ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.40 by root, Sat Nov 24 10:15:16 2007 UTC vs.
Revision 1.292 by sf-exg, Mon Mar 22 09:57:01 2010 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
9	=head1 DESCRIPTION	9	=head2 EXAMPLE PROGRAM
		10
		11	// a single header file is required
		12	#include <ev.h>
		13
		14	#include <stdio.h> // for puts
		15
		16	// every watcher type has its own typedef'd struct
		17	// with the name ev_TYPE
		18	ev_io stdin_watcher;
		19	ev_timer timeout_watcher;
		20
		21	// all watcher callbacks have a similar signature
		22	// this callback is called when data is readable on stdin
		23	static void
		24	stdin_cb (EV_P_ ev_io *w, int revents)
		25	{
		26	puts ("stdin ready");
		27	// for one-shot events, one must manually stop the watcher
		28	// with its corresponding stop function.
		29	ev_io_stop (EV_A_ w);
		30
		31	// this causes all nested ev_loop's to stop iterating
		32	ev_unloop (EV_A_ EVUNLOOP_ALL);
		33	}
		34
		35	// another callback, this time for a time-out
		36	static void
		37	timeout_cb (EV_P_ ev_timer *w, int revents)
		38	{
		39	puts ("timeout");
		40	// this causes the innermost ev_loop to stop iterating
		41	ev_unloop (EV_A_ EVUNLOOP_ONE);
		42	}
		43
		44	int
		45	main (void)
		46	{
		47	// use the default event loop unless you have special needs
		48	struct ev_loop *loop = ev_default_loop (0);
		49
		50	// initialise an io watcher, then start it
		51	// this one will watch for stdin to become readable
		52	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		53	ev_io_start (loop, &stdin_watcher);
		54
		55	// initialise a timer watcher, then start it
		56	// simple non-repeating 5.5 second timeout
		57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		58	ev_timer_start (loop, &timeout_watcher);
		59
		60	// now wait for events to arrive
		61	ev_loop (loop, 0);
		62
		63	// unloop was called, so exit
		64	return 0;
		65	}
		66
		67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
		70
		71	The newest version of this document is also available as an html-formatted
		72	web page you might find easier to navigate when reading it for the first
		73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 ABOUT LIBEV
10		84
11	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
12	file descriptor being readable or a timeout occuring), and it will manage	86	file descriptor being readable or a timeout occurring), and it will manage
13	these event sources and provide your program with events.	87	these event sources and provide your program with events.
14		88
15	To do this, it must take more or less complete control over your process	89	To do this, it must take more or less complete control over your process
16	(or thread) by executing the I<event loop> handler, and will then	90	(or thread) by executing the I<event loop> handler, and will then
17	communicate events via a callback mechanism.	91	communicate events via a callback mechanism.
…		…
19	You register interest in certain events by registering so-called I<event	93	You register interest in certain events by registering so-called I<event
20	watchers>, which are relatively small C structures you initialise with the	94	watchers>, which are relatively small C structures you initialise with the
21	details of the event, and then hand it over to libev by I<starting> the	95	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	96	watcher.
23		97
24	=head1 FEATURES	98	=head2 FEATURES
25		99
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	100	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	101	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	102	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
		103	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
		104	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
		105	timers (C<ev_timer>), absolute timers with customised rescheduling
		106	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
29	events (related to SIGCHLD), and event watchers dealing with the event	107	change events (C<ev_child>), and event watchers dealing with the event
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	108	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
		109	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		110	limited support for fork events (C<ev_fork>).
		111
		112	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	113	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	114	for example).
33		115
34	=head1 CONVENTIONS	116	=head2 CONVENTIONS
35		117
36	Libev is very configurable. In this manual the default configuration	118	Libev is very configurable. In this manual the default (and most common)
37	will be described, which supports multiple event loops. For more info	119	configuration will be described, which supports multiple event loops. For
38	about various configuration options please have a look at the file	120	more info about various configuration options please have a look at
39	F<README.embed> in the libev distribution. If libev was configured without	121	B<EMBED> section in this manual. If libev was configured without support
40	support for multiple event loops, then all functions taking an initial	122	for multiple event loops, then all functions taking an initial argument of
41	argument of name C<loop> (which is always of type C<struct ev_loop *>)	123	name C<loop> (which is always of type C<struct ev_loop *>) will not have
42	will not have this argument.	124	this argument.
43		125
44	=head1 TIME REPRESENTATION	126	=head2 TIME REPRESENTATION
45		127
46	Libev represents time as a single floating point number, representing the	128	Libev represents time as a single floating point number, representing
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	129	the (fractional) number of seconds since the (POSIX) epoch (somewhere
48	the beginning of 1970, details are complicated, don't ask). This type is	130	near the beginning of 1970, details are complicated, don't ask). This
49	called C<ev_tstamp>, which is what you should use too. It usually aliases	131	type is called C<ev_tstamp>, which is what you should use too. It usually
50	to the C<double> type in C, and when you need to do any calculations on	132	aliases to the C<double> type in C. When you need to do any calculations
51	it, you should treat it as such.	133	on it, you should treat it as some floating point value. Unlike the name
		134	component C<stamp> might indicate, it is also used for time differences
		135	throughout libev.
		136
		137	=head1 ERROR HANDLING
		138
		139	Libev knows three classes of errors: operating system errors, usage errors
		140	and internal errors (bugs).
		141
		142	When libev catches an operating system error it cannot handle (for example
		143	a system call indicating a condition libev cannot fix), it calls the callback
		144	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		145	abort. The default is to print a diagnostic message and to call C<abort
		146	()>.
		147
		148	When libev detects a usage error such as a negative timer interval, then
		149	it will print a diagnostic message and abort (via the C<assert> mechanism,
		150	so C<NDEBUG> will disable this checking): these are programming errors in
		151	the libev caller and need to be fixed there.
		152
		153	Libev also has a few internal error-checking C<assert>ions, and also has
		154	extensive consistency checking code. These do not trigger under normal
		155	circumstances, as they indicate either a bug in libev or worse.
52		156
53		157
54	=head1 GLOBAL FUNCTIONS	158	=head1 GLOBAL FUNCTIONS
55		159
56	These functions can be called anytime, even before initialising the	160	These functions can be called anytime, even before initialising the
…		…
62		166
63	Returns the current time as libev would use it. Please note that the	167	Returns the current time as libev would use it. Please note that the
64	C<ev_now> function is usually faster and also often returns the timestamp	168	C<ev_now> function is usually faster and also often returns the timestamp
65	you actually want to know.	169	you actually want to know.
66		170
		171	=item ev_sleep (ev_tstamp interval)
		172
		173	Sleep for the given interval: The current thread will be blocked until
		174	either it is interrupted or the given time interval has passed. Basically
		175	this is a sub-second-resolution C<sleep ()>.
		176
67	=item int ev_version_major ()	177	=item int ev_version_major ()
68		178
69	=item int ev_version_minor ()	179	=item int ev_version_minor ()
70		180
71	You can find out the major and minor version numbers of the library	181	You can find out the major and minor ABI version numbers of the library
72	you linked against by calling the functions C<ev_version_major> and	182	you linked against by calling the functions C<ev_version_major> and
73	C<ev_version_minor>. If you want, you can compare against the global	183	C<ev_version_minor>. If you want, you can compare against the global
74	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	184	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
75	version of the library your program was compiled against.	185	version of the library your program was compiled against.
76		186
		187	These version numbers refer to the ABI version of the library, not the
		188	release version.
		189
77	Usually, it's a good idea to terminate if the major versions mismatch,	190	Usually, it's a good idea to terminate if the major versions mismatch,
78	as this indicates an incompatible change. Minor versions are usually	191	as this indicates an incompatible change. Minor versions are usually
79	compatible to older versions, so a larger minor version alone is usually	192	compatible to older versions, so a larger minor version alone is usually
80	not a problem.	193	not a problem.
81		194
82	Example: make sure we haven't accidentally been linked against the wrong	195	Example: Make sure we haven't accidentally been linked against the wrong
83	version:	196	version.
84		197
85	assert (("libev version mismatch",	198	assert (("libev version mismatch",
86	ev_version_major () == EV_VERSION_MAJOR	199	ev_version_major () == EV_VERSION_MAJOR
87	&& ev_version_minor () >= EV_VERSION_MINOR));	200	&& ev_version_minor () >= EV_VERSION_MINOR));
88		201
89	=item unsigned int ev_supported_backends ()	202	=item unsigned int ev_supported_backends ()
90		203
91	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	204	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
92	value) compiled into this binary of libev (independent of their	205	value) compiled into this binary of libev (independent of their
…		…
94	a description of the set values.	207	a description of the set values.
95		208
96	Example: make sure we have the epoll method, because yeah this is cool and	209	Example: make sure we have the epoll method, because yeah this is cool and
97	a must have and can we have a torrent of it please!!!11	210	a must have and can we have a torrent of it please!!!11
98		211
99	assert (("sorry, no epoll, no sex",	212	assert (("sorry, no epoll, no sex",
100	ev_supported_backends () & EVBACKEND_EPOLL));	213	ev_supported_backends () & EVBACKEND_EPOLL));
101		214
102	=item unsigned int ev_recommended_backends ()	215	=item unsigned int ev_recommended_backends ()
103		216
104	Return the set of all backends compiled into this binary of libev and also	217	Return the set of all backends compiled into this binary of libev and also
105	recommended for this platform. This set is often smaller than the one	218	recommended for this platform. This set is often smaller than the one
106	returned by C<ev_supported_backends>, as for example kqueue is broken on	219	returned by C<ev_supported_backends>, as for example kqueue is broken on
107	most BSDs and will not be autodetected unless you explicitly request it	220	most BSDs and will not be auto-detected unless you explicitly request it
108	(assuming you know what you are doing). This is the set of backends that	221	(assuming you know what you are doing). This is the set of backends that
109	libev will probe for if you specify no backends explicitly.	222	libev will probe for if you specify no backends explicitly.
110		223
111	=item unsigned int ev_embeddable_backends ()	224	=item unsigned int ev_embeddable_backends ()
112		225
…		…
116	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	229	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
117	recommended ones.	230	recommended ones.
118		231
119	See the description of C<ev_embed> watchers for more info.	232	See the description of C<ev_embed> watchers for more info.
120		233
121	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	234	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
122		235
123	Sets the allocation function to use (the prototype is similar to the	236	Sets the allocation function to use (the prototype is similar - the
124	realloc C function, the semantics are identical). It is used to allocate	237	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
125	and free memory (no surprises here). If it returns zero when memory	238	used to allocate and free memory (no surprises here). If it returns zero
126	needs to be allocated, the library might abort or take some potentially	239	when memory needs to be allocated (C<size != 0>), the library might abort
127	destructive action. The default is your system realloc function.	240	or take some potentially destructive action.
		241
		242	Since some systems (at least OpenBSD and Darwin) fail to implement
		243	correct C<realloc> semantics, libev will use a wrapper around the system
		244	C<realloc> and C<free> functions by default.
128		245
129	You could override this function in high-availability programs to, say,	246	You could override this function in high-availability programs to, say,
130	free some memory if it cannot allocate memory, to use a special allocator,	247	free some memory if it cannot allocate memory, to use a special allocator,
131	or even to sleep a while and retry until some memory is available.	248	or even to sleep a while and retry until some memory is available.
132		249
133	Example: replace the libev allocator with one that waits a bit and then	250	Example: Replace the libev allocator with one that waits a bit and then
134	retries: better than mine).	251	retries (example requires a standards-compliant C<realloc>).
135		252
136	static void *	253	static void *
137	persistent_realloc (void *ptr, long size)	254	persistent_realloc (void *ptr, size_t size)
138	{	255	{
139	for (;;)	256	for (;;)
140	{	257	{
141	void *newptr = realloc (ptr, size);	258	void *newptr = realloc (ptr, size);
142		259
…		…
148	}	265	}
149		266
150	...	267	...
151	ev_set_allocator (persistent_realloc);	268	ev_set_allocator (persistent_realloc);
152		269
153	=item ev_set_syserr_cb (void (*cb)(const char *msg));	270	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
154		271
155	Set the callback function to call on a retryable syscall error (such	272	Set the callback function to call on a retryable system call error (such
156	as failed select, poll, epoll_wait). The message is a printable string	273	as failed select, poll, epoll_wait). The message is a printable string
157	indicating the system call or subsystem causing the problem. If this	274	indicating the system call or subsystem causing the problem. If this
158	callback is set, then libev will expect it to remedy the sitution, no	275	callback is set, then libev will expect it to remedy the situation, no
159	matter what, when it returns. That is, libev will generally retry the	276	matter what, when it returns. That is, libev will generally retry the
160	requested operation, or, if the condition doesn't go away, do bad stuff	277	requested operation, or, if the condition doesn't go away, do bad stuff
161	(such as abort).	278	(such as abort).
162		279
163	Example: do the same thing as libev does internally:	280	Example: This is basically the same thing that libev does internally, too.
164		281
165	static void	282	static void
166	fatal_error (const char *msg)	283	fatal_error (const char *msg)
167	{	284	{
168	perror (msg);	285	perror (msg);
…		…
174		291
175	=back	292	=back
176		293
177	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	294	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
178		295
179	An event loop is described by a C<struct ev_loop *>. The library knows two	296	An event loop is described by a C<struct ev_loop *> (the C<struct>
180	types of such loops, the I<default> loop, which supports signals and child	297	is I<not> optional in this case, as there is also an C<ev_loop>
181	events, and dynamically created loops which do not.	298	I<function>).
182		299
183	If you use threads, a common model is to run the default event loop	300	The library knows two types of such loops, the I<default> loop, which
184	in your main thread (or in a separate thread) and for each thread you	301	supports signals and child events, and dynamically created loops which do
185	create, you also create another event loop. Libev itself does no locking	302	not.
186	whatsoever, so if you mix calls to the same event loop in different
187	threads, make sure you lock (this is usually a bad idea, though, even if
188	done correctly, because it's hideous and inefficient).
189		303
190	=over 4	304	=over 4
191		305
192	=item struct ev_loop *ev_default_loop (unsigned int flags)	306	=item struct ev_loop *ev_default_loop (unsigned int flags)
193		307
…		…
197	flags. If that is troubling you, check C<ev_backend ()> afterwards).	311	flags. If that is troubling you, check C<ev_backend ()> afterwards).
198		312
199	If you don't know what event loop to use, use the one returned from this	313	If you don't know what event loop to use, use the one returned from this
200	function.	314	function.
201		315
		316	Note that this function is I<not> thread-safe, so if you want to use it
		317	from multiple threads, you have to lock (note also that this is unlikely,
		318	as loops cannot be shared easily between threads anyway).
		319
		320	The default loop is the only loop that can handle C<ev_signal> and
		321	C<ev_child> watchers, and to do this, it always registers a handler
		322	for C<SIGCHLD>. If this is a problem for your application you can either
		323	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		324	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		325	C<ev_default_init>.
		326
202	The flags argument can be used to specify special behaviour or specific	327	The flags argument can be used to specify special behaviour or specific
203	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	328	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
204		329
205	The following flags are supported:	330	The following flags are supported:
206		331
…		…
211	The default flags value. Use this if you have no clue (it's the right	336	The default flags value. Use this if you have no clue (it's the right
212	thing, believe me).	337	thing, believe me).
213		338
214	=item C<EVFLAG_NOENV>	339	=item C<EVFLAG_NOENV>
215		340
216	If this flag bit is ored into the flag value (or the program runs setuid	341	If this flag bit is or'ed into the flag value (or the program runs setuid
217	or setgid) then libev will I<not> look at the environment variable	342	or setgid) then libev will I<not> look at the environment variable
218	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	343	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
219	override the flags completely if it is found in the environment. This is	344	override the flags completely if it is found in the environment. This is
220	useful to try out specific backends to test their performance, or to work	345	useful to try out specific backends to test their performance, or to work
221	around bugs.	346	around bugs.
222		347
		348	=item C<EVFLAG_FORKCHECK>
		349
		350	Instead of calling C<ev_loop_fork> manually after a fork, you can also
		351	make libev check for a fork in each iteration by enabling this flag.
		352
		353	This works by calling C<getpid ()> on every iteration of the loop,
		354	and thus this might slow down your event loop if you do a lot of loop
		355	iterations and little real work, but is usually not noticeable (on my
		356	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
		357	without a system call and thus I<very> fast, but my GNU/Linux system also has
		358	C<pthread_atfork> which is even faster).
		359
		360	The big advantage of this flag is that you can forget about fork (and
		361	forget about forgetting to tell libev about forking) when you use this
		362	flag.
		363
		364	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
		365	environment variable.
		366
		367	=item C<EVFLAG_NOINOTIFY>
		368
		369	When this flag is specified, then libev will not attempt to use the
		370	I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and
		371	testing, this flag can be useful to conserve inotify file descriptors, as
		372	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		373
		374	=item C<EVFLAG_SIGNALFD>
		375
		376	When this flag is specified, then libev will attempt to use the
		377	I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This API
		378	delivers signals synchronously, which makes it both faster and might make
		379	it possible to get the queued signal data. It can also simplify signal
		380	handling with threads, as long as you properly block signals in your
		381	threads that are not interested in handling them.
		382
		383	Signalfd will not be used by default as this changes your signal mask, and
		384	there are a lot of shoddy libraries and programs (glib's threadpool for
		385	example) that can't properly initialise their signal masks.
		386
223	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	387	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
224		388
225	This is your standard select(2) backend. Not I<completely> standard, as	389	This is your standard select(2) backend. Not I<completely> standard, as
226	libev tries to roll its own fd_set with no limits on the number of fds,	390	libev tries to roll its own fd_set with no limits on the number of fds,
227	but if that fails, expect a fairly low limit on the number of fds when	391	but if that fails, expect a fairly low limit on the number of fds when
228	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	392	using this backend. It doesn't scale too well (O(highest_fd)), but its
229	the fastest backend for a low number of fds.	393	usually the fastest backend for a low number of (low-numbered :) fds.
		394
		395	To get good performance out of this backend you need a high amount of
		396	parallelism (most of the file descriptors should be busy). If you are
		397	writing a server, you should C<accept ()> in a loop to accept as many
		398	connections as possible during one iteration. You might also want to have
		399	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		400	readiness notifications you get per iteration.
		401
		402	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		403	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		404	C<exceptfds> set on that platform).
230		405
231	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	406	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
232		407
233	And this is your standard poll(2) backend. It's more complicated than	408	And this is your standard poll(2) backend. It's more complicated
234	select, but handles sparse fds better and has no artificial limit on the	409	than select, but handles sparse fds better and has no artificial
235	number of fds you can use (except it will slow down considerably with a	410	limit on the number of fds you can use (except it will slow down
236	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	411	considerably with a lot of inactive fds). It scales similarly to select,
		412	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		413	performance tips.
		414
		415	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		416	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
237		417
238	=item C<EVBACKEND_EPOLL> (value 4, Linux)	418	=item C<EVBACKEND_EPOLL> (value 4, Linux)
239		419
		420	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		421	kernels).
		422
240	For few fds, this backend is a bit little slower than poll and select,	423	For few fds, this backend is a bit little slower than poll and select,
241	but it scales phenomenally better. While poll and select usually scale like	424	but it scales phenomenally better. While poll and select usually scale
242	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	425	like O(total_fds) where n is the total number of fds (or the highest fd),
243	either O(1) or O(active_fds).	426	epoll scales either O(1) or O(active_fds).
244		427
		428	The epoll mechanism deserves honorable mention as the most misdesigned
		429	of the more advanced event mechanisms: mere annoyances include silently
		430	dropping file descriptors, requiring a system call per change per file
		431	descriptor (and unnecessary guessing of parameters), problems with dup and
		432	so on. The biggest issue is fork races, however - if a program forks then
		433	I<both> parent and child process have to recreate the epoll set, which can
		434	take considerable time (one syscall per file descriptor) and is of course
		435	hard to detect.
		436
		437	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		438	of course I<doesn't>, and epoll just loves to report events for totally
		439	I<different> file descriptors (even already closed ones, so one cannot
		440	even remove them from the set) than registered in the set (especially
		441	on SMP systems). Libev tries to counter these spurious notifications by
		442	employing an additional generation counter and comparing that against the
		443	events to filter out spurious ones, recreating the set when required.
		444
245	While stopping and starting an I/O watcher in the same iteration will	445	While stopping, setting and starting an I/O watcher in the same iteration
246	result in some caching, there is still a syscall per such incident	446	will result in some caching, there is still a system call per such
247	(because the fd could point to a different file description now), so its	447	incident (because the same I<file descriptor> could point to a different
248	best to avoid that. Also, dup()ed file descriptors might not work very	448	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
249	well if you register events for both fds.	449	file descriptors might not work very well if you register events for both
		450	file descriptors.
250		451
251	Please note that epoll sometimes generates spurious notifications, so you	452	Best performance from this backend is achieved by not unregistering all
252	need to use non-blocking I/O or other means to avoid blocking when no data	453	watchers for a file descriptor until it has been closed, if possible,
253	(or space) is available.	454	i.e. keep at least one watcher active per fd at all times. Stopping and
		455	starting a watcher (without re-setting it) also usually doesn't cause
		456	extra overhead. A fork can both result in spurious notifications as well
		457	as in libev having to destroy and recreate the epoll object, which can
		458	take considerable time and thus should be avoided.
		459
		460	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		461	faster than epoll for maybe up to a hundred file descriptors, depending on
		462	the usage. So sad.
		463
		464	While nominally embeddable in other event loops, this feature is broken in
		465	all kernel versions tested so far.
		466
		467	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		468	C<EVBACKEND_POLL>.
254		469
255	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	470	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
256		471
257	Kqueue deserves special mention, as at the time of this writing, it	472	Kqueue deserves special mention, as at the time of this writing, it
258	was broken on all BSDs except NetBSD (usually it doesn't work with	473	was broken on all BSDs except NetBSD (usually it doesn't work reliably
259	anything but sockets and pipes, except on Darwin, where of course its	474	with anything but sockets and pipes, except on Darwin, where of course
260	completely useless). For this reason its not being "autodetected"	475	it's completely useless). Unlike epoll, however, whose brokenness
		476	is by design, these kqueue bugs can (and eventually will) be fixed
		477	without API changes to existing programs. For this reason it's not being
261	unless you explicitly specify it explicitly in the flags (i.e. using	478	"auto-detected" unless you explicitly specify it in the flags (i.e. using
262	C<EVBACKEND_KQUEUE>).	479	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		480	system like NetBSD.
		481
		482	You still can embed kqueue into a normal poll or select backend and use it
		483	only for sockets (after having made sure that sockets work with kqueue on
		484	the target platform). See C<ev_embed> watchers for more info.
263		485
264	It scales in the same way as the epoll backend, but the interface to the	486	It scales in the same way as the epoll backend, but the interface to the
265	kernel is more efficient (which says nothing about its actual speed, of	487	kernel is more efficient (which says nothing about its actual speed, of
266	course). While starting and stopping an I/O watcher does not cause an	488	course). While stopping, setting and starting an I/O watcher does never
267	extra syscall as with epoll, it still adds up to four event changes per	489	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
268	incident, so its best to avoid that.	490	two event changes per incident. Support for C<fork ()> is very bad (but
		491	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		492	cases
		493
		494	This backend usually performs well under most conditions.
		495
		496	While nominally embeddable in other event loops, this doesn't work
		497	everywhere, so you might need to test for this. And since it is broken
		498	almost everywhere, you should only use it when you have a lot of sockets
		499	(for which it usually works), by embedding it into another event loop
		500	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
		501	also broken on OS X)) and, did I mention it, using it only for sockets.
		502
		503	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		504	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		505	C<NOTE_EOF>.
269		506
270	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	507	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
271		508
272	This is not implemented yet (and might never be).	509	This is not implemented yet (and might never be, unless you send me an
		510	implementation). According to reports, C</dev/poll> only supports sockets
		511	and is not embeddable, which would limit the usefulness of this backend
		512	immensely.
273		513
274	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	514	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
275		515
276	This uses the Solaris 10 port mechanism. As with everything on Solaris,	516	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
277	it's really slow, but it still scales very well (O(active_fds)).	517	it's really slow, but it still scales very well (O(active_fds)).
278		518
279	Please note that solaris ports can result in a lot of spurious	519	Please note that Solaris event ports can deliver a lot of spurious
280	notifications, so you need to use non-blocking I/O or other means to avoid	520	notifications, so you need to use non-blocking I/O or other means to avoid
281	blocking when no data (or space) is available.	521	blocking when no data (or space) is available.
		522
		523	While this backend scales well, it requires one system call per active
		524	file descriptor per loop iteration. For small and medium numbers of file
		525	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		526	might perform better.
		527
		528	On the positive side, with the exception of the spurious readiness
		529	notifications, this backend actually performed fully to specification
		530	in all tests and is fully embeddable, which is a rare feat among the
		531	OS-specific backends (I vastly prefer correctness over speed hacks).
		532
		533	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		534	C<EVBACKEND_POLL>.
282		535
283	=item C<EVBACKEND_ALL>	536	=item C<EVBACKEND_ALL>
284		537
285	Try all backends (even potentially broken ones that wouldn't be tried	538	Try all backends (even potentially broken ones that wouldn't be tried
286	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	539	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
287	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	540	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
288		541
		542	It is definitely not recommended to use this flag.
		543
289	=back	544	=back
290		545
291	If one or more of these are ored into the flags value, then only these	546	If one or more of the backend flags are or'ed into the flags value,
292	backends will be tried (in the reverse order as given here). If none are	547	then only these backends will be tried (in the reverse order as listed
293	specified, most compiled-in backend will be tried, usually in reverse	548	here). If none are specified, all backends in C<ev_recommended_backends
294	order of their flag values :)	549	()> will be tried.
295		550
296	The most typical usage is like this:	551	Example: This is the most typical usage.
297		552
298	if (!ev_default_loop (0))	553	if (!ev_default_loop (0))
299	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	554	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
300		555
301	Restrict libev to the select and poll backends, and do not allow	556	Example: Restrict libev to the select and poll backends, and do not allow
302	environment settings to be taken into account:	557	environment settings to be taken into account:
303		558
304	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	559	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
305		560
306	Use whatever libev has to offer, but make sure that kqueue is used if	561	Example: Use whatever libev has to offer, but make sure that kqueue is
307	available (warning, breaks stuff, best use only with your own private	562	used if available (warning, breaks stuff, best use only with your own
308	event loop and only if you know the OS supports your types of fds):	563	private event loop and only if you know the OS supports your types of
		564	fds):
309		565
310	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	566	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
311		567
312	=item struct ev_loop *ev_loop_new (unsigned int flags)	568	=item struct ev_loop *ev_loop_new (unsigned int flags)
313		569
314	Similar to C<ev_default_loop>, but always creates a new event loop that is	570	Similar to C<ev_default_loop>, but always creates a new event loop that is
315	always distinct from the default loop. Unlike the default loop, it cannot	571	always distinct from the default loop.
316	handle signal and child watchers, and attempts to do so will be greeted by
317	undefined behaviour (or a failed assertion if assertions are enabled).
