item ev_fork_init (ev_fork *, callback)
2516		3243
2517	Initialises and configures the fork watcher - it has no parameters of any	3244	Initialises and configures the fork watcher - it has no parameters of any
2518	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3245	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
2519	believe me.	3246	really.
2520		3247
2521	=back	3248	=back
2522		3249
2523		3250
		3251	=head2 C<ev_cleanup> - even the best things end
		3252
		3253	Cleanup watchers are called just before the event loop is being destroyed
		3254	by a call to C<ev_loop_destroy>.
		3255
		3256	While there is no guarantee that the event loop gets destroyed, cleanup
		3257	watchers provide a convenient method to install cleanup hooks for your
		3258	program, worker threads and so on - you just to make sure to destroy the
		3259	loop when you want them to be invoked.
		3260
		3261	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3262	all other watchers, they do not keep a reference to the event loop (which
		3263	makes a lot of sense if you think about it). Like all other watchers, you
		3264	can call libev functions in the callback, except C<ev_cleanup_start>.
		3265
		3266	=head3 Watcher-Specific Functions and Data Members
		3267
		3268	=over 4
		3269
		3270	=item ev_cleanup_init (ev_cleanup *, callback)
		3271
		3272	Initialises and configures the cleanup watcher - it has no parameters of
		3273	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3274	pointless, I assure you.
		3275
		3276	=back
		3277
		3278	Example: Register an atexit handler to destroy the default loop, so any
		3279	cleanup functions are called.
		3280
		3281	static void
		3282	program_exits (void)
		3283	{
		3284	ev_loop_destroy (EV_DEFAULT_UC);
		3285	}
		3286
		3287	...
		3288	atexit (program_exits);
		3289
		3290
2524	=head2 C<ev_async> - how to wake up another event loop	3291	=head2 C<ev_async> - how to wake up an event loop
2525		3292
2526	In general, you cannot use an C<ev_loop> from multiple threads or other	3293	In general, you cannot use an C<ev_loop> from multiple threads or other
2527	asynchronous sources such as signal handlers (as opposed to multiple event	3294	asynchronous sources such as signal handlers (as opposed to multiple event
2528	loops - those are of course safe to use in different threads).	3295	loops - those are of course safe to use in different threads).
2529		3296
2530	Sometimes, however, you need to wake up another event loop you do not	3297	Sometimes, however, you need to wake up an event loop you do not control,
2531	control, for example because it belongs to another thread. This is what	3298	for example because it belongs to another thread. This is what C<ev_async>
2532	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3299	watchers do: as long as the C<ev_async> watcher is active, you can signal
2533	can signal it by calling C<ev_async_send>, which is thread- and signal	3300	it by calling C<ev_async_send>, which is thread- and signal safe.
2534	safe.
2535		3301
2536	This functionality is very similar to C<ev_signal> watchers, as signals,	3302	This functionality is very similar to C<ev_signal> watchers, as signals,
2537	too, are asynchronous in nature, and signals, too, will be compressed	3303	too, are asynchronous in nature, and signals, too, will be compressed
2538	(i.e. the number of callback invocations may be less than the number of	3304	(i.e. the number of callback invocations may be less than the number of
2539	C<ev_async_sent> calls).	3305	C<ev_async_sent> calls). In fact, you could use signal watchers as a kind
2540		3306	of "global async watchers" by using a watcher on an otherwise unused
2541	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not	3307	signal, and C<ev_feed_signal> to signal this watcher from another thread,
2542	just the default loop.	3308	even without knowing which loop owns the signal.
2543		3309
2544	=head3 Queueing	3310	=head3 Queueing
2545		3311
2546	C<ev_async> does not support queueing of data in any way. The reason	3312	C<ev_async> does not support queueing of data in any way. The reason
2547	is that the author does not know of a simple (or any) algorithm for a	3313	is that the author does not know of a simple (or any) algorithm for a
2548	multiple-writer-single-reader queue that works in all cases and doesn't	3314	multiple-writer-single-reader queue that works in all cases and doesn't
2549	need elaborate support such as pthreads.	3315	need elaborate support such as pthreads or unportable memory access
		3316	semantics.
2550		3317
2551	That means that if you want to queue data, you have to provide your own	3318	That means that if you want to queue data, you have to provide your own
2552	queue. But at least I can tell you how to implement locking around your	3319	queue. But at least I can tell you how to implement locking around your
2553	queue:	3320	queue:
2554		3321
…		…
2632	=over 4	3399	=over 4
2633		3400
2634	=item ev_async_init (ev_async *, callback)	3401	=item ev_async_init (ev_async *, callback)
2635		3402
2636	Initialises and configures the async watcher - it has no parameters of any	3403	Initialises and configures the async watcher - it has no parameters of any
2637	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3404	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2638	trust me.	3405	trust me.
2639		3406
2640	=item ev_async_send (loop, ev_async *)	3407	=item ev_async_send (loop, ev_async *)
2641		3408
2642	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3409	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2643	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3410	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3411	returns.
		3412
2644	C<ev_feed_event>, this call is safe to do from other threads, signal or	3413	Unlike C<ev_feed_event>, this call is safe to do from other threads,
2645	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3414	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
2646	section below on what exactly this means).	3415	embedding section below on what exactly this means).
2647		3416
2648	This call incurs the overhead of a system call only once per loop iteration,	3417	Note that, as with other watchers in libev, multiple events might get
2649	so while the overhead might be noticeable, it doesn't apply to repeated	3418	compressed into a single callback invocation (another way to look at
2650	calls to C<ev_async_send>.	3419	this is that C<ev_async> watchers are level-triggered: they are set on
		3420	C<ev_async_send>, reset when the event loop detects that).
		3421
		3422	This call incurs the overhead of at most one extra system call per event
		3423	loop iteration, if the event loop is blocked, and no syscall at all if
		3424	the event loop (or your program) is processing events. That means that
		3425	repeated calls are basically free (there is no need to avoid calls for
		3426	performance reasons) and that the overhead becomes smaller (typically
		3427	zero) under load.
2651		3428
2652	=item bool = ev_async_pending (ev_async *)	3429	=item bool = ev_async_pending (ev_async *)
2653		3430
2654	Returns a non-zero value when C<ev_async_send> has been called on the	3431	Returns a non-zero value when C<ev_async_send> has been called on the
2655	watcher but the event has not yet been processed (or even noted) by the	3432	watcher but the event has not yet been processed (or even noted) by the
…		…
2658	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3435	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2659	the loop iterates next and checks for the watcher to have become active,	3436	the loop iterates next and checks for the watcher to have become active,
2660	it will reset the flag again. C<ev_async_pending> can be used to very	3437	it will reset the flag again. C<ev_async_pending> can be used to very
2661	quickly check whether invoking the loop might be a good idea.	3438	quickly check whether invoking the loop might be a good idea.
2662		3439
2663	Not that this does I<not> check whether the watcher itself is pending, only	3440	Not that this does I<not> check whether the watcher itself is pending,
2664	whether it has been requested to make this watcher pending.	3441	only whether it has been requested to make this watcher pending: there
		3442	is a time window between the event loop checking and resetting the async
		3443	notification, and the callback being invoked.
2665		3444
2666	=back	3445	=back
2667		3446
2668		3447
2669	=head1 OTHER FUNCTIONS	3448	=head1 OTHER FUNCTIONS
…		…
2686		3465
2687	If C<timeout> is less than 0, then no timeout watcher will be	3466	If C<timeout> is less than 0, then no timeout watcher will be
2688	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3467	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2689	repeat = 0) will be started. C<0> is a valid timeout.	3468	repeat = 0) will be started. C<0> is a valid timeout.
2690		3469
2691	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3470	The callback has the type C<void (*cb)(int revents, void *arg)> and is
2692	passed an C<revents> set like normal event callbacks (a combination of	3471	passed an C<revents> set like normal event callbacks (a combination of
2693	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3472	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
2694	value passed to C<ev_once>. Note that it is possible to receive I<both>	3473	value passed to C<ev_once>. Note that it is possible to receive I<both>
2695	a timeout and an io event at the same time - you probably should give io	3474	a timeout and an io event at the same time - you probably should give io
2696	events precedence.	3475	events precedence.
2697		3476
2698	Example: wait up to ten seconds for data to appear on STDIN_FILENO.	3477	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2699		3478
2700	static void stdin_ready (int revents, void *arg)	3479	static void stdin_ready (int revents, void *arg)
2701	{	3480	{
2702	if (revents & EV_READ)	3481	if (revents & EV_READ)
2703	/* stdin might have data for us, joy! */;	3482	/* stdin might have data for us, joy! */;
2704	else if (revents & EV_TIMEOUT)	3483	else if (revents & EV_TIMER)
2705	/* doh, nothing entered */;	3484	/* doh, nothing entered */;
2706	}	3485	}
2707		3486
2708	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3487	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2709		3488
2710	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
2711
2712	Feeds the given event set into the event loop, as if the specified event
2713	had happened for the specified watcher (which must be a pointer to an
2714	initialised but not necessarily started event watcher).
2715
2716	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)	3489	=item ev_feed_fd_event (loop, int fd, int revents)
2717		3490
2718	Feed an event on the given fd, as if a file descriptor backend detected	3491	Feed an event on the given fd, as if a file descriptor backend detected
2719	the given events it.	3492	the given events.
2720		3493
2721	=item ev_feed_signal_event (struct ev_loop *loop, int signum)	3494	=item ev_feed_signal_event (loop, int signum)
2722		3495
2723	Feed an event as if the given signal occurred (C<loop> must be the default	3496	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
2724	loop!).	3497	which is async-safe.
2725		3498
2726	=back	3499	=back
		3500
		3501
		3502	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3503
		3504	This section explains some common idioms that are not immediately
		3505	obvious. Note that examples are sprinkled over the whole manual, and this
		3506	section only contains stuff that wouldn't fit anywhere else.
