ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.27 by root, Wed Nov 14 05:02:07 2007 UTC vs.
Revision 1.85 by root, Mon Dec 17 07:24:12 2007 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
		9	=head1 EXAMPLE PROGRAM
		10
		11	#include <ev.h>
		12
		13	ev_io stdin_watcher;
		14	ev_timer timeout_watcher;
		15
		16	/* called when data readable on stdin */
		17	static void
		18	stdin_cb (EV_P_ struct ev_io *w, int revents)
		19	{
		20	/* puts ("stdin ready"); */
		21	ev_io_stop (EV_A_ w); /* just a syntax example */
		22	ev_unloop (EV_A_ EVUNLOOP_ALL); /* leave all loop calls */
		23	}
		24
		25	static void
		26	timeout_cb (EV_P_ struct ev_timer *w, int revents)
		27	{
		28	/* puts ("timeout"); */
		29	ev_unloop (EV_A_ EVUNLOOP_ONE); /* leave one loop call */
		30	}
		31
		32	int
		33	main (void)
		34	{
		35	struct ev_loop *loop = ev_default_loop (0);
		36
		37	/* initialise an io watcher, then start it */
		38	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		39	ev_io_start (loop, &stdin_watcher);
		40
		41	/* simple non-repeating 5.5 second timeout */
		42	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		43	ev_timer_start (loop, &timeout_watcher);
		44
		45	/* loop till timeout or data ready */
		46	ev_loop (loop, 0);
		47
		48	return 0;
		49	}
		50
9	=head1 DESCRIPTION	51	=head1 DESCRIPTION
		52
		53	The newest version of this document is also available as a html-formatted
		54	web page you might find easier to navigate when reading it for the first
		55	time: L<http://cvs.schmorp.de/libev/ev.html>.
10		56
11	Libev is an event loop: you register interest in certain events (such as a	57	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	58	file descriptor being readable or a timeout occuring), and it will manage
13	these event sources and provide your program with events.	59	these event sources and provide your program with events.
14		60
…		…
21	details of the event, and then hand it over to libev by I<starting> the	67	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	68	watcher.
23		69
24	=head1 FEATURES	70	=head1 FEATURES
25		71
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	72	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	73	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	74	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
29	events (related to SIGCHLD), and event watchers dealing with the event	75	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	76	with customised rescheduling (C<ev_periodic>), synchronous signals
		77	(C<ev_signal>), process status change events (C<ev_child>), and event
		78	watchers dealing with the event loop mechanism itself (C<ev_idle>,
		79	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as
		80	file watchers (C<ev_stat>) and even limited support for fork events
		81	(C<ev_fork>).
		82
		83	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	84	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	85	for example).
33		86
34	=head1 CONVENTIONS	87	=head1 CONVENTIONS
35		88
36	Libev is very configurable. In this manual the default configuration	89	Libev is very configurable. In this manual the default configuration will
37	will be described, which supports multiple event loops. For more info	90	be described, which supports multiple event loops. For more info about
38	about various configuration options please have a look at the file	91	various configuration options please have a look at B<EMBED> section in
39	F<README.embed> in the libev distribution. If libev was configured without	92	this manual. If libev was configured without support for multiple event
40	support for multiple event loops, then all functions taking an initial	93	loops, then all functions taking an initial argument of name C<loop>
41	argument of name C<loop> (which is always of type C<struct ev_loop *>)	94	(which is always of type C<struct ev_loop *>) will not have this argument.
42	will not have this argument.
43		95
44	=head1 TIME REPRESENTATION	96	=head1 TIME REPRESENTATION
45		97
46	Libev represents time as a single floating point number, representing the	98	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	99	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	100	the beginning of 1970, details are complicated, don't ask). This type is
49	called C<ev_tstamp>, which is what you should use too. It usually aliases	101	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the double type in C.	102	to the C<double> type in C, and when you need to do any calculations on
		103	it, you should treat it as such.
51		104
52	=head1 GLOBAL FUNCTIONS	105	=head1 GLOBAL FUNCTIONS
53		106
54	These functions can be called anytime, even before initialising the	107	These functions can be called anytime, even before initialising the
55	library in any way.	108	library in any way.
…		…
64		117
65	=item int ev_version_major ()	118	=item int ev_version_major ()
66		119
67	=item int ev_version_minor ()	120	=item int ev_version_minor ()
68		121
69	You can find out the major and minor version numbers of the library	122	You can find out the major and minor ABI version numbers of the library
70	you linked against by calling the functions C<ev_version_major> and	123	you linked against by calling the functions C<ev_version_major> and
71	C<ev_version_minor>. If you want, you can compare against the global	124	C<ev_version_minor>. If you want, you can compare against the global
72	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	125	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
73	version of the library your program was compiled against.	126	version of the library your program was compiled against.
74		127
		128	These version numbers refer to the ABI version of the library, not the
		129	release version.
		130
75	Usually, it's a good idea to terminate if the major versions mismatch,	131	Usually, it's a good idea to terminate if the major versions mismatch,
76	as this indicates an incompatible change. Minor versions are usually	132	as this indicates an incompatible change. Minor versions are usually
77	compatible to older versions, so a larger minor version alone is usually	133	compatible to older versions, so a larger minor version alone is usually
78	not a problem.	134	not a problem.
79		135
		136	Example: Make sure we haven't accidentally been linked against the wrong
		137	version.
		138
		139	assert (("libev version mismatch",
		140	ev_version_major () == EV_VERSION_MAJOR
		141	&& ev_version_minor () >= EV_VERSION_MINOR));
		142
		143	=item unsigned int ev_supported_backends ()
		144
		145	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		146	value) compiled into this binary of libev (independent of their
		147	availability on the system you are running on). See C<ev_default_loop> for
		148	a description of the set values.
		149
		150	Example: make sure we have the epoll method, because yeah this is cool and
		151	a must have and can we have a torrent of it please!!!11
		152
		153	assert (("sorry, no epoll, no sex",
		154	ev_supported_backends () & EVBACKEND_EPOLL));
		155
		156	=item unsigned int ev_recommended_backends ()
		157
		158	Return the set of all backends compiled into this binary of libev and also
		159	recommended for this platform. This set is often smaller than the one
		160	returned by C<ev_supported_backends>, as for example kqueue is broken on
		161	most BSDs and will not be autodetected unless you explicitly request it
		162	(assuming you know what you are doing). This is the set of backends that
		163	libev will probe for if you specify no backends explicitly.
		164
		165	=item unsigned int ev_embeddable_backends ()
		166
		167	Returns the set of backends that are embeddable in other event loops. This
		168	is the theoretical, all-platform, value. To find which backends
		169	might be supported on the current system, you would need to look at
		170	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
		171	recommended ones.
		172
		173	See the description of C<ev_embed> watchers for more info.
		174
80	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	175	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
81		176
82	Sets the allocation function to use (the prototype is similar to the	177	Sets the allocation function to use (the prototype is similar - the
83	realloc C function, the semantics are identical). It is used to allocate	178	semantics is identical - to the realloc C function). It is used to
84	and free memory (no surprises here). If it returns zero when memory	179	allocate and free memory (no surprises here). If it returns zero when
85	needs to be allocated, the library might abort or take some potentially	180	memory needs to be allocated, the library might abort or take some
86	destructive action. The default is your system realloc function.	181	potentially destructive action. The default is your system realloc
		182	function.
87		183
88	You could override this function in high-availability programs to, say,	184	You could override this function in high-availability programs to, say,
89	free some memory if it cannot allocate memory, to use a special allocator,	185	free some memory if it cannot allocate memory, to use a special allocator,
90	or even to sleep a while and retry until some memory is available.	186	or even to sleep a while and retry until some memory is available.
		187
		188	Example: Replace the libev allocator with one that waits a bit and then
		189	retries).
		190
		191	static void *
		192	persistent_realloc (void *ptr, size_t size)
		193	{
		194	for (;;)
		195	{
		196	void *newptr = realloc (ptr, size);
		197
		198	if (newptr)
		199	return newptr;
		200
		201	sleep (60);
		202	}
		203	}
		204
		205	...