318		572
		573	Note that this function I<is> thread-safe, and one common way to use
		574	libev with threads is indeed to create one loop per thread, and using the
		575	default loop in the "main" or "initial" thread.
		576
319	Example: try to create a event loop that uses epoll and nothing else.	577	Example: Try to create a event loop that uses epoll and nothing else.
320		578
321	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	579	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
322	if (!epoller)	580	if (!epoller)
323	fatal ("no epoll found here, maybe it hides under your chair");	581	fatal ("no epoll found here, maybe it hides under your chair");
324		582
325	=item ev_default_destroy ()	583	=item ev_default_destroy ()
326		584
327	Destroys the default loop again (frees all memory and kernel state	585	Destroys the default loop (frees all memory and kernel state etc.). None
328	etc.). None of the active event watchers will be stopped in the normal	586	of the active event watchers will be stopped in the normal sense, so
329	sense, so e.g. C<ev_is_active> might still return true. It is your	587	e.g. C<ev_is_active> might still return true. It is your responsibility to
330	responsibility to either stop all watchers cleanly yoursef I<before>	588	either stop all watchers cleanly yourself I<before> calling this function,
331	calling this function, or cope with the fact afterwards (which is usually	589	or cope with the fact afterwards (which is usually the easiest thing, you
332	the easiest thing, youc na just ignore the watchers and/or C<free ()> them	590	can just ignore the watchers and/or C<free ()> them for example).
333	for example).	591
		592	Note that certain global state, such as signal state (and installed signal
		593	handlers), will not be freed by this function, and related watchers (such
		594	as signal and child watchers) would need to be stopped manually.
		595
		596	In general it is not advisable to call this function except in the
		597	rare occasion where you really need to free e.g. the signal handling
		598	pipe fds. If you need dynamically allocated loops it is better to use
		599	C<ev_loop_new> and C<ev_loop_destroy>.
334		600
335	=item ev_loop_destroy (loop)	601	=item ev_loop_destroy (loop)
336		602
337	Like C<ev_default_destroy>, but destroys an event loop created by an	603	Like C<ev_default_destroy>, but destroys an event loop created by an
338	earlier call to C<ev_loop_new>.	604	earlier call to C<ev_loop_new>.
339		605
340	=item ev_default_fork ()	606	=item ev_default_fork ()
341		607
		608	This function sets a flag that causes subsequent C<ev_loop> iterations
342	This function reinitialises the kernel state for backends that have	609	to reinitialise the kernel state for backends that have one. Despite the
343	one. Despite the name, you can call it anytime, but it makes most sense	610	name, you can call it anytime, but it makes most sense after forking, in
344	after forking, in either the parent or child process (or both, but that	611	the child process (or both child and parent, but that again makes little
345	again makes little sense).	612	sense). You I<must> call it in the child before using any of the libev
		613	functions, and it will only take effect at the next C<ev_loop> iteration.
346		614
347	You I<must> call this function in the child process after forking if and	615	Again, you I<have> to call it on I<any> loop that you want to re-use after
348	only if you want to use the event library in both processes. If you just	616	a fork, I<even if you do not plan to use the loop in the parent>. This is
349	fork+exec, you don't have to call it.	617	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		618	during fork.
		619
		620	On the other hand, you only need to call this function in the child
		621	process if and only if you want to use the event loop in the child. If you
		622	just fork+exec or create a new loop in the child, you don't have to call
		623	it at all.
350		624
351	The function itself is quite fast and it's usually not a problem to call	625	The function itself is quite fast and it's usually not a problem to call
352	it just in case after a fork. To make this easy, the function will fit in	626	it just in case after a fork. To make this easy, the function will fit in
353	quite nicely into a call to C<pthread_atfork>:	627	quite nicely into a call to C<pthread_atfork>:
354		628
355	pthread_atfork (0, 0, ev_default_fork);	629	pthread_atfork (0, 0, ev_default_fork);
356		630
357	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
358	without calling this function, so if you force one of those backends you
359	do not need to care.
360
361	=item ev_loop_fork (loop)	631	=item ev_loop_fork (loop)
362		632
363	Like C<ev_default_fork>, but acts on an event loop created by	633	Like C<ev_default_fork>, but acts on an event loop created by
364	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	634	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
365	after fork, and how you do this is entirely your own problem.	635	after fork that you want to re-use in the child, and how you keep track of
		636	them is entirely your own problem.
		637
		638	=item int ev_is_default_loop (loop)
		639
		640	Returns true when the given loop is, in fact, the default loop, and false
		641	otherwise.
		642
		643	=item unsigned int ev_iteration (loop)
		644
		645	Returns the current iteration count for the loop, which is identical to
		646	the number of times libev did poll for new events. It starts at C<0> and
		647	happily wraps around with enough iterations.
		648
		649	This value can sometimes be useful as a generation counter of sorts (it
		650	"ticks" the number of loop iterations), as it roughly corresponds with
		651	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		652	prepare and check phases.
		653
		654	=item unsigned int ev_depth (loop)
		655
		656	Returns the number of times C<ev_loop> was entered minus the number of
		657	times C<ev_loop> was exited, in other words, the recursion depth.
		658
		659	Outside C<ev_loop>, this number is zero. In a callback, this number is
		660	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		661	in which case it is higher.
		662
		663	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		664	etc.), doesn't count as "exit" - consider this as a hint to avoid such
		665	ungentleman behaviour unless it's really convenient.
366		666
367	=item unsigned int ev_backend (loop)	667	=item unsigned int ev_backend (loop)
368		668
369	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	669	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
370	use.	670	use.
…		…
373		673
374	Returns the current "event loop time", which is the time the event loop	674	Returns the current "event loop time", which is the time the event loop
375	received events and started processing them. This timestamp does not	675	received events and started processing them. This timestamp does not
376	change as long as callbacks are being processed, and this is also the base	676	change as long as callbacks are being processed, and this is also the base
377	time used for relative timers. You can treat it as the timestamp of the	677	time used for relative timers. You can treat it as the timestamp of the
378	event occuring (or more correctly, libev finding out about it).	678	event occurring (or more correctly, libev finding out about it).
		679
		680	=item ev_now_update (loop)
		681
		682	Establishes the current time by querying the kernel, updating the time
		683	returned by C<ev_now ()> in the progress. This is a costly operation and
		684	is usually done automatically within C<ev_loop ()>.
		685
		686	This function is rarely useful, but when some event callback runs for a
		687	very long time without entering the event loop, updating libev's idea of
		688	the current time is a good idea.
		689
		690	See also L<The special problem of time updates> in the C<ev_timer> section.
		691
		692	=item ev_suspend (loop)
		693
		694	=item ev_resume (loop)
		695
		696	These two functions suspend and resume a loop, for use when the loop is
		697	not used for a while and timeouts should not be processed.
		698
		699	A typical use case would be an interactive program such as a game: When
		700	the user presses C<^Z> to suspend the game and resumes it an hour later it
		701	would be best to handle timeouts as if no time had actually passed while
		702	the program was suspended. This can be achieved by calling C<ev_suspend>
		703	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		704	C<ev_resume> directly afterwards to resume timer processing.
		705
		706	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		707	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		708	will be rescheduled (that is, they will lose any events that would have
		709	occured while suspended).
		710
		711	After calling C<ev_suspend> you B<must not> call I<any> function on the
		712	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		713	without a previous call to C<ev_suspend>.
		714
		715	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		716	event loop time (see C<ev_now_update>).
379		717
380	=item ev_loop (loop, int flags)	718	=item ev_loop (loop, int flags)
381		719
382	Finally, this is it, the event handler. This function usually is called	720	Finally, this is it, the event handler. This function usually is called
383	after you initialised all your watchers and you want to start handling	721	after you have initialised all your watchers and you want to start
384	events.	722	handling events.
385		723
386	If the flags argument is specified as C<0>, it will not return until	724	If the flags argument is specified as C<0>, it will not return until
387	either no event watchers are active anymore or C<ev_unloop> was called.	725	either no event watchers are active anymore or C<ev_unloop> was called.
388		726
389	Please note that an explicit C<ev_unloop> is usually better than	727	Please note that an explicit C<ev_unloop> is usually better than
390	relying on all watchers to be stopped when deciding when a program has	728	relying on all watchers to be stopped when deciding when a program has
391	finished (especially in interactive programs), but having a program that	729	finished (especially in interactive programs), but having a program
392	automatically loops as long as it has to and no longer by virtue of	730	that automatically loops as long as it has to and no longer by virtue
393	relying on its watchers stopping correctly is a thing of beauty.	731	of relying on its watchers stopping correctly, that is truly a thing of
		732	beauty.
394		733
395	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	734	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
396	those events and any outstanding ones, but will not block your process in	735	those events and any already outstanding ones, but will not block your
397	case there are no events and will return after one iteration of the loop.	736	process in case there are no events and will return after one iteration of
		737	the loop.
398		738
399	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	739	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
400	neccessary) and will handle those and any outstanding ones. It will block	740	necessary) and will handle those and any already outstanding ones. It
401	your process until at least one new event arrives, and will return after	741	will block your process until at least one new event arrives (which could
402	one iteration of the loop. This is useful if you are waiting for some	742	be an event internal to libev itself, so there is no guarantee that a
403	external event in conjunction with something not expressible using other	743	user-registered callback will be called), and will return after one
		744	iteration of the loop.
		745
		746	This is useful if you are waiting for some external event in conjunction
		747	with something not expressible using other libev watchers (i.e. "roll your
404	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	748	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
405	usually a better approach for this kind of thing.	749	usually a better approach for this kind of thing.
406		750
407	Here are the gory details of what C<ev_loop> does:	751	Here are the gory details of what C<ev_loop> does:
408		752
409	* If there are no active watchers (reference count is zero), return.	753	- Before the first iteration, call any pending watchers.
410	- Queue prepare watchers and then call all outstanding watchers.	754	* If EVFLAG_FORKCHECK was used, check for a fork.
		755	- If a fork was detected (by any means), queue and call all fork watchers.
		756	- Queue and call all prepare watchers.
411	- If we have been forked, recreate the kernel state.	757	- If we have been forked, detach and recreate the kernel state
		758	as to not disturb the other process.
412	- Update the kernel state with all outstanding changes.	759	- Update the kernel state with all outstanding changes.
413	- Update the "event loop time".	760	- Update the "event loop time" (ev_now ()).
414	- Calculate for how long to block.	761	- Calculate for how long to sleep or block, if at all
		762	(active idle watchers, EVLOOP_NONBLOCK or not having
		763	any active watchers at all will result in not sleeping).
		764	- Sleep if the I/O and timer collect interval say so.
415	- Block the process, waiting for any events.	765	- Block the process, waiting for any events.
416	- Queue all outstanding I/O (fd) events.	766	- Queue all outstanding I/O (fd) events.
417	- Update the "event loop time" and do time jump handling.	767	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
418	- Queue all outstanding timers.	768	- Queue all expired timers.
419	- Queue all outstanding periodics.	769	- Queue all expired periodics.
420	- If no events are pending now, queue all idle watchers.	770	- Unless any events are pending now, queue all idle watchers.
421	- Queue all check watchers.	771	- Queue all check watchers.
422	- Call all queued watchers in reverse order (i.e. check watchers first).	772	- Call all queued watchers in reverse order (i.e. check watchers first).
423	Signals and child watchers are implemented as I/O watchers, and will	773	Signals and child watchers are implemented as I/O watchers, and will
424	be handled here by queueing them when their watcher gets executed.	774	be handled here by queueing them when their watcher gets executed.
425	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	775	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
426	were used, return, otherwise continue with step *.	776	were used, or there are no active watchers, return, otherwise
		777	continue with step *.
427		778
428	Example: queue some jobs and then loop until no events are outsanding	779	Example: Queue some jobs and then loop until no events are outstanding
429	anymore.	780	anymore.
430		781
431	... queue jobs here, make sure they register event watchers as long	782	... queue jobs here, make sure they register event watchers as long
432	... as they still have work to do (even an idle watcher will do..)	783	... as they still have work to do (even an idle watcher will do..)
433	ev_loop (my_loop, 0);	784	ev_loop (my_loop, 0);
434	... jobs done. yeah!	785	... jobs done or somebody called unloop. yeah!
435		786
436	=item ev_unloop (loop, how)	787	=item ev_unloop (loop, how)
437		788
438	Can be used to make a call to C<ev_loop> return early (but only after it	789	Can be used to make a call to C<ev_loop> return early (but only after it
439	has processed all outstanding events). The C<how> argument must be either	790	has processed all outstanding events). The C<how> argument must be either
440	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	791	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
441	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	792	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
442		793
		794	This "unloop state" will be cleared when entering C<ev_loop> again.
		795
		796	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		797
443	=item ev_ref (loop)	798	=item ev_ref (loop)
444		799
445	=item ev_unref (loop)	800	=item ev_unref (loop)
446		801
447	Ref/unref can be used to add or remove a reference count on the event	802	Ref/unref can be used to add or remove a reference count on the event
448	loop: Every watcher keeps one reference, and as long as the reference	803	loop: Every watcher keeps one reference, and as long as the reference
449	count is nonzero, C<ev_loop> will not return on its own. If you have	804	count is nonzero, C<ev_loop> will not return on its own.
450	a watcher you never unregister that should not keep C<ev_loop> from	805
451	returning, ev_unref() after starting, and ev_ref() before stopping it. For	806	This is useful when you have a watcher that you never intend to
		807	unregister, but that nevertheless should not keep C<ev_loop> from
		808	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		809	before stopping it.
		810
452	example, libev itself uses this for its internal signal pipe: It is not	811	As an example, libev itself uses this for its internal signal pipe: It
453	visible to the libev user and should not keep C<ev_loop> from exiting if	812	is not visible to the libev user and should not keep C<ev_loop> from
454	no event watchers registered by it are active. It is also an excellent	813	exiting if no event watchers registered by it are active. It is also an
455	way to do this for generic recurring timers or from within third-party	814	excellent way to do this for generic recurring timers or from within
456	libraries. Just remember to I<unref after start> and I<ref before stop>.	815	third-party libraries. Just remember to I<unref after start> and I<ref
		816	before stop> (but only if the watcher wasn't active before, or was active
		817	before, respectively. Note also that libev might stop watchers itself
		818	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		819	in the callback).
457		820
458	Example: create a signal watcher, but keep it from keeping C<ev_loop>	821	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
459	running when nothing else is active.	822	running when nothing else is active.
460		823
461	struct dv_signal exitsig;	824	ev_signal exitsig;
462	ev_signal_init (&exitsig, sig_cb, SIGINT);	825	ev_signal_init (&exitsig, sig_cb, SIGINT);
463	ev_signal_start (myloop, &exitsig);	826	ev_signal_start (loop, &exitsig);
464	evf_unref (myloop);	827	evf_unref (loop);
465		828
466	Example: for some weird reason, unregister the above signal handler again.	829	Example: For some weird reason, unregister the above signal handler again.
467		830
468	ev_ref (myloop);	831	ev_ref (loop);
469	ev_signal_stop (myloop, &exitsig);	832	ev_signal_stop (loop, &exitsig);
		833
		834	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		835
		836	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		837
		838	These advanced functions influence the time that libev will spend waiting
		839	for events. Both time intervals are by default C<0>, meaning that libev
		840	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		841	latency.
		842
		843	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		844	allows libev to delay invocation of I/O and timer/periodic callbacks
		845	to increase efficiency of loop iterations (or to increase power-saving
		846	opportunities).
		847
		848	The idea is that sometimes your program runs just fast enough to handle
		849	one (or very few) event(s) per loop iteration. While this makes the
		850	program responsive, it also wastes a lot of CPU time to poll for new
		851	events, especially with backends like C<select ()> which have a high
		852	overhead for the actual polling but can deliver many events at once.
		853
		854	By setting a higher I<io collect interval> you allow libev to spend more
		855	time collecting I/O events, so you can handle more events per iteration,
		856	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		857	C<ev_timer>) will be not affected. Setting this to a non-null value will
		858	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		859	sleep time ensures that libev will not poll for I/O events more often then
		860	once per this interval, on average.
		861
		862	Likewise, by setting a higher I<timeout collect interval> you allow libev
		863	to spend more time collecting timeouts, at the expense of increased
		864	latency/jitter/inexactness (the watcher callback will be called
		865	later). C<ev_io> watchers will not be affected. Setting this to a non-null
		866	value will not introduce any overhead in libev.
		867
		868	Many (busy) programs can usually benefit by setting the I/O collect
		869	interval to a value near C<0.1> or so, which is often enough for
		870	interactive servers (of course not for games), likewise for timeouts. It
		871	usually doesn't make much sense to set it to a lower value than C<0.01>,
		872	as this approaches the timing granularity of most systems. Note that if
		873	you do transactions with the outside world and you can't increase the
		874	parallelity, then this setting will limit your transaction rate (if you
		875	need to poll once per transaction and the I/O collect interval is 0.01,
		876	then you can't do more than 100 transations per second).
		877
		878	Setting the I<timeout collect interval> can improve the opportunity for
		879	saving power, as the program will "bundle" timer callback invocations that
		880	are "near" in time together, by delaying some, thus reducing the number of
		881	times the process sleeps and wakes up again. Another useful technique to
		882	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		883	they fire on, say, one-second boundaries only.
		884
		885	Example: we only need 0.1s timeout granularity, and we wish not to poll
		886	more often than 100 times per second:
		887
		888	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		889	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		890
		891	=item ev_invoke_pending (loop)
		892
		893	This call will simply invoke all pending watchers while resetting their
		894	pending state. Normally, C<ev_loop> does this automatically when required,
		895	but when overriding the invoke callback this call comes handy.
		896
		897	=item int ev_pending_count (loop)
		898
		899	Returns the number of pending watchers - zero indicates that no watchers
		900	are pending.
		901
		902	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		903
		904	This overrides the invoke pending functionality of the loop: Instead of
		905	invoking all pending watchers when there are any, C<ev_loop> will call
		906	this callback instead. This is useful, for example, when you want to
		907	invoke the actual watchers inside another context (another thread etc.).
		908
		909	If you want to reset the callback, use C<ev_invoke_pending> as new
		910	callback.
		911
		912	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		913
		914	Sometimes you want to share the same loop between multiple threads. This
		915	can be done relatively simply by putting mutex_lock/unlock calls around
		916	each call to a libev function.
		917
		918	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		919	wait for it to return. One way around this is to wake up the loop via
		920	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		921	and I<acquire> callbacks on the loop.
		922
		923	When set, then C<release> will be called just before the thread is
		924	suspended waiting for new events, and C<acquire> is called just
		925	afterwards.
		926
		927	Ideally, C<release> will just call your mutex_unlock function, and
		928	C<acquire> will just call the mutex_lock function again.
		929
		930	While event loop modifications are allowed between invocations of
		931	C<release> and C<acquire> (that's their only purpose after all), no
		932	modifications done will affect the event loop, i.e. adding watchers will
		933	have no effect on the set of file descriptors being watched, or the time
		934	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it
		935	to take note of any changes you made.
		936
		937	In theory, threads executing C<ev_loop> will be async-cancel safe between
		938	invocations of C<release> and C<acquire>.
		939
		940	See also the locking example in the C<THREADS> section later in this
		941	document.
		942
		943	=item ev_set_userdata (loop, void *data)
		944
		945	=item ev_userdata (loop)
		946
		947	Set and retrieve a single C<void *> associated with a loop. When
		948	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		949	C<0.>
		950
		951	These two functions can be used to associate arbitrary data with a loop,
		952	and are intended solely for the C<invoke_pending_cb>, C<release> and
		953	C<acquire> callbacks described above, but of course can be (ab-)used for
		954	any other purpose as well.
		955
		956	=item ev_loop_verify (loop)
		957
		958	This function only does something when C<EV_VERIFY> support has been
		959	compiled in, which is the default for non-minimal builds. It tries to go
		960	through all internal structures and checks them for validity. If anything
		961	is found to be inconsistent, it will print an error message to standard
		962	error and call C<abort ()>.
		963
		964	This can be used to catch bugs inside libev itself: under normal
		965	circumstances, this function will never abort as of course libev keeps its
		966	data structures consistent.
470		967
471	=back	968	=back
472		969
		970
473	=head1 ANATOMY OF A WATCHER	971	=head1 ANATOMY OF A WATCHER
		972
		973	In the following description, uppercase C<TYPE> in names stands for the
		974	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		975	watchers and C<ev_io_start> for I/O watchers.
474		976
475	A watcher is a structure that you create and register to record your	977	A watcher is a structure that you create and register to record your
476	interest in some event. For instance, if you want to wait for STDIN to	978	interest in some event. For instance, if you want to wait for STDIN to
477	become readable, you would create an C<ev_io> watcher for that:	979	become readable, you would create an C<ev_io> watcher for that:
478		980
479	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	981	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
480	{	982	{
481	ev_io_stop (w);	983	ev_io_stop (w);
482	ev_unloop (loop, EVUNLOOP_ALL);	984	ev_unloop (loop, EVUNLOOP_ALL);
483	}	985	}
484		986
485	struct ev_loop *loop = ev_default_loop (0);	987	struct ev_loop *loop = ev_default_loop (0);
		988
486	struct ev_io stdin_watcher;	989	ev_io stdin_watcher;
		990
487	ev_init (&stdin_watcher, my_cb);	991	ev_init (&stdin_watcher, my_cb);
488	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	992	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
489	ev_io_start (loop, &stdin_watcher);	993	ev_io_start (loop, &stdin_watcher);
		994
490	ev_loop (loop, 0);	995	ev_loop (loop, 0);
491		996
492	As you can see, you are responsible for allocating the memory for your	997	As you can see, you are responsible for allocating the memory for your
493	watcher structures (and it is usually a bad idea to do this on the stack,	998	watcher structures (and it is I<usually> a bad idea to do this on the
494	although this can sometimes be quite valid).	999	stack).
		1000
		1001	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1002	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
495		1003
496	Each watcher structure must be initialised by a call to C<ev_init	1004	Each watcher structure must be initialised by a call to C<ev_init
497	(watcher *, callback)>, which expects a callback to be provided. This	1005	(watcher *, callback)>, which expects a callback to be provided. This
498	callback gets invoked each time the event occurs (or, in the case of io	1006	callback gets invoked each time the event occurs (or, in the case of I/O
499	watchers, each time the event loop detects that the file descriptor given	1007	watchers, each time the event loop detects that the file descriptor given
500	is readable and/or writable).	1008	is readable and/or writable).
501		1009
502	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1010	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
503	with arguments specific to this watcher type. There is also a macro	1011	macro to configure it, with arguments specific to the watcher type. There
504	to combine initialisation and setting in one call: C<< ev_<type>_init	1012	is also a macro to combine initialisation and setting in one call: C<<
505	(watcher *, callback, ...) >>.	1013	ev_TYPE_init (watcher *, callback, ...) >>.
506		1014
507	To make the watcher actually watch out for events, you have to start it	1015	To make the watcher actually watch out for events, you have to start it
508	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1016	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
509	*) >>), and you can stop watching for events at any time by calling the	1017	*) >>), and you can stop watching for events at any time by calling the
510	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1018	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
511		1019
512	As long as your watcher is active (has been started but not stopped) you	1020	As long as your watcher is active (has been started but not stopped) you
513	must not touch the values stored in it. Most specifically you must never	1021	must not touch the values stored in it. Most specifically you must never
514	reinitialise it or call its C<set> macro.	1022	reinitialise it or call its C<ev_TYPE_set> macro.
515		1023
516	Each and every callback receives the event loop pointer as first, the	1024	Each and every callback receives the event loop pointer as first, the
517	registered watcher structure as second, and a bitset of received events as	1025	registered watcher structure as second, and a bitset of received events as
518	third argument.	1026	third argument.
519		1027
…		…
528	=item C<EV_WRITE>	1036	=item C<EV_WRITE>
529		1037
530	The file descriptor in the C<ev_io> watcher has become readable and/or	1038	The file descriptor in the C<ev_io> watcher has become readable and/or
531	writable.	1039	writable.
532		1040
533	=item C<EV_TIMEOUT>	1041	=item C<EV_TIMER>
534		1042
535	The C<ev_timer> watcher has timed out.	1043	The C<ev_timer> watcher has timed out.
536		1044
537	=item C<EV_PERIODIC>	1045	=item C<EV_PERIODIC>
538		1046
…		…
543	The signal specified in the C<ev_signal> watcher has been received by a thread.	1051	The signal specified in the C<ev_signal> watcher has been received by a thread.
544		1052
545	=item C<EV_CHILD>	1053	=item C<EV_CHILD>
546		1054
547	The pid specified in the C<ev_child> watcher has received a status change.	1055	The pid specified in the C<ev_child> watcher has received a status change.
		1056
		1057	=item C<EV_STAT>
		1058
		1059	The path specified in the C<ev_stat> watcher changed its attributes somehow.
548		1060
549	=item C<EV_IDLE>	1061	=item C<EV_IDLE>
550		1062
551	The C<ev_idle> watcher has determined that you have nothing better to do.	1063	The C<ev_idle> watcher has determined that you have nothing better to do.
552		1064
…		…
560	received events. Callbacks of both watcher types can start and stop as	1072	received events. Callbacks of both watcher types can start and stop as
561	many watchers as they want, and all of them will be taken into account	1073	many watchers as they want, and all of them will be taken into account
562	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1074	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
563	C<ev_loop> from blocking).	1075	C<ev_loop> from blocking).
564		1076
		1077	=item C<EV_EMBED>
		1078
		1079	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		1080
		1081	=item C<EV_FORK>
		1082
		1083	The event loop has been resumed in the child process after fork (see
		1084	C<ev_fork>).
		1085
		1086	=item C<EV_ASYNC>
		1087
		1088	The given async watcher has been asynchronously notified (see C<ev_async>).
		1089
		1090	=item C<EV_CUSTOM>
		1091
		1092	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1093	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1094
565	=item C<EV_ERROR>	1095	=item C<EV_ERROR>
566		1096
567	An unspecified error has occured, the watcher has been stopped. This might	1097	An unspecified error has occurred, the watcher has been stopped. This might
568	happen because the watcher could not be properly started because libev	1098	happen because the watcher could not be properly started because libev
569	ran out of memory, a file descriptor was found to be closed or any other	1099	ran out of memory, a file descriptor was found to be closed or any other
		1100	problem. Libev considers these application bugs.
		1101
570	problem. You best act on it by reporting the problem and somehow coping	1102	You best act on it by reporting the problem and somehow coping with the
571	with the watcher being stopped.	1103	watcher being stopped. Note that well-written programs should not receive
		1104	an error ever, so when your watcher receives it, this usually indicates a
		1105	bug in your program.
572		1106
573	Libev will usually signal a few "dummy" events together with an error,	1107	Libev will usually signal a few "dummy" events together with an error, for
574	for example it might indicate that a fd is readable or writable, and if	1108	example it might indicate that a fd is readable or writable, and if your
575	your callbacks is well-written it can just attempt the operation and cope	1109	callbacks is well-written it can just attempt the operation and cope with
576	with the error from read() or write(). This will not work in multithreaded	1110	the error from read() or write(). This will not work in multi-threaded
577	programs, though, so beware.	1111	programs, though, as the fd could already be closed and reused for another
		1112	thing, so beware.