		3507
		3508	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3509
		3510	Each watcher has, by default, a C<void *data> member that you can read
		3511	or modify at any time: libev will completely ignore it. This can be used
		3512	to associate arbitrary data with your watcher. If you need more data and
		3513	don't want to allocate memory separately and store a pointer to it in that
		3514	data member, you can also "subclass" the watcher type and provide your own
		3515	data:
		3516
		3517	struct my_io
		3518	{
		3519	ev_io io;
		3520	int otherfd;
		3521	void *somedata;
		3522	struct whatever *mostinteresting;
		3523	};
		3524
		3525	...
		3526	struct my_io w;
		3527	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3528
		3529	And since your callback will be called with a pointer to the watcher, you
		3530	can cast it back to your own type:
		3531
		3532	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3533	{
		3534	struct my_io *w = (struct my_io *)w_;
		3535	...
		3536	}
		3537
		3538	More interesting and less C-conformant ways of casting your callback
		3539	function type instead have been omitted.
		3540
		3541	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3542
		3543	Another common scenario is to use some data structure with multiple
		3544	embedded watchers, in effect creating your own watcher that combines
		3545	multiple libev event sources into one "super-watcher":
		3546
		3547	struct my_biggy
		3548	{
		3549	int some_data;
		3550	ev_timer t1;
		3551	ev_timer t2;
		3552	}
		3553
		3554	In this case getting the pointer to C<my_biggy> is a bit more
		3555	complicated: Either you store the address of your C<my_biggy> struct in
		3556	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3557	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3558	real programmers):
		3559
		3560	#include <stddef.h>
		3561
		3562	static void
		3563	t1_cb (EV_P_ ev_timer *w, int revents)
		3564	{
		3565	struct my_biggy big = (struct my_biggy *)
		3566	(((char *)w) - offsetof (struct my_biggy, t1));
		3567	}
		3568
		3569	static void
		3570	t2_cb (EV_P_ ev_timer *w, int revents)
		3571	{
		3572	struct my_biggy big = (struct my_biggy *)
		3573	(((char *)w) - offsetof (struct my_biggy, t2));
		3574	}
		3575
		3576	=head2 AVOIDING FINISHING BEFORE RETURNING
		3577
		3578	Often you have structures like this in event-based programs:
		3579
		3580	callback ()
		3581	{
		3582	free (request);
		3583	}
		3584
		3585	request = start_new_request (..., callback);
		3586
		3587	The intent is to start some "lengthy" operation. The C<request> could be
		3588	used to cancel the operation, or do other things with it.
		3589
		3590	It's not uncommon to have code paths in C<start_new_request> that
		3591	immediately invoke the callback, for example, to report errors. Or you add
		3592	some caching layer that finds that it can skip the lengthy aspects of the
		3593	operation and simply invoke the callback with the result.
		3594
		3595	The problem here is that this will happen I<before> C<start_new_request>
		3596	has returned, so C<request> is not set.
		3597
		3598	Even if you pass the request by some safer means to the callback, you
		3599	might want to do something to the request after starting it, such as
		3600	canceling it, which probably isn't working so well when the callback has
		3601	already been invoked.
		3602
		3603	A common way around all these issues is to make sure that
		3604	C<start_new_request> I<always> returns before the callback is invoked. If
		3605	C<start_new_request> immediately knows the result, it can artificially
		3606	delay invoking the callback by e.g. using a C<prepare> or C<idle> watcher
		3607	for example, or more sneakily, by reusing an existing (stopped) watcher
		3608	and pushing it into the pending queue:
		3609
		3610	ev_set_cb (watcher, callback);
		3611	ev_feed_event (EV_A_ watcher, 0);
		3612
		3613	This way, C<start_new_request> can safely return before the callback is
		3614	invoked, while not delaying callback invocation too much.
		3615
		3616	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3617
		3618	Often (especially in GUI toolkits) there are places where you have
		3619	I<modal> interaction, which is most easily implemented by recursively
		3620	invoking C<ev_run>.
		3621
		3622	This brings the problem of exiting - a callback might want to finish the
		3623	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3624	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3625	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3626	other combination: In these cases, C<ev_break> will not work alone.
		3627
		3628	The solution is to maintain "break this loop" variable for each C<ev_run>
		3629	invocation, and use a loop around C<ev_run> until the condition is
		3630	triggered, using C<EVRUN_ONCE>:
		3631
		3632	// main loop
		3633	int exit_main_loop = 0;
		3634
		3635	while (!exit_main_loop)
		3636	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3637
		3638	// in a model watcher
		3639	int exit_nested_loop = 0;
		3640
		3641	while (!exit_nested_loop)
		3642	ev_run (EV_A_ EVRUN_ONCE);
		3643
		3644	To exit from any of these loops, just set the corresponding exit variable:
		3645
		3646	// exit modal loop
		3647	exit_nested_loop = 1;
		3648
		3649	// exit main program, after modal loop is finished
		3650	exit_main_loop = 1;
		3651
		3652	// exit both
		3653	exit_main_loop = exit_nested_loop = 1;
		3654
		3655	=head2 THREAD LOCKING EXAMPLE
		3656
		3657	Here is a fictitious example of how to run an event loop in a different
		3658	thread from where callbacks are being invoked and watchers are
		3659	created/added/removed.
		3660
		3661	For a real-world example, see the C<EV::Loop::Async> perl module,
		3662	which uses exactly this technique (which is suited for many high-level
		3663	languages).
		3664
		3665	The example uses a pthread mutex to protect the loop data, a condition
		3666	variable to wait for callback invocations, an async watcher to notify the
		3667	event loop thread and an unspecified mechanism to wake up the main thread.
		3668
		3669	First, you need to associate some data with the event loop:
		3670
		3671	typedef struct {
		3672	mutex_t lock; /* global loop lock */
		3673	ev_async async_w;
		3674	thread_t tid;
		3675	cond_t invoke_cv;
		3676	} userdata;
		3677
		3678	void prepare_loop (EV_P)
		3679	{
		3680	// for simplicity, we use a static userdata struct.
		3681	static userdata u;
		3682
		3683	ev_async_init (&u->async_w, async_cb);
		3684	ev_async_start (EV_A_ &u->async_w);
		3685
		3686	pthread_mutex_init (&u->lock, 0);
		3687	pthread_cond_init (&u->invoke_cv, 0);
		3688
		3689	// now associate this with the loop
		3690	ev_set_userdata (EV_A_ u);
		3691	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3692	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3693
		3694	// then create the thread running ev_run
		3695	pthread_create (&u->tid, 0, l_run, EV_A);
		3696	}
		3697
		3698	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3699	solely to wake up the event loop so it takes notice of any new watchers
		3700	that might have been added:
		3701
		3702	static void
		3703	async_cb (EV_P_ ev_async *w, int revents)
		3704	{
		3705	// just used for the side effects
		3706	}
		3707
		3708	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3709	protecting the loop data, respectively.
		3710
		3711	static void
		3712	l_release (EV_P)
		3713	{
		3714	userdata *u = ev_userdata (EV_A);
		3715	pthread_mutex_unlock (&u->lock);
		3716	}
		3717
		3718	static void
		3719	l_acquire (EV_P)
		3720	{
		3721	userdata *u = ev_userdata (EV_A);
		3722	pthread_mutex_lock (&u->lock);
		3723	}
		3724
		3725	The event loop thread first acquires the mutex, and then jumps straight
		3726	into C<ev_run>:
		3727
		3728	void *
		3729	l_run (void *thr_arg)
		3730	{
		3731	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3732
		3733	l_acquire (EV_A);
		3734	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3735	ev_run (EV_A_ 0);
		3736	l_release (EV_A);
		3737
		3738	return 0;
		3739	}
		3740
		3741	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3742	signal the main thread via some unspecified mechanism (signals? pipe
		3743	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3744	have been called (in a while loop because a) spurious wakeups are possible
		3745	and b) skipping inter-thread-communication when there are no pending
		3746	watchers is very beneficial):
		3747
		3748	static void
		3749	l_invoke (EV_P)
		3750	{
		3751	userdata *u = ev_userdata (EV_A);
		3752
		3753	while (ev_pending_count (EV_A))
		3754	{
		3755	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3756	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3757	}
		3758	}
		3759
		3760	Now, whenever the main thread gets told to invoke pending watchers, it
		3761	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3762	thread to continue:
		3763
		3764	static void
		3765	real_invoke_pending (EV_P)
		3766	{
		3767	userdata *u = ev_userdata (EV_A);
		3768
		3769	pthread_mutex_lock (&u->lock);
		3770	ev_invoke_pending (EV_A);
		3771	pthread_cond_signal (&u->invoke_cv);
		3772	pthread_mutex_unlock (&u->lock);
		3773	}
		3774
		3775	Whenever you want to start/stop a watcher or do other modifications to an
		3776	event loop, you will now have to lock:
		3777
		3778	ev_timer timeout_watcher;
		3779	userdata *u = ev_userdata (EV_A);
		3780
		3781	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3782
		3783	pthread_mutex_lock (&u->lock);
		3784	ev_timer_start (EV_A_ &timeout_watcher);
		3785	ev_async_send (EV_A_ &u->async_w);
		3786	pthread_mutex_unlock (&u->lock);
		3787
		3788	Note that sending the C<ev_async> watcher is required because otherwise
		3789	an event loop currently blocking in the kernel will have no knowledge
		3790	about the newly added timer. By waking up the loop it will pick up any new
		3791	watchers in the next event loop iteration.
		3792
		3793	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3794
		3795	While the overhead of a callback that e.g. schedules a thread is small, it
		3796	is still an overhead. If you embed libev, and your main usage is with some
		3797	kind of threads or coroutines, you might want to customise libev so that
		3798	doesn't need callbacks anymore.
		3799
		3800	Imagine you have coroutines that you can switch to using a function
		3801	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3802	and that due to some magic, the currently active coroutine is stored in a
		3803	global called C<current_coro>. Then you can build your own "wait for libev
		3804	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3805	the differing C<;> conventions):
		3806
		3807	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3808	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3809
		3810	That means instead of having a C callback function, you store the
		3811	coroutine to switch to in each watcher, and instead of having libev call
		3812	your callback, you instead have it switch to that coroutine.