		206	ev_set_allocator (persistent_realloc);
91		207
92	=item ev_set_syserr_cb (void (*cb)(const char *msg));	208	=item ev_set_syserr_cb (void (*cb)(const char *msg));
93		209
94	Set the callback function to call on a retryable syscall error (such	210	Set the callback function to call on a retryable syscall error (such
95	as failed select, poll, epoll_wait). The message is a printable string	211	as failed select, poll, epoll_wait). The message is a printable string
…		…
97	callback is set, then libev will expect it to remedy the sitution, no	213	callback is set, then libev will expect it to remedy the sitution, no
98	matter what, when it returns. That is, libev will generally retry the	214	matter what, when it returns. That is, libev will generally retry the
99	requested operation, or, if the condition doesn't go away, do bad stuff	215	requested operation, or, if the condition doesn't go away, do bad stuff
100	(such as abort).	216	(such as abort).
101		217
		218	Example: This is basically the same thing that libev does internally, too.
		219
		220	static void
		221	fatal_error (const char *msg)
		222	{
		223	perror (msg);
		224	abort ();
		225	}
		226
		227	...
		228	ev_set_syserr_cb (fatal_error);
		229
102	=back	230	=back
103		231
104	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	232	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
105		233
106	An event loop is described by a C<struct ev_loop *>. The library knows two	234	An event loop is described by a C<struct ev_loop *>. The library knows two
…		…
119	=item struct ev_loop *ev_default_loop (unsigned int flags)	247	=item struct ev_loop *ev_default_loop (unsigned int flags)
120		248
121	This will initialise the default event loop if it hasn't been initialised	249	This will initialise the default event loop if it hasn't been initialised
122	yet and return it. If the default loop could not be initialised, returns	250	yet and return it. If the default loop could not be initialised, returns
123	false. If it already was initialised it simply returns it (and ignores the	251	false. If it already was initialised it simply returns it (and ignores the
124	flags).	252	flags. If that is troubling you, check C<ev_backend ()> afterwards).
125		253
126	If you don't know what event loop to use, use the one returned from this	254	If you don't know what event loop to use, use the one returned from this
127	function.	255	function.
128		256
129	The flags argument can be used to specify special behaviour or specific	257	The flags argument can be used to specify special behaviour or specific
130	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	258	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
131		259
132	It supports the following flags:	260	The following flags are supported:
133		261
134	=over 4	262	=over 4
135		263
136	=item C<EVFLAG_AUTO>	264	=item C<EVFLAG_AUTO>
137		265
…		…
145	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	273	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
146	override the flags completely if it is found in the environment. This is	274	override the flags completely if it is found in the environment. This is
147	useful to try out specific backends to test their performance, or to work	275	useful to try out specific backends to test their performance, or to work
148	around bugs.	276	around bugs.
149		277
		278	=item C<EVFLAG_FORKCHECK>
		279
		280	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		281	a fork, you can also make libev check for a fork in each iteration by
		282	enabling this flag.
		283
		284	This works by calling C<getpid ()> on every iteration of the loop,
		285	and thus this might slow down your event loop if you do a lot of loop
		286	iterations and little real work, but is usually not noticeable (on my
		287	Linux system for example, C<getpid> is actually a simple 5-insn sequence
		288	without a syscall and thus I<very> fast, but my Linux system also has
		289	C<pthread_atfork> which is even faster).
		290
		291	The big advantage of this flag is that you can forget about fork (and
		292	forget about forgetting to tell libev about forking) when you use this
		293	flag.
		294
		295	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		296	environment variable.
		297
150	=item C<EVMETHOD_SELECT> (portable select backend)	298	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
151		299
		300	This is your standard select(2) backend. Not I<completely> standard, as
		301	libev tries to roll its own fd_set with no limits on the number of fds,
		302	but if that fails, expect a fairly low limit on the number of fds when
		303	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
		304	the fastest backend for a low number of fds.
		305
152	=item C<EVMETHOD_POLL> (poll backend, available everywhere except on windows)	306	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
153		307
154	=item C<EVMETHOD_EPOLL> (linux only)	308	And this is your standard poll(2) backend. It's more complicated than
		309	select, but handles sparse fds better and has no artificial limit on the
		310	number of fds you can use (except it will slow down considerably with a
		311	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
155		312
156	=item C<EVMETHOD_KQUEUE> (some bsds only)	313	=item C<EVBACKEND_EPOLL> (value 4, Linux)
157		314
158	=item C<EVMETHOD_DEVPOLL> (solaris 8 only)	315	For few fds, this backend is a bit little slower than poll and select,
		316	but it scales phenomenally better. While poll and select usually scale like
		317	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
		318	either O(1) or O(active_fds).
159		319
160	=item C<EVMETHOD_PORT> (solaris 10 only)	320	While stopping and starting an I/O watcher in the same iteration will
		321	result in some caching, there is still a syscall per such incident
		322	(because the fd could point to a different file description now), so its
		323	best to avoid that. Also, dup()ed file descriptors might not work very
		324	well if you register events for both fds.
		325
		326	Please note that epoll sometimes generates spurious notifications, so you
		327	need to use non-blocking I/O or other means to avoid blocking when no data
		328	(or space) is available.
		329
		330	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
		331
		332	Kqueue deserves special mention, as at the time of this writing, it
		333	was broken on all BSDs except NetBSD (usually it doesn't work with
		334	anything but sockets and pipes, except on Darwin, where of course its
		335	completely useless). For this reason its not being "autodetected"
		336	unless you explicitly specify it explicitly in the flags (i.e. using
		337	C<EVBACKEND_KQUEUE>).
		338
		339	It scales in the same way as the epoll backend, but the interface to the
		340	kernel is more efficient (which says nothing about its actual speed, of
		341	course). While starting and stopping an I/O watcher does not cause an
		342	extra syscall as with epoll, it still adds up to four event changes per
		343	incident, so its best to avoid that.
		344
		345	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
		346
		347	This is not implemented yet (and might never be).
		348
		349	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
		350
		351	This uses the Solaris 10 port mechanism. As with everything on Solaris,
		352	it's really slow, but it still scales very well (O(active_fds)).
		353
		354	Please note that solaris ports can result in a lot of spurious
		355	notifications, so you need to use non-blocking I/O or other means to avoid
		356	blocking when no data (or space) is available.
		357
		358	=item C<EVBACKEND_ALL>
		359
		360	Try all backends (even potentially broken ones that wouldn't be tried
		361	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
		362	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
		363
		364	=back
161		365
162	If one or more of these are ored into the flags value, then only these	366	If one or more of these are ored into the flags value, then only these
163	backends will be tried (in the reverse order as given here). If one are	367	backends will be tried (in the reverse order as given here). If none are
164	specified, any backend will do.	368	specified, most compiled-in backend will be tried, usually in reverse
		369	order of their flag values :)
165		370
166	=back	371	The most typical usage is like this:
		372
		373	if (!ev_default_loop (0))
		374	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		375
		376	Restrict libev to the select and poll backends, and do not allow
		377	environment settings to be taken into account:
		378
		379	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		380
		381	Use whatever libev has to offer, but make sure that kqueue is used if
		382	available (warning, breaks stuff, best use only with your own private
		383	event loop and only if you know the OS supports your types of fds):
		384
		385	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
167		386
168	=item struct ev_loop *ev_loop_new (unsigned int flags)	387	=item struct ev_loop *ev_loop_new (unsigned int flags)
169		388
170	Similar to C<ev_default_loop>, but always creates a new event loop that is	389	Similar to C<ev_default_loop>, but always creates a new event loop that is
171	always distinct from the default loop. Unlike the default loop, it cannot	390	always distinct from the default loop. Unlike the default loop, it cannot
172	handle signal and child watchers, and attempts to do so will be greeted by	391	handle signal and child watchers, and attempts to do so will be greeted by
173	undefined behaviour (or a failed assertion if assertions are enabled).	392	undefined behaviour (or a failed assertion if assertions are enabled).
174		393
		394	Example: Try to create a event loop that uses epoll and nothing else.
		395
		396	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
		397	if (!epoller)
		398	fatal ("no epoll found here, maybe it hides under your chair");
		399
175	=item ev_default_destroy ()	400	=item ev_default_destroy ()
176		401
177	Destroys the default loop again (frees all memory and kernel state	402	Destroys the default loop again (frees all memory and kernel state
178	etc.). This stops all registered event watchers (by not touching them in	403	etc.). None of the active event watchers will be stopped in the normal
179	any way whatsoever, although you cannot rely on this :).	404	sense, so e.g. C<ev_is_active> might still return true. It is your
		405	responsibility to either stop all watchers cleanly yoursef I<before>
		406	calling this function, or cope with the fact afterwards (which is usually
		407	the easiest thing, youc na just ignore the watchers and/or C<free ()> them
		408	for example).