578		1113
579	=back	1114	=back
580		1115
581	=head2 SUMMARY OF GENERIC WATCHER FUNCTIONS	1116	=head2 GENERIC WATCHER FUNCTIONS
582
583	In the following description, C<TYPE> stands for the watcher type,
584	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
585		1117
586	=over 4	1118	=over 4
587		1119
588	=item C<ev_init> (ev_TYPE *watcher, callback)	1120	=item C<ev_init> (ev_TYPE *watcher, callback)
589		1121
…		…
595	which rolls both calls into one.	1127	which rolls both calls into one.
596		1128
597	You can reinitialise a watcher at any time as long as it has been stopped	1129	You can reinitialise a watcher at any time as long as it has been stopped
598	(or never started) and there are no pending events outstanding.	1130	(or never started) and there are no pending events outstanding.
599		1131
600	The callbakc is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1132	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
601	int revents)>.	1133	int revents)>.
602		1134
		1135	Example: Initialise an C<ev_io> watcher in two steps.
		1136
		1137	ev_io w;
		1138	ev_init (&w, my_cb);
		1139	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1140
603	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1141	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
604		1142
605	This macro initialises the type-specific parts of a watcher. You need to	1143	This macro initialises the type-specific parts of a watcher. You need to
606	call C<ev_init> at least once before you call this macro, but you can	1144	call C<ev_init> at least once before you call this macro, but you can
607	call C<ev_TYPE_set> any number of times. You must not, however, call this	1145	call C<ev_TYPE_set> any number of times. You must not, however, call this
608	macro on a watcher that is active (it can be pending, however, which is a	1146	macro on a watcher that is active (it can be pending, however, which is a
609	difference to the C<ev_init> macro).	1147	difference to the C<ev_init> macro).
610		1148
611	Although some watcher types do not have type-specific arguments	1149	Although some watcher types do not have type-specific arguments
612	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1150	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
613		1151
		1152	See C<ev_init>, above, for an example.
		1153
614	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1154	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
615		1155
616	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1156	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
617	calls into a single call. This is the most convinient method to initialise	1157	calls into a single call. This is the most convenient method to initialise
618	a watcher. The same limitations apply, of course.	1158	a watcher. The same limitations apply, of course.
619		1159
		1160	Example: Initialise and set an C<ev_io> watcher in one step.
		1161
		1162	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1163
620	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1164	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
621		1165
622	Starts (activates) the given watcher. Only active watchers will receive	1166	Starts (activates) the given watcher. Only active watchers will receive
623	events. If the watcher is already active nothing will happen.	1167	events. If the watcher is already active nothing will happen.
624		1168
		1169	Example: Start the C<ev_io> watcher that is being abused as example in this
		1170	whole section.
		1171
		1172	ev_io_start (EV_DEFAULT_UC, &w);
		1173
625	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1174	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
626		1175
627	Stops the given watcher again (if active) and clears the pending	1176	Stops the given watcher if active, and clears the pending status (whether
		1177	the watcher was active or not).
		1178
628	status. It is possible that stopped watchers are pending (for example,	1179	It is possible that stopped watchers are pending - for example,
629	non-repeating timers are being stopped when they become pending), but	1180	non-repeating timers are being stopped when they become pending - but
630	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1181	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
631	you want to free or reuse the memory used by the watcher it is therefore a	1182	pending. If you want to free or reuse the memory used by the watcher it is
632	good idea to always call its C<ev_TYPE_stop> function.	1183	therefore a good idea to always call its C<ev_TYPE_stop> function.
633		1184
634	=item bool ev_is_active (ev_TYPE *watcher)	1185	=item bool ev_is_active (ev_TYPE *watcher)
635		1186
636	Returns a true value iff the watcher is active (i.e. it has been started	1187	Returns a true value iff the watcher is active (i.e. it has been started
637	and not yet been stopped). As long as a watcher is active you must not modify	1188	and not yet been stopped). As long as a watcher is active you must not modify
…		…
640	=item bool ev_is_pending (ev_TYPE *watcher)	1191	=item bool ev_is_pending (ev_TYPE *watcher)
641		1192
642	Returns a true value iff the watcher is pending, (i.e. it has outstanding	1193	Returns a true value iff the watcher is pending, (i.e. it has outstanding
643	events but its callback has not yet been invoked). As long as a watcher	1194	events but its callback has not yet been invoked). As long as a watcher
644	is pending (but not active) you must not call an init function on it (but	1195	is pending (but not active) you must not call an init function on it (but
645	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	1196	C<ev_TYPE_set> is safe), you must not change its priority, and you must
646	libev (e.g. you cnanot C<free ()> it).	1197	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		1198	it).
647		1199
648	=item callback = ev_cb (ev_TYPE *watcher)	1200	=item callback ev_cb (ev_TYPE *watcher)
649		1201
650	Returns the callback currently set on the watcher.	1202	Returns the callback currently set on the watcher.
651		1203
652	=item ev_cb_set (ev_TYPE *watcher, callback)	1204	=item ev_cb_set (ev_TYPE *watcher, callback)
653		1205
654	Change the callback. You can change the callback at virtually any time	1206	Change the callback. You can change the callback at virtually any time
655	(modulo threads).	1207	(modulo threads).
656		1208
		1209	=item ev_set_priority (ev_TYPE *watcher, int priority)
		1210
		1211	=item int ev_priority (ev_TYPE *watcher)
		1212
		1213	Set and query the priority of the watcher. The priority is a small
		1214	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		1215	(default: C<-2>). Pending watchers with higher priority will be invoked
		1216	before watchers with lower priority, but priority will not keep watchers
		1217	from being executed (except for C<ev_idle> watchers).
		1218
		1219	If you need to suppress invocation when higher priority events are pending
		1220	you need to look at C<ev_idle> watchers, which provide this functionality.
		1221
		1222	You I<must not> change the priority of a watcher as long as it is active or
		1223	pending.
		1224
		1225	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1226	fine, as long as you do not mind that the priority value you query might
		1227	or might not have been clamped to the valid range.
		1228
		1229	The default priority used by watchers when no priority has been set is
		1230	always C<0>, which is supposed to not be too high and not be too low :).
		1231
		1232	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
		1233	priorities.
		1234
		1235	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		1236
		1237	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		1238	C<loop> nor C<revents> need to be valid as long as the watcher callback
		1239	can deal with that fact, as both are simply passed through to the
		1240	callback.
		1241
		1242	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		1243
		1244	If the watcher is pending, this function clears its pending status and
		1245	returns its C<revents> bitset (as if its callback was invoked). If the
		1246	watcher isn't pending it does nothing and returns C<0>.
		1247
		1248	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1249	callback to be invoked, which can be accomplished with this function.
		1250
		1251	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1252
		1253	Feeds the given event set into the event loop, as if the specified event
		1254	had happened for the specified watcher (which must be a pointer to an
		1255	initialised but not necessarily started event watcher). Obviously you must
		1256	not free the watcher as long as it has pending events.
		1257
		1258	Stopping the watcher, letting libev invoke it, or calling
		1259	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1260	not started in the first place.
		1261
		1262	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1263	functions that do not need a watcher.
		1264
657	=back	1265	=back
658		1266
659		1267
660	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1268	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
661		1269
662	Each watcher has, by default, a member C<void *data> that you can change	1270	Each watcher has, by default, a member C<void *data> that you can change
663	and read at any time, libev will completely ignore it. This can be used	1271	and read at any time: libev will completely ignore it. This can be used
664	to associate arbitrary data with your watcher. If you need more data and	1272	to associate arbitrary data with your watcher. If you need more data and
665	don't want to allocate memory and store a pointer to it in that data	1273	don't want to allocate memory and store a pointer to it in that data
666	member, you can also "subclass" the watcher type and provide your own	1274	member, you can also "subclass" the watcher type and provide your own
667	data:	1275	data:
668		1276
669	struct my_io	1277	struct my_io
670	{	1278	{
671	struct ev_io io;	1279	ev_io io;
672	int otherfd;	1280	int otherfd;
673	void *somedata;	1281	void *somedata;
674	struct whatever *mostinteresting;	1282	struct whatever *mostinteresting;
675	}	1283	};
		1284
		1285	...
		1286	struct my_io w;
		1287	ev_io_init (&w.io, my_cb, fd, EV_READ);
676		1288
677	And since your callback will be called with a pointer to the watcher, you	1289	And since your callback will be called with a pointer to the watcher, you
678	can cast it back to your own type:	1290	can cast it back to your own type:
679		1291
680	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1292	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
681	{	1293	{
682	struct my_io *w = (struct my_io *)w_;	1294	struct my_io *w = (struct my_io *)w_;
683	...	1295	...
684	}	1296	}
685		1297
686	More interesting and less C-conformant ways of catsing your callback type	1298	More interesting and less C-conformant ways of casting your callback type
687	have been omitted....	1299	instead have been omitted.
		1300
		1301	Another common scenario is to use some data structure with multiple
		1302	embedded watchers:
		1303
		1304	struct my_biggy
		1305	{
		1306	int some_data;
		1307	ev_timer t1;
		1308	ev_timer t2;
		1309	}
		1310
		1311	In this case getting the pointer to C<my_biggy> is a bit more
		1312	complicated: Either you store the address of your C<my_biggy> struct
		1313	in the C<data> member of the watcher (for woozies), or you need to use
		1314	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1315	programmers):
		1316
		1317	#include <stddef.h>
		1318
		1319	static void
		1320	t1_cb (EV_P_ ev_timer *w, int revents)
		1321	{
		1322	struct my_biggy big = (struct my_biggy *)
		1323	(((char *)w) - offsetof (struct my_biggy, t1));
		1324	}
		1325
		1326	static void
		1327	t2_cb (EV_P_ ev_timer *w, int revents)
		1328	{
		1329	struct my_biggy big = (struct my_biggy *)
		1330	(((char *)w) - offsetof (struct my_biggy, t2));
		1331	}
		1332
		1333	=head2 WATCHER PRIORITY MODELS
		1334
		1335	Many event loops support I<watcher priorities>, which are usually small
		1336	integers that influence the ordering of event callback invocation
		1337	between watchers in some way, all else being equal.
		1338
		1339	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1340	description for the more technical details such as the actual priority
		1341	range.
		1342
		1343	There are two common ways how these these priorities are being interpreted
		1344	by event loops:
		1345
		1346	In the more common lock-out model, higher priorities "lock out" invocation
		1347	of lower priority watchers, which means as long as higher priority
		1348	watchers receive events, lower priority watchers are not being invoked.
		1349
		1350	The less common only-for-ordering model uses priorities solely to order
		1351	callback invocation within a single event loop iteration: Higher priority
		1352	watchers are invoked before lower priority ones, but they all get invoked
		1353	before polling for new events.
		1354
		1355	Libev uses the second (only-for-ordering) model for all its watchers
		1356	except for idle watchers (which use the lock-out model).
		1357
		1358	The rationale behind this is that implementing the lock-out model for
		1359	watchers is not well supported by most kernel interfaces, and most event
		1360	libraries will just poll for the same events again and again as long as
		1361	their callbacks have not been executed, which is very inefficient in the
		1362	common case of one high-priority watcher locking out a mass of lower
		1363	priority ones.
		1364
		1365	Static (ordering) priorities are most useful when you have two or more
		1366	watchers handling the same resource: a typical usage example is having an
		1367	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1368	timeouts. Under load, data might be received while the program handles
		1369	other jobs, but since timers normally get invoked first, the timeout
		1370	handler will be executed before checking for data. In that case, giving
		1371	the timer a lower priority than the I/O watcher ensures that I/O will be
		1372	handled first even under adverse conditions (which is usually, but not
		1373	always, what you want).
		1374
		1375	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1376	will only be executed when no same or higher priority watchers have
		1377	received events, they can be used to implement the "lock-out" model when
		1378	required.
		1379
		1380	For example, to emulate how many other event libraries handle priorities,
		1381	you can associate an C<ev_idle> watcher to each such watcher, and in
		1382	the normal watcher callback, you just start the idle watcher. The real
		1383	processing is done in the idle watcher callback. This causes libev to
		1384	continously poll and process kernel event data for the watcher, but when
		1385	the lock-out case is known to be rare (which in turn is rare :), this is
		1386	workable.
		1387
		1388	Usually, however, the lock-out model implemented that way will perform
		1389	miserably under the type of load it was designed to handle. In that case,
		1390	it might be preferable to stop the real watcher before starting the
		1391	idle watcher, so the kernel will not have to process the event in case
		1392	the actual processing will be delayed for considerable time.
		1393
		1394	Here is an example of an I/O watcher that should run at a strictly lower
		1395	priority than the default, and which should only process data when no
		1396	other events are pending:
		1397
		1398	ev_idle idle; // actual processing watcher
		1399	ev_io io; // actual event watcher
		1400
		1401	static void
		1402	io_cb (EV_P_ ev_io *w, int revents)
		1403	{
		1404	// stop the I/O watcher, we received the event, but
		1405	// are not yet ready to handle it.
		1406	ev_io_stop (EV_A_ w);
		1407
		1408	// start the idle watcher to ahndle the actual event.
		1409	// it will not be executed as long as other watchers
		1410	// with the default priority are receiving events.
		1411	ev_idle_start (EV_A_ &idle);
		1412	}
		1413
		1414	static void
		1415	idle_cb (EV_P_ ev_idle *w, int revents)
		1416	{
		1417	// actual processing
		1418	read (STDIN_FILENO, ...);
		1419
		1420	// have to start the I/O watcher again, as
		1421	// we have handled the event
		1422	ev_io_start (EV_P_ &io);
		1423	}
		1424
		1425	// initialisation
		1426	ev_idle_init (&idle, idle_cb);
		1427	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1428	ev_io_start (EV_DEFAULT_ &io);
		1429
		1430	In the "real" world, it might also be beneficial to start a timer, so that
		1431	low-priority connections can not be locked out forever under load. This
		1432	enables your program to keep a lower latency for important connections
		1433	during short periods of high load, while not completely locking out less
		1434	important ones.
688		1435
689		1436
690	=head1 WATCHER TYPES	1437	=head1 WATCHER TYPES
691		1438
692	This section describes each watcher in detail, but will not repeat	1439	This section describes each watcher in detail, but will not repeat
693	information given in the last section.	1440	information given in the last section. Any initialisation/set macros,
		1441	functions and members specific to the watcher type are explained.
694		1442
		1443	Members are additionally marked with either I<[read-only]>, meaning that,
		1444	while the watcher is active, you can look at the member and expect some
		1445	sensible content, but you must not modify it (you can modify it while the
		1446	watcher is stopped to your hearts content), or I<[read-write]>, which
		1447	means you can expect it to have some sensible content while the watcher
		1448	is active, but you can also modify it. Modifying it may not do something
		1449	sensible or take immediate effect (or do anything at all), but libev will
		1450	not crash or malfunction in any way.
695		1451
		1452
696	=head2 C<ev_io> - is this file descriptor readable or writable	1453	=head2 C<ev_io> - is this file descriptor readable or writable?
697		1454
698	I/O watchers check whether a file descriptor is readable or writable	1455	I/O watchers check whether a file descriptor is readable or writable
699	in each iteration of the event loop (This behaviour is called	1456	in each iteration of the event loop, or, more precisely, when reading
700	level-triggering because you keep receiving events as long as the	1457	would not block the process and writing would at least be able to write
701	condition persists. Remember you can stop the watcher if you don't want to	1458	some data. This behaviour is called level-triggering because you keep
702	act on the event and neither want to receive future events).	1459	receiving events as long as the condition persists. Remember you can stop
		1460	the watcher if you don't want to act on the event and neither want to
		1461	receive future events.
703		1462
704	In general you can register as many read and/or write event watchers per	1463	In general you can register as many read and/or write event watchers per
705	fd as you want (as long as you don't confuse yourself). Setting all file	1464	fd as you want (as long as you don't confuse yourself). Setting all file
706	descriptors to non-blocking mode is also usually a good idea (but not	1465	descriptors to non-blocking mode is also usually a good idea (but not
707	required if you know what you are doing).	1466	required if you know what you are doing).
708		1467
709	You have to be careful with dup'ed file descriptors, though. Some backends	1468	If you cannot use non-blocking mode, then force the use of a
710	(the linux epoll backend is a notable example) cannot handle dup'ed file	1469	known-to-be-good backend (at the time of this writing, this includes only
711	descriptors correctly if you register interest in two or more fds pointing	1470	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
712	to the same underlying file/socket etc. description (that is, they share	1471	descriptors for which non-blocking operation makes no sense (such as
713	the same underlying "file open").	1472	files) - libev doesn't guarentee any specific behaviour in that case.
714		1473
715	If you must do this, then force the use of a known-to-be-good backend	1474	Another thing you have to watch out for is that it is quite easy to
716	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1475	receive "spurious" readiness notifications, that is your callback might
		1476	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		1477	because there is no data. Not only are some backends known to create a
		1478	lot of those (for example Solaris ports), it is very easy to get into
		1479	this situation even with a relatively standard program structure. Thus
		1480	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		1481	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		1482
		1483	If you cannot run the fd in non-blocking mode (for example you should
		1484	not play around with an Xlib connection), then you have to separately
		1485	re-test whether a file descriptor is really ready with a known-to-be good
		1486	interface such as poll (fortunately in our Xlib example, Xlib already
		1487	does this on its own, so its quite safe to use). Some people additionally
		1488	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1489	indefinitely.
		1490
		1491	But really, best use non-blocking mode.
		1492
		1493	=head3 The special problem of disappearing file descriptors
		1494
		1495	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1496	descriptor (either due to calling C<close> explicitly or any other means,
		1497	such as C<dup2>). The reason is that you register interest in some file
		1498	descriptor, but when it goes away, the operating system will silently drop
		1499	this interest. If another file descriptor with the same number then is
		1500	registered with libev, there is no efficient way to see that this is, in
		1501	fact, a different file descriptor.
		1502
		1503	To avoid having to explicitly tell libev about such cases, libev follows
		1504	the following policy: Each time C<ev_io_set> is being called, libev
		1505	will assume that this is potentially a new file descriptor, otherwise
		1506	it is assumed that the file descriptor stays the same. That means that
		1507	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1508	descriptor even if the file descriptor number itself did not change.
		1509
		1510	This is how one would do it normally anyway, the important point is that
		1511	the libev application should not optimise around libev but should leave
		1512	optimisations to libev.
		1513
		1514	=head3 The special problem of dup'ed file descriptors
		1515
		1516	Some backends (e.g. epoll), cannot register events for file descriptors,
		1517	but only events for the underlying file descriptions. That means when you
		1518	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1519	events for them, only one file descriptor might actually receive events.
		1520
		1521	There is no workaround possible except not registering events
		1522	for potentially C<dup ()>'ed file descriptors, or to resort to
		1523	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1524
		1525	=head3 The special problem of fork
		1526
		1527	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1528	useless behaviour. Libev fully supports fork, but needs to be told about
		1529	it in the child.
		1530
		1531	To support fork in your programs, you either have to call
		1532	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
		1533	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
717	C<EVBACKEND_POLL>).	1534	C<EVBACKEND_POLL>.
		1535
		1536	=head3 The special problem of SIGPIPE
		1537
		1538	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1539	when writing to a pipe whose other end has been closed, your program gets
		1540	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1541	this is sensible behaviour, for daemons, this is usually undesirable.
		1542
		1543	So when you encounter spurious, unexplained daemon exits, make sure you
		1544	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1545	somewhere, as that would have given you a big clue).
		1546
		1547	=head3 The special problem of accept()ing when you can't
		1548
		1549	Many implementations of the POSIX C<accept> function (for example,
		1550	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1551	connection from the pending queue in all error cases.
		1552
		1553	For example, larger servers often run out of file descriptors (because
		1554	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1555	rejecting the connection, leading to libev signalling readiness on
		1556	the next iteration again (the connection still exists after all), and
		1557	typically causing the program to loop at 100% CPU usage.
		1558
		1559	Unfortunately, the set of errors that cause this issue differs between
		1560	operating systems, there is usually little the app can do to remedy the
		1561	situation, and no known thread-safe method of removing the connection to
		1562	cope with overload is known (to me).
		1563
		1564	One of the easiest ways to handle this situation is to just ignore it
		1565	- when the program encounters an overload, it will just loop until the
		1566	situation is over. While this is a form of busy waiting, no OS offers an
		1567	event-based way to handle this situation, so it's the best one can do.
		1568
		1569	A better way to handle the situation is to log any errors other than
		1570	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1571	messages, and continue as usual, which at least gives the user an idea of
		1572	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1573	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1574	usage.
		1575
		1576	If your program is single-threaded, then you could also keep a dummy file
		1577	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1578	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1579	close that fd, and create a new dummy fd. This will gracefully refuse
		1580	clients under typical overload conditions.
		1581
		1582	The last way to handle it is to simply log the error and C<exit>, as
		1583	is often done with C<malloc> failures, but this results in an easy
		1584	opportunity for a DoS attack.
		1585
		1586	=head3 Watcher-Specific Functions
718		1587
719	=over 4	1588	=over 4
720		1589
721	=item ev_io_init (ev_io *, callback, int fd, int events)	1590	=item ev_io_init (ev_io *, callback, int fd, int events)
722		1591
723	=item ev_io_set (ev_io *, int fd, int events)	1592	=item ev_io_set (ev_io *, int fd, int events)
724		1593
725	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	1594	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
726	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	1595	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
727	EV_WRITE> to receive the given events.	1596	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
728		1597
729	Please note that most of the more scalable backend mechanisms (for example	1598	=item int fd [read-only]
730	epoll and solaris ports) can result in spurious readyness notifications	1599
731	for file descriptors, so you practically need to use non-blocking I/O (and	1600	The file descriptor being watched.
732	treat callback invocation as hint only), or retest separately with a safe	1601
733	interface before doing I/O (XLib can do this), or force the use of either	1602	=item int events [read-only]
734	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>, which don't suffer from this	1603
735	problem. Also note that it is quite easy to have your callback invoked	1604	The events being watched.
736	when the readyness condition is no longer valid even when employing
737	typical ways of handling events, so its a good idea to use non-blocking
738	I/O unconditionally.
739		1605
740	=back	1606	=back
741		1607
		1608	=head3 Examples
		1609
742	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	1610	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
743	readable, but only once. Since it is likely line-buffered, you could	1611	readable, but only once. Since it is likely line-buffered, you could
744	attempt to read a whole line in the callback:	1612	attempt to read a whole line in the callback.
745		1613
746	static void	1614	static void
747	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1615	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
748	{	1616	{
749	ev_io_stop (loop, w);	1617	ev_io_stop (loop, w);
750	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1618	.. read from stdin here (or from w->fd) and handle any I/O errors
751	}	1619	}
752		1620
753	...	1621	...
754	struct ev_loop *loop = ev_default_init (0);	1622	struct ev_loop *loop = ev_default_init (0);
755	struct ev_io stdin_readable;	1623	ev_io stdin_readable;
756	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1624	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
757	ev_io_start (loop, &stdin_readable);	1625	ev_io_start (loop, &stdin_readable);
758	ev_loop (loop, 0);	1626	ev_loop (loop, 0);
759		1627
760		1628
761	=head2 C<ev_timer> - relative and optionally recurring timeouts	1629	=head2 C<ev_timer> - relative and optionally repeating timeouts
762		1630
763	Timer watchers are simple relative timers that generate an event after a	1631	Timer watchers are simple relative timers that generate an event after a
764	given time, and optionally repeating in regular intervals after that.	1632	given time, and optionally repeating in regular intervals after that.
765		1633
766	The timers are based on real time, that is, if you register an event that	1634	The timers are based on real time, that is, if you register an event that
767	times out after an hour and you reset your system clock to last years	1635	times out after an hour and you reset your system clock to January last
768	time, it will still time out after (roughly) and hour. "Roughly" because	1636	year, it will still time out after (roughly) one hour. "Roughly" because
769	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1637	detecting time jumps is hard, and some inaccuracies are unavoidable (the
770	monotonic clock option helps a lot here).	1638	monotonic clock option helps a lot here).
		1639
		1640	The callback is guaranteed to be invoked only I<after> its timeout has
		1641	passed (not I<at>, so on systems with very low-resolution clocks this
		1642	might introduce a small delay). If multiple timers become ready during the
		1643	same loop iteration then the ones with earlier time-out values are invoked
		1644	before ones of the same priority with later time-out values (but this is
		1645	no longer true when a callback calls C<ev_loop> recursively).
		1646
		1647	=head3 Be smart about timeouts
		1648
		1649	Many real-world problems involve some kind of timeout, usually for error
		1650	recovery. A typical example is an HTTP request - if the other side hangs,
		1651	you want to raise some error after a while.
		1652
		1653	What follows are some ways to handle this problem, from obvious and
		1654	inefficient to smart and efficient.
		1655
		1656	In the following, a 60 second activity timeout is assumed - a timeout that
		1657	gets reset to 60 seconds each time there is activity (e.g. each time some
		1658	data or other life sign was received).
		1659
		1660	=over 4
		1661
		1662	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1663
		1664	This is the most obvious, but not the most simple way: In the beginning,
		1665	start the watcher:
		1666
		1667	ev_timer_init (timer, callback, 60., 0.);
		1668	ev_timer_start (loop, timer);
		1669
		1670	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1671	and start it again:
		1672
		1673	ev_timer_stop (loop, timer);
		1674	ev_timer_set (timer, 60., 0.);
		1675	ev_timer_start (loop, timer);
		1676
		1677	This is relatively simple to implement, but means that each time there is
		1678	some activity, libev will first have to remove the timer from its internal
		1679	data structure and then add it again. Libev tries to be fast, but it's
		1680	still not a constant-time operation.
		1681
		1682	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1683
		1684	This is the easiest way, and involves using C<ev_timer_again> instead of
		1685	C<ev_timer_start>.
		1686
		1687	To implement this, configure an C<ev_timer> with a C<repeat> value
		1688	of C<60> and then call C<ev_timer_again> at start and each time you
		1689	successfully read or write some data. If you go into an idle state where
		1690	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1691	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1692
		1693	That means you can ignore both the C<ev_timer_start> function and the
		1694	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1695	member and C<ev_timer_again>.
		1696
		1697	At start:
		1698
		1699	ev_init (timer, callback);
		1700	timer->repeat = 60.;
		1701	ev_timer_again (loop, timer);
		1702
		1703	Each time there is some activity:
		1704
		1705	ev_timer_again (loop, timer);
		1706
		1707	It is even possible to change the time-out on the fly, regardless of
		1708	whether the watcher is active or not:
		1709
		1710	timer->repeat = 30.;
		1711	ev_timer_again (loop, timer);
		1712
		1713	This is slightly more efficient then stopping/starting the timer each time
		1714	you want to modify its timeout value, as libev does not have to completely
		1715	remove and re-insert the timer from/into its internal data structure.