		3813
		3814	A coroutine might now wait for an event with a function called
		3815	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3816	matter when, or whether the watcher is active or not when this function is
		3817	called):
		3818
		3819	void
		3820	wait_for_event (ev_watcher *w)
		3821	{
		3822	ev_cb_set (w) = current_coro;
		3823	switch_to (libev_coro);
		3824	}
		3825
		3826	That basically suspends the coroutine inside C<wait_for_event> and
		3827	continues the libev coroutine, which, when appropriate, switches back to
		3828	this or any other coroutine. I am sure if you sue this your own :)
		3829
		3830	You can do similar tricks if you have, say, threads with an event queue -
		3831	instead of storing a coroutine, you store the queue object and instead of
		3832	switching to a coroutine, you push the watcher onto the queue and notify
		3833	any waiters.
		3834
		3835	To embed libev, see L<EMBEDDING>, but in short, it's easiest to create two
		3836	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3837
		3838	// my_ev.h
		3839	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3840	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb);
		3841	#include "../libev/ev.h"
		3842
		3843	// my_ev.c
		3844	#define EV_H "my_ev.h"
		3845	#include "../libev/ev.c"
		3846
		3847	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3848	F<my_ev.c> into your project. When properly specifying include paths, you
		3849	can even use F<ev.h> as header file name directly.
2727		3850
2728		3851
2729	=head1 LIBEVENT EMULATION	3852	=head1 LIBEVENT EMULATION
2730		3853
2731	Libev offers a compatibility emulation layer for libevent. It cannot	3854	Libev offers a compatibility emulation layer for libevent. It cannot
2732	emulate the internals of libevent, so here are some usage hints:	3855	emulate the internals of libevent, so here are some usage hints:
2733		3856
2734	=over 4	3857	=over 4
		3858
		3859	=item * Only the libevent-1.4.1-beta API is being emulated.
		3860
		3861	This was the newest libevent version available when libev was implemented,
		3862	and is still mostly unchanged in 2010.
2735		3863
2736	=item * Use it by including <event.h>, as usual.	3864	=item * Use it by including <event.h>, as usual.
2737		3865
2738	=item * The following members are fully supported: ev_base, ev_callback,	3866	=item * The following members are fully supported: ev_base, ev_callback,
2739	ev_arg, ev_fd, ev_res, ev_events.	3867	ev_arg, ev_fd, ev_res, ev_events.
…		…
2745	=item * Priorities are not currently supported. Initialising priorities	3873	=item * Priorities are not currently supported. Initialising priorities
2746	will fail and all watchers will have the same priority, even though there	3874	will fail and all watchers will have the same priority, even though there
2747	is an ev_pri field.	3875	is an ev_pri field.
2748		3876
2749	=item * In libevent, the last base created gets the signals, in libev, the	3877	=item * In libevent, the last base created gets the signals, in libev, the
2750	first base created (== the default loop) gets the signals.	3878	base that registered the signal gets the signals.
2751		3879
2752	=item * Other members are not supported.	3880	=item * Other members are not supported.
2753		3881
2754	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3882	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2755	to use the libev header file and library.	3883	to use the libev header file and library.
…		…
2774	Care has been taken to keep the overhead low. The only data member the C++	3902	Care has been taken to keep the overhead low. The only data member the C++
2775	classes add (compared to plain C-style watchers) is the event loop pointer	3903	classes add (compared to plain C-style watchers) is the event loop pointer
2776	that the watcher is associated with (or no additional members at all if	3904	that the watcher is associated with (or no additional members at all if
2777	you disable C<EV_MULTIPLICITY> when embedding libev).	3905	you disable C<EV_MULTIPLICITY> when embedding libev).
2778		3906
2779	Currently, functions, and static and non-static member functions can be	3907	Currently, functions, static and non-static member functions and classes
2780	used as callbacks. Other types should be easy to add as long as they only	3908	with C<operator ()> can be used as callbacks. Other types should be easy
2781	need one additional pointer for context. If you need support for other	3909	to add as long as they only need one additional pointer for context. If
2782	types of functors please contact the author (preferably after implementing	3910	you need support for other types of functors please contact the author
2783	it).	3911	(preferably after implementing it).
2784		3912
2785	Here is a list of things available in the C<ev> namespace:	3913	Here is a list of things available in the C<ev> namespace:
2786		3914
2787	=over 4	3915	=over 4
2788		3916
…		…
2806		3934
2807	=over 4	3935	=over 4
2808		3936
2809	=item ev::TYPE::TYPE ()	3937	=item ev::TYPE::TYPE ()
2810		3938
2811	=item ev::TYPE::TYPE (struct ev_loop *)	3939	=item ev::TYPE::TYPE (loop)
2812		3940
2813	=item ev::TYPE::~TYPE	3941	=item ev::TYPE::~TYPE
2814		3942
2815	The constructor (optionally) takes an event loop to associate the watcher	3943	The constructor (optionally) takes an event loop to associate the watcher
2816	with. If it is omitted, it will use C<EV_DEFAULT>.	3944	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2848		3976
2849	myclass obj;	3977	myclass obj;
2850	ev::io iow;	3978	ev::io iow;
2851	iow.set <myclass, &myclass::io_cb> (&obj);	3979	iow.set <myclass, &myclass::io_cb> (&obj);
2852		3980
		3981	=item w->set (object *)
		3982
		3983	This is a variation of a method callback - leaving out the method to call
		3984	will default the method to C<operator ()>, which makes it possible to use
		3985	functor objects without having to manually specify the C<operator ()> all
		3986	the time. Incidentally, you can then also leave out the template argument
		3987	list.
		3988
		3989	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3990	int revents)>.
		3991
		3992	See the method-C<set> above for more details.
		3993
		3994	Example: use a functor object as callback.
		3995
		3996	struct myfunctor
		3997	{
		3998	void operator() (ev::io &w, int revents)
		3999	{
		4000	...
		4001	}
		4002	}
		4003
		4004	myfunctor f;
		4005
		4006	ev::io w;
		4007	w.set (&f);
		4008
2853	=item w->set<function> (void *data = 0)	4009	=item w->set<function> (void *data = 0)
2854		4010
2855	Also sets a callback, but uses a static method or plain function as	4011	Also sets a callback, but uses a static method or plain function as
2856	callback. The optional C<data> argument will be stored in the watcher's	4012	callback. The optional C<data> argument will be stored in the watcher's
2857	C<data> member and is free for you to use.	4013	C<data> member and is free for you to use.
…		…
2863	Example: Use a plain function as callback.	4019	Example: Use a plain function as callback.
2864		4020
2865	static void io_cb (ev::io &w, int revents) { }	4021	static void io_cb (ev::io &w, int revents) { }
2866	iow.set <io_cb> ();	4022	iow.set <io_cb> ();
2867		4023
2868	=item w->set (struct ev_loop *)	4024	=item w->set (loop)
2869		4025
2870	Associates a different C<struct ev_loop> with this watcher. You can only	4026	Associates a different C<struct ev_loop> with this watcher. You can only
2871	do this when the watcher is inactive (and not pending either).	4027	do this when the watcher is inactive (and not pending either).
2872		4028
2873	=item w->set ([arguments])	4029	=item w->set ([arguments])
2874		4030
2875	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be	4031	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
2876	called at least once. Unlike the C counterpart, an active watcher gets	4032	method or a suitable start method must be called at least once. Unlike the
2877	automatically stopped and restarted when reconfiguring it with this	4033	C counterpart, an active watcher gets automatically stopped and restarted
2878	method.	4034	when reconfiguring it with this method.
2879		4035
2880	=item w->start ()	4036	=item w->start ()
2881		4037
2882	Starts the watcher. Note that there is no C<loop> argument, as the	4038	Starts the watcher. Note that there is no C<loop> argument, as the
2883	constructor already stores the event loop.	4039	constructor already stores the event loop.
2884		4040
		4041	=item w->start ([arguments])
		4042
		4043	Instead of calling C<set> and C<start> methods separately, it is often
		4044	convenient to wrap them in one call. Uses the same type of arguments as
		4045	the configure C<set> method of the watcher.
		4046
2885	=item w->stop ()	4047	=item w->stop ()
2886		4048
2887	Stops the watcher if it is active. Again, no C<loop> argument.	4049	Stops the watcher if it is active. Again, no C<loop> argument.
2888		4050
2889	=item w->again () (C<ev::timer>, C<ev::periodic> only)	4051	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
2901		4063
2902	=back	4064	=back
2903		4065
2904	=back	4066	=back
2905		4067
2906	Example: Define a class with an IO and idle watcher, start one of them in	4068	Example: Define a class with two I/O and idle watchers, start the I/O
2907	the constructor.	4069	watchers in the constructor.
2908		4070
2909	class myclass	4071	class myclass
2910	{	4072	{
2911	ev::io io ; void io_cb (ev::io &w, int revents);	4073	ev::io io ; void io_cb (ev::io &w, int revents);
		4074	ev::io io2 ; void io2_cb (ev::io &w, int revents);
2912	ev::idle idle; void idle_cb (ev::idle &w, int revents);	4075	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2913		4076
2914	myclass (int fd)	4077	myclass (int fd)
2915	{	4078	{
2916	io .set <myclass, &myclass::io_cb > (this);	4079	io .set <myclass, &myclass::io_cb > (this);
		4080	io2 .set <myclass, &myclass::io2_cb > (this);
2917	idle.set <myclass, &myclass::idle_cb> (this);	4081	idle.set <myclass, &myclass::idle_cb> (this);
2918		4082
2919	io.start (fd, ev::READ);	4083	io.set (fd, ev::WRITE); // configure the watcher
		4084	io.start (); // start it whenever convenient
		4085
		4086	io2.start (fd, ev::READ); // set + start in one call
2920	}	4087	}
2921	};	4088	};
2922		4089
2923		4090
2924	=head1 OTHER LANGUAGE BINDINGS	4091	=head1 OTHER LANGUAGE BINDINGS
…		…
2943	L<http://software.schmorp.de/pkg/EV>.	4110	L<http://software.schmorp.de/pkg/EV>.
2944		4111
2945	=item Python	4112	=item Python
2946		4113
2947	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	4114	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2948	seems to be quite complete and well-documented. Note, however, that the	4115	seems to be quite complete and well-documented.