180		409
181	=item ev_loop_destroy (loop)	410	=item ev_loop_destroy (loop)
182		411
183	Like C<ev_default_destroy>, but destroys an event loop created by an	412	Like C<ev_default_destroy>, but destroys an event loop created by an
184	earlier call to C<ev_loop_new>.	413	earlier call to C<ev_loop_new>.
…		…
188	This function reinitialises the kernel state for backends that have	417	This function reinitialises the kernel state for backends that have
189	one. Despite the name, you can call it anytime, but it makes most sense	418	one. Despite the name, you can call it anytime, but it makes most sense
190	after forking, in either the parent or child process (or both, but that	419	after forking, in either the parent or child process (or both, but that
191	again makes little sense).	420	again makes little sense).
192		421
193	You I<must> call this function after forking if and only if you want to	422	You I<must> call this function in the child process after forking if and
194	use the event library in both processes. If you just fork+exec, you don't	423	only if you want to use the event library in both processes. If you just
195	have to call it.	424	fork+exec, you don't have to call it.
196		425
197	The function itself is quite fast and it's usually not a problem to call	426	The function itself is quite fast and it's usually not a problem to call
198	it just in case after a fork. To make this easy, the function will fit in	427	it just in case after a fork. To make this easy, the function will fit in
199	quite nicely into a call to C<pthread_atfork>:	428	quite nicely into a call to C<pthread_atfork>:
200		429
201	pthread_atfork (0, 0, ev_default_fork);	430	pthread_atfork (0, 0, ev_default_fork);
202		431
		432	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
		433	without calling this function, so if you force one of those backends you
		434	do not need to care.
		435
203	=item ev_loop_fork (loop)	436	=item ev_loop_fork (loop)
204		437
205	Like C<ev_default_fork>, but acts on an event loop created by	438	Like C<ev_default_fork>, but acts on an event loop created by
206	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	439	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
207	after fork, and how you do this is entirely your own problem.	440	after fork, and how you do this is entirely your own problem.
208		441
		442	=item unsigned int ev_loop_count (loop)
		443
		444	Returns the count of loop iterations for the loop, which is identical to
		445	the number of times libev did poll for new events. It starts at C<0> and
		446	happily wraps around with enough iterations.
		447
		448	This value can sometimes be useful as a generation counter of sorts (it
		449	"ticks" the number of loop iterations), as it roughly corresponds with
		450	C<ev_prepare> and C<ev_check> calls.
		451
209	=item unsigned int ev_method (loop)	452	=item unsigned int ev_backend (loop)
210		453
211	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	454	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
212	use.	455	use.
213		456
214	=item ev_tstamp ev_now (loop)	457	=item ev_tstamp ev_now (loop)
215		458
216	Returns the current "event loop time", which is the time the event loop	459	Returns the current "event loop time", which is the time the event loop
217	got events and started processing them. This timestamp does not change	460	received events and started processing them. This timestamp does not
218	as long as callbacks are being processed, and this is also the base time	461	change as long as callbacks are being processed, and this is also the base
219	used for relative timers. You can treat it as the timestamp of the event	462	time used for relative timers. You can treat it as the timestamp of the
220	occuring (or more correctly, the mainloop finding out about it).	463	event occuring (or more correctly, libev finding out about it).
221		464
222	=item ev_loop (loop, int flags)	465	=item ev_loop (loop, int flags)
223		466
224	Finally, this is it, the event handler. This function usually is called	467	Finally, this is it, the event handler. This function usually is called
225	after you initialised all your watchers and you want to start handling	468	after you initialised all your watchers and you want to start handling
226	events.	469	events.
227		470
228	If the flags argument is specified as 0, it will not return until either	471	If the flags argument is specified as C<0>, it will not return until
229	no event watchers are active anymore or C<ev_unloop> was called.	472	either no event watchers are active anymore or C<ev_unloop> was called.
		473
		474	Please note that an explicit C<ev_unloop> is usually better than
		475	relying on all watchers to be stopped when deciding when a program has
		476	finished (especially in interactive programs), but having a program that
		477	automatically loops as long as it has to and no longer by virtue of
		478	relying on its watchers stopping correctly is a thing of beauty.
230		479
231	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	480	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
232	those events and any outstanding ones, but will not block your process in	481	those events and any outstanding ones, but will not block your process in
233	case there are no events and will return after one iteration of the loop.	482	case there are no events and will return after one iteration of the loop.
234		483
235	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	484	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
236	neccessary) and will handle those and any outstanding ones. It will block	485	neccessary) and will handle those and any outstanding ones. It will block
237	your process until at least one new event arrives, and will return after	486	your process until at least one new event arrives, and will return after
238	one iteration of the loop.	487	one iteration of the loop. This is useful if you are waiting for some
		488	external event in conjunction with something not expressible using other
		489	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
		490	usually a better approach for this kind of thing.
239		491
240	This flags value could be used to implement alternative looping
241	constructs, but the C<prepare> and C<check> watchers provide a better and
242	more generic mechanism.
243
244	Here are the gory details of what ev_loop does:	492	Here are the gory details of what C<ev_loop> does:
245		493
		494	- Before the first iteration, call any pending watchers.
246	1. If there are no active watchers (reference count is zero), return.	495	* If there are no active watchers (reference count is zero), return.
247	2. Queue and immediately call all prepare watchers.	496	- Queue all prepare watchers and then call all outstanding watchers.
248	3. If we have been forked, recreate the kernel state.	497	- If we have been forked, recreate the kernel state.
249	4. Update the kernel state with all outstanding changes.	498	- Update the kernel state with all outstanding changes.
250	5. Update the "event loop time".	499	- Update the "event loop time".
251	6. Calculate for how long to block.	500	- Calculate for how long to block.
252	7. Block the process, waiting for events.	501	- Block the process, waiting for any events.
		502	- Queue all outstanding I/O (fd) events.
253	8. Update the "event loop time" and do time jump handling.	503	- Update the "event loop time" and do time jump handling.
254	9. Queue all outstanding timers.	504	- Queue all outstanding timers.
255	10. Queue all outstanding periodics.	505	- Queue all outstanding periodics.
256	11. If no events are pending now, queue all idle watchers.	506	- If no events are pending now, queue all idle watchers.
257	12. Queue all check watchers.	507	- Queue all check watchers.
258	13. Call all queued watchers in reverse order (i.e. check watchers first).	508	- Call all queued watchers in reverse order (i.e. check watchers first).
		509	Signals and child watchers are implemented as I/O watchers, and will
		510	be handled here by queueing them when their watcher gets executed.
259	14. If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	511	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
260	was used, return, otherwise continue with step #1.	512	were used, return, otherwise continue with step *.
		513
		514	Example: Queue some jobs and then loop until no events are outsanding
		515	anymore.
		516
		517	... queue jobs here, make sure they register event watchers as long
		518	... as they still have work to do (even an idle watcher will do..)
		519	ev_loop (my_loop, 0);
		520	... jobs done. yeah!
261		521
262	=item ev_unloop (loop, how)	522	=item ev_unloop (loop, how)
263		523
264	Can be used to make a call to C<ev_loop> return early (but only after it	524	Can be used to make a call to C<ev_loop> return early (but only after it
265	has processed all outstanding events). The C<how> argument must be either	525	has processed all outstanding events). The C<how> argument must be either
…		…
279	visible to the libev user and should not keep C<ev_loop> from exiting if	539	visible to the libev user and should not keep C<ev_loop> from exiting if
280	no event watchers registered by it are active. It is also an excellent	540	no event watchers registered by it are active. It is also an excellent
281	way to do this for generic recurring timers or from within third-party	541	way to do this for generic recurring timers or from within third-party
282	libraries. Just remember to I<unref after start> and I<ref before stop>.	542	libraries. Just remember to I<unref after start> and I<ref before stop>.
283		543
		544	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
		545	running when nothing else is active.
		546
		547	struct ev_signal exitsig;
		548	ev_signal_init (&exitsig, sig_cb, SIGINT);
		549	ev_signal_start (loop, &exitsig);
		550	evf_unref (loop);
		551
		552	Example: For some weird reason, unregister the above signal handler again.
		553
		554	ev_ref (loop);
		555	ev_signal_stop (loop, &exitsig);
		556
284	=back	557	=back
		558
285		559
286	=head1 ANATOMY OF A WATCHER	560	=head1 ANATOMY OF A WATCHER
287		561
288	A watcher is a structure that you create and register to record your	562	A watcher is a structure that you create and register to record your
289	interest in some event. For instance, if you want to wait for STDIN to	563	interest in some event. For instance, if you want to wait for STDIN to
…		…
322	*) >>), and you can stop watching for events at any time by calling the	596	*) >>), and you can stop watching for events at any time by calling the
323	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	597	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.