		1716
		1717	It is, however, even simpler than the "obvious" way to do it.
		1718
		1719	=item 3. Let the timer time out, but then re-arm it as required.
		1720
		1721	This method is more tricky, but usually most efficient: Most timeouts are
		1722	relatively long compared to the intervals between other activity - in
		1723	our example, within 60 seconds, there are usually many I/O events with
		1724	associated activity resets.
		1725
		1726	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1727	but remember the time of last activity, and check for a real timeout only
		1728	within the callback:
		1729
		1730	ev_tstamp last_activity; // time of last activity
		1731
		1732	static void
		1733	callback (EV_P_ ev_timer *w, int revents)
		1734	{
		1735	ev_tstamp now = ev_now (EV_A);
		1736	ev_tstamp timeout = last_activity + 60.;
		1737
		1738	// if last_activity + 60. is older than now, we did time out
		1739	if (timeout < now)
		1740	{
		1741	// timeout occured, take action
		1742	}
		1743	else
		1744	{
		1745	// callback was invoked, but there was some activity, re-arm
		1746	// the watcher to fire in last_activity + 60, which is
		1747	// guaranteed to be in the future, so "again" is positive:
		1748	w->repeat = timeout - now;
		1749	ev_timer_again (EV_A_ w);
		1750	}
		1751	}
		1752
		1753	To summarise the callback: first calculate the real timeout (defined
		1754	as "60 seconds after the last activity"), then check if that time has
		1755	been reached, which means something I<did>, in fact, time out. Otherwise
		1756	the callback was invoked too early (C<timeout> is in the future), so
		1757	re-schedule the timer to fire at that future time, to see if maybe we have
		1758	a timeout then.
		1759
		1760	Note how C<ev_timer_again> is used, taking advantage of the
		1761	C<ev_timer_again> optimisation when the timer is already running.
		1762
		1763	This scheme causes more callback invocations (about one every 60 seconds
		1764	minus half the average time between activity), but virtually no calls to
		1765	libev to change the timeout.
		1766
		1767	To start the timer, simply initialise the watcher and set C<last_activity>
		1768	to the current time (meaning we just have some activity :), then call the
		1769	callback, which will "do the right thing" and start the timer:
		1770
		1771	ev_init (timer, callback);
		1772	last_activity = ev_now (loop);
		1773	callback (loop, timer, EV_TIMER);
		1774
		1775	And when there is some activity, simply store the current time in
		1776	C<last_activity>, no libev calls at all:
		1777
		1778	last_actiivty = ev_now (loop);
		1779
		1780	This technique is slightly more complex, but in most cases where the
		1781	time-out is unlikely to be triggered, much more efficient.
		1782
		1783	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1784	callback :) - just change the timeout and invoke the callback, which will
		1785	fix things for you.
		1786
		1787	=item 4. Wee, just use a double-linked list for your timeouts.
		1788
		1789	If there is not one request, but many thousands (millions...), all
		1790	employing some kind of timeout with the same timeout value, then one can
		1791	do even better:
		1792
		1793	When starting the timeout, calculate the timeout value and put the timeout
		1794	at the I<end> of the list.
		1795
		1796	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1797	the list is expected to fire (for example, using the technique #3).
		1798
		1799	When there is some activity, remove the timer from the list, recalculate
		1800	the timeout, append it to the end of the list again, and make sure to
		1801	update the C<ev_timer> if it was taken from the beginning of the list.
		1802
		1803	This way, one can manage an unlimited number of timeouts in O(1) time for
		1804	starting, stopping and updating the timers, at the expense of a major
		1805	complication, and having to use a constant timeout. The constant timeout
		1806	ensures that the list stays sorted.
		1807
		1808	=back
		1809
		1810	So which method the best?
		1811
		1812	Method #2 is a simple no-brain-required solution that is adequate in most
		1813	situations. Method #3 requires a bit more thinking, but handles many cases
		1814	better, and isn't very complicated either. In most case, choosing either
		1815	one is fine, with #3 being better in typical situations.
		1816
		1817	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1818	rather complicated, but extremely efficient, something that really pays
		1819	off after the first million or so of active timers, i.e. it's usually
		1820	overkill :)
		1821
		1822	=head3 The special problem of time updates
		1823
		1824	Establishing the current time is a costly operation (it usually takes at
		1825	least two system calls): EV therefore updates its idea of the current
		1826	time only before and after C<ev_loop> collects new events, which causes a
		1827	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1828	lots of events in one iteration.
771		1829
772	The relative timeouts are calculated relative to the C<ev_now ()>	1830	The relative timeouts are calculated relative to the C<ev_now ()>
773	time. This is usually the right thing as this timestamp refers to the time	1831	time. This is usually the right thing as this timestamp refers to the time
774	of the event triggering whatever timeout you are modifying/starting. If	1832	of the event triggering whatever timeout you are modifying/starting. If
775	you suspect event processing to be delayed and you I<need> to base the timeout	1833	you suspect event processing to be delayed and you I<need> to base the
776	on the current time, use something like this to adjust for this:	1834	timeout on the current time, use something like this to adjust for this:
777		1835
778	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1836	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
779		1837
780	The callback is guarenteed to be invoked only when its timeout has passed,	1838	If the event loop is suspended for a long time, you can also force an
781	but if multiple timers become ready during the same loop iteration then	1839	update of the time returned by C<ev_now ()> by calling C<ev_now_update
782	order of execution is undefined.	1840	()>.
		1841
		1842	=head3 The special problems of suspended animation
		1843
		1844	When you leave the server world it is quite customary to hit machines that
		1845	can suspend/hibernate - what happens to the clocks during such a suspend?
		1846
		1847	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1848	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1849	to run until the system is suspended, but they will not advance while the
		1850	system is suspended. That means, on resume, it will be as if the program
		1851	was frozen for a few seconds, but the suspend time will not be counted
		1852	towards C<ev_timer> when a monotonic clock source is used. The real time
		1853	clock advanced as expected, but if it is used as sole clocksource, then a
		1854	long suspend would be detected as a time jump by libev, and timers would
		1855	be adjusted accordingly.
		1856
		1857	I would not be surprised to see different behaviour in different between
		1858	operating systems, OS versions or even different hardware.
		1859
		1860	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1861	time jump in the monotonic clocks and the realtime clock. If the program
		1862	is suspended for a very long time, and monotonic clock sources are in use,
		1863	then you can expect C<ev_timer>s to expire as the full suspension time
		1864	will be counted towards the timers. When no monotonic clock source is in
		1865	use, then libev will again assume a timejump and adjust accordingly.
		1866
		1867	It might be beneficial for this latter case to call C<ev_suspend>
		1868	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1869	deterministic behaviour in this case (you can do nothing against
		1870	C<SIGSTOP>).
		1871
		1872	=head3 Watcher-Specific Functions and Data Members
783		1873
784	=over 4	1874	=over 4
785		1875
786	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1876	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
787		1877
788	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1878	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
789		1879
790	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1880	Configure the timer to trigger after C<after> seconds. If C<repeat>
791	C<0.>, then it will automatically be stopped. If it is positive, then the	1881	is C<0.>, then it will automatically be stopped once the timeout is
792	timer will automatically be configured to trigger again C<repeat> seconds	1882	reached. If it is positive, then the timer will automatically be
793	later, again, and again, until stopped manually.	1883	configured to trigger again C<repeat> seconds later, again, and again,
		1884	until stopped manually.
794		1885
795	The timer itself will do a best-effort at avoiding drift, that is, if you	1886	The timer itself will do a best-effort at avoiding drift, that is, if
796	configure a timer to trigger every 10 seconds, then it will trigger at	1887	you configure a timer to trigger every 10 seconds, then it will normally
797	exactly 10 second intervals. If, however, your program cannot keep up with	1888	trigger at exactly 10 second intervals. If, however, your program cannot
798	the timer (because it takes longer than those 10 seconds to do stuff) the	1889	keep up with the timer (because it takes longer than those 10 seconds to
799	timer will not fire more than once per event loop iteration.	1890	do stuff) the timer will not fire more than once per event loop iteration.
800		1891
801	=item ev_timer_again (loop)	1892	=item ev_timer_again (loop, ev_timer *)
802		1893
803	This will act as if the timer timed out and restart it again if it is	1894	This will act as if the timer timed out and restart it again if it is
804	repeating. The exact semantics are:	1895	repeating. The exact semantics are:
805		1896
		1897	If the timer is pending, its pending status is cleared.
		1898
806	If the timer is started but nonrepeating, stop it.	1899	If the timer is started but non-repeating, stop it (as if it timed out).
807		1900
808	If the timer is repeating, either start it if necessary (with the repeat	1901	If the timer is repeating, either start it if necessary (with the
809	value), or reset the running timer to the repeat value.	1902	C<repeat> value), or reset the running timer to the C<repeat> value.
810		1903
811	This sounds a bit complicated, but here is a useful and typical	1904	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
812	example: Imagine you have a tcp connection and you want a so-called idle	1905	usage example.
813	timeout, that is, you want to be called when there have been, say, 60	1906
814	seconds of inactivity on the socket. The easiest way to do this is to	1907	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
815	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	1908
816	time you successfully read or write some data. If you go into an idle	1909	Returns the remaining time until a timer fires. If the timer is active,
817	state where you do not expect data to travel on the socket, you can stop	1910	then this time is relative to the current event loop time, otherwise it's
818	the timer, and again will automatically restart it if need be.	1911	the timeout value currently configured.
		1912
		1913	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
		1914	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		1915	will return C<4>. When the timer expires and is restarted, it will return
		1916	roughly C<7> (likely slightly less as callback invocation takes some time,
		1917	too), and so on.
		1918
		1919	=item ev_tstamp repeat [read-write]
		1920
		1921	The current C<repeat> value. Will be used each time the watcher times out
		1922	or C<ev_timer_again> is called, and determines the next timeout (if any),
		1923	which is also when any modifications are taken into account.
819		1924
820	=back	1925	=back
821		1926
		1927	=head3 Examples
		1928
822	Example: create a timer that fires after 60 seconds.	1929	Example: Create a timer that fires after 60 seconds.
823		1930
824	static void	1931	static void
825	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1932	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
826	{	1933	{
827	.. one minute over, w is actually stopped right here	1934	.. one minute over, w is actually stopped right here
828	}	1935	}
829		1936
830	struct ev_timer mytimer;	1937	ev_timer mytimer;
831	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1938	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
832	ev_timer_start (loop, &mytimer);	1939	ev_timer_start (loop, &mytimer);
833		1940
834	Example: create a timeout timer that times out after 10 seconds of	1941	Example: Create a timeout timer that times out after 10 seconds of
835	inactivity.	1942	inactivity.
836		1943
837	static void	1944	static void
838	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1945	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
839	{	1946	{
840	.. ten seconds without any activity	1947	.. ten seconds without any activity
841	}	1948	}
842		1949
843	struct ev_timer mytimer;	1950	ev_timer mytimer;
844	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1951	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
845	ev_timer_again (&mytimer); /* start timer */	1952	ev_timer_again (&mytimer); /* start timer */
846	ev_loop (loop, 0);	1953	ev_loop (loop, 0);
847		1954
848	// and in some piece of code that gets executed on any "activity":	1955	// and in some piece of code that gets executed on any "activity":
849	// reset the timeout to start ticking again at 10 seconds	1956	// reset the timeout to start ticking again at 10 seconds
850	ev_timer_again (&mytimer);	1957	ev_timer_again (&mytimer);
851		1958
852		1959
853	=head2 C<ev_periodic> - to cron or not to cron	1960	=head2 C<ev_periodic> - to cron or not to cron?
854		1961
855	Periodic watchers are also timers of a kind, but they are very versatile	1962	Periodic watchers are also timers of a kind, but they are very versatile
856	(and unfortunately a bit complex).	1963	(and unfortunately a bit complex).
857		1964
858	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1965	Unlike C<ev_timer>, periodic watchers are not based on real time (or
859	but on wallclock time (absolute time). You can tell a periodic watcher	1966	relative time, the physical time that passes) but on wall clock time
860	to trigger "at" some specific point in time. For example, if you tell a	1967	(absolute time, the thing you can read on your calender or clock). The
861	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1968	difference is that wall clock time can run faster or slower than real
862	+ 10.>) and then reset your system clock to the last year, then it will	1969	time, and time jumps are not uncommon (e.g. when you adjust your
863	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1970	wrist-watch).
864	roughly 10 seconds later and of course not if you reset your system time
865	again).
866		1971
867	They can also be used to implement vastly more complex timers, such as	1972	You can tell a periodic watcher to trigger after some specific point
		1973	in time: for example, if you tell a periodic watcher to trigger "in 10
		1974	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1975	not a delay) and then reset your system clock to January of the previous
		1976	year, then it will take a year or more to trigger the event (unlike an
		1977	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1978	it, as it uses a relative timeout).
		1979
		1980	C<ev_periodic> watchers can also be used to implement vastly more complex
868	triggering an event on eahc midnight, local time.	1981	timers, such as triggering an event on each "midnight, local time", or
		1982	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1983	those cannot react to time jumps.
869		1984
870	As with timers, the callback is guarenteed to be invoked only when the	1985	As with timers, the callback is guaranteed to be invoked only when the
871	time (C<at>) has been passed, but if multiple periodic timers become ready	1986	point in time where it is supposed to trigger has passed. If multiple
872	during the same loop iteration then order of execution is undefined.	1987	timers become ready during the same loop iteration then the ones with
		1988	earlier time-out values are invoked before ones with later time-out values
		1989	(but this is no longer true when a callback calls C<ev_loop> recursively).
		1990
		1991	=head3 Watcher-Specific Functions and Data Members
873		1992
874	=over 4	1993	=over 4
875		1994
876	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1995	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
877		1996
878	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1997	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
879		1998
880	Lots of arguments, lets sort it out... There are basically three modes of	1999	Lots of arguments, let's sort it out... There are basically three modes of
881	operation, and we will explain them from simplest to complex:	2000	operation, and we will explain them from simplest to most complex:
882		2001
883	=over 4	2002	=over 4
884		2003
885	=item * absolute timer (interval = reschedule_cb = 0)	2004	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
886		2005
887	In this configuration the watcher triggers an event at the wallclock time	2006	In this configuration the watcher triggers an event after the wall clock
888	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	2007	time C<offset> has passed. It will not repeat and will not adjust when a
889	that is, if it is to be run at January 1st 2011 then it will run when the	2008	time jump occurs, that is, if it is to be run at January 1st 2011 then it
890	system time reaches or surpasses this time.	2009	will be stopped and invoked when the system clock reaches or surpasses
		2010	this point in time.
891		2011
892	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	2012	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
893		2013
894	In this mode the watcher will always be scheduled to time out at the next	2014	In this mode the watcher will always be scheduled to time out at the next
895	C<at + N * interval> time (for some integer N) and then repeat, regardless	2015	C<offset + N * interval> time (for some integer N, which can also be
896	of any time jumps.	2016	negative) and then repeat, regardless of any time jumps. The C<offset>
		2017	argument is merely an offset into the C<interval> periods.
897		2018
898	This can be used to create timers that do not drift with respect to system	2019	This can be used to create timers that do not drift with respect to the
899	time:	2020	system clock, for example, here is an C<ev_periodic> that triggers each
		2021	hour, on the hour (with respect to UTC):
900		2022
901	ev_periodic_set (&periodic, 0., 3600., 0);	2023	ev_periodic_set (&periodic, 0., 3600., 0);
902		2024
903	This doesn't mean there will always be 3600 seconds in between triggers,	2025	This doesn't mean there will always be 3600 seconds in between triggers,
904	but only that the the callback will be called when the system time shows a	2026	but only that the callback will be called when the system time shows a
905	full hour (UTC), or more correctly, when the system time is evenly divisible	2027	full hour (UTC), or more correctly, when the system time is evenly divisible
906	by 3600.	2028	by 3600.
907		2029
908	Another way to think about it (for the mathematically inclined) is that	2030	Another way to think about it (for the mathematically inclined) is that
909	C<ev_periodic> will try to run the callback in this mode at the next possible	2031	C<ev_periodic> will try to run the callback in this mode at the next possible
910	time where C<time = at (mod interval)>, regardless of any time jumps.	2032	time where C<time = offset (mod interval)>, regardless of any time jumps.
911		2033
		2034	For numerical stability it is preferable that the C<offset> value is near
		2035	C<ev_now ()> (the current time), but there is no range requirement for
		2036	this value, and in fact is often specified as zero.
		2037
		2038	Note also that there is an upper limit to how often a timer can fire (CPU
		2039	speed for example), so if C<interval> is very small then timing stability
		2040	will of course deteriorate. Libev itself tries to be exact to be about one
		2041	millisecond (if the OS supports it and the machine is fast enough).
		2042
912	=item * manual reschedule mode (reschedule_cb = callback)	2043	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
913		2044
914	In this mode the values for C<interval> and C<at> are both being	2045	In this mode the values for C<interval> and C<offset> are both being
915	ignored. Instead, each time the periodic watcher gets scheduled, the	2046	ignored. Instead, each time the periodic watcher gets scheduled, the
916	reschedule callback will be called with the watcher as first, and the	2047	reschedule callback will be called with the watcher as first, and the
917	current time as second argument.	2048	current time as second argument.
918		2049
919	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2050	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
920	ever, or make any event loop modifications>. If you need to stop it,	2051	or make ANY other event loop modifications whatsoever, unless explicitly
921	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	2052	allowed by documentation here>.
922	starting a prepare watcher).
923		2053
		2054	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		2055	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		2056	only event loop modification you are allowed to do).
		2057
924	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	2058	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
925	ev_tstamp now)>, e.g.:	2059	*w, ev_tstamp now)>, e.g.:
926		2060
		2061	static ev_tstamp
927	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2062	my_rescheduler (ev_periodic *w, ev_tstamp now)
928	{	2063	{
929	return now + 60.;	2064	return now + 60.;
930	}	2065	}
931		2066
932	It must return the next time to trigger, based on the passed time value	2067	It must return the next time to trigger, based on the passed time value
933	(that is, the lowest time value larger than to the second argument). It	2068	(that is, the lowest time value larger than to the second argument). It
934	will usually be called just before the callback will be triggered, but	2069	will usually be called just before the callback will be triggered, but
935	might be called at other times, too.	2070	might be called at other times, too.
936		2071
937	NOTE: I<< This callback must always return a time that is later than the	2072	NOTE: I<< This callback must always return a time that is higher than or
938	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	2073	equal to the passed C<now> value >>.
939		2074
940	This can be used to create very complex timers, such as a timer that	2075	This can be used to create very complex timers, such as a timer that
941	triggers on each midnight, local time. To do this, you would calculate the	2076	triggers on "next midnight, local time". To do this, you would calculate the
942	next midnight after C<now> and return the timestamp value for this. How	2077	next midnight after C<now> and return the timestamp value for this. How
943	you do this is, again, up to you (but it is not trivial, which is the main	2078	you do this is, again, up to you (but it is not trivial, which is the main
944	reason I omitted it as an example).	2079	reason I omitted it as an example).
945		2080
946	=back	2081	=back
…		…
950	Simply stops and restarts the periodic watcher again. This is only useful	2085	Simply stops and restarts the periodic watcher again. This is only useful
951	when you changed some parameters or the reschedule callback would return	2086	when you changed some parameters or the reschedule callback would return
952	a different time than the last time it was called (e.g. in a crond like	2087	a different time than the last time it was called (e.g. in a crond like
953	program when the crontabs have changed).	2088	program when the crontabs have changed).
954		2089
		2090	=item ev_tstamp ev_periodic_at (ev_periodic *)
		2091
		2092	When active, returns the absolute time that the watcher is supposed
		2093	to trigger next. This is not the same as the C<offset> argument to
		2094	C<ev_periodic_set>, but indeed works even in interval and manual
		2095	rescheduling modes.
		2096
		2097	=item ev_tstamp offset [read-write]
		2098
		2099	When repeating, this contains the offset value, otherwise this is the
		2100	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2101	although libev might modify this value for better numerical stability).
		2102
		2103	Can be modified any time, but changes only take effect when the periodic
		2104	timer fires or C<ev_periodic_again> is being called.
		2105
		2106	=item ev_tstamp interval [read-write]
		2107
		2108	The current interval value. Can be modified any time, but changes only
		2109	take effect when the periodic timer fires or C<ev_periodic_again> is being
		2110	called.
		2111
		2112	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
		2113
		2114	The current reschedule callback, or C<0>, if this functionality is
		2115	switched off. Can be changed any time, but changes only take effect when
		2116	the periodic timer fires or C<ev_periodic_again> is being called.
		2117
955	=back	2118	=back
956		2119
		2120	=head3 Examples
		2121
957	Example: call a callback every hour, or, more precisely, whenever the	2122	Example: Call a callback every hour, or, more precisely, whenever the
958	system clock is divisible by 3600. The callback invocation times have	2123	system time is divisible by 3600. The callback invocation times have
959	potentially a lot of jittering, but good long-term stability.	2124	potentially a lot of jitter, but good long-term stability.
960		2125
961	static void	2126	static void
962	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2127	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
963	{	2128	{
964	... its now a full hour (UTC, or TAI or whatever your clock follows)	2129	... its now a full hour (UTC, or TAI or whatever your clock follows)
965	}	2130	}
966		2131
967	struct ev_periodic hourly_tick;	2132	ev_periodic hourly_tick;
968	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2133	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
969	ev_periodic_start (loop, &hourly_tick);	2134	ev_periodic_start (loop, &hourly_tick);
970		2135
971	Example: the same as above, but use a reschedule callback to do it:	2136	Example: The same as above, but use a reschedule callback to do it:
972		2137
973	#include <math.h>	2138	#include <math.h>
974		2139
975	static ev_tstamp	2140	static ev_tstamp
976	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2141	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
977	{	2142	{
978	return fmod (now, 3600.) + 3600.;	2143	return now + (3600. - fmod (now, 3600.));
979	}	2144	}
980		2145
981	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2146	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
982		2147
983	Example: call a callback every hour, starting now:	2148	Example: Call a callback every hour, starting now:
984		2149
985	struct ev_periodic hourly_tick;	2150	ev_periodic hourly_tick;
986	ev_periodic_init (&hourly_tick, clock_cb,	2151	ev_periodic_init (&hourly_tick, clock_cb,
987	fmod (ev_now (loop), 3600.), 3600., 0);	2152	fmod (ev_now (loop), 3600.), 3600., 0);
988	ev_periodic_start (loop, &hourly_tick);	2153	ev_periodic_start (loop, &hourly_tick);
989		2154
990		2155
991	=head2 C<ev_signal> - signal me when a signal gets signalled	2156	=head2 C<ev_signal> - signal me when a signal gets signalled!
992		2157
993	Signal watchers will trigger an event when the process receives a specific	2158	Signal watchers will trigger an event when the process receives a specific
994	signal one or more times. Even though signals are very asynchronous, libev	2159	signal one or more times. Even though signals are very asynchronous, libev
995	will try it's best to deliver signals synchronously, i.e. as part of the	2160	will try it's best to deliver signals synchronously, i.e. as part of the
996	normal event processing, like any other event.	2161	normal event processing, like any other event.
997		2162
		2163	If you want signals to be delivered truly asynchronously, just use
		2164	C<sigaction> as you would do without libev and forget about sharing
		2165	the signal. You can even use C<ev_async> from a signal handler to
		2166	synchronously wake up an event loop.
		2167
998	You can configure as many watchers as you like per signal. Only when the	2168	You can configure as many watchers as you like for the same signal, but
		2169	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2170	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2171	C<SIGINT> in both the default loop and another loop at the same time. At
		2172	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2173
999	first watcher gets started will libev actually register a signal watcher	2174	When the first watcher gets started will libev actually register something
1000	with the kernel (thus it coexists with your own signal handlers as long	2175	with the kernel (thus it coexists with your own signal handlers as long as
1001	as you don't register any with libev). Similarly, when the last signal	2176	you don't register any with libev for the same signal).
1002	watcher for a signal is stopped libev will reset the signal handler to	2177
1003	SIG_DFL (regardless of what it was set to before).	2178	If possible and supported, libev will install its handlers with
		2179	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
		2180	not be unduly interrupted. If you have a problem with system calls getting
		2181	interrupted by signals you can block all signals in an C<ev_check> watcher
		2182	and unblock them in an C<ev_prepare> watcher.
		2183
		2184	=head3 The special problem of inheritance over fork/execve/pthread_create
		2185
		2186	Both the signal mask (C<sigprocmask>) and the signal disposition
		2187	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2188	stopping it again), that is, libev might or might not block the signal,
		2189	and might or might not set or restore the installed signal handler.
		2190
		2191	While this does not matter for the signal disposition (libev never
		2192	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2193	C<execve>), this matters for the signal mask: many programs do not expect
		2194	certain signals to be blocked.
		2195
		2196	This means that before calling C<exec> (from the child) you should reset
		2197	the signal mask to whatever "default" you expect (all clear is a good
		2198	choice usually).
		2199
		2200	The simplest way to ensure that the signal mask is reset in the child is
		2201	to install a fork handler with C<pthread_atfork> that resets it. That will
		2202	catch fork calls done by libraries (such as the libc) as well.
		2203
		2204	In current versions of libev, the signal will not be blocked indefinitely
		2205	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2206	the window of opportunity for problems, it will not go away, as libev
		2207	I<has> to modify the signal mask, at least temporarily.
		2208
		2209	So I can't stress this enough: I<If you do not reset your signal mask when
		2210	you expect it to be empty, you have a race condition in your code>. This
		2211	is not a libev-specific thing, this is true for most event libraries.
		2212
		2213	=head3 Watcher-Specific Functions and Data Members
1004		2214
1005	=over 4	2215	=over 4
1006		2216
1007	=item ev_signal_init (ev_signal *, callback, int signum)	2217	=item ev_signal_init (ev_signal *, callback, int signum)
1008		2218
1009	=item ev_signal_set (ev_signal *, int signum)	2219	=item ev_signal_set (ev_signal *, int signum)
1010		2220
1011	Configures the watcher to trigger on the given signal number (usually one	2221	Configures the watcher to trigger on the given signal number (usually one
1012	of the C<SIGxxx> constants).	2222	of the C<SIGxxx> constants).
1013		2223
		2224	=item int signum [read-only]
		2225
		2226	The signal the watcher watches out for.
		2227
1014	=back	2228	=back
1015		2229
		2230	=head3 Examples
1016		2231
		2232	Example: Try to exit cleanly on SIGINT.
		2233
		2234	static void
		2235	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		2236	{
		2237	ev_unloop (loop, EVUNLOOP_ALL);
		2238	}
		2239
		2240	ev_signal signal_watcher;
		2241	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		2242	ev_signal_start (loop, &signal_watcher);
		2243
		2244
1017	=head2 C<ev_child> - wait for pid status changes	2245	=head2 C<ev_child> - watch out for process status changes
1018		2246
1019	Child watchers trigger when your process receives a SIGCHLD in response to	2247	Child watchers trigger when your process receives a SIGCHLD in response to
1020	some child status changes (most typically when a child of yours dies).	2248	some child status changes (most typically when a child of yours dies or
		2249	exits). It is permissible to install a child watcher I<after> the child
		2250	has been forked (which implies it might have already exited), as long
		2251	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2252	forking and then immediately registering a watcher for the child is fine,
		2253	but forking and registering a watcher a few event loop iterations later or
		2254	in the next callback invocation is not.