2949	patch they require for libev is outright dangerous as it breaks the ABI
2950	for everybody else, and therefore, should never be applied in an installed
2951	libev (if python requires an incompatible ABI then it needs to embed
2952	libev).
2953		4116
2954	=item Ruby	4117	=item Ruby
2955		4118
2956	Tony Arcieri has written a ruby extension that offers access to a subset	4119	Tony Arcieri has written a ruby extension that offers access to a subset
2957	of the libev API and adds file handle abstractions, asynchronous DNS and	4120	of the libev API and adds file handle abstractions, asynchronous DNS and
2958	more on top of it. It can be found via gem servers. Its homepage is at	4121	more on top of it. It can be found via gem servers. Its homepage is at
2959	L<http://rev.rubyforge.org/>.	4122	L<http://rev.rubyforge.org/>.
2960		4123
		4124	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		4125	makes rev work even on mingw.
		4126
		4127	=item Haskell
		4128
		4129	A haskell binding to libev is available at
		4130	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		4131
2961	=item D	4132	=item D
2962		4133
2963	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	4134	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2964	be found at L<http://proj.llucax.com.ar/wiki/evd>.	4135	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
2965		4136
2966	=item Ocaml	4137	=item Ocaml
2967		4138
2968	Erkki Seppala has written Ocaml bindings for libev, to be found at	4139	Erkki Seppala has written Ocaml bindings for libev, to be found at
2969	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	4140	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		4141
		4142	=item Lua
		4143
		4144	Brian Maher has written a partial interface to libev for lua (at the
		4145	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		4146	L<http://github.com/brimworks/lua-ev>.
2970		4147
2971	=back	4148	=back
2972		4149
2973		4150
2974	=head1 MACRO MAGIC	4151	=head1 MACRO MAGIC
…		…
2988	loop argument"). The C<EV_A> form is used when this is the sole argument,	4165	loop argument"). The C<EV_A> form is used when this is the sole argument,
2989	C<EV_A_> is used when other arguments are following. Example:	4166	C<EV_A_> is used when other arguments are following. Example:
2990		4167
2991	ev_unref (EV_A);	4168	ev_unref (EV_A);
2992	ev_timer_add (EV_A_ watcher);	4169	ev_timer_add (EV_A_ watcher);
2993	ev_loop (EV_A_ 0);	4170	ev_run (EV_A_ 0);
2994		4171
2995	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	4172	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
2996	which is often provided by the following macro.	4173	which is often provided by the following macro.
2997		4174
2998	=item C<EV_P>, C<EV_P_>	4175	=item C<EV_P>, C<EV_P_>
…		…
3011	suitable for use with C<EV_A>.	4188	suitable for use with C<EV_A>.
3012		4189
3013	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4190	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3014		4191
3015	Similar to the other two macros, this gives you the value of the default	4192	Similar to the other two macros, this gives you the value of the default
3016	loop, if multiple loops are supported ("ev loop default").	4193	loop, if multiple loops are supported ("ev loop default"). The default loop
		4194	will be initialised if it isn't already initialised.
		4195
		4196	For non-multiplicity builds, these macros do nothing, so you always have
		4197	to initialise the loop somewhere.
3017		4198
3018	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>	4199	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3019		4200
3020	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the	4201	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3021	default loop has been initialised (C<UC> == unchecked). Their behaviour	4202	default loop has been initialised (C<UC> == unchecked). Their behaviour
…		…
3038	}	4219	}
3039		4220
3040	ev_check check;	4221	ev_check check;
3041	ev_check_init (&check, check_cb);	4222	ev_check_init (&check, check_cb);
3042	ev_check_start (EV_DEFAULT_ &check);	4223	ev_check_start (EV_DEFAULT_ &check);
3043	ev_loop (EV_DEFAULT_ 0);	4224	ev_run (EV_DEFAULT_ 0);
3044		4225
3045	=head1 EMBEDDING	4226	=head1 EMBEDDING
3046		4227
3047	Libev can (and often is) directly embedded into host	4228	Libev can (and often is) directly embedded into host
3048	applications. Examples of applications that embed it include the Deliantra	4229	applications. Examples of applications that embed it include the Deliantra
…		…
3075		4256
3076	#define EV_STANDALONE 1	4257	#define EV_STANDALONE 1
3077	#include "ev.h"	4258	#include "ev.h"
3078		4259
3079	Both header files and implementation files can be compiled with a C++	4260	Both header files and implementation files can be compiled with a C++
3080	compiler (at least, thats a stated goal, and breakage will be treated	4261	compiler (at least, that's a stated goal, and breakage will be treated
3081	as a bug).	4262	as a bug).
3082		4263
3083	You need the following files in your source tree, or in a directory	4264	You need the following files in your source tree, or in a directory
3084	in your include path (e.g. in libev/ when using -Ilibev):	4265	in your include path (e.g. in libev/ when using -Ilibev):
3085		4266
…		…
3128	libev.m4	4309	libev.m4
3129		4310
3130	=head2 PREPROCESSOR SYMBOLS/MACROS	4311	=head2 PREPROCESSOR SYMBOLS/MACROS
3131		4312
3132	Libev can be configured via a variety of preprocessor symbols you have to	4313	Libev can be configured via a variety of preprocessor symbols you have to
3133	define before including any of its files. The default in the absence of	4314	define before including (or compiling) any of its files. The default in
3134	autoconf is documented for every option.	4315	the absence of autoconf is documented for every option.
		4316
		4317	Symbols marked with "(h)" do not change the ABI, and can have different
		4318	values when compiling libev vs. including F<ev.h>, so it is permissible
		4319	to redefine them before including F<ev.h> without breaking compatibility
		4320	to a compiled library. All other symbols change the ABI, which means all
		4321	users of libev and the libev code itself must be compiled with compatible
		4322	settings.
3135		4323
3136	=over 4	4324	=over 4
3137		4325
		4326	=item EV_COMPAT3 (h)
		4327
		4328	Backwards compatibility is a major concern for libev. This is why this
		4329	release of libev comes with wrappers for the functions and symbols that
		4330	have been renamed between libev version 3 and 4.
		4331
		4332	You can disable these wrappers (to test compatibility with future
		4333	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4334	sources. This has the additional advantage that you can drop the C<struct>
		4335	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4336	typedef in that case.
		4337
		4338	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4339	and in some even more future version the compatibility code will be
		4340	removed completely.
		4341
3138	=item EV_STANDALONE	4342	=item EV_STANDALONE (h)
3139		4343
3140	Must always be C<1> if you do not use autoconf configuration, which	4344	Must always be C<1> if you do not use autoconf configuration, which
3141	keeps libev from including F<config.h>, and it also defines dummy	4345	keeps libev from including F<config.h>, and it also defines dummy
3142	implementations for some libevent functions (such as logging, which is not	4346	implementations for some libevent functions (such as logging, which is not
3143	supported). It will also not define any of the structs usually found in	4347	supported). It will also not define any of the structs usually found in
3144	F<event.h> that are not directly supported by the libev core alone.	4348	F<event.h> that are not directly supported by the libev core alone.
3145		4349
		4350	In standalone mode, libev will still try to automatically deduce the
		4351	configuration, but has to be more conservative.
		4352
		4353	=item EV_USE_FLOOR
		4354
		4355	If defined to be C<1>, libev will use the C<floor ()> function for its
		4356	periodic reschedule calculations, otherwise libev will fall back on a
		4357	portable (slower) implementation. If you enable this, you usually have to
		4358	link against libm or something equivalent. Enabling this when the C<floor>
		4359	function is not available will fail, so the safe default is to not enable
		4360	this.
		4361
3146	=item EV_USE_MONOTONIC	4362	=item EV_USE_MONOTONIC
3147		4363
3148	If defined to be C<1>, libev will try to detect the availability of the	4364	If defined to be C<1>, libev will try to detect the availability of the
3149	monotonic clock option at both compile time and runtime. Otherwise no use	4365	monotonic clock option at both compile time and runtime. Otherwise no
3150	of the monotonic clock option will be attempted. If you enable this, you	4366	use of the monotonic clock option will be attempted. If you enable this,
3151	usually have to link against librt or something similar. Enabling it when	4367	you usually have to link against librt or something similar. Enabling it
3152	the functionality isn't available is safe, though, although you have	4368	when the functionality isn't available is safe, though, although you have
3153	to make sure you link against any libraries where the C<clock_gettime>	4369	to make sure you link against any libraries where the C<clock_gettime>
3154	function is hiding in (often F<-lrt>).	4370	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
3155		4371
3156	=item EV_USE_REALTIME	4372	=item EV_USE_REALTIME
3157		4373
3158	If defined to be C<1>, libev will try to detect the availability of the	4374	If defined to be C<1>, libev will try to detect the availability of the
3159	real-time clock option at compile time (and assume its availability at	4375	real-time clock option at compile time (and assume its availability
3160	runtime if successful). Otherwise no use of the real-time clock option will	4376	at runtime if successful). Otherwise no use of the real-time clock
3161	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	4377	option will be attempted. This effectively replaces C<gettimeofday>
3162	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	4378	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
3163	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	4379	correctness. See the note about libraries in the description of
		4380	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		4381	C<EV_USE_CLOCK_SYSCALL>.
		4382
		4383	=item EV_USE_CLOCK_SYSCALL
		4384
		4385	If defined to be C<1>, libev will try to use a direct syscall instead
		4386	of calling the system-provided C<clock_gettime> function. This option
		4387	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		4388	unconditionally pulls in C<libpthread>, slowing down single-threaded
		4389	programs needlessly. Using a direct syscall is slightly slower (in
		4390	theory), because no optimised vdso implementation can be used, but avoids
		4391	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		4392	higher, as it simplifies linking (no need for C<-lrt>).