324		598
325	As long as your watcher is active (has been started but not stopped) you	599	As long as your watcher is active (has been started but not stopped) you
326	must not touch the values stored in it. Most specifically you must never	600	must not touch the values stored in it. Most specifically you must never
327	reinitialise it or call its set method.	601	reinitialise it or call its C<set> macro.
328
329	You can check whether an event is active by calling the C<ev_is_active
330	(watcher *)> macro. To see whether an event is outstanding (but the
331	callback for it has not been called yet) you can use the C<ev_is_pending
332	(watcher *)> macro.
333		602
334	Each and every callback receives the event loop pointer as first, the	603	Each and every callback receives the event loop pointer as first, the
335	registered watcher structure as second, and a bitset of received events as	604	registered watcher structure as second, and a bitset of received events as
336	third argument.	605	third argument.
337		606
…		…
361	The signal specified in the C<ev_signal> watcher has been received by a thread.	630	The signal specified in the C<ev_signal> watcher has been received by a thread.
362		631
363	=item C<EV_CHILD>	632	=item C<EV_CHILD>
364		633
365	The pid specified in the C<ev_child> watcher has received a status change.	634	The pid specified in the C<ev_child> watcher has received a status change.
		635
		636	=item C<EV_STAT>
		637
		638	The path specified in the C<ev_stat> watcher changed its attributes somehow.
366		639
367	=item C<EV_IDLE>	640	=item C<EV_IDLE>
368		641
369	The C<ev_idle> watcher has determined that you have nothing better to do.	642	The C<ev_idle> watcher has determined that you have nothing better to do.
370		643
…		…
378	received events. Callbacks of both watcher types can start and stop as	651	received events. Callbacks of both watcher types can start and stop as
379	many watchers as they want, and all of them will be taken into account	652	many watchers as they want, and all of them will be taken into account
380	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	653	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
381	C<ev_loop> from blocking).	654	C<ev_loop> from blocking).
382		655
		656	=item C<EV_EMBED>
		657
		658	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		659
		660	=item C<EV_FORK>
		661
		662	The event loop has been resumed in the child process after fork (see
		663	C<ev_fork>).
		664
383	=item C<EV_ERROR>	665	=item C<EV_ERROR>
384		666
385	An unspecified error has occured, the watcher has been stopped. This might	667	An unspecified error has occured, the watcher has been stopped. This might
386	happen because the watcher could not be properly started because libev	668	happen because the watcher could not be properly started because libev
387	ran out of memory, a file descriptor was found to be closed or any other	669	ran out of memory, a file descriptor was found to be closed or any other
…		…
393	your callbacks is well-written it can just attempt the operation and cope	675	your callbacks is well-written it can just attempt the operation and cope
394	with the error from read() or write(). This will not work in multithreaded	676	with the error from read() or write(). This will not work in multithreaded
395	programs, though, so beware.	677	programs, though, so beware.
396		678
397	=back	679	=back
		680
		681	=head2 GENERIC WATCHER FUNCTIONS
		682
		683	In the following description, C<TYPE> stands for the watcher type,
		684	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
		685
		686	=over 4
		687
		688	=item C<ev_init> (ev_TYPE *watcher, callback)
		689
		690	This macro initialises the generic portion of a watcher. The contents
		691	of the watcher object can be arbitrary (so C<malloc> will do). Only
		692	the generic parts of the watcher are initialised, you I<need> to call
		693	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		694	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		695	which rolls both calls into one.
		696
		697	You can reinitialise a watcher at any time as long as it has been stopped
		698	(or never started) and there are no pending events outstanding.
		699
		700	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,
		701	int revents)>.
		702
		703	=item C<ev_TYPE_set> (ev_TYPE *, [args])
		704
		705	This macro initialises the type-specific parts of a watcher. You need to
		706	call C<ev_init> at least once before you call this macro, but you can
		707	call C<ev_TYPE_set> any number of times. You must not, however, call this
		708	macro on a watcher that is active (it can be pending, however, which is a
		709	difference to the C<ev_init> macro).
		710
		711	Although some watcher types do not have type-specific arguments
		712	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		713
		714	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		715
		716	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		717	calls into a single call. This is the most convinient method to initialise
		718	a watcher. The same limitations apply, of course.
		719
		720	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
		721
		722	Starts (activates) the given watcher. Only active watchers will receive
		723	events. If the watcher is already active nothing will happen.
		724
		725	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
		726
		727	Stops the given watcher again (if active) and clears the pending
		728	status. It is possible that stopped watchers are pending (for example,
		729	non-repeating timers are being stopped when they become pending), but
		730	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If
		731	you want to free or reuse the memory used by the watcher it is therefore a
		732	good idea to always call its C<ev_TYPE_stop> function.
		733
		734	=item bool ev_is_active (ev_TYPE *watcher)
		735
		736	Returns a true value iff the watcher is active (i.e. it has been started
		737	and not yet been stopped). As long as a watcher is active you must not modify
		738	it.
		739
		740	=item bool ev_is_pending (ev_TYPE *watcher)
		741
		742	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		743	events but its callback has not yet been invoked). As long as a watcher
		744	is pending (but not active) you must not call an init function on it (but
		745	C<ev_TYPE_set> is safe), you must not change its priority, and you must
		746	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		747	it).
		748
		749	=item callback ev_cb (ev_TYPE *watcher)
		750
		751	Returns the callback currently set on the watcher.
		752
		753	=item ev_cb_set (ev_TYPE *watcher, callback)
		754
		755	Change the callback. You can change the callback at virtually any time
		756	(modulo threads).
		757
		758	=item ev_set_priority (ev_TYPE *watcher, priority)
		759
		760	=item int ev_priority (ev_TYPE *watcher)
		761
		762	Set and query the priority of the watcher. The priority is a small
		763	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		764	(default: C<-2>). Pending watchers with higher priority will be invoked
		765	before watchers with lower priority, but priority will not keep watchers
		766	from being executed (except for C<ev_idle> watchers).
		767
		768	This means that priorities are I<only> used for ordering callback
		769	invocation after new events have been received. This is useful, for
		770	example, to reduce latency after idling, or more often, to bind two
		771	watchers on the same event and make sure one is called first.
		772
		773	If you need to suppress invocation when higher priority events are pending
		774	you need to look at C<ev_idle> watchers, which provide this functionality.
		775
		776	You I<must not> change the priority of a watcher as long as it is active or
		777	pending.
		778
		779	The default priority used by watchers when no priority has been set is
		780	always C<0>, which is supposed to not be too high and not be too low :).
		781
		782	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		783	fine, as long as you do not mind that the priority value you query might
		784	or might not have been adjusted to be within valid range.
		785
		786	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		787
		788	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		789	C<loop> nor C<revents> need to be valid as long as the watcher callback
		790	can deal with that fact.
		791
		792	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		793
		794	If the watcher is pending, this function returns clears its pending status
		795	and returns its C<revents> bitset (as if its callback was invoked). If the
		796	watcher isn't pending it does nothing and returns C<0>.
		797
		798	=back
		799
398		800
399	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	801	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
400		802
401	Each watcher has, by default, a member C<void *data> that you can change	803	Each watcher has, by default, a member C<void *data> that you can change
402	and read at any time, libev will completely ignore it. This can be used	804	and read at any time, libev will completely ignore it. This can be used
…		…
420	{	822	{
421	struct my_io *w = (struct my_io *)w_;	823	struct my_io *w = (struct my_io *)w_;
422	...	824	...
423	}	825	}
424		826
425	More interesting and less C-conformant ways of catsing your callback type	827	More interesting and less C-conformant ways of casting your callback type
426	have been omitted....	828	instead have been omitted.