		2255
		2256	Only the default event loop is capable of handling signals, and therefore
		2257	you can only register child watchers in the default event loop.
		2258
		2259	Due to some design glitches inside libev, child watchers will always be
		2260	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2261	libev)
		2262
		2263	=head3 Process Interaction
		2264
		2265	Libev grabs C<SIGCHLD> as soon as the default event loop is
		2266	initialised. This is necessary to guarantee proper behaviour even if the
		2267	first child watcher is started after the child exits. The occurrence
		2268	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		2269	synchronously as part of the event loop processing. Libev always reaps all
		2270	children, even ones not watched.
		2271
		2272	=head3 Overriding the Built-In Processing
		2273
		2274	Libev offers no special support for overriding the built-in child
		2275	processing, but if your application collides with libev's default child
		2276	handler, you can override it easily by installing your own handler for
		2277	C<SIGCHLD> after initialising the default loop, and making sure the
		2278	default loop never gets destroyed. You are encouraged, however, to use an
		2279	event-based approach to child reaping and thus use libev's support for
		2280	that, so other libev users can use C<ev_child> watchers freely.
		2281
		2282	=head3 Stopping the Child Watcher
		2283
		2284	Currently, the child watcher never gets stopped, even when the
		2285	child terminates, so normally one needs to stop the watcher in the
		2286	callback. Future versions of libev might stop the watcher automatically
		2287	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2288	problem).
		2289
		2290	=head3 Watcher-Specific Functions and Data Members
1021		2291
1022	=over 4	2292	=over 4
1023		2293
1024	=item ev_child_init (ev_child *, callback, int pid)	2294	=item ev_child_init (ev_child *, callback, int pid, int trace)
1025		2295
1026	=item ev_child_set (ev_child *, int pid)	2296	=item ev_child_set (ev_child *, int pid, int trace)
1027		2297
1028	Configures the watcher to wait for status changes of process C<pid> (or	2298	Configures the watcher to wait for status changes of process C<pid> (or
1029	I<any> process if C<pid> is specified as C<0>). The callback can look	2299	I<any> process if C<pid> is specified as C<0>). The callback can look
1030	at the C<rstatus> member of the C<ev_child> watcher structure to see	2300	at the C<rstatus> member of the C<ev_child> watcher structure to see
1031	the status word (use the macros from C<sys/wait.h> and see your systems	2301	the status word (use the macros from C<sys/wait.h> and see your systems
1032	C<waitpid> documentation). The C<rpid> member contains the pid of the	2302	C<waitpid> documentation). The C<rpid> member contains the pid of the
1033	process causing the status change.	2303	process causing the status change. C<trace> must be either C<0> (only
		2304	activate the watcher when the process terminates) or C<1> (additionally
		2305	activate the watcher when the process is stopped or continued).
		2306
		2307	=item int pid [read-only]
		2308
		2309	The process id this watcher watches out for, or C<0>, meaning any process id.
		2310
		2311	=item int rpid [read-write]
		2312
		2313	The process id that detected a status change.
		2314
		2315	=item int rstatus [read-write]
		2316
		2317	The process exit/trace status caused by C<rpid> (see your systems
		2318	C<waitpid> and C<sys/wait.h> documentation for details).
1034		2319
1035	=back	2320	=back
1036		2321
1037	Example: try to exit cleanly on SIGINT and SIGTERM.	2322	=head3 Examples
1038		2323
		2324	Example: C<fork()> a new process and install a child handler to wait for
		2325	its completion.
		2326
		2327	ev_child cw;
		2328
1039	static void	2329	static void
1040	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2330	child_cb (EV_P_ ev_child *w, int revents)
1041	{	2331	{
1042	ev_unloop (loop, EVUNLOOP_ALL);	2332	ev_child_stop (EV_A_ w);
		2333	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1043	}	2334	}
1044		2335
1045	struct ev_signal signal_watcher;	2336	pid_t pid = fork ();
1046	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1047	ev_signal_start (loop, &sigint_cb);
1048		2337
		2338	if (pid < 0)
		2339	// error
		2340	else if (pid == 0)
		2341	{
		2342	// the forked child executes here
		2343	exit (1);
		2344	}
		2345	else
		2346	{
		2347	ev_child_init (&cw, child_cb, pid, 0);
		2348	ev_child_start (EV_DEFAULT_ &cw);
		2349	}
1049		2350
		2351
		2352	=head2 C<ev_stat> - did the file attributes just change?
		2353
		2354	This watches a file system path for attribute changes. That is, it calls
		2355	C<stat> on that path in regular intervals (or when the OS says it changed)
		2356	and sees if it changed compared to the last time, invoking the callback if
		2357	it did.
		2358
		2359	The path does not need to exist: changing from "path exists" to "path does
		2360	not exist" is a status change like any other. The condition "path does not
		2361	exist" (or more correctly "path cannot be stat'ed") is signified by the
		2362	C<st_nlink> field being zero (which is otherwise always forced to be at
		2363	least one) and all the other fields of the stat buffer having unspecified
		2364	contents.
		2365
		2366	The path I<must not> end in a slash or contain special components such as
		2367	C<.> or C<..>. The path I<should> be absolute: If it is relative and
		2368	your working directory changes, then the behaviour is undefined.
		2369
		2370	Since there is no portable change notification interface available, the
		2371	portable implementation simply calls C<stat(2)> regularly on the path
		2372	to see if it changed somehow. You can specify a recommended polling
		2373	interval for this case. If you specify a polling interval of C<0> (highly
		2374	recommended!) then a I<suitable, unspecified default> value will be used
		2375	(which you can expect to be around five seconds, although this might
		2376	change dynamically). Libev will also impose a minimum interval which is
		2377	currently around C<0.1>, but that's usually overkill.
		2378
		2379	This watcher type is not meant for massive numbers of stat watchers,
		2380	as even with OS-supported change notifications, this can be
		2381	resource-intensive.
		2382
		2383	At the time of this writing, the only OS-specific interface implemented
		2384	is the Linux inotify interface (implementing kqueue support is left as an
		2385	exercise for the reader. Note, however, that the author sees no way of
		2386	implementing C<ev_stat> semantics with kqueue, except as a hint).
		2387
		2388	=head3 ABI Issues (Largefile Support)
		2389
		2390	Libev by default (unless the user overrides this) uses the default
		2391	compilation environment, which means that on systems with large file
		2392	support disabled by default, you get the 32 bit version of the stat
		2393	structure. When using the library from programs that change the ABI to
		2394	use 64 bit file offsets the programs will fail. In that case you have to
		2395	compile libev with the same flags to get binary compatibility. This is
		2396	obviously the case with any flags that change the ABI, but the problem is
		2397	most noticeably displayed with ev_stat and large file support.
		2398
		2399	The solution for this is to lobby your distribution maker to make large
		2400	file interfaces available by default (as e.g. FreeBSD does) and not
		2401	optional. Libev cannot simply switch on large file support because it has
		2402	to exchange stat structures with application programs compiled using the
		2403	default compilation environment.
		2404
		2405	=head3 Inotify and Kqueue
		2406
		2407	When C<inotify (7)> support has been compiled into libev and present at
		2408	runtime, it will be used to speed up change detection where possible. The
		2409	inotify descriptor will be created lazily when the first C<ev_stat>
		2410	watcher is being started.
		2411
		2412	Inotify presence does not change the semantics of C<ev_stat> watchers
		2413	except that changes might be detected earlier, and in some cases, to avoid
		2414	making regular C<stat> calls. Even in the presence of inotify support
		2415	there are many cases where libev has to resort to regular C<stat> polling,
		2416	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2417	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2418	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2419	xfs are fully working) libev usually gets away without polling.
		2420
		2421	There is no support for kqueue, as apparently it cannot be used to
		2422	implement this functionality, due to the requirement of having a file
		2423	descriptor open on the object at all times, and detecting renames, unlinks
		2424	etc. is difficult.
		2425
		2426	=head3 C<stat ()> is a synchronous operation
		2427
		2428	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2429	the process. The exception are C<ev_stat> watchers - those call C<stat
		2430	()>, which is a synchronous operation.
		2431
		2432	For local paths, this usually doesn't matter: unless the system is very
		2433	busy or the intervals between stat's are large, a stat call will be fast,
		2434	as the path data is usually in memory already (except when starting the
		2435	watcher).
		2436
		2437	For networked file systems, calling C<stat ()> can block an indefinite
		2438	time due to network issues, and even under good conditions, a stat call
		2439	often takes multiple milliseconds.
		2440
		2441	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2442	paths, although this is fully supported by libev.
		2443
		2444	=head3 The special problem of stat time resolution
		2445
		2446	The C<stat ()> system call only supports full-second resolution portably,
		2447	and even on systems where the resolution is higher, most file systems
		2448	still only support whole seconds.
		2449
		2450	That means that, if the time is the only thing that changes, you can
		2451	easily miss updates: on the first update, C<ev_stat> detects a change and
		2452	calls your callback, which does something. When there is another update
		2453	within the same second, C<ev_stat> will be unable to detect unless the
		2454	stat data does change in other ways (e.g. file size).
		2455
		2456	The solution to this is to delay acting on a change for slightly more
		2457	than a second (or till slightly after the next full second boundary), using
		2458	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		2459	ev_timer_again (loop, w)>).
		2460
		2461	The C<.02> offset is added to work around small timing inconsistencies
		2462	of some operating systems (where the second counter of the current time
		2463	might be be delayed. One such system is the Linux kernel, where a call to
		2464	C<gettimeofday> might return a timestamp with a full second later than
		2465	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2466	update file times then there will be a small window where the kernel uses
		2467	the previous second to update file times but libev might already execute
		2468	the timer callback).
		2469
		2470	=head3 Watcher-Specific Functions and Data Members
		2471
		2472	=over 4
		2473
		2474	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		2475
		2476	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		2477
		2478	Configures the watcher to wait for status changes of the given
		2479	C<path>. The C<interval> is a hint on how quickly a change is expected to
		2480	be detected and should normally be specified as C<0> to let libev choose
		2481	a suitable value. The memory pointed to by C<path> must point to the same
		2482	path for as long as the watcher is active.
		2483
		2484	The callback will receive an C<EV_STAT> event when a change was detected,
		2485	relative to the attributes at the time the watcher was started (or the
		2486	last change was detected).
		2487
		2488	=item ev_stat_stat (loop, ev_stat *)
		2489
		2490	Updates the stat buffer immediately with new values. If you change the
		2491	watched path in your callback, you could call this function to avoid
		2492	detecting this change (while introducing a race condition if you are not
		2493	the only one changing the path). Can also be useful simply to find out the
		2494	new values.
		2495
		2496	=item ev_statdata attr [read-only]
		2497
		2498	The most-recently detected attributes of the file. Although the type is
		2499	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		2500	suitable for your system, but you can only rely on the POSIX-standardised
		2501	members to be present. If the C<st_nlink> member is C<0>, then there was
		2502	some error while C<stat>ing the file.
		2503
		2504	=item ev_statdata prev [read-only]
		2505
		2506	The previous attributes of the file. The callback gets invoked whenever
		2507	C<prev> != C<attr>, or, more precisely, one or more of these members
		2508	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2509	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
		2510
		2511	=item ev_tstamp interval [read-only]
		2512
		2513	The specified interval.
		2514
		2515	=item const char *path [read-only]
		2516
		2517	The file system path that is being watched.
		2518
		2519	=back
		2520
		2521	=head3 Examples
		2522
		2523	Example: Watch C</etc/passwd> for attribute changes.
		2524
		2525	static void
		2526	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		2527	{
		2528	/* /etc/passwd changed in some way */
		2529	if (w->attr.st_nlink)
		2530	{
		2531	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		2532	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		2533	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		2534	}
		2535	else
		2536	/* you shalt not abuse printf for puts */
		2537	puts ("wow, /etc/passwd is not there, expect problems. "
		2538	"if this is windows, they already arrived\n");
		2539	}
		2540
		2541	...
		2542	ev_stat passwd;
		2543
		2544	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
		2545	ev_stat_start (loop, &passwd);
		2546
		2547	Example: Like above, but additionally use a one-second delay so we do not
		2548	miss updates (however, frequent updates will delay processing, too, so
		2549	one might do the work both on C<ev_stat> callback invocation I<and> on
		2550	C<ev_timer> callback invocation).
		2551
		2552	static ev_stat passwd;
		2553	static ev_timer timer;
		2554
		2555	static void
		2556	timer_cb (EV_P_ ev_timer *w, int revents)
		2557	{
		2558	ev_timer_stop (EV_A_ w);
		2559
		2560	/* now it's one second after the most recent passwd change */
		2561	}
		2562
		2563	static void
		2564	stat_cb (EV_P_ ev_stat *w, int revents)
		2565	{
		2566	/* reset the one-second timer */
		2567	ev_timer_again (EV_A_ &timer);
		2568	}
		2569
		2570	...
		2571	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		2572	ev_stat_start (loop, &passwd);
		2573	ev_timer_init (&timer, timer_cb, 0., 1.02);
		2574
		2575
1050	=head2 C<ev_idle> - when you've got nothing better to do	2576	=head2 C<ev_idle> - when you've got nothing better to do...
1051		2577
1052	Idle watchers trigger events when there are no other events are pending	2578	Idle watchers trigger events when no other events of the same or higher
1053	(prepare, check and other idle watchers do not count). That is, as long	2579	priority are pending (prepare, check and other idle watchers do not count
1054	as your process is busy handling sockets or timeouts (or even signals,	2580	as receiving "events").
1055	imagine) it will not be triggered. But when your process is idle all idle	2581
1056	watchers are being called again and again, once per event loop iteration -	2582	That is, as long as your process is busy handling sockets or timeouts
		2583	(or even signals, imagine) of the same or higher priority it will not be
		2584	triggered. But when your process is idle (or only lower-priority watchers
		2585	are pending), the idle watchers are being called once per event loop
1057	until stopped, that is, or your process receives more events and becomes	2586	iteration - until stopped, that is, or your process receives more events
1058	busy.	2587	and becomes busy again with higher priority stuff.
1059		2588
1060	The most noteworthy effect is that as long as any idle watchers are	2589	The most noteworthy effect is that as long as any idle watchers are
1061	active, the process will not block when waiting for new events.	2590	active, the process will not block when waiting for new events.
1062		2591
1063	Apart from keeping your process non-blocking (which is a useful	2592	Apart from keeping your process non-blocking (which is a useful
1064	effect on its own sometimes), idle watchers are a good place to do	2593	effect on its own sometimes), idle watchers are a good place to do
1065	"pseudo-background processing", or delay processing stuff to after the	2594	"pseudo-background processing", or delay processing stuff to after the
1066	event loop has handled all outstanding events.	2595	event loop has handled all outstanding events.
1067		2596
		2597	=head3 Watcher-Specific Functions and Data Members
		2598
1068	=over 4	2599	=over 4
1069		2600
1070	=item ev_idle_init (ev_signal *, callback)	2601	=item ev_idle_init (ev_idle *, callback)
1071		2602
1072	Initialises and configures the idle watcher - it has no parameters of any	2603	Initialises and configures the idle watcher - it has no parameters of any
1073	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2604	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1074	believe me.	2605	believe me.
1075		2606
1076	=back	2607	=back
1077		2608
		2609	=head3 Examples
		2610
1078	Example: dynamically allocate an C<ev_idle>, start it, and in the	2611	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1079	callback, free it. Alos, use no error checking, as usual.	2612	callback, free it. Also, use no error checking, as usual.
1080		2613
1081	static void	2614	static void
1082	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2615	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1083	{	2616	{
1084	free (w);	2617	free (w);
1085	// now do something you wanted to do when the program has	2618	// now do something you wanted to do when the program has
1086	// no longer asnything immediate to do.	2619	// no longer anything immediate to do.
1087	}	2620	}
1088		2621
1089	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2622	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1090	ev_idle_init (idle_watcher, idle_cb);	2623	ev_idle_init (idle_watcher, idle_cb);
1091	ev_idle_start (loop, idle_cb);	2624	ev_idle_start (loop, idle_watcher);
1092		2625
1093		2626
1094	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	2627	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1095		2628
1096	Prepare and check watchers are usually (but not always) used in tandem:	2629	Prepare and check watchers are usually (but not always) used in pairs:
1097	prepare watchers get invoked before the process blocks and check watchers	2630	prepare watchers get invoked before the process blocks and check watchers
1098	afterwards.	2631	afterwards.
1099		2632
		2633	You I<must not> call C<ev_loop> or similar functions that enter
		2634	the current event loop from either C<ev_prepare> or C<ev_check>
		2635	watchers. Other loops than the current one are fine, however. The
		2636	rationale behind this is that you do not need to check for recursion in
		2637	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		2638	C<ev_check> so if you have one watcher of each kind they will always be
		2639	called in pairs bracketing the blocking call.
		2640
1100	Their main purpose is to integrate other event mechanisms into libev and	2641	Their main purpose is to integrate other event mechanisms into libev and
1101	their use is somewhat advanced. This could be used, for example, to track	2642	their use is somewhat advanced. They could be used, for example, to track
1102	variable changes, implement your own watchers, integrate net-snmp or a	2643	variable changes, implement your own watchers, integrate net-snmp or a
1103	coroutine library and lots more.	2644	coroutine library and lots more. They are also occasionally useful if
		2645	you cache some data and want to flush it before blocking (for example,
		2646	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		2647	watcher).
1104		2648
1105	This is done by examining in each prepare call which file descriptors need	2649	This is done by examining in each prepare call which file descriptors
1106	to be watched by the other library, registering C<ev_io> watchers for	2650	need to be watched by the other library, registering C<ev_io> watchers
1107	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2651	for them and starting an C<ev_timer> watcher for any timeouts (many
1108	provide just this functionality). Then, in the check watcher you check for	2652	libraries provide exactly this functionality). Then, in the check watcher,
1109	any events that occured (by checking the pending status of all watchers	2653	you check for any events that occurred (by checking the pending status
1110	and stopping them) and call back into the library. The I/O and timer	2654	of all watchers and stopping them) and call back into the library. The
1111	callbacks will never actually be called (but must be valid nevertheless,	2655	I/O and timer callbacks will never actually be called (but must be valid
1112	because you never know, you know?).	2656	nevertheless, because you never know, you know?).
1113		2657
1114	As another example, the Perl Coro module uses these hooks to integrate	2658	As another example, the Perl Coro module uses these hooks to integrate
1115	coroutines into libev programs, by yielding to other active coroutines	2659	coroutines into libev programs, by yielding to other active coroutines
1116	during each prepare and only letting the process block if no coroutines	2660	during each prepare and only letting the process block if no coroutines
1117	are ready to run (it's actually more complicated: it only runs coroutines	2661	are ready to run (it's actually more complicated: it only runs coroutines
1118	with priority higher than or equal to the event loop and one coroutine	2662	with priority higher than or equal to the event loop and one coroutine
1119	of lower priority, but only once, using idle watchers to keep the event	2663	of lower priority, but only once, using idle watchers to keep the event
1120	loop from blocking if lower-priority coroutines are active, thus mapping	2664	loop from blocking if lower-priority coroutines are active, thus mapping
1121	low-priority coroutines to idle/background tasks).	2665	low-priority coroutines to idle/background tasks).
1122		2666
		2667	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		2668	priority, to ensure that they are being run before any other watchers
		2669	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2670
		2671	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
		2672	activate ("feed") events into libev. While libev fully supports this, they
		2673	might get executed before other C<ev_check> watchers did their job. As
		2674	C<ev_check> watchers are often used to embed other (non-libev) event
		2675	loops those other event loops might be in an unusable state until their
		2676	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		2677	others).
		2678
		2679	=head3 Watcher-Specific Functions and Data Members
		2680
1123	=over 4	2681	=over 4
1124		2682
1125	=item ev_prepare_init (ev_prepare *, callback)	2683	=item ev_prepare_init (ev_prepare *, callback)
1126		2684
1127	=item ev_check_init (ev_check *, callback)	2685	=item ev_check_init (ev_check *, callback)
1128		2686
1129	Initialises and configures the prepare or check watcher - they have no	2687	Initialises and configures the prepare or check watcher - they have no
1130	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2688	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1131	macros, but using them is utterly, utterly and completely pointless.	2689	macros, but using them is utterly, utterly, utterly and completely
		2690	pointless.
1132		2691
1133	=back	2692	=back
1134		2693
1135	Example: *TODO*.	2694	=head3 Examples
1136		2695
		2696	There are a number of principal ways to embed other event loops or modules
		2697	into libev. Here are some ideas on how to include libadns into libev
		2698	(there is a Perl module named C<EV::ADNS> that does this, which you could
		2699	use as a working example. Another Perl module named C<EV::Glib> embeds a
		2700	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		2701	Glib event loop).
1137		2702
		2703	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		2704	and in a check watcher, destroy them and call into libadns. What follows
		2705	is pseudo-code only of course. This requires you to either use a low
		2706	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		2707	the callbacks for the IO/timeout watchers might not have been called yet.
		2708
		2709	static ev_io iow [nfd];
		2710	static ev_timer tw;
		2711
		2712	static void
		2713	io_cb (struct ev_loop *loop, ev_io *w, int revents)
		2714	{
		2715	}
		2716
		2717	// create io watchers for each fd and a timer before blocking
		2718	static void
		2719	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
		2720	{
		2721	int timeout = 3600000;
		2722	struct pollfd fds [nfd];
		2723	// actual code will need to loop here and realloc etc.
		2724	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		2725
		2726	/* the callback is illegal, but won't be called as we stop during check */
		2727	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
		2728	ev_timer_start (loop, &tw);
		2729
		2730	// create one ev_io per pollfd
		2731	for (int i = 0; i < nfd; ++i)
		2732	{
		2733	ev_io_init (iow + i, io_cb, fds [i].fd,
		2734	((fds [i].events & POLLIN ? EV_READ : 0)
		2735	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		2736
		2737	fds [i].revents = 0;
		2738	ev_io_start (loop, iow + i);
		2739	}
		2740	}
		2741
		2742	// stop all watchers after blocking
		2743	static void
		2744	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
		2745	{
		2746	ev_timer_stop (loop, &tw);
		2747
		2748	for (int i = 0; i < nfd; ++i)
		2749	{
		2750	// set the relevant poll flags
		2751	// could also call adns_processreadable etc. here
		2752	struct pollfd *fd = fds + i;
		2753	int revents = ev_clear_pending (iow + i);
		2754	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		2755	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		2756
		2757	// now stop the watcher
		2758	ev_io_stop (loop, iow + i);
		2759	}
		2760
		2761	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		2762	}
		2763
		2764	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		2765	in the prepare watcher and would dispose of the check watcher.
		2766
		2767	Method 3: If the module to be embedded supports explicit event
		2768	notification (libadns does), you can also make use of the actual watcher
		2769	callbacks, and only destroy/create the watchers in the prepare watcher.
		2770
		2771	static void
		2772	timer_cb (EV_P_ ev_timer *w, int revents)
		2773	{
		2774	adns_state ads = (adns_state)w->data;
		2775	update_now (EV_A);
		2776
		2777	adns_processtimeouts (ads, &tv_now);
		2778	}
		2779
		2780	static void
		2781	io_cb (EV_P_ ev_io *w, int revents)
		2782	{
		2783	adns_state ads = (adns_state)w->data;
		2784	update_now (EV_A);
		2785
		2786	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		2787	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		2788	}
		2789
		2790	// do not ever call adns_afterpoll
		2791
		2792	Method 4: Do not use a prepare or check watcher because the module you
		2793	want to embed is not flexible enough to support it. Instead, you can
		2794	override their poll function. The drawback with this solution is that the
		2795	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
		2796	this approach, effectively embedding EV as a client into the horrible
		2797	libglib event loop.
		2798
		2799	static gint
		2800	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		2801	{
		2802	int got_events = 0;
		2803
		2804	for (n = 0; n < nfds; ++n)
		2805	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		2806
		2807	if (timeout >= 0)
		2808	// create/start timer
		2809
		2810	// poll
		2811	ev_loop (EV_A_ 0);
		2812
		2813	// stop timer again
		2814	if (timeout >= 0)
		2815	ev_timer_stop (EV_A_ &to);
		2816
		2817	// stop io watchers again - their callbacks should have set
		2818	for (n = 0; n < nfds; ++n)
		2819	ev_io_stop (EV_A_ iow [n]);
		2820
		2821	return got_events;
		2822	}
		2823
		2824
1138	=head2 C<ev_embed> - when one backend isn't enough	2825	=head2 C<ev_embed> - when one backend isn't enough...
1139		2826
1140	This is a rather advanced watcher type that lets you embed one event loop	2827	This is a rather advanced watcher type that lets you embed one event loop
1141	into another (currently only C<ev_io> events are supported in the embedded	2828	into another (currently only C<ev_io> events are supported in the embedded
1142	loop, other types of watchers might be handled in a delayed or incorrect	2829	loop, other types of watchers might be handled in a delayed or incorrect
1143	fashion and must not be used).	2830	fashion and must not be used).
…		…
1146	prioritise I/O.	2833	prioritise I/O.
1147		2834
1148	As an example for a bug workaround, the kqueue backend might only support	2835	As an example for a bug workaround, the kqueue backend might only support
1149	sockets on some platform, so it is unusable as generic backend, but you	2836	sockets on some platform, so it is unusable as generic backend, but you
1150	still want to make use of it because you have many sockets and it scales	2837	still want to make use of it because you have many sockets and it scales
1151	so nicely. In this case, you would create a kqueue-based loop and embed it	2838	so nicely. In this case, you would create a kqueue-based loop and embed
1152	into your default loop (which might use e.g. poll). Overall operation will	2839	it into your default loop (which might use e.g. poll). Overall operation
1153	be a bit slower because first libev has to poll and then call kevent, but	2840	will be a bit slower because first libev has to call C<poll> and then
1154	at least you can use both at what they are best.	2841	C<kevent>, but at least you can use both mechanisms for what they are
		2842	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1155		2843
1156	As for prioritising I/O: rarely you have the case where some fds have	2844	As for prioritising I/O: under rare circumstances you have the case where
1157	to be watched and handled very quickly (with low latency), and even	2845	some fds have to be watched and handled very quickly (with low latency),
1158	priorities and idle watchers might have too much overhead. In this case	2846	and even priorities and idle watchers might have too much overhead. In
1159	you would put all the high priority stuff in one loop and all the rest in	2847	this case you would put all the high priority stuff in one loop and all
1160	a second one, and embed the second one in the first.	2848	the rest in a second one, and embed the second one in the first.