3164		4393
3165	=item EV_USE_NANOSLEEP	4394	=item EV_USE_NANOSLEEP
3166		4395
3167	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	4396	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
3168	and will use it for delays. Otherwise it will use C<select ()>.	4397	and will use it for delays. Otherwise it will use C<select ()>.
…		…
3184		4413
3185	=item EV_SELECT_USE_FD_SET	4414	=item EV_SELECT_USE_FD_SET
3186		4415
3187	If defined to C<1>, then the select backend will use the system C<fd_set>	4416	If defined to C<1>, then the select backend will use the system C<fd_set>
3188	structure. This is useful if libev doesn't compile due to a missing	4417	structure. This is useful if libev doesn't compile due to a missing
3189	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	4418	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
3190	exotic systems. This usually limits the range of file descriptors to some	4419	on exotic systems. This usually limits the range of file descriptors to
3191	low limit such as 1024 or might have other limitations (winsocket only	4420	some low limit such as 1024 or might have other limitations (winsocket
3192	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	4421	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
3193	influence the size of the C<fd_set> used.	4422	configures the maximum size of the C<fd_set>.
3194		4423
3195	=item EV_SELECT_IS_WINSOCKET	4424	=item EV_SELECT_IS_WINSOCKET
3196		4425
3197	When defined to C<1>, the select backend will assume that	4426	When defined to C<1>, the select backend will assume that
3198	select/socket/connect etc. don't understand file descriptors but	4427	select/socket/connect etc. don't understand file descriptors but
…		…
3200	be used is the winsock select). This means that it will call	4429	be used is the winsock select). This means that it will call
3201	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	4430	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
3202	it is assumed that all these functions actually work on fds, even	4431	it is assumed that all these functions actually work on fds, even
3203	on win32. Should not be defined on non-win32 platforms.	4432	on win32. Should not be defined on non-win32 platforms.
3204		4433
3205	=item EV_FD_TO_WIN32_HANDLE	4434	=item EV_FD_TO_WIN32_HANDLE(fd)
3206		4435
3207	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	4436	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
3208	file descriptors to socket handles. When not defining this symbol (the	4437	file descriptors to socket handles. When not defining this symbol (the
3209	default), then libev will call C<_get_osfhandle>, which is usually	4438	default), then libev will call C<_get_osfhandle>, which is usually
3210	correct. In some cases, programs use their own file descriptor management,	4439	correct. In some cases, programs use their own file descriptor management,
3211	in which case they can provide this function to map fds to socket handles.	4440	in which case they can provide this function to map fds to socket handles.
		4441
		4442	=item EV_WIN32_HANDLE_TO_FD(handle)
		4443
		4444	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		4445	using the standard C<_open_osfhandle> function. For programs implementing
		4446	their own fd to handle mapping, overwriting this function makes it easier
		4447	to do so. This can be done by defining this macro to an appropriate value.
		4448
		4449	=item EV_WIN32_CLOSE_FD(fd)
		4450
		4451	If programs implement their own fd to handle mapping on win32, then this
		4452	macro can be used to override the C<close> function, useful to unregister
		4453	file descriptors again. Note that the replacement function has to close
		4454	the underlying OS handle.
3212		4455
3213	=item EV_USE_POLL	4456	=item EV_USE_POLL
3214		4457
3215	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4458	If defined to be C<1>, libev will compile in support for the C<poll>(2)
3216	backend. Otherwise it will be enabled on non-win32 platforms. It	4459	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
3255	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4498	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3256		4499
3257	=item EV_ATOMIC_T	4500	=item EV_ATOMIC_T
3258		4501
3259	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	4502	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
3260	access is atomic with respect to other threads or signal contexts. No such	4503	access is atomic and serialised with respect to other threads or signal
3261	type is easily found in the C language, so you can provide your own type	4504	contexts. No such type is easily found in the C language, so you can
3262	that you know is safe for your purposes. It is used both for signal handler "locking"	4505	provide your own type that you know is safe for your purposes. It is used
3263	as well as for signal and thread safety in C<ev_async> watchers.	4506	both for signal handler "locking" as well as for signal and thread safety
		4507	in C<ev_async> watchers.
3264		4508
3265	In the absence of this define, libev will use C<sig_atomic_t volatile>	4509	In the absence of this define, libev will use C<sig_atomic_t volatile>
3266	(from F<signal.h>), which is usually good enough on most platforms.	4510	(from F<signal.h>), which is usually good enough on most platforms,
		4511	although strictly speaking using a type that also implies a memory fence
		4512	is required.
3267		4513
3268	=item EV_H	4514	=item EV_H (h)
3269		4515
3270	The name of the F<ev.h> header file used to include it. The default if	4516	The name of the F<ev.h> header file used to include it. The default if
3271	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	4517	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
3272	used to virtually rename the F<ev.h> header file in case of conflicts.	4518	used to virtually rename the F<ev.h> header file in case of conflicts.
3273		4519
3274	=item EV_CONFIG_H	4520	=item EV_CONFIG_H (h)
3275		4521
3276	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	4522	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
3277	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	4523	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
3278	C<EV_H>, above.	4524	C<EV_H>, above.
3279		4525
3280	=item EV_EVENT_H	4526	=item EV_EVENT_H (h)
3281		4527
3282	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	4528	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
3283	of how the F<event.h> header can be found, the default is C<"event.h">.	4529	of how the F<event.h> header can be found, the default is C<"event.h">.
3284		4530
3285	=item EV_PROTOTYPES	4531	=item EV_PROTOTYPES (h)
3286		4532
3287	If defined to be C<0>, then F<ev.h> will not define any function	4533	If defined to be C<0>, then F<ev.h> will not define any function
3288	prototypes, but still define all the structs and other symbols. This is	4534	prototypes, but still define all the structs and other symbols. This is
3289	occasionally useful if you want to provide your own wrapper functions	4535	occasionally useful if you want to provide your own wrapper functions
3290	around libev functions.	4536	around libev functions.
…		…
3295	will have the C<struct ev_loop *> as first argument, and you can create	4541	will have the C<struct ev_loop *> as first argument, and you can create
3296	additional independent event loops. Otherwise there will be no support	4542	additional independent event loops. Otherwise there will be no support
3297	for multiple event loops and there is no first event loop pointer	4543	for multiple event loops and there is no first event loop pointer
3298	argument. Instead, all functions act on the single default loop.	4544	argument. Instead, all functions act on the single default loop.
3299		4545
		4546	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4547	default loop when multiplicity is switched off - you always have to
		4548	initialise the loop manually in this case.
		4549
3300	=item EV_MINPRI	4550	=item EV_MINPRI
3301		4551
3302	=item EV_MAXPRI	4552	=item EV_MAXPRI
3303		4553
3304	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4554	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
3312	fine.	4562	fine.
3313		4563
3314	If your embedding application does not need any priorities, defining these	4564	If your embedding application does not need any priorities, defining these
3315	both to C<0> will save some memory and CPU.	4565	both to C<0> will save some memory and CPU.
3316		4566
3317	=item EV_PERIODIC_ENABLE	4567	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4568	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4569	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3318		4570
3319	If undefined or defined to be C<1>, then periodic timers are supported. If	4571	If undefined or defined to be C<1> (and the platform supports it), then
3320	defined to be C<0>, then they are not. Disabling them saves a few kB of	4572	the respective watcher type is supported. If defined to be C<0>, then it
3321	code.	4573	is not. Disabling watcher types mainly saves code size.
3322		4574
3323	=item EV_IDLE_ENABLE	4575	=item EV_FEATURES
3324
3325	If undefined or defined to be C<1>, then idle watchers are supported. If
3326	defined to be C<0>, then they are not. Disabling them saves a few kB of
3327	code.
3328
3329	=item EV_EMBED_ENABLE
3330
3331	If undefined or defined to be C<1>, then embed watchers are supported. If
3332	defined to be C<0>, then they are not. Embed watchers rely on most other
3333	watcher types, which therefore must not be disabled.
3334
3335	=item EV_STAT_ENABLE
3336
3337	If undefined or defined to be C<1>, then stat watchers are supported. If
3338	defined to be C<0>, then they are not.
3339
3340	=item EV_FORK_ENABLE
3341
3342	If undefined or defined to be C<1>, then fork watchers are supported. If
3343	defined to be C<0>, then they are not.
3344
3345	=item EV_ASYNC_ENABLE
3346
3347	If undefined or defined to be C<1>, then async watchers are supported. If
3348	defined to be C<0>, then they are not.
3349
3350	=item EV_MINIMAL
3351		4576
3352	If you need to shave off some kilobytes of code at the expense of some	4577	If you need to shave off some kilobytes of code at the expense of some
3353	speed, define this symbol to C<1>. Currently this is used to override some	4578	speed (but with the full API), you can define this symbol to request
3354	inlining decisions, saves roughly 30% code size on amd64. It also selects a	4579	certain subsets of functionality. The default is to enable all features
3355	much smaller 2-heap for timer management over the default 4-heap.	4580	that can be enabled on the platform.
		4581
		4582	A typical way to use this symbol is to define it to C<0> (or to a bitset
		4583	with some broad features you want) and then selectively re-enable
		4584	additional parts you want, for example if you want everything minimal,
		4585	but multiple event loop support, async and child watchers and the poll
		4586	backend, use this:
		4587
		4588	#define EV_FEATURES 0
		4589	#define EV_MULTIPLICITY 1
		4590	#define EV_USE_POLL 1
		4591	#define EV_CHILD_ENABLE 1
		4592	#define EV_ASYNC_ENABLE 1
		4593
		4594	The actual value is a bitset, it can be a combination of the following
		4595	values:
		4596
		4597	=over 4
		4598
		4599	=item C<1> - faster/larger code
		4600
		4601	Use larger code to speed up some operations.
		4602
		4603	Currently this is used to override some inlining decisions (enlarging the
		4604	code size by roughly 30% on amd64).