		829
		830	Another common scenario is having some data structure with multiple
		831	watchers:
		832
		833	struct my_biggy
		834	{
		835	int some_data;
		836	ev_timer t1;
		837	ev_timer t2;
		838	}
		839
		840	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		841	you need to use C<offsetof>:
		842
		843	#include <stddef.h>
		844
		845	static void
		846	t1_cb (EV_P_ struct ev_timer *w, int revents)
		847	{
		848	struct my_biggy big = (struct my_biggy *
		849	(((char *)w) - offsetof (struct my_biggy, t1));
		850	}
		851
		852	static void
		853	t2_cb (EV_P_ struct ev_timer *w, int revents)
		854	{
		855	struct my_biggy big = (struct my_biggy *
		856	(((char *)w) - offsetof (struct my_biggy, t2));
		857	}
427		858
428		859
429	=head1 WATCHER TYPES	860	=head1 WATCHER TYPES
430		861
431	This section describes each watcher in detail, but will not repeat	862	This section describes each watcher in detail, but will not repeat
432	information given in the last section.	863	information given in the last section. Any initialisation/set macros,
		864	functions and members specific to the watcher type are explained.
433		865
		866	Members are additionally marked with either I<[read-only]>, meaning that,
		867	while the watcher is active, you can look at the member and expect some
		868	sensible content, but you must not modify it (you can modify it while the
		869	watcher is stopped to your hearts content), or I<[read-write]>, which
		870	means you can expect it to have some sensible content while the watcher
		871	is active, but you can also modify it. Modifying it may not do something
		872	sensible or take immediate effect (or do anything at all), but libev will
		873	not crash or malfunction in any way.
		874
		875
434	=head2 C<ev_io> - is this file descriptor readable or writable	876	=head2 C<ev_io> - is this file descriptor readable or writable?
435		877
436	I/O watchers check whether a file descriptor is readable or writable	878	I/O watchers check whether a file descriptor is readable or writable
437	in each iteration of the event loop (This behaviour is called	879	in each iteration of the event loop, or, more precisely, when reading
438	level-triggering because you keep receiving events as long as the	880	would not block the process and writing would at least be able to write
439	condition persists. Remember you can stop the watcher if you don't want to	881	some data. This behaviour is called level-triggering because you keep
440	act on the event and neither want to receive future events).	882	receiving events as long as the condition persists. Remember you can stop
		883	the watcher if you don't want to act on the event and neither want to
		884	receive future events.
441		885
442	In general you can register as many read and/or write event watchers per	886	In general you can register as many read and/or write event watchers per
443	fd as you want (as long as you don't confuse yourself). Setting all file	887	fd as you want (as long as you don't confuse yourself). Setting all file
444	descriptors to non-blocking mode is also usually a good idea (but not	888	descriptors to non-blocking mode is also usually a good idea (but not
445	required if you know what you are doing).	889	required if you know what you are doing).
446		890
447	You have to be careful with dup'ed file descriptors, though. Some backends	891	You have to be careful with dup'ed file descriptors, though. Some backends
448	(the linux epoll backend is a notable example) cannot handle dup'ed file	892	(the linux epoll backend is a notable example) cannot handle dup'ed file
449	descriptors correctly if you register interest in two or more fds pointing	893	descriptors correctly if you register interest in two or more fds pointing
450	to the same underlying file/socket etc. description (that is, they share	894	to the same underlying file/socket/etc. description (that is, they share
451	the same underlying "file open").	895	the same underlying "file open").
452		896
453	If you must do this, then force the use of a known-to-be-good backend	897	If you must do this, then force the use of a known-to-be-good backend
454	(at the time of this writing, this includes only EVMETHOD_SELECT and	898	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
455	EVMETHOD_POLL).	899	C<EVBACKEND_POLL>).
		900
		901	Another thing you have to watch out for is that it is quite easy to
		902	receive "spurious" readyness notifications, that is your callback might
		903	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		904	because there is no data. Not only are some backends known to create a
		905	lot of those (for example solaris ports), it is very easy to get into
		906	this situation even with a relatively standard program structure. Thus
		907	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		908	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		909
		910	If you cannot run the fd in non-blocking mode (for example you should not
		911	play around with an Xlib connection), then you have to seperately re-test
		912	whether a file descriptor is really ready with a known-to-be good interface
		913	such as poll (fortunately in our Xlib example, Xlib already does this on
		914	its own, so its quite safe to use).
		915
		916	=head3 The special problem of disappearing file descriptors
		917
		918	Some backends (e.g kqueue, epoll) need to be told about closing a file
		919	descriptor (either by calling C<close> explicitly or by any other means,
		920	such as C<dup>). The reason is that you register interest in some file
		921	descriptor, but when it goes away, the operating system will silently drop
		922	this interest. If another file descriptor with the same number then is
		923	registered with libev, there is no efficient way to see that this is, in
		924	fact, a different file descriptor.
		925
		926	To avoid having to explicitly tell libev about such cases, libev follows
		927	the following policy: Each time C<ev_io_set> is being called, libev
		928	will assume that this is potentially a new file descriptor, otherwise
		929	it is assumed that the file descriptor stays the same. That means that
		930	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		931	descriptor even if the file descriptor number itself did not change.
		932
		933	This is how one would do it normally anyway, the important point is that
		934	the libev application should not optimise around libev but should leave
		935	optimisations to libev.
		936
		937
		938	=head3 Watcher-Specific Functions
456		939
457	=over 4	940	=over 4
458		941
459	=item ev_io_init (ev_io *, callback, int fd, int events)	942	=item ev_io_init (ev_io *, callback, int fd, int events)
460		943
461	=item ev_io_set (ev_io *, int fd, int events)	944	=item ev_io_set (ev_io *, int fd, int events)
462		945
463	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	946	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
464	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	947	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or
465	EV_WRITE> to receive the given events.	948	C<EV_READ | EV_WRITE> to receive the given events.
		949
		950	=item int fd [read-only]
		951
		952	The file descriptor being watched.
		953
		954	=item int events [read-only]
		955
		956	The events being watched.
466		957
467	=back	958	=back
468		959
		960	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
		961	readable, but only once. Since it is likely line-buffered, you could
		962	attempt to read a whole line in the callback.
		963
		964	static void
		965	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		966	{
		967	ev_io_stop (loop, w);
		968	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		969	}
		970
		971	...
		972	struct ev_loop *loop = ev_default_init (0);
		973	struct ev_io stdin_readable;
		974	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		975	ev_io_start (loop, &stdin_readable);
		976	ev_loop (loop, 0);
		977
		978
469	=head2 C<ev_timer> - relative and optionally recurring timeouts	979	=head2 C<ev_timer> - relative and optionally repeating timeouts
470		980
471	Timer watchers are simple relative timers that generate an event after a	981	Timer watchers are simple relative timers that generate an event after a
472	given time, and optionally repeating in regular intervals after that.	982	given time, and optionally repeating in regular intervals after that.
473		983
474	The timers are based on real time, that is, if you register an event that	984	The timers are based on real time, that is, if you register an event that
475	times out after an hour and you reset your system clock to last years	985	times out after an hour and you reset your system clock to last years
476	time, it will still time out after (roughly) and hour. "Roughly" because	986	time, it will still time out after (roughly) and hour. "Roughly" because
477	detecting time jumps is hard, and soem inaccuracies are unavoidable (the	987	detecting time jumps is hard, and some inaccuracies are unavoidable (the
478	monotonic clock option helps a lot here).	988	monotonic clock option helps a lot here).
479		989
480	The relative timeouts are calculated relative to the C<ev_now ()>	990	The relative timeouts are calculated relative to the C<ev_now ()>
481	time. This is usually the right thing as this timestamp refers to the time	991	time. This is usually the right thing as this timestamp refers to the time
482	of the event triggering whatever timeout you are modifying/starting. If	992	of the event triggering whatever timeout you are modifying/starting. If
483	you suspect event processing to be delayed and you *need* to base the timeout	993	you suspect event processing to be delayed and you I<need> to base the timeout
484	on the current time, use something like this to adjust for this:	994	on the current time, use something like this to adjust for this:
485		995
486	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	996	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
		997
		998	The callback is guarenteed to be invoked only when its timeout has passed,
		999	but if multiple timers become ready during the same loop iteration then
		1000	order of execution is undefined.
		1001
		1002	=head3 Watcher-Specific Functions and Data Members
487		1003
488	=over 4	1004	=over 4
489		1005
490	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1006	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
491		1007
…		…
505	=item ev_timer_again (loop)	1021	=item ev_timer_again (loop)
506		1022
507	This will act as if the timer timed out and restart it again if it is	1023	This will act as if the timer timed out and restart it again if it is
508	repeating. The exact semantics are:	1024	repeating. The exact semantics are:
509		1025
		1026	If the timer is pending, its pending status is cleared.
		1027
510	If the timer is started but nonrepeating, stop it.	1028	If the timer is started but nonrepeating, stop it (as if it timed out).