1161		2849
1162	As long as the watcher is active, the callback will be invoked every time	2850	As long as the watcher is active, the callback will be invoked every
1163	there might be events pending in the embedded loop. The callback must then	2851	time there might be events pending in the embedded loop. The callback
1164	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2852	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
1165	their callbacks (you could also start an idle watcher to give the embedded	2853	sweep and invoke their callbacks (the callback doesn't need to invoke the
1166	loop strictly lower priority for example). You can also set the callback	2854	C<ev_embed_sweep> function directly, it could also start an idle watcher
1167	to C<0>, in which case the embed watcher will automatically execute the	2855	to give the embedded loop strictly lower priority for example).
1168	embedded loop sweep.
1169		2856
1170	As long as the watcher is started it will automatically handle events. The	2857	You can also set the callback to C<0>, in which case the embed watcher
1171	callback will be invoked whenever some events have been handled. You can	2858	will automatically execute the embedded loop sweep whenever necessary.
1172	set the callback to C<0> to avoid having to specify one if you are not
1173	interested in that.
1174		2859
1175	Also, there have not currently been made special provisions for forking:	2860	Fork detection will be handled transparently while the C<ev_embed> watcher
1176	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2861	is active, i.e., the embedded loop will automatically be forked when the
1177	but you will also have to stop and restart any C<ev_embed> watchers	2862	embedding loop forks. In other cases, the user is responsible for calling
1178	yourself.	2863	C<ev_loop_fork> on the embedded loop.
1179		2864
1180	Unfortunately, not all backends are embeddable, only the ones returned by	2865	Unfortunately, not all backends are embeddable: only the ones returned by
1181	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2866	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1182	portable one.	2867	portable one.
1183		2868
1184	So when you want to use this feature you will always have to be prepared	2869	So when you want to use this feature you will always have to be prepared
1185	that you cannot get an embeddable loop. The recommended way to get around	2870	that you cannot get an embeddable loop. The recommended way to get around
1186	this is to have a separate variables for your embeddable loop, try to	2871	this is to have a separate variables for your embeddable loop, try to
1187	create it, and if that fails, use the normal loop for everything:	2872	create it, and if that fails, use the normal loop for everything.
1188		2873
1189	struct ev_loop *loop_hi = ev_default_init (0);	2874	=head3 C<ev_embed> and fork
1190	struct ev_loop *loop_lo = 0;
1191	struct ev_embed embed;
1192
1193	// see if there is a chance of getting one that works
1194	// (remember that a flags value of 0 means autodetection)
1195	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1196	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1197	: 0;
1198		2875
1199	// if we got one, then embed it, otherwise default to loop_hi	2876	While the C<ev_embed> watcher is running, forks in the embedding loop will
1200	if (loop_lo)	2877	automatically be applied to the embedded loop as well, so no special
1201	{	2878	fork handling is required in that case. When the watcher is not running,
1202	ev_embed_init (&embed, 0, loop_lo);	2879	however, it is still the task of the libev user to call C<ev_loop_fork ()>
1203	ev_embed_start (loop_hi, &embed);	2880	as applicable.
1204	}	2881
1205	else	2882	=head3 Watcher-Specific Functions and Data Members
1206	loop_lo = loop_hi;
1207		2883
1208	=over 4	2884	=over 4
1209		2885
1210	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2886	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
1211		2887
…		…
1213		2889
1214	Configures the watcher to embed the given loop, which must be	2890	Configures the watcher to embed the given loop, which must be
1215	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	2891	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1216	invoked automatically, otherwise it is the responsibility of the callback	2892	invoked automatically, otherwise it is the responsibility of the callback
1217	to invoke it (it will continue to be called until the sweep has been done,	2893	to invoke it (it will continue to be called until the sweep has been done,
1218	if you do not want thta, you need to temporarily stop the embed watcher).	2894	if you do not want that, you need to temporarily stop the embed watcher).
1219		2895
1220	=item ev_embed_sweep (loop, ev_embed *)	2896	=item ev_embed_sweep (loop, ev_embed *)
1221		2897
1222	Make a single, non-blocking sweep over the embedded loop. This works	2898	Make a single, non-blocking sweep over the embedded loop. This works
1223	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	2899	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1224	apropriate way for embedded loops.	2900	appropriate way for embedded loops.
		2901
		2902	=item struct ev_loop *other [read-only]
		2903
		2904	The embedded event loop.
1225		2905
1226	=back	2906	=back
1227		2907
		2908	=head3 Examples
		2909
		2910	Example: Try to get an embeddable event loop and embed it into the default
		2911	event loop. If that is not possible, use the default loop. The default
		2912	loop is stored in C<loop_hi>, while the embeddable loop is stored in
		2913	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
		2914	used).
		2915
		2916	struct ev_loop *loop_hi = ev_default_init (0);
		2917	struct ev_loop *loop_lo = 0;
		2918	ev_embed embed;
		2919
		2920	// see if there is a chance of getting one that works
		2921	// (remember that a flags value of 0 means autodetection)
		2922	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		2923	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		2924	: 0;
		2925
		2926	// if we got one, then embed it, otherwise default to loop_hi
		2927	if (loop_lo)
		2928	{
		2929	ev_embed_init (&embed, 0, loop_lo);
		2930	ev_embed_start (loop_hi, &embed);
		2931	}
		2932	else
		2933	loop_lo = loop_hi;
		2934
		2935	Example: Check if kqueue is available but not recommended and create
		2936	a kqueue backend for use with sockets (which usually work with any
		2937	kqueue implementation). Store the kqueue/socket-only event loop in
		2938	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		2939
		2940	struct ev_loop *loop = ev_default_init (0);
		2941	struct ev_loop *loop_socket = 0;
		2942	ev_embed embed;
		2943
		2944	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2945	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2946	{
		2947	ev_embed_init (&embed, 0, loop_socket);
		2948	ev_embed_start (loop, &embed);
		2949	}
		2950
		2951	if (!loop_socket)
		2952	loop_socket = loop;
		2953
		2954	// now use loop_socket for all sockets, and loop for everything else
		2955
		2956
		2957	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		2958
		2959	Fork watchers are called when a C<fork ()> was detected (usually because
		2960	whoever is a good citizen cared to tell libev about it by calling
		2961	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		2962	event loop blocks next and before C<ev_check> watchers are being called,
		2963	and only in the child after the fork. If whoever good citizen calling
		2964	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		2965	handlers will be invoked, too, of course.
		2966
		2967	=head3 The special problem of life after fork - how is it possible?
		2968
		2969	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2970	up/change the process environment, followed by a call to C<exec()>. This
		2971	sequence should be handled by libev without any problems.
		2972
		2973	This changes when the application actually wants to do event handling
		2974	in the child, or both parent in child, in effect "continuing" after the
		2975	fork.
		2976
		2977	The default mode of operation (for libev, with application help to detect
		2978	forks) is to duplicate all the state in the child, as would be expected
		2979	when I<either> the parent I<or> the child process continues.
		2980
		2981	When both processes want to continue using libev, then this is usually the
		2982	wrong result. In that case, usually one process (typically the parent) is
		2983	supposed to continue with all watchers in place as before, while the other
		2984	process typically wants to start fresh, i.e. without any active watchers.
		2985
		2986	The cleanest and most efficient way to achieve that with libev is to
		2987	simply create a new event loop, which of course will be "empty", and
		2988	use that for new watchers. This has the advantage of not touching more
		2989	memory than necessary, and thus avoiding the copy-on-write, and the
		2990	disadvantage of having to use multiple event loops (which do not support
		2991	signal watchers).
		2992
		2993	When this is not possible, or you want to use the default loop for
		2994	other reasons, then in the process that wants to start "fresh", call
		2995	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2996	the default loop will "orphan" (not stop) all registered watchers, so you
		2997	have to be careful not to execute code that modifies those watchers. Note
		2998	also that in that case, you have to re-register any signal watchers.
		2999
		3000	=head3 Watcher-Specific Functions and Data Members
		3001
		3002	=over 4
		3003
		3004	=item ev_fork_init (ev_signal *, callback)
		3005
		3006	Initialises and configures the fork watcher - it has no parameters of any
		3007	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		3008	believe me.
		3009
		3010	=back
		3011
		3012
		3013	=head2 C<ev_async> - how to wake up another event loop
		3014
		3015	In general, you cannot use an C<ev_loop> from multiple threads or other
		3016	asynchronous sources such as signal handlers (as opposed to multiple event
		3017	loops - those are of course safe to use in different threads).
		3018
		3019	Sometimes, however, you need to wake up another event loop you do not
		3020	control, for example because it belongs to another thread. This is what
		3021	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		3022	can signal it by calling C<ev_async_send>, which is thread- and signal
		3023	safe.
		3024
		3025	This functionality is very similar to C<ev_signal> watchers, as signals,
		3026	too, are asynchronous in nature, and signals, too, will be compressed
		3027	(i.e. the number of callback invocations may be less than the number of
		3028	C<ev_async_sent> calls).
		3029
		3030	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		3031	just the default loop.
		3032
		3033	=head3 Queueing
		3034
		3035	C<ev_async> does not support queueing of data in any way. The reason
		3036	is that the author does not know of a simple (or any) algorithm for a
		3037	multiple-writer-single-reader queue that works in all cases and doesn't
		3038	need elaborate support such as pthreads or unportable memory access
		3039	semantics.
		3040
		3041	That means that if you want to queue data, you have to provide your own
		3042	queue. But at least I can tell you how to implement locking around your
		3043	queue:
		3044
		3045	=over 4
		3046
		3047	=item queueing from a signal handler context
		3048
		3049	To implement race-free queueing, you simply add to the queue in the signal
		3050	handler but you block the signal handler in the watcher callback. Here is
		3051	an example that does that for some fictitious SIGUSR1 handler:
		3052
		3053	static ev_async mysig;
		3054
		3055	static void
		3056	sigusr1_handler (void)
		3057	{
		3058	sometype data;
		3059
		3060	// no locking etc.
		3061	queue_put (data);
		3062	ev_async_send (EV_DEFAULT_ &mysig);
		3063	}
		3064
		3065	static void
		3066	mysig_cb (EV_P_ ev_async *w, int revents)
		3067	{
		3068	sometype data;
		3069	sigset_t block, prev;
		3070
		3071	sigemptyset (&block);
		3072	sigaddset (&block, SIGUSR1);
		3073	sigprocmask (SIG_BLOCK, &block, &prev);
		3074
		3075	while (queue_get (&data))
		3076	process (data);
		3077
		3078	if (sigismember (&prev, SIGUSR1)
		3079	sigprocmask (SIG_UNBLOCK, &block, 0);
		3080	}
		3081
		3082	(Note: pthreads in theory requires you to use C<pthread_setmask>
		3083	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		3084	either...).
		3085
		3086	=item queueing from a thread context
		3087
		3088	The strategy for threads is different, as you cannot (easily) block
		3089	threads but you can easily preempt them, so to queue safely you need to
		3090	employ a traditional mutex lock, such as in this pthread example:
		3091
		3092	static ev_async mysig;
		3093	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		3094
		3095	static void
		3096	otherthread (void)
		3097	{
		3098	// only need to lock the actual queueing operation
		3099	pthread_mutex_lock (&mymutex);
		3100	queue_put (data);
		3101	pthread_mutex_unlock (&mymutex);
		3102
		3103	ev_async_send (EV_DEFAULT_ &mysig);
		3104	}
		3105
		3106	static void
		3107	mysig_cb (EV_P_ ev_async *w, int revents)
		3108	{
		3109	pthread_mutex_lock (&mymutex);
		3110
		3111	while (queue_get (&data))
		3112	process (data);
		3113
		3114	pthread_mutex_unlock (&mymutex);
		3115	}
		3116
		3117	=back
		3118
		3119
		3120	=head3 Watcher-Specific Functions and Data Members
		3121
		3122	=over 4
		3123
		3124	=item ev_async_init (ev_async *, callback)
		3125
		3126	Initialises and configures the async watcher - it has no parameters of any
		3127	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
		3128	trust me.
		3129
		3130	=item ev_async_send (loop, ev_async *)
		3131
		3132	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		3133	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		3134	C<ev_feed_event>, this call is safe to do from other threads, signal or
		3135	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
		3136	section below on what exactly this means).
		3137
		3138	Note that, as with other watchers in libev, multiple events might get
		3139	compressed into a single callback invocation (another way to look at this
		3140	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3141	reset when the event loop detects that).
		3142
		3143	This call incurs the overhead of a system call only once per event loop
		3144	iteration, so while the overhead might be noticeable, it doesn't apply to
		3145	repeated calls to C<ev_async_send> for the same event loop.
		3146
		3147	=item bool = ev_async_pending (ev_async *)
		3148
		3149	Returns a non-zero value when C<ev_async_send> has been called on the
		3150	watcher but the event has not yet been processed (or even noted) by the
		3151	event loop.
		3152
		3153	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		3154	the loop iterates next and checks for the watcher to have become active,
		3155	it will reset the flag again. C<ev_async_pending> can be used to very
		3156	quickly check whether invoking the loop might be a good idea.
		3157
		3158	Not that this does I<not> check whether the watcher itself is pending,
		3159	only whether it has been requested to make this watcher pending: there
		3160	is a time window between the event loop checking and resetting the async
		3161	notification, and the callback being invoked.
		3162
		3163	=back
		3164
1228		3165
1229	=head1 OTHER FUNCTIONS	3166	=head1 OTHER FUNCTIONS
1230		3167
1231	There are some other functions of possible interest. Described. Here. Now.	3168	There are some other functions of possible interest. Described. Here. Now.
1232		3169
1233	=over 4	3170	=over 4
1234		3171
1235	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3172	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
1236		3173
1237	This function combines a simple timer and an I/O watcher, calls your	3174	This function combines a simple timer and an I/O watcher, calls your
1238	callback on whichever event happens first and automatically stop both	3175	callback on whichever event happens first and automatically stops both
1239	watchers. This is useful if you want to wait for a single event on an fd	3176	watchers. This is useful if you want to wait for a single event on an fd
1240	or timeout without having to allocate/configure/start/stop/free one or	3177	or timeout without having to allocate/configure/start/stop/free one or
1241	more watchers yourself.	3178	more watchers yourself.
1242		3179
1243	If C<fd> is less than 0, then no I/O watcher will be started and events	3180	If C<fd> is less than 0, then no I/O watcher will be started and the
1244	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3181	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
1245	C<events> set will be craeted and started.	3182	the given C<fd> and C<events> set will be created and started.
1246		3183
1247	If C<timeout> is less than 0, then no timeout watcher will be	3184	If C<timeout> is less than 0, then no timeout watcher will be
1248	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3185	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
1249	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3186	repeat = 0) will be started. C<0> is a valid timeout.
1250	dubious value.
1251		3187
1252	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3188	The callback has the type C<void (*cb)(int revents, void *arg)> and is
1253	passed an C<revents> set like normal event callbacks (a combination of	3189	passed an C<revents> set like normal event callbacks (a combination of
1254	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3190	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
1255	value passed to C<ev_once>:	3191	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3192	a timeout and an io event at the same time - you probably should give io
		3193	events precedence.
1256		3194
		3195	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3196
1257	static void stdin_ready (int revents, void *arg)	3197	static void stdin_ready (int revents, void *arg)
1258	{	3198	{
1259	if (revents & EV_TIMEOUT)
1260	/* doh, nothing entered */;
1261	else if (revents & EV_READ)	3199	if (revents & EV_READ)
1262	/* stdin might have data for us, joy! */;	3200	/* stdin might have data for us, joy! */;
		3201	else if (revents & EV_TIMER)
		3202	/* doh, nothing entered */;
1263	}	3203	}
1264		3204
1265	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3205	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
1266		3206
1267	=item ev_feed_event (ev_loop *, watcher *, int revents)
1268
1269	Feeds the given event set into the event loop, as if the specified event
1270	had happened for the specified watcher (which must be a pointer to an
1271	initialised but not necessarily started event watcher).
1272
1273	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3207	=item ev_feed_fd_event (loop, int fd, int revents)
1274		3208
1275	Feed an event on the given fd, as if a file descriptor backend detected	3209	Feed an event on the given fd, as if a file descriptor backend detected
1276	the given events it.	3210	the given events it.
1277		3211
1278	=item ev_feed_signal_event (ev_loop *loop, int signum)	3212	=item ev_feed_signal_event (loop, int signum)
1279		3213
1280	Feed an event as if the given signal occured (C<loop> must be the default	3214	Feed an event as if the given signal occurred (C<loop> must be the default
1281	loop!).	3215	loop!).
1282		3216
1283	=back	3217	=back
1284		3218
1285		3219
…		…
1301		3235
1302	=item * Priorities are not currently supported. Initialising priorities	3236	=item * Priorities are not currently supported. Initialising priorities
1303	will fail and all watchers will have the same priority, even though there	3237	will fail and all watchers will have the same priority, even though there
1304	is an ev_pri field.	3238	is an ev_pri field.
1305		3239
		3240	=item * In libevent, the last base created gets the signals, in libev, the
		3241	first base created (== the default loop) gets the signals.
		3242
1306	=item * Other members are not supported.	3243	=item * Other members are not supported.
1307		3244
1308	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3245	=item * The libev emulation is I<not> ABI compatible to libevent, you need
1309	to use the libev header file and library.	3246	to use the libev header file and library.
1310		3247
1311	=back	3248	=back
1312		3249
1313	=head1 C++ SUPPORT	3250	=head1 C++ SUPPORT
1314		3251
1315	Libev comes with some simplistic wrapper classes for C++ that mainly allow	3252	Libev comes with some simplistic wrapper classes for C++ that mainly allow
1316	you to use some convinience methods to start/stop watchers and also change	3253	you to use some convenience methods to start/stop watchers and also change
1317	the callback model to a model using method callbacks on objects.	3254	the callback model to a model using method callbacks on objects.
1318		3255
1319	To use it,	3256	To use it,
1320		3257
1321	#include <ev++.h>	3258	#include <ev++.h>
1322		3259
1323	(it is not installed by default). This automatically includes F<ev.h>	3260	This automatically includes F<ev.h> and puts all of its definitions (many
1324	and puts all of its definitions (many of them macros) into the global	3261	of them macros) into the global namespace. All C++ specific things are
1325	namespace. All C++ specific things are put into the C<ev> namespace.	3262	put into the C<ev> namespace. It should support all the same embedding
		3263	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1326		3264
1327	It should support all the same embedding options as F<ev.h>, most notably	3265	Care has been taken to keep the overhead low. The only data member the C++
1328	C<EV_MULTIPLICITY>.	3266	classes add (compared to plain C-style watchers) is the event loop pointer
		3267	that the watcher is associated with (or no additional members at all if
		3268	you disable C<EV_MULTIPLICITY> when embedding libev).
		3269
		3270	Currently, functions, and static and non-static member functions can be
		3271	used as callbacks. Other types should be easy to add as long as they only
		3272	need one additional pointer for context. If you need support for other
		3273	types of functors please contact the author (preferably after implementing
		3274	it).
1329		3275
1330	Here is a list of things available in the C<ev> namespace:	3276	Here is a list of things available in the C<ev> namespace:
1331		3277
1332	=over 4	3278	=over 4
1333		3279
…		…
1349		3295
1350	All of those classes have these methods:	3296	All of those classes have these methods:
1351		3297
1352	=over 4	3298	=over 4
1353		3299
1354	=item ev::TYPE::TYPE (object *, object::method *)	3300	=item ev::TYPE::TYPE ()
1355		3301
1356	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)	3302	=item ev::TYPE::TYPE (loop)
1357		3303
1358	=item ev::TYPE::~TYPE	3304	=item ev::TYPE::~TYPE
1359		3305
1360	The constructor takes a pointer to an object and a method pointer to	3306	The constructor (optionally) takes an event loop to associate the watcher
1361	the event handler callback to call in this class. The constructor calls	3307	with. If it is omitted, it will use C<EV_DEFAULT>.
1362	C<ev_init> for you, which means you have to call the C<set> method	3308
1363	before starting it. If you do not specify a loop then the constructor	3309	The constructor calls C<ev_init> for you, which means you have to call the
1364	automatically associates the default loop with this watcher.	3310	C<set> method before starting it.
		3311
		3312	It will not set a callback, however: You have to call the templated C<set>
		3313	method to set a callback before you can start the watcher.
		3314
		3315	(The reason why you have to use a method is a limitation in C++ which does
		3316	not allow explicit template arguments for constructors).
1365		3317
1366	The destructor automatically stops the watcher if it is active.	3318	The destructor automatically stops the watcher if it is active.
1367		3319
		3320	=item w->set<class, &class::method> (object *)
		3321
		3322	This method sets the callback method to call. The method has to have a
		3323	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		3324	first argument and the C<revents> as second. The object must be given as
		3325	parameter and is stored in the C<data> member of the watcher.
		3326
		3327	This method synthesizes efficient thunking code to call your method from
		3328	the C callback that libev requires. If your compiler can inline your
		3329	callback (i.e. it is visible to it at the place of the C<set> call and
		3330	your compiler is good :), then the method will be fully inlined into the
		3331	thunking function, making it as fast as a direct C callback.
		3332
		3333	Example: simple class declaration and watcher initialisation
		3334
		3335	struct myclass
		3336	{
		3337	void io_cb (ev::io &w, int revents) { }
		3338	}
		3339
		3340	myclass obj;
		3341	ev::io iow;
		3342	iow.set <myclass, &myclass::io_cb> (&obj);
		3343
		3344	=item w->set (object *)
		3345
		3346	This is an B<experimental> feature that might go away in a future version.
		3347
		3348	This is a variation of a method callback - leaving out the method to call
		3349	will default the method to C<operator ()>, which makes it possible to use
		3350	functor objects without having to manually specify the C<operator ()> all
		3351	the time. Incidentally, you can then also leave out the template argument
		3352	list.
		3353
		3354	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3355	int revents)>.
		3356
		3357	See the method-C<set> above for more details.
		3358
		3359	Example: use a functor object as callback.
		3360
		3361	struct myfunctor
		3362	{
		3363	void operator() (ev::io &w, int revents)
		3364	{
		3365	...
		3366	}
		3367	}
		3368
		3369	myfunctor f;
		3370
		3371	ev::io w;
		3372	w.set (&f);
		3373
		3374	=item w->set<function> (void *data = 0)
		3375
		3376	Also sets a callback, but uses a static method or plain function as
		3377	callback. The optional C<data> argument will be stored in the watcher's
		3378	C<data> member and is free for you to use.
		3379
		3380	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		3381
		3382	See the method-C<set> above for more details.
		3383
		3384	Example: Use a plain function as callback.
		3385
		3386	static void io_cb (ev::io &w, int revents) { }
		3387	iow.set <io_cb> ();
		3388
1368	=item w->set (struct ev_loop *)	3389	=item w->set (loop)
1369		3390
1370	Associates a different C<struct ev_loop> with this watcher. You can only	3391	Associates a different C<struct ev_loop> with this watcher. You can only
1371	do this when the watcher is inactive (and not pending either).	3392	do this when the watcher is inactive (and not pending either).
1372		3393
1373	=item w->set ([args])	3394	=item w->set ([arguments])
1374		3395
1375	Basically the same as C<ev_TYPE_set>, with the same args. Must be	3396	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be
1376	called at least once. Unlike the C counterpart, an active watcher gets	3397	called at least once. Unlike the C counterpart, an active watcher gets
1377	automatically stopped and restarted.	3398	automatically stopped and restarted when reconfiguring it with this
		3399	method.
1378		3400
1379	=item w->start ()	3401	=item w->start ()
1380		3402
1381	Starts the watcher. Note that there is no C<loop> argument as the	3403	Starts the watcher. Note that there is no C<loop> argument, as the
1382	constructor already takes the loop.	3404	constructor already stores the event loop.
1383		3405
1384	=item w->stop ()	3406	=item w->stop ()
1385		3407
1386	Stops the watcher if it is active. Again, no C<loop> argument.	3408	Stops the watcher if it is active. Again, no C<loop> argument.
1387		3409
1388	=item w->again () C<ev::timer>, C<ev::periodic> only	3410	=item w->again () (C<ev::timer>, C<ev::periodic> only)
1389		3411
1390	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding	3412	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
1391	C<ev_TYPE_again> function.	3413	C<ev_TYPE_again> function.
1392		3414
1393	=item w->sweep () C<ev::embed> only	3415	=item w->sweep () (C<ev::embed> only)
1394		3416
1395	Invokes C<ev_embed_sweep>.	3417	Invokes C<ev_embed_sweep>.
		3418
		3419	=item w->update () (C<ev::stat> only)
		3420
		3421	Invokes C<ev_stat_stat>.
1396		3422
1397	=back	3423	=back
1398		3424
1399	=back	3425	=back
1400		3426
1401	Example: Define a class with an IO and idle watcher, start one of them in	3427	Example: Define a class with an IO and idle watcher, start one of them in
1402	the constructor.	3428	the constructor.
1403		3429
1404	class myclass	3430	class myclass
1405	{	3431	{
1406	ev_io io; void io_cb (ev::io &w, int revents);	3432	ev::io io ; void io_cb (ev::io &w, int revents);
1407	ev_idle idle void idle_cb (ev::idle &w, int revents);	3433	ev::idle idle; void idle_cb (ev::idle &w, int revents);
1408		3434
1409	myclass ();	3435	myclass (int fd)
1410	}	3436	{
		3437	io .set <myclass, &myclass::io_cb > (this);
		3438	idle.set <myclass, &myclass::idle_cb> (this);
1411		3439
1412	myclass::myclass (int fd)
1413	: io (this, &myclass::io_cb),
1414	idle (this, &myclass::idle_cb)
1415	{
1416	io.start (fd, ev::READ);	3440	io.start (fd, ev::READ);
		3441	}
		3442	};
		3443
		3444
		3445	=head1 OTHER LANGUAGE BINDINGS
		3446
		3447	Libev does not offer other language bindings itself, but bindings for a
		3448	number of languages exist in the form of third-party packages. If you know
		3449	any interesting language binding in addition to the ones listed here, drop
		3450	me a note.
		3451
		3452	=over 4
		3453
		3454	=item Perl
		3455
		3456	The EV module implements the full libev API and is actually used to test
		3457	libev. EV is developed together with libev. Apart from the EV core module,
		3458	there are additional modules that implement libev-compatible interfaces
		3459	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		3460	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3461	and C<EV::Glib>).
		3462
		3463	It can be found and installed via CPAN, its homepage is at
		3464	L<http://software.schmorp.de/pkg/EV>.
		3465
		3466	=item Python
		3467
		3468	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		3469	seems to be quite complete and well-documented.
		3470
		3471	=item Ruby
		3472
		3473	Tony Arcieri has written a ruby extension that offers access to a subset
		3474	of the libev API and adds file handle abstractions, asynchronous DNS and
		3475	more on top of it. It can be found via gem servers. Its homepage is at
		3476	L<http://rev.rubyforge.org/>.
		3477
		3478	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3479	makes rev work even on mingw.
		3480
		3481	=item Haskell
		3482
		3483	A haskell binding to libev is available at
		3484	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3485
		3486	=item D
		3487
		3488	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		3489	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3490
		3491	=item Ocaml
		3492
		3493	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3494	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3495
		3496	=item Lua
		3497
		3498	Brian Maher has written a partial interface to libev for lua (at the
		3499	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		3500	L<http://github.com/brimworks/lua-ev>.