		4605
		4606	When optimising for size, use of compiler flags such as C<-Os> with
		4607	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4608	assertions.
		4609
		4610	=item C<2> - faster/larger data structures
		4611
		4612	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4613	hash table sizes and so on. This will usually further increase code size
		4614	and can additionally have an effect on the size of data structures at
		4615	runtime.
		4616
		4617	=item C<4> - full API configuration
		4618
		4619	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4620	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4621
		4622	=item C<8> - full API
		4623
		4624	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4625	details on which parts of the API are still available without this
		4626	feature, and do not complain if this subset changes over time.
		4627
		4628	=item C<16> - enable all optional watcher types
		4629
		4630	Enables all optional watcher types. If you want to selectively enable
		4631	only some watcher types other than I/O and timers (e.g. prepare,
		4632	embed, async, child...) you can enable them manually by defining
		4633	C<EV_watchertype_ENABLE> to C<1> instead.
		4634
		4635	=item C<32> - enable all backends
		4636
		4637	This enables all backends - without this feature, you need to enable at
		4638	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4639
		4640	=item C<64> - enable OS-specific "helper" APIs
		4641
		4642	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4643	default.
		4644
		4645	=back
		4646
		4647	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4648	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4649	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4650	watchers, timers and monotonic clock support.
		4651
		4652	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4653	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4654	your program might be left out as well - a binary starting a timer and an
		4655	I/O watcher then might come out at only 5Kb.
		4656
		4657	=item EV_API_STATIC
		4658
		4659	If this symbol is defined (by default it is not), then all identifiers
		4660	will have static linkage. This means that libev will not export any
		4661	identifiers, and you cannot link against libev anymore. This can be useful
		4662	when you embed libev, only want to use libev functions in a single file,
		4663	and do not want its identifiers to be visible.
		4664
		4665	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4666	wants to use libev.
		4667
		4668	=item EV_AVOID_STDIO
		4669
		4670	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4671	functions (printf, scanf, perror etc.). This will increase the code size
		4672	somewhat, but if your program doesn't otherwise depend on stdio and your
		4673	libc allows it, this avoids linking in the stdio library which is quite
		4674	big.
		4675
		4676	Note that error messages might become less precise when this option is
		4677	enabled.
		4678
		4679	=item EV_NSIG
		4680
		4681	The highest supported signal number, +1 (or, the number of
		4682	signals): Normally, libev tries to deduce the maximum number of signals
		4683	automatically, but sometimes this fails, in which case it can be
		4684	specified. Also, using a lower number than detected (C<32> should be
		4685	good for about any system in existence) can save some memory, as libev
		4686	statically allocates some 12-24 bytes per signal number.
3356		4687
3357	=item EV_PID_HASHSIZE	4688	=item EV_PID_HASHSIZE
3358		4689
3359	C<ev_child> watchers use a small hash table to distribute workload by	4690	C<ev_child> watchers use a small hash table to distribute workload by
3360	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	4691	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3361	than enough. If you need to manage thousands of children you might want to	4692	usually more than enough. If you need to manage thousands of children you
3362	increase this value (I<must> be a power of two).	4693	might want to increase this value (I<must> be a power of two).
3363		4694
3364	=item EV_INOTIFY_HASHSIZE	4695	=item EV_INOTIFY_HASHSIZE
3365		4696
3366	C<ev_stat> watchers use a small hash table to distribute workload by	4697	C<ev_stat> watchers use a small hash table to distribute workload by
3367	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4698	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3368	usually more than enough. If you need to manage thousands of C<ev_stat>	4699	disabled), usually more than enough. If you need to manage thousands of
3369	watchers you might want to increase this value (I<must> be a power of	4700	C<ev_stat> watchers you might want to increase this value (I<must> be a
3370	two).	4701	power of two).
3371		4702
3372	=item EV_USE_4HEAP	4703	=item EV_USE_4HEAP
3373		4704
3374	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4705	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3375	timer and periodics heaps, libev uses a 4-heap when this symbol is defined	4706	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3376	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably	4707	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3377	faster performance with many (thousands) of watchers.	4708	faster performance with many (thousands) of watchers.
3378		4709
3379	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4710	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3380	(disabled).	4711	will be C<0>.
3381		4712
3382	=item EV_HEAP_CACHE_AT	4713	=item EV_HEAP_CACHE_AT
3383		4714
3384	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4715	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3385	timer and periodics heaps, libev can cache the timestamp (I<at>) within	4716	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3386	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4717	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3387	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4718	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3388	but avoids random read accesses on heap changes. This improves performance	4719	but avoids random read accesses on heap changes. This improves performance
3389	noticeably with many (hundreds) of watchers.	4720	noticeably with many (hundreds) of watchers.
3390		4721
3391	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4722	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3392	(disabled).	4723	will be C<0>.
3393		4724
3394	=item EV_VERIFY	4725	=item EV_VERIFY
3395		4726
3396	Controls how much internal verification (see C<ev_loop_verify ()>) will	4727	Controls how much internal verification (see C<ev_verify ()>) will
3397	be done: If set to C<0>, no internal verification code will be compiled	4728	be done: If set to C<0>, no internal verification code will be compiled
3398	in. If set to C<1>, then verification code will be compiled in, but not	4729	in. If set to C<1>, then verification code will be compiled in, but not
3399	called. If set to C<2>, then the internal verification code will be	4730	called. If set to C<2>, then the internal verification code will be
3400	called once per loop, which can slow down libev. If set to C<3>, then the	4731	called once per loop, which can slow down libev. If set to C<3>, then the
3401	verification code will be called very frequently, which will slow down	4732	verification code will be called very frequently, which will slow down
3402	libev considerably.	4733	libev considerably.
3403		4734
3404	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4735	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3405	C<0>.	4736	will be C<0>.
3406		4737
3407	=item EV_COMMON	4738	=item EV_COMMON
3408		4739
3409	By default, all watchers have a C<void *data> member. By redefining	4740	By default, all watchers have a C<void *data> member. By redefining
3410	this macro to a something else you can include more and other types of	4741	this macro to something else you can include more and other types of
3411	members. You have to define it each time you include one of the files,	4742	members. You have to define it each time you include one of the files,
3412	though, and it must be identical each time.	4743	though, and it must be identical each time.
3413		4744
3414	For example, the perl EV module uses something like this:	4745	For example, the perl EV module uses something like this:
3415		4746
…		…
3468	file.	4799	file.
3469		4800
3470	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4801	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3471	that everybody includes and which overrides some configure choices:	4802	that everybody includes and which overrides some configure choices:
3472		4803
3473	#define EV_MINIMAL 1	4804	#define EV_FEATURES 8
3474	#define EV_USE_POLL 0	4805	#define EV_USE_SELECT 1
3475	#define EV_MULTIPLICITY 0
3476	#define EV_PERIODIC_ENABLE 0	4806	#define EV_PREPARE_ENABLE 1
		4807	#define EV_IDLE_ENABLE 1
3477	#define EV_STAT_ENABLE 0	4808	#define EV_SIGNAL_ENABLE 1
3478	#define EV_FORK_ENABLE 0	4809	#define EV_CHILD_ENABLE 1
		4810	#define EV_USE_STDEXCEPT 0
3479	#define EV_CONFIG_H <config.h>	4811	#define EV_CONFIG_H <config.h>
3480	#define EV_MINPRI 0
3481	#define EV_MAXPRI 0
3482		4812
3483	#include "ev++.h"	4813	#include "ev++.h"
3484		4814
3485	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4815	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3486		4816
3487	#include "ev_cpp.h"	4817	#include "ev_cpp.h"
3488	#include "ev.c"	4818	#include "ev.c"
3489		4819
3490	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES	4820	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
3491		4821
3492	=head2 THREADS AND COROUTINES	4822	=head2 THREADS AND COROUTINES
3493		4823
3494	=head3 THREADS	4824	=head3 THREADS
3495		4825
…		…
3546	default loop and triggering an C<ev_async> watcher from the default loop	4876	default loop and triggering an C<ev_async> watcher from the default loop
3547	watcher callback into the event loop interested in the signal.	4877	watcher callback into the event loop interested in the signal.
3548		4878
3549	=back	4879	=back
3550		4880
		4881	See also L<THREAD LOCKING EXAMPLE>.
		4882
3551	=head3 COROUTINES	4883	=head3 COROUTINES
3552		4884
3553	Libev is very accommodating to coroutines ("cooperative threads"):	4885	Libev is very accommodating to coroutines ("cooperative threads"):
3554	libev fully supports nesting calls to its functions from different	4886	libev fully supports nesting calls to its functions from different
3555	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4887	coroutines (e.g. you can call C<ev_run> on the same loop from two
3556	different coroutines, and switch freely between both coroutines running the	4888	different coroutines, and switch freely between both coroutines running
3557	loop, as long as you don't confuse yourself). The only exception is that	4889	the loop, as long as you don't confuse yourself). The only exception is
3558	you must not do this from C<ev_periodic> reschedule callbacks.	4890	that you must not do this from C<ev_periodic> reschedule callbacks.
3559		4891
3560	Care has been taken to ensure that libev does not keep local state inside	4892	Care has been taken to ensure that libev does not keep local state inside
3561	C<ev_loop>, and other calls do not usually allow for coroutine switches as	4893	C<ev_run>, and other calls do not usually allow for coroutine switches as
3562	they do not clal any callbacks.	4894	they do not call any callbacks.
3563		4895
3564	=head2 COMPILER WARNINGS	4896	=head2 COMPILER WARNINGS
3565		4897
3566	Depending on your compiler and compiler settings, you might get no or a	4898	Depending on your compiler and compiler settings, you might get no or a
3567	lot of warnings when compiling libev code. Some people are apparently	4899	lot of warnings when compiling libev code. Some people are apparently
…		…
3577	maintainable.	4909	maintainable.