511		1029
512	If the timer is repeating, either start it if necessary (with the repeat	1030	If the timer is repeating, either start it if necessary (with the
513	value), or reset the running timer to the repeat value.	1031	C<repeat> value), or reset the running timer to the C<repeat> value.
514		1032
515	This sounds a bit complicated, but here is a useful and typical	1033	This sounds a bit complicated, but here is a useful and typical
516	example: Imagine you have a tcp connection and you want a so-called idle	1034	example: Imagine you have a tcp connection and you want a so-called idle
517	timeout, that is, you want to be called when there have been, say, 60	1035	timeout, that is, you want to be called when there have been, say, 60
518	seconds of inactivity on the socket. The easiest way to do this is to	1036	seconds of inactivity on the socket. The easiest way to do this is to
519	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	1037	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
520	time you successfully read or write some data. If you go into an idle	1038	C<ev_timer_again> each time you successfully read or write some data. If
521	state where you do not expect data to travel on the socket, you can stop	1039	you go into an idle state where you do not expect data to travel on the
522	the timer, and again will automatically restart it if need be.	1040	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
		1041	automatically restart it if need be.
		1042
		1043	That means you can ignore the C<after> value and C<ev_timer_start>
		1044	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
		1045
		1046	ev_timer_init (timer, callback, 0., 5.);
		1047	ev_timer_again (loop, timer);
		1048	...
		1049	timer->again = 17.;
		1050	ev_timer_again (loop, timer);
		1051	...
		1052	timer->again = 10.;
		1053	ev_timer_again (loop, timer);
		1054
		1055	This is more slightly efficient then stopping/starting the timer each time
		1056	you want to modify its timeout value.
		1057
		1058	=item ev_tstamp repeat [read-write]
		1059
		1060	The current C<repeat> value. Will be used each time the watcher times out
		1061	or C<ev_timer_again> is called and determines the next timeout (if any),
		1062	which is also when any modifications are taken into account.
523		1063
524	=back	1064	=back
525		1065
		1066	Example: Create a timer that fires after 60 seconds.
		1067
		1068	static void
		1069	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1070	{
		1071	.. one minute over, w is actually stopped right here
		1072	}
		1073
		1074	struct ev_timer mytimer;
		1075	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		1076	ev_timer_start (loop, &mytimer);
		1077
		1078	Example: Create a timeout timer that times out after 10 seconds of
		1079	inactivity.
		1080
		1081	static void
		1082	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1083	{
		1084	.. ten seconds without any activity
		1085	}
		1086
		1087	struct ev_timer mytimer;
		1088	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		1089	ev_timer_again (&mytimer); /* start timer */
		1090	ev_loop (loop, 0);
		1091
		1092	// and in some piece of code that gets executed on any "activity":
		1093	// reset the timeout to start ticking again at 10 seconds
		1094	ev_timer_again (&mytimer);
		1095
		1096
526	=head2 C<ev_periodic> - to cron or not to cron	1097	=head2 C<ev_periodic> - to cron or not to cron?
527		1098
528	Periodic watchers are also timers of a kind, but they are very versatile	1099	Periodic watchers are also timers of a kind, but they are very versatile
529	(and unfortunately a bit complex).	1100	(and unfortunately a bit complex).
530		1101
531	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1102	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
532	but on wallclock time (absolute time). You can tell a periodic watcher	1103	but on wallclock time (absolute time). You can tell a periodic watcher
533	to trigger "at" some specific point in time. For example, if you tell a	1104	to trigger "at" some specific point in time. For example, if you tell a
534	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	1105	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
535	+ 10.>) and then reset your system clock to the last year, then it will	1106	+ 10.>) and then reset your system clock to the last year, then it will
536	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1107	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
537	roughly 10 seconds later and of course not if you reset your system time	1108	roughly 10 seconds later).
538	again).
539		1109
540	They can also be used to implement vastly more complex timers, such as	1110	They can also be used to implement vastly more complex timers, such as
541	triggering an event on eahc midnight, local time.	1111	triggering an event on each midnight, local time or other, complicated,
		1112	rules.
		1113
		1114	As with timers, the callback is guarenteed to be invoked only when the
		1115	time (C<at>) has been passed, but if multiple periodic timers become ready
		1116	during the same loop iteration then order of execution is undefined.
		1117
		1118	=head3 Watcher-Specific Functions and Data Members
542		1119
543	=over 4	1120	=over 4
544		1121
545	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1122	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
546		1123
547	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1124	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
548		1125
549	Lots of arguments, lets sort it out... There are basically three modes of	1126	Lots of arguments, lets sort it out... There are basically three modes of
550	operation, and we will explain them from simplest to complex:	1127	operation, and we will explain them from simplest to complex:
551		1128
552
553	=over 4	1129	=over 4
554		1130
555	=item * absolute timer (interval = reschedule_cb = 0)	1131	=item * absolute timer (at = time, interval = reschedule_cb = 0)
556		1132
557	In this configuration the watcher triggers an event at the wallclock time	1133	In this configuration the watcher triggers an event at the wallclock time
558	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1134	C<at> and doesn't repeat. It will not adjust when a time jump occurs,
559	that is, if it is to be run at January 1st 2011 then it will run when the	1135	that is, if it is to be run at January 1st 2011 then it will run when the
560	system time reaches or surpasses this time.	1136	system time reaches or surpasses this time.
561		1137
562	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1138	=item * non-repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
563		1139
564	In this mode the watcher will always be scheduled to time out at the next	1140	In this mode the watcher will always be scheduled to time out at the next
565	C<at + N * interval> time (for some integer N) and then repeat, regardless	1141	C<at + N * interval> time (for some integer N, which can also be negative)
566	of any time jumps.	1142	and then repeat, regardless of any time jumps.
567		1143
568	This can be used to create timers that do not drift with respect to system	1144	This can be used to create timers that do not drift with respect to system
569	time:	1145	time:
570		1146
571	ev_periodic_set (&periodic, 0., 3600., 0);	1147	ev_periodic_set (&periodic, 0., 3600., 0);
…		…
577		1153
578	Another way to think about it (for the mathematically inclined) is that	1154	Another way to think about it (for the mathematically inclined) is that
579	C<ev_periodic> will try to run the callback in this mode at the next possible	1155	C<ev_periodic> will try to run the callback in this mode at the next possible
580	time where C<time = at (mod interval)>, regardless of any time jumps.	1156	time where C<time = at (mod interval)>, regardless of any time jumps.
581		1157
		1158	For numerical stability it is preferable that the C<at> value is near
		1159	C<ev_now ()> (the current time), but there is no range requirement for
		1160	this value.
		1161
582	=item * manual reschedule mode (reschedule_cb = callback)	1162	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
583		1163
584	In this mode the values for C<interval> and C<at> are both being	1164	In this mode the values for C<interval> and C<at> are both being
585	ignored. Instead, each time the periodic watcher gets scheduled, the	1165	ignored. Instead, each time the periodic watcher gets scheduled, the
586	reschedule callback will be called with the watcher as first, and the	1166	reschedule callback will be called with the watcher as first, and the
587	current time as second argument.	1167	current time as second argument.
588		1168
589	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1169	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
590	ever, or make any event loop modifications>. If you need to stop it,	1170	ever, or make any event loop modifications>. If you need to stop it,
591	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	1171	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
592	starting a prepare watcher).	1172	starting an C<ev_prepare> watcher, which is legal).
593		1173
594	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1174	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
595	ev_tstamp now)>, e.g.:	1175	ev_tstamp now)>, e.g.:
596		1176
597	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1177	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
…		…
620	Simply stops and restarts the periodic watcher again. This is only useful	1200	Simply stops and restarts the periodic watcher again. This is only useful
621	when you changed some parameters or the reschedule callback would return	1201	when you changed some parameters or the reschedule callback would return
622	a different time than the last time it was called (e.g. in a crond like	1202	a different time than the last time it was called (e.g. in a crond like
623	program when the crontabs have changed).	1203	program when the crontabs have changed).
624		1204
		1205	=item ev_tstamp offset [read-write]
		1206
		1207	When repeating, this contains the offset value, otherwise this is the
		1208	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1209
		1210	Can be modified any time, but changes only take effect when the periodic
		1211	timer fires or C<ev_periodic_again> is being called.
		1212
		1213	=item ev_tstamp interval [read-write]
		1214
		1215	The current interval value. Can be modified any time, but changes only
		1216	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1217	called.