		3501
		3502	=back
		3503
		3504
		3505	=head1 MACRO MAGIC
		3506
		3507	Libev can be compiled with a variety of options, the most fundamental
		3508	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		3509	functions and callbacks have an initial C<struct ev_loop *> argument.
		3510
		3511	To make it easier to write programs that cope with either variant, the
		3512	following macros are defined:
		3513
		3514	=over 4
		3515
		3516	=item C<EV_A>, C<EV_A_>
		3517
		3518	This provides the loop I<argument> for functions, if one is required ("ev
		3519	loop argument"). The C<EV_A> form is used when this is the sole argument,
		3520	C<EV_A_> is used when other arguments are following. Example:
		3521
		3522	ev_unref (EV_A);
		3523	ev_timer_add (EV_A_ watcher);
		3524	ev_loop (EV_A_ 0);
		3525
		3526	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		3527	which is often provided by the following macro.
		3528
		3529	=item C<EV_P>, C<EV_P_>
		3530
		3531	This provides the loop I<parameter> for functions, if one is required ("ev
		3532	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		3533	C<EV_P_> is used when other parameters are following. Example:
		3534
		3535	// this is how ev_unref is being declared
		3536	static void ev_unref (EV_P);
		3537
		3538	// this is how you can declare your typical callback
		3539	static void cb (EV_P_ ev_timer *w, int revents)
		3540
		3541	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		3542	suitable for use with C<EV_A>.
		3543
		3544	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		3545
		3546	Similar to the other two macros, this gives you the value of the default
		3547	loop, if multiple loops are supported ("ev loop default").
		3548
		3549	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		3550
		3551	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		3552	default loop has been initialised (C<UC> == unchecked). Their behaviour
		3553	is undefined when the default loop has not been initialised by a previous
		3554	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		3555
		3556	It is often prudent to use C<EV_DEFAULT> when initialising the first
		3557	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		3558
		3559	=back
		3560
		3561	Example: Declare and initialise a check watcher, utilising the above
		3562	macros so it will work regardless of whether multiple loops are supported
		3563	or not.
		3564
		3565	static void
		3566	check_cb (EV_P_ ev_timer *w, int revents)
		3567	{
		3568	ev_check_stop (EV_A_ w);
1417	}	3569	}
		3570
		3571	ev_check check;
		3572	ev_check_init (&check, check_cb);
		3573	ev_check_start (EV_DEFAULT_ &check);
		3574	ev_loop (EV_DEFAULT_ 0);
1418		3575
1419	=head1 EMBEDDING	3576	=head1 EMBEDDING
1420		3577
1421	Libev can (and often is) directly embedded into host	3578	Libev can (and often is) directly embedded into host
1422	applications. Examples of applications that embed it include the Deliantra	3579	applications. Examples of applications that embed it include the Deliantra
1423	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)	3580	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
1424	and rxvt-unicode.	3581	and rxvt-unicode.
1425		3582
1426	The goal is to enable you to just copy the neecssary files into your	3583	The goal is to enable you to just copy the necessary files into your
1427	source directory without having to change even a single line in them, so	3584	source directory without having to change even a single line in them, so
1428	you can easily upgrade by simply copying (or having a checked-out copy of	3585	you can easily upgrade by simply copying (or having a checked-out copy of
1429	libev somewhere in your source tree).	3586	libev somewhere in your source tree).
1430		3587
1431	=head2 FILESETS	3588	=head2 FILESETS
1432		3589
1433	Depending on what features you need you need to include one or more sets of files	3590	Depending on what features you need you need to include one or more sets of files
1434	in your app.	3591	in your application.
1435		3592
1436	=head3 CORE EVENT LOOP	3593	=head3 CORE EVENT LOOP
1437		3594
1438	To include only the libev core (all the C<ev_*> functions), with manual	3595	To include only the libev core (all the C<ev_*> functions), with manual
1439	configuration (no autoconf):	3596	configuration (no autoconf):
1440		3597
1441	#define EV_STANDALONE 1	3598	#define EV_STANDALONE 1
1442	#include "ev.c"	3599	#include "ev.c"
1443		3600
1444	This will automatically include F<ev.h>, too, and should be done in a	3601	This will automatically include F<ev.h>, too, and should be done in a
1445	single C source file only to provide the function implementations. To use	3602	single C source file only to provide the function implementations. To use
1446	it, do the same for F<ev.h> in all files wishing to use this API (best	3603	it, do the same for F<ev.h> in all files wishing to use this API (best
1447	done by writing a wrapper around F<ev.h> that you can include instead and	3604	done by writing a wrapper around F<ev.h> that you can include instead and
1448	where you can put other configuration options):	3605	where you can put other configuration options):
1449		3606
1450	#define EV_STANDALONE 1	3607	#define EV_STANDALONE 1
1451	#include "ev.h"	3608	#include "ev.h"
1452		3609
1453	Both header files and implementation files can be compiled with a C++	3610	Both header files and implementation files can be compiled with a C++
1454	compiler (at least, thats a stated goal, and breakage will be treated	3611	compiler (at least, that's a stated goal, and breakage will be treated
1455	as a bug).	3612	as a bug).
1456		3613
1457	You need the following files in your source tree, or in a directory	3614	You need the following files in your source tree, or in a directory
1458	in your include path (e.g. in libev/ when using -Ilibev):	3615	in your include path (e.g. in libev/ when using -Ilibev):
1459		3616
1460	ev.h	3617	ev.h
1461	ev.c	3618	ev.c
1462	ev_vars.h	3619	ev_vars.h
1463	ev_wrap.h	3620	ev_wrap.h
1464		3621
1465	ev_win32.c required on win32 platforms only	3622	ev_win32.c required on win32 platforms only
1466		3623
1467	ev_select.c only when select backend is enabled (which is is by default)	3624	ev_select.c only when select backend is enabled (which is enabled by default)
1468	ev_poll.c only when poll backend is enabled (disabled by default)	3625	ev_poll.c only when poll backend is enabled (disabled by default)
1469	ev_epoll.c only when the epoll backend is enabled (disabled by default)	3626	ev_epoll.c only when the epoll backend is enabled (disabled by default)
1470	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	3627	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1471	ev_port.c only when the solaris port backend is enabled (disabled by default)	3628	ev_port.c only when the solaris port backend is enabled (disabled by default)
1472		3629
1473	F<ev.c> includes the backend files directly when enabled, so you only need	3630	F<ev.c> includes the backend files directly when enabled, so you only need
1474	to compile a single file.	3631	to compile this single file.
1475		3632
1476	=head3 LIBEVENT COMPATIBILITY API	3633	=head3 LIBEVENT COMPATIBILITY API
1477		3634
1478	To include the libevent compatibility API, also include:	3635	To include the libevent compatibility API, also include:
1479		3636
1480	#include "event.c"	3637	#include "event.c"
1481		3638
1482	in the file including F<ev.c>, and:	3639	in the file including F<ev.c>, and:
1483		3640
1484	#include "event.h"	3641	#include "event.h"
1485		3642
1486	in the files that want to use the libevent API. This also includes F<ev.h>.	3643	in the files that want to use the libevent API. This also includes F<ev.h>.
1487		3644
1488	You need the following additional files for this:	3645	You need the following additional files for this:
1489		3646
1490	event.h	3647	event.h
1491	event.c	3648	event.c
1492		3649
1493	=head3 AUTOCONF SUPPORT	3650	=head3 AUTOCONF SUPPORT
1494		3651
1495	Instead of using C<EV_STANDALONE=1> and providing your config in	3652	Instead of using C<EV_STANDALONE=1> and providing your configuration in
1496	whatever way you want, you can also C<m4_include([libev.m4])> in your	3653	whatever way you want, you can also C<m4_include([libev.m4])> in your
1497	F<configure.ac> and leave C<EV_STANDALONE> off. F<ev.c> will then include	3654	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
1498	F<config.h> and configure itself accordingly.	3655	include F<config.h> and configure itself accordingly.
1499		3656
1500	For this of course you need the m4 file:	3657	For this of course you need the m4 file:
1501		3658
1502	libev.m4	3659	libev.m4
1503		3660
1504	=head2 PREPROCESSOR SYMBOLS/MACROS	3661	=head2 PREPROCESSOR SYMBOLS/MACROS
1505		3662
1506	Libev can be configured via a variety of preprocessor symbols you have to define	3663	Libev can be configured via a variety of preprocessor symbols you have to
1507	before including any of its files. The default is not to build for multiplicity	3664	define before including (or compiling) any of its files. The default in
1508	and only include the select backend.	3665	the absence of autoconf is documented for every option.
		3666
		3667	Symbols marked with "(h)" do not change the ABI, and can have different
		3668	values when compiling libev vs. including F<ev.h>, so it is permissible
		3669	to redefine them before including F<ev.h> without breaking compatibility
		3670	to a compiled library. All other symbols change the ABI, which means all
		3671	users of libev and the libev code itself must be compiled with compatible
		3672	settings.
1509		3673
1510	=over 4	3674	=over 4
1511		3675
1512	=item EV_STANDALONE	3676	=item EV_STANDALONE (h)
1513		3677
1514	Must always be C<1> if you do not use autoconf configuration, which	3678	Must always be C<1> if you do not use autoconf configuration, which
1515	keeps libev from including F<config.h>, and it also defines dummy	3679	keeps libev from including F<config.h>, and it also defines dummy
1516	implementations for some libevent functions (such as logging, which is not	3680	implementations for some libevent functions (such as logging, which is not
1517	supported). It will also not define any of the structs usually found in	3681	supported). It will also not define any of the structs usually found in
1518	F<event.h> that are not directly supported by the libev core alone.	3682	F<event.h> that are not directly supported by the libev core alone.
1519		3683
		3684	In standalone mode, libev will still try to automatically deduce the
		3685	configuration, but has to be more conservative.
		3686
1520	=item EV_USE_MONOTONIC	3687	=item EV_USE_MONOTONIC
1521		3688
1522	If defined to be C<1>, libev will try to detect the availability of the	3689	If defined to be C<1>, libev will try to detect the availability of the
1523	monotonic clock option at both compiletime and runtime. Otherwise no use	3690	monotonic clock option at both compile time and runtime. Otherwise no
1524	of the monotonic clock option will be attempted. If you enable this, you	3691	use of the monotonic clock option will be attempted. If you enable this,
1525	usually have to link against librt or something similar. Enabling it when	3692	you usually have to link against librt or something similar. Enabling it
1526	the functionality isn't available is safe, though, althoguh you have	3693	when the functionality isn't available is safe, though, although you have
1527	to make sure you link against any libraries where the C<clock_gettime>	3694	to make sure you link against any libraries where the C<clock_gettime>
1528	function is hiding in (often F<-lrt>).	3695	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
1529		3696
1530	=item EV_USE_REALTIME	3697	=item EV_USE_REALTIME
1531		3698
1532	If defined to be C<1>, libev will try to detect the availability of the	3699	If defined to be C<1>, libev will try to detect the availability of the
1533	realtime clock option at compiletime (and assume its availability at	3700	real-time clock option at compile time (and assume its availability
1534	runtime if successful). Otherwise no use of the realtime clock option will	3701	at runtime if successful). Otherwise no use of the real-time clock
1535	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3702	option will be attempted. This effectively replaces C<gettimeofday>
1536	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries	3703	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
1537	in the description of C<EV_USE_MONOTONIC>, though.	3704	correctness. See the note about libraries in the description of
		3705	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3706	C<EV_USE_CLOCK_SYSCALL>.
		3707
		3708	=item EV_USE_CLOCK_SYSCALL
		3709
		3710	If defined to be C<1>, libev will try to use a direct syscall instead
		3711	of calling the system-provided C<clock_gettime> function. This option
		3712	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3713	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3714	programs needlessly. Using a direct syscall is slightly slower (in
		3715	theory), because no optimised vdso implementation can be used, but avoids
		3716	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3717	higher, as it simplifies linking (no need for C<-lrt>).
		3718
		3719	=item EV_USE_NANOSLEEP
		3720
		3721	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		3722	and will use it for delays. Otherwise it will use C<select ()>.
		3723
		3724	=item EV_USE_EVENTFD
		3725
		3726	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		3727	available and will probe for kernel support at runtime. This will improve
		3728	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		3729	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		3730	2.7 or newer, otherwise disabled.
1538		3731
1539	=item EV_USE_SELECT	3732	=item EV_USE_SELECT
1540		3733
1541	If undefined or defined to be C<1>, libev will compile in support for the	3734	If undefined or defined to be C<1>, libev will compile in support for the
1542	C<select>(2) backend. No attempt at autodetection will be done: if no	3735	C<select>(2) backend. No attempt at auto-detection will be done: if no
1543	other method takes over, select will be it. Otherwise the select backend	3736	other method takes over, select will be it. Otherwise the select backend
1544	will not be compiled in.	3737	will not be compiled in.
1545		3738
1546	=item EV_SELECT_USE_FD_SET	3739	=item EV_SELECT_USE_FD_SET
1547		3740
1548	If defined to C<1>, then the select backend will use the system C<fd_set>	3741	If defined to C<1>, then the select backend will use the system C<fd_set>
1549	structure. This is useful if libev doesn't compile due to a missing	3742	structure. This is useful if libev doesn't compile due to a missing
1550	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	3743	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
1551	exotic systems. This usually limits the range of file descriptors to some	3744	on exotic systems. This usually limits the range of file descriptors to
1552	low limit such as 1024 or might have other limitations (winsocket only	3745	some low limit such as 1024 or might have other limitations (winsocket
1553	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3746	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
1554	influence the size of the C<fd_set> used.	3747	configures the maximum size of the C<fd_set>.
1555		3748
1556	=item EV_SELECT_IS_WINSOCKET	3749	=item EV_SELECT_IS_WINSOCKET
1557		3750
1558	When defined to C<1>, the select backend will assume that	3751	When defined to C<1>, the select backend will assume that
1559	select/socket/connect etc. don't understand file descriptors but	3752	select/socket/connect etc. don't understand file descriptors but
…		…
1561	be used is the winsock select). This means that it will call	3754	be used is the winsock select). This means that it will call
1562	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3755	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
1563	it is assumed that all these functions actually work on fds, even	3756	it is assumed that all these functions actually work on fds, even
1564	on win32. Should not be defined on non-win32 platforms.	3757	on win32. Should not be defined on non-win32 platforms.
1565		3758
		3759	=item EV_FD_TO_WIN32_HANDLE(fd)
		3760
		3761	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		3762	file descriptors to socket handles. When not defining this symbol (the
		3763	default), then libev will call C<_get_osfhandle>, which is usually
		3764	correct. In some cases, programs use their own file descriptor management,
		3765	in which case they can provide this function to map fds to socket handles.
		3766
		3767	=item EV_WIN32_HANDLE_TO_FD(handle)
		3768
		3769	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3770	using the standard C<_open_osfhandle> function. For programs implementing
		3771	their own fd to handle mapping, overwriting this function makes it easier
		3772	to do so. This can be done by defining this macro to an appropriate value.
		3773
		3774	=item EV_WIN32_CLOSE_FD(fd)
		3775
		3776	If programs implement their own fd to handle mapping on win32, then this
		3777	macro can be used to override the C<close> function, useful to unregister
		3778	file descriptors again. Note that the replacement function has to close
		3779	the underlying OS handle.
		3780
1566	=item EV_USE_POLL	3781	=item EV_USE_POLL
1567		3782
1568	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3783	If defined to be C<1>, libev will compile in support for the C<poll>(2)
1569	backend. Otherwise it will be enabled on non-win32 platforms. It	3784	backend. Otherwise it will be enabled on non-win32 platforms. It
1570	takes precedence over select.	3785	takes precedence over select.
1571		3786
1572	=item EV_USE_EPOLL	3787	=item EV_USE_EPOLL
1573		3788
1574	If defined to be C<1>, libev will compile in support for the Linux	3789	If defined to be C<1>, libev will compile in support for the Linux
1575	C<epoll>(7) backend. Its availability will be detected at runtime,	3790	C<epoll>(7) backend. Its availability will be detected at runtime,
1576	otherwise another method will be used as fallback. This is the	3791	otherwise another method will be used as fallback. This is the preferred
1577	preferred backend for GNU/Linux systems.	3792	backend for GNU/Linux systems. If undefined, it will be enabled if the
		3793	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
1578		3794
1579	=item EV_USE_KQUEUE	3795	=item EV_USE_KQUEUE
1580		3796
1581	If defined to be C<1>, libev will compile in support for the BSD style	3797	If defined to be C<1>, libev will compile in support for the BSD style
1582	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	3798	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
1583	otherwise another method will be used as fallback. This is the preferred	3799	otherwise another method will be used as fallback. This is the preferred
1584	backend for BSD and BSD-like systems, although on most BSDs kqueue only	3800	backend for BSD and BSD-like systems, although on most BSDs kqueue only
1585	supports some types of fds correctly (the only platform we found that	3801	supports some types of fds correctly (the only platform we found that
1586	supports ptys for example was NetBSD), so kqueue might be compiled in, but	3802	supports ptys for example was NetBSD), so kqueue might be compiled in, but
1587	not be used unless explicitly requested. The best way to use it is to find	3803	not be used unless explicitly requested. The best way to use it is to find
1588	out wether kqueue supports your type of fd properly and use an embedded	3804	out whether kqueue supports your type of fd properly and use an embedded
1589	kqueue loop.	3805	kqueue loop.
1590		3806
1591	=item EV_USE_PORT	3807	=item EV_USE_PORT
1592		3808
1593	If defined to be C<1>, libev will compile in support for the Solaris	3809	If defined to be C<1>, libev will compile in support for the Solaris
…		…
1595	otherwise another method will be used as fallback. This is the preferred	3811	otherwise another method will be used as fallback. This is the preferred
1596	backend for Solaris 10 systems.	3812	backend for Solaris 10 systems.
1597		3813
1598	=item EV_USE_DEVPOLL	3814	=item EV_USE_DEVPOLL
1599		3815
1600	reserved for future expansion, works like the USE symbols above.	3816	Reserved for future expansion, works like the USE symbols above.
1601		3817
		3818	=item EV_USE_INOTIFY
		3819
		3820	If defined to be C<1>, libev will compile in support for the Linux inotify
		3821	interface to speed up C<ev_stat> watchers. Its actual availability will
		3822	be detected at runtime. If undefined, it will be enabled if the headers
		3823	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		3824
		3825	=item EV_ATOMIC_T
		3826
		3827	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		3828	access is atomic with respect to other threads or signal contexts. No such
		3829	type is easily found in the C language, so you can provide your own type
		3830	that you know is safe for your purposes. It is used both for signal handler "locking"
		3831	as well as for signal and thread safety in C<ev_async> watchers.
		3832
		3833	In the absence of this define, libev will use C<sig_atomic_t volatile>
		3834	(from F<signal.h>), which is usually good enough on most platforms.
		3835
1602	=item EV_H	3836	=item EV_H (h)
1603		3837
1604	The name of the F<ev.h> header file used to include it. The default if	3838	The name of the F<ev.h> header file used to include it. The default if
1605	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This	3839	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
1606	can be used to virtually rename the F<ev.h> header file in case of conflicts.	3840	used to virtually rename the F<ev.h> header file in case of conflicts.
1607		3841
1608	=item EV_CONFIG_H	3842	=item EV_CONFIG_H (h)
1609		3843
1610	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	3844	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
1611	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	3845	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
1612	C<EV_H>, above.	3846	C<EV_H>, above.
1613		3847
1614	=item EV_EVENT_H	3848	=item EV_EVENT_H (h)
1615		3849
1616	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	3850	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
1617	of how the F<event.h> header can be found.	3851	of how the F<event.h> header can be found, the default is C<"event.h">.
1618		3852
1619	=item EV_PROTOTYPES	3853	=item EV_PROTOTYPES (h)
1620		3854
1621	If defined to be C<0>, then F<ev.h> will not define any function	3855	If defined to be C<0>, then F<ev.h> will not define any function
1622	prototypes, but still define all the structs and other symbols. This is	3856	prototypes, but still define all the structs and other symbols. This is
1623	occasionally useful if you want to provide your own wrapper functions	3857	occasionally useful if you want to provide your own wrapper functions
1624	around libev functions.	3858	around libev functions.
…		…
1629	will have the C<struct ev_loop *> as first argument, and you can create	3863	will have the C<struct ev_loop *> as first argument, and you can create
1630	additional independent event loops. Otherwise there will be no support	3864	additional independent event loops. Otherwise there will be no support
1631	for multiple event loops and there is no first event loop pointer	3865	for multiple event loops and there is no first event loop pointer
1632	argument. Instead, all functions act on the single default loop.	3866	argument. Instead, all functions act on the single default loop.
1633		3867
1634	=item EV_PERIODICS	3868	=item EV_MINPRI
1635		3869
1636	If undefined or defined to be C<1>, then periodic timers are supported,	3870	=item EV_MAXPRI
1637	otherwise not. This saves a few kb of code.	3871
		3872	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		3873	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		3874	provide for more priorities by overriding those symbols (usually defined
		3875	to be C<-2> and C<2>, respectively).
		3876
		3877	When doing priority-based operations, libev usually has to linearly search
		3878	all the priorities, so having many of them (hundreds) uses a lot of space
		3879	and time, so using the defaults of five priorities (-2 .. +2) is usually
		3880	fine.
		3881
		3882	If your embedding application does not need any priorities, defining these
		3883	both to C<0> will save some memory and CPU.
		3884
		3885	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		3886	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		3887	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
		3888
		3889	If undefined or defined to be C<1> (and the platform supports it), then
		3890	the respective watcher type is supported. If defined to be C<0>, then it
		3891	is not. Disabling watcher types mainly saves codesize.
		3892
		3893	=item EV_FEATURES
		3894
		3895	If you need to shave off some kilobytes of code at the expense of some
		3896	speed (but with the full API), you can define this symbol to request
		3897	certain subsets of functionality. The default is to enable all features
		3898	that can be enabled on the platform.
		3899
		3900	A typical way to use this symbol is to define it to C<0> (or to a bitset
		3901	with some broad features you want) and then selectively re-enable
		3902	additional parts you want, for example if you want everything minimal,
		3903	but multiple event loop support, async and child watchers and the poll
		3904	backend, use this:
		3905
		3906	#define EV_FEATURES 0
		3907	#define EV_MULTIPLICITY 1
		3908	#define EV_USE_POLL 1
		3909	#define EV_CHILD_ENABLE 1
		3910	#define EV_ASYNC_ENABLE 1
		3911
		3912	The actual value is a bitset, it can be a combination of the following
		3913	values:
		3914
		3915	=over 4
		3916
		3917	=item C<1> - faster/larger code
		3918
		3919	Use larger code to speed up some operations.
		3920
		3921	Currently this is used to override some inlining decisions (enlarging the roughly
		3922	30% code size on amd64.
		3923
		3924	When optimising for size, use of compiler flags such as C<-Os> with
		3925	gcc recommended, as well as C<-DNDEBUG>, as libev contains a number of
		3926	assertions.
		3927
		3928	=item C<2> - faster/larger data structures
		3929
		3930	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		3931	hash table sizes and so on. This will usually further increase codesize
		3932	and can additionally have an effect on the size of data structures at
		3933	runtime.
		3934
		3935	=item C<4> - full API configuration
		3936
		3937	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		3938	enables multiplicity (C<EV_MULTIPLICITY>=1).
		3939
		3940	=item C<8> - full API
		3941
		3942	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		3943	details on which parts of the API are still available without this
		3944	feature, and do not complain if this subset changes over time.
		3945
		3946	=item C<16> - enable all optional watcher types
		3947
		3948	Enables all optional watcher types. If you want to selectively enable
		3949	only some watcher types other than I/O and timers (e.g. prepare,
		3950	embed, async, child...) you can enable them manually by defining
		3951	C<EV_watchertype_ENABLE> to C<1> instead.
		3952
		3953	=item C<32> - enable all backends
		3954
		3955	This enables all backends - without this feature, you need to enable at
		3956	least one backend manually (C<EV_USE_SELECT> is a good choice).
		3957
		3958	=item C<64> - enable OS-specific "helper" APIs
		3959
		3960	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		3961	default.
		3962
		3963	=back
		3964
		3965	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		3966	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		3967	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		3968	watchers, timers and monotonic clock support.
		3969
		3970	With an intelligent-enough linker (gcc+binutils are intelligent enough
		3971	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		3972	your program might be left out as well - a binary starting a timer and an
		3973	I/O watcher then might come out at only 5Kb.
		3974
		3975	=item EV_AVOID_STDIO
		3976
		3977	If this is set to C<1> at compiletime, then libev will avoid using stdio
		3978	functions (printf, scanf, perror etc.). This will increase the codesize
		3979	somewhat, but if your program doesn't otherwise depend on stdio and your
		3980	libc allows it, this avoids linking in the stdio library which is quite
		3981	big.
		3982
		3983	Note that error messages might become less precise when this option is
		3984	enabled.
		3985
		3986	=item EV_NSIG
		3987
		3988	The highest supported signal number, +1 (or, the number of
		3989	signals): Normally, libev tries to deduce the maximum number of signals
		3990	automatically, but sometimes this fails, in which case it can be
		3991	specified. Also, using a lower number than detected (C<32> should be
		3992	good for about any system in existance) can save some memory, as libev
		3993	statically allocates some 12-24 bytes per signal number.
		3994
		3995	=item EV_PID_HASHSIZE
		3996
		3997	C<ev_child> watchers use a small hash table to distribute workload by
		3998	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
		3999	usually more than enough. If you need to manage thousands of children you
		4000	might want to increase this value (I<must> be a power of two).
		4001
		4002	=item EV_INOTIFY_HASHSIZE
		4003
		4004	C<ev_stat> watchers use a small hash table to distribute workload by
		4005	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
		4006	disabled), usually more than enough. If you need to manage thousands of
		4007	C<ev_stat> watchers you might want to increase this value (I<must> be a
		4008	power of two).
		4009
		4010	=item EV_USE_4HEAP
		4011
		4012	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4013	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		4014	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		4015	faster performance with many (thousands) of watchers.
		4016
		4017	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4018	will be C<0>.
		4019
		4020	=item EV_HEAP_CACHE_AT
		4021
		4022	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4023	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		4024	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		4025	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		4026	but avoids random read accesses on heap changes. This improves performance
		4027	noticeably with many (hundreds) of watchers.
		4028
		4029	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4030	will be C<0>.
		4031
		4032	=item EV_VERIFY
		4033
		4034	Controls how much internal verification (see C<ev_loop_verify ()>) will
		4035	be done: If set to C<0>, no internal verification code will be compiled
		4036	in. If set to C<1>, then verification code will be compiled in, but not
		4037	called. If set to C<2>, then the internal verification code will be
		4038	called once per loop, which can slow down libev. If set to C<3>, then the
		4039	verification code will be called very frequently, which will slow down
		4040	libev considerably.
		4041
		4042	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4043	will be C<0>.