3578		4910
3579	And of course, some compiler warnings are just plain stupid, or simply	4911	And of course, some compiler warnings are just plain stupid, or simply
3580	wrong (because they don't actually warn about the condition their message	4912	wrong (because they don't actually warn about the condition their message
3581	seems to warn about). For example, certain older gcc versions had some	4913	seems to warn about). For example, certain older gcc versions had some
3582	warnings that resulted an extreme number of false positives. These have	4914	warnings that resulted in an extreme number of false positives. These have
3583	been fixed, but some people still insist on making code warn-free with	4915	been fixed, but some people still insist on making code warn-free with
3584	such buggy versions.	4916	such buggy versions.
3585		4917
3586	While libev is written to generate as few warnings as possible,	4918	While libev is written to generate as few warnings as possible,
3587	"warn-free" code is not a goal, and it is recommended not to build libev	4919	"warn-free" code is not a goal, and it is recommended not to build libev
…		…
3601	==2274== definitely lost: 0 bytes in 0 blocks.	4933	==2274== definitely lost: 0 bytes in 0 blocks.
3602	==2274== possibly lost: 0 bytes in 0 blocks.	4934	==2274== possibly lost: 0 bytes in 0 blocks.
3603	==2274== still reachable: 256 bytes in 1 blocks.	4935	==2274== still reachable: 256 bytes in 1 blocks.
3604		4936
3605	Then there is no memory leak, just as memory accounted to global variables	4937	Then there is no memory leak, just as memory accounted to global variables
3606	is not a memleak - the memory is still being refernced, and didn't leak.	4938	is not a memleak - the memory is still being referenced, and didn't leak.
3607		4939
3608	Similarly, under some circumstances, valgrind might report kernel bugs	4940	Similarly, under some circumstances, valgrind might report kernel bugs
3609	as if it were a bug in libev (e.g. in realloc or in the poll backend,	4941	as if it were a bug in libev (e.g. in realloc or in the poll backend,
3610	although an acceptable workaround has been found here), or it might be	4942	although an acceptable workaround has been found here), or it might be
3611	confused.	4943	confused.
…		…
3623	I suggest using suppression lists.	4955	I suggest using suppression lists.
3624		4956
3625		4957
3626	=head1 PORTABILITY NOTES	4958	=head1 PORTABILITY NOTES
3627		4959
		4960	=head2 GNU/LINUX 32 BIT LIMITATIONS
		4961
		4962	GNU/Linux is the only common platform that supports 64 bit file/large file
		4963	interfaces but I<disables> them by default.
		4964
		4965	That means that libev compiled in the default environment doesn't support
		4966	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		4967
		4968	Unfortunately, many programs try to work around this GNU/Linux issue
		4969	by enabling the large file API, which makes them incompatible with the
		4970	standard libev compiled for their system.
		4971
		4972	Likewise, libev cannot enable the large file API itself as this would
		4973	suddenly make it incompatible to the default compile time environment,
		4974	i.e. all programs not using special compile switches.
		4975
		4976	=head2 OS/X AND DARWIN BUGS
		4977
		4978	The whole thing is a bug if you ask me - basically any system interface
		4979	you touch is broken, whether it is locales, poll, kqueue or even the
		4980	OpenGL drivers.
		4981
		4982	=head3 C<kqueue> is buggy
		4983
		4984	The kqueue syscall is broken in all known versions - most versions support
		4985	only sockets, many support pipes.
		4986
		4987	Libev tries to work around this by not using C<kqueue> by default on this
		4988	rotten platform, but of course you can still ask for it when creating a
		4989	loop - embedding a socket-only kqueue loop into a select-based one is
		4990	probably going to work well.
		4991
		4992	=head3 C<poll> is buggy
		4993
		4994	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4995	implementation by something calling C<kqueue> internally around the 10.5.6
		4996	release, so now C<kqueue> I<and> C<poll> are broken.
		4997
		4998	Libev tries to work around this by not using C<poll> by default on
		4999	this rotten platform, but of course you can still ask for it when creating
		5000	a loop.
		5001
		5002	=head3 C<select> is buggy
		5003
		5004	All that's left is C<select>, and of course Apple found a way to fuck this
		5005	one up as well: On OS/X, C<select> actively limits the number of file
		5006	descriptors you can pass in to 1024 - your program suddenly crashes when
		5007	you use more.
		5008
		5009	There is an undocumented "workaround" for this - defining
		5010	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		5011	work on OS/X.
		5012
		5013	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		5014
		5015	=head3 C<errno> reentrancy
		5016
		5017	The default compile environment on Solaris is unfortunately so
		5018	thread-unsafe that you can't even use components/libraries compiled
		5019	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		5020	defined by default. A valid, if stupid, implementation choice.
		5021
		5022	If you want to use libev in threaded environments you have to make sure
		5023	it's compiled with C<_REENTRANT> defined.
		5024
		5025	=head3 Event port backend
		5026
		5027	The scalable event interface for Solaris is called "event
		5028	ports". Unfortunately, this mechanism is very buggy in all major
		5029	releases. If you run into high CPU usage, your program freezes or you get
		5030	a large number of spurious wakeups, make sure you have all the relevant
		5031	and latest kernel patches applied. No, I don't know which ones, but there
		5032	are multiple ones to apply, and afterwards, event ports actually work
		5033	great.
		5034
		5035	If you can't get it to work, you can try running the program by setting
		5036	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		5037	C<select> backends.
		5038
		5039	=head2 AIX POLL BUG
		5040
		5041	AIX unfortunately has a broken C<poll.h> header. Libev works around
		5042	this by trying to avoid the poll backend altogether (i.e. it's not even
		5043	compiled in), which normally isn't a big problem as C<select> works fine
		5044	with large bitsets on AIX, and AIX is dead anyway.
		5045
3628	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	5046	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		5047
		5048	=head3 General issues
3629		5049
3630	Win32 doesn't support any of the standards (e.g. POSIX) that libev	5050	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3631	requires, and its I/O model is fundamentally incompatible with the POSIX	5051	requires, and its I/O model is fundamentally incompatible with the POSIX
3632	model. Libev still offers limited functionality on this platform in	5052	model. Libev still offers limited functionality on this platform in
3633	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5053	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
3634	descriptors. This only applies when using Win32 natively, not when using	5054	descriptors. This only applies when using Win32 natively, not when using
3635	e.g. cygwin.	5055	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		5056	as every compiler comes with a slightly differently broken/incompatible
		5057	environment.
3636		5058
3637	Lifting these limitations would basically require the full	5059	Lifting these limitations would basically require the full
3638	re-implementation of the I/O system. If you are into these kinds of	5060	re-implementation of the I/O system. If you are into this kind of thing,
3639	things, then note that glib does exactly that for you in a very portable	5061	then note that glib does exactly that for you in a very portable way (note
3640	way (note also that glib is the slowest event library known to man).	5062	also that glib is the slowest event library known to man).
3641		5063
3642	There is no supported compilation method available on windows except	5064	There is no supported compilation method available on windows except
3643	embedding it into other applications.	5065	embedding it into other applications.
		5066
		5067	Sensible signal handling is officially unsupported by Microsoft - libev
		5068	tries its best, but under most conditions, signals will simply not work.
3644		5069
3645	Not a libev limitation but worth mentioning: windows apparently doesn't	5070	Not a libev limitation but worth mentioning: windows apparently doesn't
3646	accept large writes: instead of resulting in a partial write, windows will	5071	accept large writes: instead of resulting in a partial write, windows will
3647	either accept everything or return C<ENOBUFS> if the buffer is too large,	5072	either accept everything or return C<ENOBUFS> if the buffer is too large,
3648	so make sure you only write small amounts into your sockets (less than a	5073	so make sure you only write small amounts into your sockets (less than a
…		…
3653	the abysmal performance of winsockets, using a large number of sockets	5078	the abysmal performance of winsockets, using a large number of sockets
3654	is not recommended (and not reasonable). If your program needs to use	5079	is not recommended (and not reasonable). If your program needs to use
3655	more than a hundred or so sockets, then likely it needs to use a totally	5080	more than a hundred or so sockets, then likely it needs to use a totally
3656	different implementation for windows, as libev offers the POSIX readiness	5081	different implementation for windows, as libev offers the POSIX readiness
3657	notification model, which cannot be implemented efficiently on windows	5082	notification model, which cannot be implemented efficiently on windows
3658	(Microsoft monopoly games).	5083	(due to Microsoft monopoly games).
3659		5084
3660	A typical way to use libev under windows is to embed it (see the embedding	5085	A typical way to use libev under windows is to embed it (see the embedding
3661	section for details) and use the following F<evwrap.h> header file instead	5086	section for details) and use the following F<evwrap.h> header file instead
3662	of F<ev.h>:	5087	of F<ev.h>:
3663		5088
…		…
3670	you do I<not> compile the F<ev.c> or any other embedded source files!):	5095	you do I<not> compile the F<ev.c> or any other embedded source files!):
3671		5096
3672	#include "evwrap.h"	5097	#include "evwrap.h"
3673	#include "ev.c"	5098	#include "ev.c"
3674		5099
3675	=over 4
3676
3677	=item The winsocket select function	5100	=head3 The winsocket C<select> function
3678		5101
3679	The winsocket C<select> function doesn't follow POSIX in that it	5102	The winsocket C<select> function doesn't follow POSIX in that it
3680	requires socket I<handles> and not socket I<file descriptors> (it is	5103	requires socket I<handles> and not socket I<file descriptors> (it is
3681	also extremely buggy). This makes select very inefficient, and also	5104	also extremely buggy). This makes select very inefficient, and also
3682	requires a mapping from file descriptors to socket handles (the Microsoft	5105	requires a mapping from file descriptors to socket handles (the Microsoft
…		…
3691	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	5114	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
3692		5115
3693	Note that winsockets handling of fd sets is O(n), so you can easily get a	5116	Note that winsockets handling of fd sets is O(n), so you can easily get a
3694	complexity in the O(n²) range when using win32.	5117	complexity in the O(n²) range when using win32.