		1218
		1219	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]
		1220
		1221	The current reschedule callback, or C<0>, if this functionality is
		1222	switched off. Can be changed any time, but changes only take effect when
		1223	the periodic timer fires or C<ev_periodic_again> is being called.
		1224
		1225	=item ev_tstamp at [read-only]
		1226
		1227	When active, contains the absolute time that the watcher is supposed to
		1228	trigger next.
		1229
625	=back	1230	=back
626		1231
		1232	Example: Call a callback every hour, or, more precisely, whenever the
		1233	system clock is divisible by 3600. The callback invocation times have
		1234	potentially a lot of jittering, but good long-term stability.
		1235
		1236	static void
		1237	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		1238	{
		1239	... its now a full hour (UTC, or TAI or whatever your clock follows)
		1240	}
		1241
		1242	struct ev_periodic hourly_tick;
		1243	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		1244	ev_periodic_start (loop, &hourly_tick);
		1245
		1246	Example: The same as above, but use a reschedule callback to do it:
		1247
		1248	#include <math.h>
		1249
		1250	static ev_tstamp
		1251	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		1252	{
		1253	return fmod (now, 3600.) + 3600.;
		1254	}
		1255
		1256	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		1257
		1258	Example: Call a callback every hour, starting now:
		1259
		1260	struct ev_periodic hourly_tick;
		1261	ev_periodic_init (&hourly_tick, clock_cb,
		1262	fmod (ev_now (loop), 3600.), 3600., 0);
		1263	ev_periodic_start (loop, &hourly_tick);
		1264
		1265
627	=head2 C<ev_signal> - signal me when a signal gets signalled	1266	=head2 C<ev_signal> - signal me when a signal gets signalled!
628		1267
629	Signal watchers will trigger an event when the process receives a specific	1268	Signal watchers will trigger an event when the process receives a specific
630	signal one or more times. Even though signals are very asynchronous, libev	1269	signal one or more times. Even though signals are very asynchronous, libev
631	will try it's best to deliver signals synchronously, i.e. as part of the	1270	will try it's best to deliver signals synchronously, i.e. as part of the
632	normal event processing, like any other event.	1271	normal event processing, like any other event.
…		…
636	with the kernel (thus it coexists with your own signal handlers as long	1275	with the kernel (thus it coexists with your own signal handlers as long
637	as you don't register any with libev). Similarly, when the last signal	1276	as you don't register any with libev). Similarly, when the last signal
638	watcher for a signal is stopped libev will reset the signal handler to	1277	watcher for a signal is stopped libev will reset the signal handler to
639	SIG_DFL (regardless of what it was set to before).	1278	SIG_DFL (regardless of what it was set to before).
640		1279
		1280	=head3 Watcher-Specific Functions and Data Members
		1281
641	=over 4	1282	=over 4
642		1283
643	=item ev_signal_init (ev_signal *, callback, int signum)	1284	=item ev_signal_init (ev_signal *, callback, int signum)
644		1285
645	=item ev_signal_set (ev_signal *, int signum)	1286	=item ev_signal_set (ev_signal *, int signum)
646		1287
647	Configures the watcher to trigger on the given signal number (usually one	1288	Configures the watcher to trigger on the given signal number (usually one
648	of the C<SIGxxx> constants).	1289	of the C<SIGxxx> constants).
649		1290
		1291	=item int signum [read-only]
		1292
		1293	The signal the watcher watches out for.
		1294
650	=back	1295	=back
651		1296
		1297
652	=head2 C<ev_child> - wait for pid status changes	1298	=head2 C<ev_child> - watch out for process status changes
653		1299
654	Child watchers trigger when your process receives a SIGCHLD in response to	1300	Child watchers trigger when your process receives a SIGCHLD in response to
655	some child status changes (most typically when a child of yours dies).	1301	some child status changes (most typically when a child of yours dies).
		1302
		1303	=head3 Watcher-Specific Functions and Data Members
656		1304
657	=over 4	1305	=over 4
658		1306
659	=item ev_child_init (ev_child *, callback, int pid)	1307	=item ev_child_init (ev_child *, callback, int pid)
660		1308
…		…
665	at the C<rstatus> member of the C<ev_child> watcher structure to see	1313	at the C<rstatus> member of the C<ev_child> watcher structure to see
666	the status word (use the macros from C<sys/wait.h> and see your systems	1314	the status word (use the macros from C<sys/wait.h> and see your systems
667	C<waitpid> documentation). The C<rpid> member contains the pid of the	1315	C<waitpid> documentation). The C<rpid> member contains the pid of the
668	process causing the status change.	1316	process causing the status change.
669		1317
		1318	=item int pid [read-only]
		1319
		1320	The process id this watcher watches out for, or C<0>, meaning any process id.
		1321
		1322	=item int rpid [read-write]
		1323
		1324	The process id that detected a status change.
		1325
		1326	=item int rstatus [read-write]
		1327
		1328	The process exit/trace status caused by C<rpid> (see your systems
		1329	C<waitpid> and C<sys/wait.h> documentation for details).
		1330
670	=back	1331	=back
671		1332
		1333	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1334
		1335	static void
		1336	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1337	{
		1338	ev_unloop (loop, EVUNLOOP_ALL);
		1339	}
		1340
		1341	struct ev_signal signal_watcher;
		1342	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1343	ev_signal_start (loop, &sigint_cb);
		1344
		1345
		1346	=head2 C<ev_stat> - did the file attributes just change?
		1347
		1348	This watches a filesystem path for attribute changes. That is, it calls
		1349	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1350	compared to the last time, invoking the callback if it did.
		1351
		1352	The path does not need to exist: changing from "path exists" to "path does
		1353	not exist" is a status change like any other. The condition "path does
		1354	not exist" is signified by the C<st_nlink> field being zero (which is
		1355	otherwise always forced to be at least one) and all the other fields of
		1356	the stat buffer having unspecified contents.
		1357
		1358	The path I<should> be absolute and I<must not> end in a slash. If it is
		1359	relative and your working directory changes, the behaviour is undefined.
		1360
		1361	Since there is no standard to do this, the portable implementation simply
		1362	calls C<stat (2)> regularly on the path to see if it changed somehow. You
		1363	can specify a recommended polling interval for this case. If you specify
		1364	a polling interval of C<0> (highly recommended!) then a I<suitable,
		1365	unspecified default> value will be used (which you can expect to be around
		1366	five seconds, although this might change dynamically). Libev will also
		1367	impose a minimum interval which is currently around C<0.1>, but thats
		1368	usually overkill.
		1369
		1370	This watcher type is not meant for massive numbers of stat watchers,
		1371	as even with OS-supported change notifications, this can be
		1372	resource-intensive.
		1373
		1374	At the time of this writing, only the Linux inotify interface is
		1375	implemented (implementing kqueue support is left as an exercise for the
		1376	reader). Inotify will be used to give hints only and should not change the
		1377	semantics of C<ev_stat> watchers, which means that libev sometimes needs
		1378	to fall back to regular polling again even with inotify, but changes are
		1379	usually detected immediately, and if the file exists there will be no
		1380	polling.
		1381
		1382	=head3 Watcher-Specific Functions and Data Members
		1383
		1384	=over 4
		1385
		1386	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		1387
		1388	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		1389
		1390	Configures the watcher to wait for status changes of the given
		1391	C<path>. The C<interval> is a hint on how quickly a change is expected to
		1392	be detected and should normally be specified as C<0> to let libev choose
		1393	a suitable value. The memory pointed to by C<path> must point to the same
		1394	path for as long as the watcher is active.
		1395
		1396	The callback will be receive C<EV_STAT> when a change was detected,
		1397	relative to the attributes at the time the watcher was started (or the
		1398	last change was detected).
		1399
		1400	=item ev_stat_stat (ev_stat *)
		1401
		1402	Updates the stat buffer immediately with new values. If you change the
		1403	watched path in your callback, you could call this fucntion to avoid
		1404	detecting this change (while introducing a race condition). Can also be
		1405	useful simply to find out the new values.
		1406
		1407	=item ev_statdata attr [read-only]
		1408
		1409	The most-recently detected attributes of the file. Although the type is of
		1410	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		1411	suitable for your system. If the C<st_nlink> member is C<0>, then there
		1412	was some error while C<stat>ing the file.