1638		4044
1639	=item EV_COMMON	4045	=item EV_COMMON
1640		4046
1641	By default, all watchers have a C<void *data> member. By redefining	4047	By default, all watchers have a C<void *data> member. By redefining
1642	this macro to a something else you can include more and other types of	4048	this macro to a something else you can include more and other types of
1643	members. You have to define it each time you include one of the files,	4049	members. You have to define it each time you include one of the files,
1644	though, and it must be identical each time.	4050	though, and it must be identical each time.
1645		4051
1646	For example, the perl EV module uses something like this:	4052	For example, the perl EV module uses something like this:
1647		4053
1648	#define EV_COMMON \	4054	#define EV_COMMON \
1649	SV *self; /* contains this struct */ \	4055	SV *self; /* contains this struct */ \
1650	SV *cb_sv, *fh /* note no trailing ";" */	4056	SV *cb_sv, *fh /* note no trailing ";" */
1651		4057
1652	=item EV_CB_DECLARE(type)	4058	=item EV_CB_DECLARE (type)
1653		4059
1654	=item EV_CB_INVOKE(watcher,revents)	4060	=item EV_CB_INVOKE (watcher, revents)
1655		4061
1656	=item ev_set_cb(ev,cb)	4062	=item ev_set_cb (ev, cb)
1657		4063
1658	Can be used to change the callback member declaration in each watcher,	4064	Can be used to change the callback member declaration in each watcher,
1659	and the way callbacks are invoked and set. Must expand to a struct member	4065	and the way callbacks are invoked and set. Must expand to a struct member
1660	definition and a statement, respectively. See the F<ev.v> header file for	4066	definition and a statement, respectively. See the F<ev.h> header file for
1661	their default definitions. One possible use for overriding these is to	4067	their default definitions. One possible use for overriding these is to
1662	avoid the ev_loop pointer as first argument in all cases, or to use method	4068	avoid the C<struct ev_loop *> as first argument in all cases, or to use
1663	calls instead of plain function calls in C++.	4069	method calls instead of plain function calls in C++.
		4070
		4071	=back
		4072
		4073	=head2 EXPORTED API SYMBOLS
		4074
		4075	If you need to re-export the API (e.g. via a DLL) and you need a list of
		4076	exported symbols, you can use the provided F<Symbol.*> files which list
		4077	all public symbols, one per line:
		4078
		4079	Symbols.ev for libev proper
		4080	Symbols.event for the libevent emulation
		4081
		4082	This can also be used to rename all public symbols to avoid clashes with
		4083	multiple versions of libev linked together (which is obviously bad in
		4084	itself, but sometimes it is inconvenient to avoid this).
		4085
		4086	A sed command like this will create wrapper C<#define>'s that you need to
		4087	include before including F<ev.h>:
		4088
		4089	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		4090
		4091	This would create a file F<wrap.h> which essentially looks like this:
		4092
		4093	#define ev_backend myprefix_ev_backend
		4094	#define ev_check_start myprefix_ev_check_start
		4095	#define ev_check_stop myprefix_ev_check_stop
		4096	...
1664		4097
1665	=head2 EXAMPLES	4098	=head2 EXAMPLES
1666		4099
1667	For a real-world example of a program the includes libev	4100	For a real-world example of a program the includes libev
1668	verbatim, you can have a look at the EV perl module	4101	verbatim, you can have a look at the EV perl module
…		…
1671	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file	4104	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
1672	will be compiled. It is pretty complex because it provides its own header	4105	will be compiled. It is pretty complex because it provides its own header
1673	file.	4106	file.
1674		4107
1675	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4108	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
1676	that everybody includes and which overrides some autoconf choices:	4109	that everybody includes and which overrides some configure choices:
1677		4110
		4111	#define EV_FEATURES 8
		4112	#define EV_USE_SELECT 1
		4113	#define EV_PREPARE_ENABLE 1
		4114	#define EV_IDLE_ENABLE 1
		4115	#define EV_SIGNAL_ENABLE 1
		4116	#define EV_CHILD_ENABLE 1
1678	#define EV_USE_POLL 0	4117	#define EV_USE_STDEXCEPT 0
1679	#define EV_MULTIPLICITY 0
1680	#define EV_PERIODICS 0
1681	#define EV_CONFIG_H <config.h>	4118	#define EV_CONFIG_H <config.h>
1682		4119
1683	#include "ev++.h"	4120	#include "ev++.h"
1684		4121
1685	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4122	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
1686		4123
1687	#include "ev_cpp.h"	4124	#include "ev_cpp.h"
1688	#include "ev.c"	4125	#include "ev.c"
		4126
		4127	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
		4128
		4129	=head2 THREADS AND COROUTINES
		4130
		4131	=head3 THREADS
		4132
		4133	All libev functions are reentrant and thread-safe unless explicitly
		4134	documented otherwise, but libev implements no locking itself. This means
		4135	that you can use as many loops as you want in parallel, as long as there
		4136	are no concurrent calls into any libev function with the same loop
		4137	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4138	of course): libev guarantees that different event loops share no data
		4139	structures that need any locking.
		4140
		4141	Or to put it differently: calls with different loop parameters can be done
		4142	concurrently from multiple threads, calls with the same loop parameter
		4143	must be done serially (but can be done from different threads, as long as
		4144	only one thread ever is inside a call at any point in time, e.g. by using
		4145	a mutex per loop).
		4146
		4147	Specifically to support threads (and signal handlers), libev implements
		4148	so-called C<ev_async> watchers, which allow some limited form of
		4149	concurrency on the same event loop, namely waking it up "from the
		4150	outside".
		4151
		4152	If you want to know which design (one loop, locking, or multiple loops
		4153	without or something else still) is best for your problem, then I cannot
		4154	help you, but here is some generic advice:
		4155
		4156	=over 4
		4157
		4158	=item * most applications have a main thread: use the default libev loop
		4159	in that thread, or create a separate thread running only the default loop.
		4160
		4161	This helps integrating other libraries or software modules that use libev
		4162	themselves and don't care/know about threading.
		4163
		4164	=item * one loop per thread is usually a good model.
		4165
		4166	Doing this is almost never wrong, sometimes a better-performance model
		4167	exists, but it is always a good start.
		4168
		4169	=item * other models exist, such as the leader/follower pattern, where one
		4170	loop is handed through multiple threads in a kind of round-robin fashion.
		4171
		4172	Choosing a model is hard - look around, learn, know that usually you can do
		4173	better than you currently do :-)
		4174
		4175	=item * often you need to talk to some other thread which blocks in the
		4176	event loop.
		4177
		4178	C<ev_async> watchers can be used to wake them up from other threads safely
		4179	(or from signal contexts...).
		4180
		4181	An example use would be to communicate signals or other events that only
		4182	work in the default loop by registering the signal watcher with the
		4183	default loop and triggering an C<ev_async> watcher from the default loop
		4184	watcher callback into the event loop interested in the signal.
		4185
		4186	=back
		4187
		4188	=head4 THREAD LOCKING EXAMPLE
		4189
		4190	Here is a fictitious example of how to run an event loop in a different
		4191	thread than where callbacks are being invoked and watchers are
		4192	created/added/removed.
		4193
		4194	For a real-world example, see the C<EV::Loop::Async> perl module,
		4195	which uses exactly this technique (which is suited for many high-level
		4196	languages).
		4197
		4198	The example uses a pthread mutex to protect the loop data, a condition
		4199	variable to wait for callback invocations, an async watcher to notify the
		4200	event loop thread and an unspecified mechanism to wake up the main thread.
		4201
		4202	First, you need to associate some data with the event loop:
		4203
		4204	typedef struct {
		4205	mutex_t lock; /* global loop lock */
		4206	ev_async async_w;
		4207	thread_t tid;
		4208	cond_t invoke_cv;
		4209	} userdata;
		4210
		4211	void prepare_loop (EV_P)
		4212	{
		4213	// for simplicity, we use a static userdata struct.
		4214	static userdata u;
		4215
		4216	ev_async_init (&u->async_w, async_cb);
		4217	ev_async_start (EV_A_ &u->async_w);
		4218
		4219	pthread_mutex_init (&u->lock, 0);
		4220	pthread_cond_init (&u->invoke_cv, 0);
		4221
		4222	// now associate this with the loop
		4223	ev_set_userdata (EV_A_ u);
		4224	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4225	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4226
		4227	// then create the thread running ev_loop
		4228	pthread_create (&u->tid, 0, l_run, EV_A);
		4229	}
		4230
		4231	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4232	solely to wake up the event loop so it takes notice of any new watchers
		4233	that might have been added:
		4234
		4235	static void
		4236	async_cb (EV_P_ ev_async *w, int revents)
		4237	{
		4238	// just used for the side effects
		4239	}
		4240
		4241	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4242	protecting the loop data, respectively.
		4243
		4244	static void
		4245	l_release (EV_P)
		4246	{
		4247	userdata *u = ev_userdata (EV_A);
		4248	pthread_mutex_unlock (&u->lock);
		4249	}
		4250
		4251	static void
		4252	l_acquire (EV_P)
		4253	{
		4254	userdata *u = ev_userdata (EV_A);
		4255	pthread_mutex_lock (&u->lock);
		4256	}
		4257
		4258	The event loop thread first acquires the mutex, and then jumps straight
		4259	into C<ev_loop>:
		4260
		4261	void *
		4262	l_run (void *thr_arg)
		4263	{
		4264	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4265
		4266	l_acquire (EV_A);
		4267	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4268	ev_loop (EV_A_ 0);
		4269	l_release (EV_A);
		4270
		4271	return 0;
		4272	}
		4273
		4274	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4275	signal the main thread via some unspecified mechanism (signals? pipe
		4276	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4277	have been called (in a while loop because a) spurious wakeups are possible
		4278	and b) skipping inter-thread-communication when there are no pending
		4279	watchers is very beneficial):
		4280
		4281	static void
		4282	l_invoke (EV_P)
		4283	{
		4284	userdata *u = ev_userdata (EV_A);
		4285
		4286	while (ev_pending_count (EV_A))
		4287	{
		4288	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4289	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4290	}
		4291	}
		4292
		4293	Now, whenever the main thread gets told to invoke pending watchers, it
		4294	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4295	thread to continue:
		4296
		4297	static void
		4298	real_invoke_pending (EV_P)
		4299	{
		4300	userdata *u = ev_userdata (EV_A);
		4301
		4302	pthread_mutex_lock (&u->lock);
		4303	ev_invoke_pending (EV_A);
		4304	pthread_cond_signal (&u->invoke_cv);
		4305	pthread_mutex_unlock (&u->lock);
		4306	}
		4307
		4308	Whenever you want to start/stop a watcher or do other modifications to an
		4309	event loop, you will now have to lock:
		4310
		4311	ev_timer timeout_watcher;
		4312	userdata *u = ev_userdata (EV_A);
		4313
		4314	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4315
		4316	pthread_mutex_lock (&u->lock);
		4317	ev_timer_start (EV_A_ &timeout_watcher);
		4318	ev_async_send (EV_A_ &u->async_w);
		4319	pthread_mutex_unlock (&u->lock);
		4320
		4321	Note that sending the C<ev_async> watcher is required because otherwise
		4322	an event loop currently blocking in the kernel will have no knowledge
		4323	about the newly added timer. By waking up the loop it will pick up any new
		4324	watchers in the next event loop iteration.
		4325
		4326	=head3 COROUTINES
		4327
		4328	Libev is very accommodating to coroutines ("cooperative threads"):
		4329	libev fully supports nesting calls to its functions from different
		4330	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		4331	different coroutines, and switch freely between both coroutines running
		4332	the loop, as long as you don't confuse yourself). The only exception is
		4333	that you must not do this from C<ev_periodic> reschedule callbacks.
		4334
		4335	Care has been taken to ensure that libev does not keep local state inside
		4336	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		4337	they do not call any callbacks.
		4338
		4339	=head2 COMPILER WARNINGS
		4340
		4341	Depending on your compiler and compiler settings, you might get no or a
		4342	lot of warnings when compiling libev code. Some people are apparently
		4343	scared by this.
		4344
		4345	However, these are unavoidable for many reasons. For one, each compiler
		4346	has different warnings, and each user has different tastes regarding
		4347	warning options. "Warn-free" code therefore cannot be a goal except when
		4348	targeting a specific compiler and compiler-version.
		4349
		4350	Another reason is that some compiler warnings require elaborate
		4351	workarounds, or other changes to the code that make it less clear and less
		4352	maintainable.
		4353
		4354	And of course, some compiler warnings are just plain stupid, or simply
		4355	wrong (because they don't actually warn about the condition their message
		4356	seems to warn about). For example, certain older gcc versions had some
		4357	warnings that resulted an extreme number of false positives. These have
		4358	been fixed, but some people still insist on making code warn-free with
		4359	such buggy versions.
		4360
		4361	While libev is written to generate as few warnings as possible,
		4362	"warn-free" code is not a goal, and it is recommended not to build libev
		4363	with any compiler warnings enabled unless you are prepared to cope with
		4364	them (e.g. by ignoring them). Remember that warnings are just that:
		4365	warnings, not errors, or proof of bugs.
		4366
		4367
		4368	=head2 VALGRIND
		4369
		4370	Valgrind has a special section here because it is a popular tool that is
		4371	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		4372
		4373	If you think you found a bug (memory leak, uninitialised data access etc.)
		4374	in libev, then check twice: If valgrind reports something like:
		4375
		4376	==2274== definitely lost: 0 bytes in 0 blocks.
		4377	==2274== possibly lost: 0 bytes in 0 blocks.
		4378	==2274== still reachable: 256 bytes in 1 blocks.
		4379
		4380	Then there is no memory leak, just as memory accounted to global variables
		4381	is not a memleak - the memory is still being referenced, and didn't leak.
		4382
		4383	Similarly, under some circumstances, valgrind might report kernel bugs
		4384	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4385	although an acceptable workaround has been found here), or it might be
		4386	confused.
		4387
		4388	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		4389	make it into some kind of religion.
		4390
		4391	If you are unsure about something, feel free to contact the mailing list
		4392	with the full valgrind report and an explanation on why you think this
		4393	is a bug in libev (best check the archives, too :). However, don't be
		4394	annoyed when you get a brisk "this is no bug" answer and take the chance
		4395	of learning how to interpret valgrind properly.
		4396
		4397	If you need, for some reason, empty reports from valgrind for your project
		4398	I suggest using suppression lists.
		4399
		4400
		4401	=head1 PORTABILITY NOTES
		4402
		4403	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4404
		4405	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		4406	requires, and its I/O model is fundamentally incompatible with the POSIX
		4407	model. Libev still offers limited functionality on this platform in
		4408	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		4409	descriptors. This only applies when using Win32 natively, not when using
		4410	e.g. cygwin.
		4411
		4412	Lifting these limitations would basically require the full
		4413	re-implementation of the I/O system. If you are into these kinds of
		4414	things, then note that glib does exactly that for you in a very portable
		4415	way (note also that glib is the slowest event library known to man).
		4416
		4417	There is no supported compilation method available on windows except
		4418	embedding it into other applications.
		4419
		4420	Sensible signal handling is officially unsupported by Microsoft - libev
		4421	tries its best, but under most conditions, signals will simply not work.
		4422
		4423	Not a libev limitation but worth mentioning: windows apparently doesn't
		4424	accept large writes: instead of resulting in a partial write, windows will
		4425	either accept everything or return C<ENOBUFS> if the buffer is too large,
		4426	so make sure you only write small amounts into your sockets (less than a
		4427	megabyte seems safe, but this apparently depends on the amount of memory
		4428	available).
		4429
		4430	Due to the many, low, and arbitrary limits on the win32 platform and
		4431	the abysmal performance of winsockets, using a large number of sockets
		4432	is not recommended (and not reasonable). If your program needs to use
		4433	more than a hundred or so sockets, then likely it needs to use a totally
		4434	different implementation for windows, as libev offers the POSIX readiness
		4435	notification model, which cannot be implemented efficiently on windows
		4436	(due to Microsoft monopoly games).
		4437
		4438	A typical way to use libev under windows is to embed it (see the embedding
		4439	section for details) and use the following F<evwrap.h> header file instead
		4440	of F<ev.h>:
		4441
		4442	#define EV_STANDALONE /* keeps ev from requiring config.h */
		4443	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		4444
		4445	#include "ev.h"
		4446
		4447	And compile the following F<evwrap.c> file into your project (make sure
		4448	you do I<not> compile the F<ev.c> or any other embedded source files!):
		4449
		4450	#include "evwrap.h"
		4451	#include "ev.c"
		4452
		4453	=over 4
		4454
		4455	=item The winsocket select function
		4456
		4457	The winsocket C<select> function doesn't follow POSIX in that it
		4458	requires socket I<handles> and not socket I<file descriptors> (it is
		4459	also extremely buggy). This makes select very inefficient, and also
		4460	requires a mapping from file descriptors to socket handles (the Microsoft
		4461	C runtime provides the function C<_open_osfhandle> for this). See the
		4462	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		4463	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		4464
		4465	The configuration for a "naked" win32 using the Microsoft runtime
		4466	libraries and raw winsocket select is:
		4467
		4468	#define EV_USE_SELECT 1
		4469	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		4470
		4471	Note that winsockets handling of fd sets is O(n), so you can easily get a
		4472	complexity in the O(n²) range when using win32.
		4473
		4474	=item Limited number of file descriptors
		4475
		4476	Windows has numerous arbitrary (and low) limits on things.
		4477
		4478	Early versions of winsocket's select only supported waiting for a maximum
		4479	of C<64> handles (probably owning to the fact that all windows kernels
		4480	can only wait for C<64> things at the same time internally; Microsoft
		4481	recommends spawning a chain of threads and wait for 63 handles and the
		4482	previous thread in each. Sounds great!).
		4483
		4484	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		4485	to some high number (e.g. C<2048>) before compiling the winsocket select
		4486	call (which might be in libev or elsewhere, for example, perl and many
		4487	other interpreters do their own select emulation on windows).
		4488
		4489	Another limit is the number of file descriptors in the Microsoft runtime
		4490	libraries, which by default is C<64> (there must be a hidden I<64>
		4491	fetish or something like this inside Microsoft). You can increase this
		4492	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
		4493	(another arbitrary limit), but is broken in many versions of the Microsoft
		4494	runtime libraries. This might get you to about C<512> or C<2048> sockets
		4495	(depending on windows version and/or the phase of the moon). To get more,
		4496	you need to wrap all I/O functions and provide your own fd management, but
		4497	the cost of calling select (O(n²)) will likely make this unworkable.
		4498
		4499	=back
		4500
		4501	=head2 PORTABILITY REQUIREMENTS
		4502
		4503	In addition to a working ISO-C implementation and of course the
		4504	backend-specific APIs, libev relies on a few additional extensions:
		4505
		4506	=over 4
		4507
		4508	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		4509	calling conventions regardless of C<ev_watcher_type *>.
		4510
		4511	Libev assumes not only that all watcher pointers have the same internal
		4512	structure (guaranteed by POSIX but not by ISO C for example), but it also
		4513	assumes that the same (machine) code can be used to call any watcher
		4514	callback: The watcher callbacks have different type signatures, but libev
		4515	calls them using an C<ev_watcher *> internally.
		4516
		4517	=item C<sig_atomic_t volatile> must be thread-atomic as well
		4518
		4519	The type C<sig_atomic_t volatile> (or whatever is defined as
		4520	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		4521	threads. This is not part of the specification for C<sig_atomic_t>, but is
		4522	believed to be sufficiently portable.
		4523
		4524	=item C<sigprocmask> must work in a threaded environment
		4525
		4526	Libev uses C<sigprocmask> to temporarily block signals. This is not
		4527	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		4528	pthread implementations will either allow C<sigprocmask> in the "main
		4529	thread" or will block signals process-wide, both behaviours would
		4530	be compatible with libev. Interaction between C<sigprocmask> and
		4531	C<pthread_sigmask> could complicate things, however.
		4532
		4533	The most portable way to handle signals is to block signals in all threads
		4534	except the initial one, and run the default loop in the initial thread as
		4535	well.
		4536
		4537	=item C<long> must be large enough for common memory allocation sizes
		4538
		4539	To improve portability and simplify its API, libev uses C<long> internally
		4540	instead of C<size_t> when allocating its data structures. On non-POSIX
		4541	systems (Microsoft...) this might be unexpectedly low, but is still at
		4542	least 31 bits everywhere, which is enough for hundreds of millions of
		4543	watchers.
		4544
		4545	=item C<double> must hold a time value in seconds with enough accuracy
		4546
		4547	The type C<double> is used to represent timestamps. It is required to
		4548	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		4549	enough for at least into the year 4000. This requirement is fulfilled by
		4550	implementations implementing IEEE 754, which is basically all existing
		4551	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4552	2200.
		4553
		4554	=back
		4555
		4556	If you know of other additional requirements drop me a note.
		4557
		4558
		4559	=head1 ALGORITHMIC COMPLEXITIES
		4560
		4561	In this section the complexities of (many of) the algorithms used inside
		4562	libev will be documented. For complexity discussions about backends see
		4563	the documentation for C<ev_default_init>.
		4564
		4565	All of the following are about amortised time: If an array needs to be
		4566	extended, libev needs to realloc and move the whole array, but this
		4567	happens asymptotically rarer with higher number of elements, so O(1) might
		4568	mean that libev does a lengthy realloc operation in rare cases, but on
		4569	average it is much faster and asymptotically approaches constant time.
		4570
		4571	=over 4
		4572
		4573	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		4574
		4575	This means that, when you have a watcher that triggers in one hour and
		4576	there are 100 watchers that would trigger before that, then inserting will
		4577	have to skip roughly seven (C<ld 100>) of these watchers.
		4578
		4579	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		4580
		4581	That means that changing a timer costs less than removing/adding them,
		4582	as only the relative motion in the event queue has to be paid for.
		4583
		4584	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		4585
		4586	These just add the watcher into an array or at the head of a list.
		4587
		4588	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		4589
		4590	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		4591
		4592	These watchers are stored in lists, so they need to be walked to find the
		4593	correct watcher to remove. The lists are usually short (you don't usually
		4594	have many watchers waiting for the same fd or signal: one is typical, two
		4595	is rare).
		4596
		4597	=item Finding the next timer in each loop iteration: O(1)
		4598
		4599	By virtue of using a binary or 4-heap, the next timer is always found at a
		4600	fixed position in the storage array.
		4601
		4602	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4603
		4604	A change means an I/O watcher gets started or stopped, which requires
		4605	libev to recalculate its status (and possibly tell the kernel, depending
		4606	on backend and whether C<ev_io_set> was used).
		4607
		4608	=item Activating one watcher (putting it into the pending state): O(1)
		4609
		4610	=item Priority handling: O(number_of_priorities)
		4611
		4612	Priorities are implemented by allocating some space for each
		4613	priority. When doing priority-based operations, libev usually has to
		4614	linearly search all the priorities, but starting/stopping and activating
		4615	watchers becomes O(1) with respect to priority handling.
		4616
		4617	=item Sending an ev_async: O(1)
		4618
		4619	=item Processing ev_async_send: O(number_of_async_watchers)
		4620
		4621	=item Processing signals: O(max_signal_number)
		4622
		4623	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4624	calls in the current loop iteration. Checking for async and signal events
		4625	involves iterating over all running async watchers or all signal numbers.
		4626
		4627	=back
		4628
		4629
		4630	=head1 PORTING FROM LIBEV 3.X TO 4.X
		4631
		4632	The major version 4 introduced some minor incompatible changes to the API.
		4633
		4634	At the moment, the C<ev.h> header file tries to implement superficial
		4635	compatibility, so most programs should still compile. Those might be
		4636	removed in later versions of libev, so better update early than late.
		4637
		4638	=over 4
		4639
		4640	=item C<ev_loop_count> renamed to C<ev_iteration>
		4641
		4642	=item C<ev_loop_depth> renamed to C<ev_depth>
		4643
		4644	=item C<ev_loop_verify> renamed to C<ev_verify>
		4645
		4646	Most functions working on C<struct ev_loop> objects don't have an
		4647	C<ev_loop_> prefix, so it was removed. Note that C<ev_loop_fork> is
		4648	still called C<ev_loop_fork> because it would otherwise clash with the
		4649	C<ev_fork> typedef.
		4650
		4651	=item C<EV_TIMEOUT> renamed to C<EV_TIMER> in C<revents>
		4652
		4653	This is a simple rename - all other watcher types use their name
		4654	as revents flag, and now C<ev_timer> does, too.
		4655
		4656	Both C<EV_TIMER> and C<EV_TIMEOUT> symbols were present in 3.x versions
		4657	and continue to be present for the forseeable future, so this is mostly a
		4658	documentation change.
		4659
		4660	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		4661
		4662	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		4663	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		4664	and work, but the library code will of course be larger.
		4665
		4666	=back
		4667
		4668
		4669	=head1 GLOSSARY
		4670
		4671	=over 4
		4672
		4673	=item active
		4674
		4675	A watcher is active as long as it has been started (has been attached to
		4676	an event loop) but not yet stopped (disassociated from the event loop).
		4677
		4678	=item application
		4679
		4680	In this document, an application is whatever is using libev.
		4681
		4682	=item callback
		4683
		4684	The address of a function that is called when some event has been
		4685	detected. Callbacks are being passed the event loop, the watcher that
		4686	received the event, and the actual event bitset.
		4687
		4688	=item callback invocation
		4689
		4690	The act of calling the callback associated with a watcher.
		4691
		4692	=item event
		4693
		4694	A change of state of some external event, such as data now being available
		4695	for reading on a file descriptor, time having passed or simply not having
		4696	any other events happening anymore.
		4697
		4698	In libev, events are represented as single bits (such as C<EV_READ> or
		4699	C<EV_TIMER>).
		4700
		4701	=item event library
		4702
		4703	A software package implementing an event model and loop.
		4704
		4705	=item event loop
		4706
		4707	An entity that handles and processes external events and converts them
		4708	into callback invocations.
		4709
		4710	=item event model
		4711
		4712	The model used to describe how an event loop handles and processes
		4713	watchers and events.
		4714
		4715	=item pending
		4716
		4717	A watcher is pending as soon as the corresponding event has been detected,
		4718	and stops being pending as soon as the watcher will be invoked or its
		4719	pending status is explicitly cleared by the application.
		4720
		4721	A watcher can be pending, but not active. Stopping a watcher also clears
		4722	its pending status.
		4723
		4724	=item real time
		4725
		4726	The physical time that is observed. It is apparently strictly monotonic :)
		4727
		4728	=item wall-clock time
		4729
		4730	The time and date as shown on clocks. Unlike real time, it can actually
		4731	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4732	clock.
		4733
		4734	=item watcher
		4735
		4736	A data structure that describes interest in certain events. Watchers need
		4737	to be started (attached to an event loop) before they can receive events.
		4738
		4739	=item watcher invocation
		4740
		4741	The act of calling the callback associated with a watcher.
		4742
		4743	=back
1689		4744
1690	=head1 AUTHOR	4745	=head1 AUTHOR
1691		4746
1692	Marc Lehmann <libev@schmorp.de>.	4747	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
1693		4748

Diff Legend

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