3695		5118
3696	=item Limited number of file descriptors	5119	=head3 Limited number of file descriptors
3697		5120
3698	Windows has numerous arbitrary (and low) limits on things.	5121	Windows has numerous arbitrary (and low) limits on things.
3699		5122
3700	Early versions of winsocket's select only supported waiting for a maximum	5123	Early versions of winsocket's select only supported waiting for a maximum
3701	of C<64> handles (probably owning to the fact that all windows kernels	5124	of C<64> handles (probably owning to the fact that all windows kernels
3702	can only wait for C<64> things at the same time internally; Microsoft	5125	can only wait for C<64> things at the same time internally; Microsoft
3703	recommends spawning a chain of threads and wait for 63 handles and the	5126	recommends spawning a chain of threads and wait for 63 handles and the
3704	previous thread in each. Great).	5127	previous thread in each. Sounds great!).
3705		5128
3706	Newer versions support more handles, but you need to define C<FD_SETSIZE>	5129	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3707	to some high number (e.g. C<2048>) before compiling the winsocket select	5130	to some high number (e.g. C<2048>) before compiling the winsocket select
3708	call (which might be in libev or elsewhere, for example, perl does its own	5131	call (which might be in libev or elsewhere, for example, perl and many
3709	select emulation on windows).	5132	other interpreters do their own select emulation on windows).
3710		5133
3711	Another limit is the number of file descriptors in the Microsoft runtime	5134	Another limit is the number of file descriptors in the Microsoft runtime
3712	libraries, which by default is C<64> (there must be a hidden I<64> fetish	5135	libraries, which by default is C<64> (there must be a hidden I<64>
3713	or something like this inside Microsoft). You can increase this by calling	5136	fetish or something like this inside Microsoft). You can increase this
3714	C<_setmaxstdio>, which can increase this limit to C<2048> (another	5137	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3715	arbitrary limit), but is broken in many versions of the Microsoft runtime	5138	(another arbitrary limit), but is broken in many versions of the Microsoft
3716	libraries.
3717
3718	This might get you to about C<512> or C<2048> sockets (depending on	5139	runtime libraries. This might get you to about C<512> or C<2048> sockets
3719	windows version and/or the phase of the moon). To get more, you need to	5140	(depending on windows version and/or the phase of the moon). To get more,
3720	wrap all I/O functions and provide your own fd management, but the cost of	5141	you need to wrap all I/O functions and provide your own fd management, but
3721	calling select (O(n²)) will likely make this unworkable.	5142	the cost of calling select (O(n²)) will likely make this unworkable.
3722
3723	=back
3724		5143
3725	=head2 PORTABILITY REQUIREMENTS	5144	=head2 PORTABILITY REQUIREMENTS
3726		5145
3727	In addition to a working ISO-C implementation and of course the	5146	In addition to a working ISO-C implementation and of course the
3728	backend-specific APIs, libev relies on a few additional extensions:	5147	backend-specific APIs, libev relies on a few additional extensions:
…		…
3735	Libev assumes not only that all watcher pointers have the same internal	5154	Libev assumes not only that all watcher pointers have the same internal
3736	structure (guaranteed by POSIX but not by ISO C for example), but it also	5155	structure (guaranteed by POSIX but not by ISO C for example), but it also
3737	assumes that the same (machine) code can be used to call any watcher	5156	assumes that the same (machine) code can be used to call any watcher
3738	callback: The watcher callbacks have different type signatures, but libev	5157	callback: The watcher callbacks have different type signatures, but libev
3739	calls them using an C<ev_watcher *> internally.	5158	calls them using an C<ev_watcher *> internally.
		5159
		5160	=item pointer accesses must be thread-atomic
		5161
		5162	Accessing a pointer value must be atomic, it must both be readable and
		5163	writable in one piece - this is the case on all current architectures.
3740		5164
3741	=item C<sig_atomic_t volatile> must be thread-atomic as well	5165	=item C<sig_atomic_t volatile> must be thread-atomic as well
3742		5166
3743	The type C<sig_atomic_t volatile> (or whatever is defined as	5167	The type C<sig_atomic_t volatile> (or whatever is defined as
3744	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different	5168	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
…		…
3767	watchers.	5191	watchers.
3768		5192
3769	=item C<double> must hold a time value in seconds with enough accuracy	5193	=item C<double> must hold a time value in seconds with enough accuracy
3770		5194
3771	The type C<double> is used to represent timestamps. It is required to	5195	The type C<double> is used to represent timestamps. It is required to
3772	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	5196	have at least 51 bits of mantissa (and 9 bits of exponent), which is
3773	enough for at least into the year 4000. This requirement is fulfilled by	5197	good enough for at least into the year 4000 with millisecond accuracy
		5198	(the design goal for libev). This requirement is overfulfilled by
3774	implementations implementing IEEE 754 (basically all existing ones).	5199	implementations using IEEE 754, which is basically all existing ones.
		5200
		5201	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5202	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5203	is either obsolete or somebody patched it to use C<long double> or
		5204	something like that, just kidding).
3775		5205
3776	=back	5206	=back
3777		5207
3778	If you know of other additional requirements drop me a note.	5208	If you know of other additional requirements drop me a note.
3779		5209
…		…
3841	=item Processing ev_async_send: O(number_of_async_watchers)	5271	=item Processing ev_async_send: O(number_of_async_watchers)
3842		5272
3843	=item Processing signals: O(max_signal_number)	5273	=item Processing signals: O(max_signal_number)
3844		5274
3845	Sending involves a system call I<iff> there were no other C<ev_async_send>	5275	Sending involves a system call I<iff> there were no other C<ev_async_send>
3846	calls in the current loop iteration. Checking for async and signal events	5276	calls in the current loop iteration and the loop is currently
		5277	blocked. Checking for async and signal events involves iterating over all
3847	involves iterating over all running async watchers or all signal numbers.	5278	running async watchers or all signal numbers.
3848		5279
3849	=back	5280	=back
3850		5281
3851		5282
		5283	=head1 PORTING FROM LIBEV 3.X TO 4.X
		5284
		5285	The major version 4 introduced some incompatible changes to the API.
		5286
		5287	At the moment, the C<ev.h> header file provides compatibility definitions
		5288	for all changes, so most programs should still compile. The compatibility
		5289	layer might be removed in later versions of libev, so better update to the
		5290	new API early than late.
		5291
		5292	=over 4
		5293
		5294	=item C<EV_COMPAT3> backwards compatibility mechanism
		5295
		5296	The backward compatibility mechanism can be controlled by
		5297	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		5298	section.
		5299
		5300	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5301
		5302	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5303
		5304	ev_loop_destroy (EV_DEFAULT_UC);
		5305	ev_loop_fork (EV_DEFAULT);
		5306
		5307	=item function/symbol renames
		5308
		5309	A number of functions and symbols have been renamed:
		5310
		5311	ev_loop => ev_run
		5312	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5313	EVLOOP_ONESHOT => EVRUN_ONCE
		5314
		5315	ev_unloop => ev_break
		5316	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5317	EVUNLOOP_ONE => EVBREAK_ONE
		5318	EVUNLOOP_ALL => EVBREAK_ALL
		5319
		5320	EV_TIMEOUT => EV_TIMER
		5321
		5322	ev_loop_count => ev_iteration
		5323	ev_loop_depth => ev_depth
		5324	ev_loop_verify => ev_verify
		5325
		5326	Most functions working on C<struct ev_loop> objects don't have an
		5327	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5328	associated constants have been renamed to not collide with the C<struct
		5329	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5330	as all other watcher types. Note that C<ev_loop_fork> is still called
		5331	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		5332	typedef.
		5333
		5334	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		5335
		5336	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		5337	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		5338	and work, but the library code will of course be larger.
		5339
		5340	=back
		5341
		5342
		5343	=head1 GLOSSARY
		5344
		5345	=over 4
		5346
		5347	=item active
		5348
		5349	A watcher is active as long as it has been started and not yet stopped.
		5350	See L<WATCHER STATES> for details.
		5351
		5352	=item application
		5353
		5354	In this document, an application is whatever is using libev.
		5355
		5356	=item backend
		5357
		5358	The part of the code dealing with the operating system interfaces.
		5359
		5360	=item callback
		5361
		5362	The address of a function that is called when some event has been
		5363	detected. Callbacks are being passed the event loop, the watcher that
		5364	received the event, and the actual event bitset.
		5365
		5366	=item callback/watcher invocation
		5367
		5368	The act of calling the callback associated with a watcher.
		5369
		5370	=item event
		5371
		5372	A change of state of some external event, such as data now being available
		5373	for reading on a file descriptor, time having passed or simply not having
		5374	any other events happening anymore.
		5375
		5376	In libev, events are represented as single bits (such as C<EV_READ> or
		5377	C<EV_TIMER>).
		5378
		5379	=item event library
		5380
		5381	A software package implementing an event model and loop.
		5382
		5383	=item event loop
		5384
		5385	An entity that handles and processes external events and converts them
		5386	into callback invocations.
		5387
		5388	=item event model
		5389
		5390	The model used to describe how an event loop handles and processes
		5391	watchers and events.
		5392
		5393	=item pending
		5394
		5395	A watcher is pending as soon as the corresponding event has been
		5396	detected. See L<WATCHER STATES> for details.
		5397
		5398	=item real time
		5399
		5400	The physical time that is observed. It is apparently strictly monotonic :)
		5401
		5402	=item wall-clock time
		5403
		5404	The time and date as shown on clocks. Unlike real time, it can actually
		5405	be wrong and jump forwards and backwards, e.g. when you adjust your
		5406	clock.
		5407
		5408	=item watcher
		5409
		5410	A data structure that describes interest in certain events. Watchers need
		5411	to be started (attached to an event loop) before they can receive events.
		5412
		5413	=back
		5414
3852	=head1 AUTHOR	5415	=head1 AUTHOR
3853		5416
3854	Marc Lehmann <libev@schmorp.de>.	5417	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5418	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
3855		5419

Diff Legend

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