		1413
		1414	=item ev_statdata prev [read-only]
		1415
		1416	The previous attributes of the file. The callback gets invoked whenever
		1417	C<prev> != C<attr>.
		1418
		1419	=item ev_tstamp interval [read-only]
		1420
		1421	The specified interval.
		1422
		1423	=item const char *path [read-only]
		1424
		1425	The filesystem path that is being watched.
		1426
		1427	=back
		1428
		1429	Example: Watch C</etc/passwd> for attribute changes.
		1430
		1431	static void
		1432	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		1433	{
		1434	/* /etc/passwd changed in some way */
		1435	if (w->attr.st_nlink)
		1436	{
		1437	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1438	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1439	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1440	}
		1441	else
		1442	/* you shalt not abuse printf for puts */
		1443	puts ("wow, /etc/passwd is not there, expect problems. "
		1444	"if this is windows, they already arrived\n");
		1445	}
		1446
		1447	...
		1448	ev_stat passwd;
		1449
		1450	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
		1451	ev_stat_start (loop, &passwd);
		1452
		1453
672	=head2 C<ev_idle> - when you've got nothing better to do	1454	=head2 C<ev_idle> - when you've got nothing better to do...
673		1455
674	Idle watchers trigger events when there are no other events are pending	1456	Idle watchers trigger events when no other events of the same or higher
675	(prepare, check and other idle watchers do not count). That is, as long	1457	priority are pending (prepare, check and other idle watchers do not
676	as your process is busy handling sockets or timeouts (or even signals,	1458	count).
677	imagine) it will not be triggered. But when your process is idle all idle	1459
678	watchers are being called again and again, once per event loop iteration -	1460	That is, as long as your process is busy handling sockets or timeouts
		1461	(or even signals, imagine) of the same or higher priority it will not be
		1462	triggered. But when your process is idle (or only lower-priority watchers
		1463	are pending), the idle watchers are being called once per event loop
679	until stopped, that is, or your process receives more events and becomes	1464	iteration - until stopped, that is, or your process receives more events
680	busy.	1465	and becomes busy again with higher priority stuff.
681		1466
682	The most noteworthy effect is that as long as any idle watchers are	1467	The most noteworthy effect is that as long as any idle watchers are
683	active, the process will not block when waiting for new events.	1468	active, the process will not block when waiting for new events.
684		1469
685	Apart from keeping your process non-blocking (which is a useful	1470	Apart from keeping your process non-blocking (which is a useful
686	effect on its own sometimes), idle watchers are a good place to do	1471	effect on its own sometimes), idle watchers are a good place to do
687	"pseudo-background processing", or delay processing stuff to after the	1472	"pseudo-background processing", or delay processing stuff to after the
688	event loop has handled all outstanding events.	1473	event loop has handled all outstanding events.
689		1474
		1475	=head3 Watcher-Specific Functions and Data Members
		1476
690	=over 4	1477	=over 4
691		1478
692	=item ev_idle_init (ev_signal *, callback)	1479	=item ev_idle_init (ev_signal *, callback)
693		1480
694	Initialises and configures the idle watcher - it has no parameters of any	1481	Initialises and configures the idle watcher - it has no parameters of any
695	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1482	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
696	believe me.	1483	believe me.
697		1484
698	=back	1485	=back
699		1486
		1487	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
		1488	callback, free it. Also, use no error checking, as usual.
		1489
		1490	static void
		1491	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
		1492	{
		1493	free (w);
		1494	// now do something you wanted to do when the program has
		1495	// no longer asnything immediate to do.
		1496	}
		1497
		1498	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1499	ev_idle_init (idle_watcher, idle_cb);
		1500	ev_idle_start (loop, idle_cb);
		1501
		1502
700	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1503	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
701		1504
702	Prepare and check watchers are usually (but not always) used in tandem:	1505	Prepare and check watchers are usually (but not always) used in tandem:
703	prepare watchers get invoked before the process blocks and check watchers	1506	prepare watchers get invoked before the process blocks and check watchers
704	afterwards.	1507	afterwards.
705		1508
		1509	You I<must not> call C<ev_loop> or similar functions that enter
		1510	the current event loop from either C<ev_prepare> or C<ev_check>
		1511	watchers. Other loops than the current one are fine, however. The
		1512	rationale behind this is that you do not need to check for recursion in
		1513	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		1514	C<ev_check> so if you have one watcher of each kind they will always be
		1515	called in pairs bracketing the blocking call.
		1516
706	Their main purpose is to integrate other event mechanisms into libev. This	1517	Their main purpose is to integrate other event mechanisms into libev and
707	could be used, for example, to track variable changes, implement your own	1518	their use is somewhat advanced. This could be used, for example, to track
708	watchers, integrate net-snmp or a coroutine library and lots more.	1519	variable changes, implement your own watchers, integrate net-snmp or a
		1520	coroutine library and lots more. They are also occasionally useful if
		1521	you cache some data and want to flush it before blocking (for example,
		1522	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		1523	watcher).
709		1524
710	This is done by examining in each prepare call which file descriptors need	1525	This is done by examining in each prepare call which file descriptors need
711	to be watched by the other library, registering C<ev_io> watchers for	1526	to be watched by the other library, registering C<ev_io> watchers for
712	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1527	them and starting an C<ev_timer> watcher for any timeouts (many libraries
713	provide just this functionality). Then, in the check watcher you check for	1528	provide just this functionality). Then, in the check watcher you check for
…		…
723	with priority higher than or equal to the event loop and one coroutine	1538	with priority higher than or equal to the event loop and one coroutine
724	of lower priority, but only once, using idle watchers to keep the event	1539	of lower priority, but only once, using idle watchers to keep the event
725	loop from blocking if lower-priority coroutines are active, thus mapping	1540	loop from blocking if lower-priority coroutines are active, thus mapping
726	low-priority coroutines to idle/background tasks).	1541	low-priority coroutines to idle/background tasks).
727		1542
		1543	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1544	priority, to ensure that they are being run before any other watchers
		1545	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1546	too) should not activate ("feed") events into libev. While libev fully
		1547	supports this, they will be called before other C<ev_check> watchers did
		1548	their job. As C<ev_check> watchers are often used to embed other event
		1549	loops those other event loops might be in an unusable state until their
		1550	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		1551	others).
		1552
		1553	=head3 Watcher-Specific Functions and Data Members
		1554
728	=over 4	1555	=over 4
729		1556
730	=item ev_prepare_init (ev_prepare *, callback)	1557	=item ev_prepare_init (ev_prepare *, callback)
731		1558
732	=item ev_check_init (ev_check *, callback)	1559	=item ev_check_init (ev_check *, callback)
…		…
734	Initialises and configures the prepare or check watcher - they have no	1561	Initialises and configures the prepare or check watcher - they have no
735	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1562	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
736	macros, but using them is utterly, utterly and completely pointless.	1563	macros, but using them is utterly, utterly and completely pointless.
737		1564
738	=back	1565	=back
		1566
		1567	There are a number of principal ways to embed other event loops or modules
		1568	into libev. Here are some ideas on how to include libadns into libev
		1569	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1570	use for an actually working example. Another Perl module named C<EV::Glib>
		1571	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV
		1572	into the Glib event loop).
		1573
		1574	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		1575	and in a check watcher, destroy them and call into libadns. What follows
		1576	is pseudo-code only of course. This requires you to either use a low
		1577	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1578	the callbacks for the IO/timeout watchers might not have been called yet.
		1579
		1580	static ev_io iow [nfd];
		1581	static ev_timer tw;
		1582
		1583	static void
		1584	io_cb (ev_loop *loop, ev_io *w, int revents)
		1585	{
		1586	}
		1587
		1588	// create io watchers for each fd and a timer before blocking
		1589	static void
		1590	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
		1591	{
		1592	int timeout = 3600000;
		1593	struct pollfd fds [nfd];
		1594	// actual code will need to loop here and realloc etc.
		1595	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1596
		1597	/* the callback is illegal, but won't be called as we stop during check */
		1598	ev_timer_init (&tw, 0, timeout * 1e-3);
		1599	ev_timer_start (loop, &tw);
		1600
		1601	// create one ev_io per pollfd
		1602	for (int i = 0; i < nfd; ++i)
		1603	{
		1604	ev_io_init (iow + i, io_cb, fds [i].fd,
		1605	((fds [i].events & POLLIN ? EV_READ : 0)
		1606	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1607
		1608	fds [i].revents = 0;
		1609	ev_io_start (loop, iow + i);
		1610	}
		1611	}
		1612
		1613	// stop all watchers after blocking
		1614	static void
		1615	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
		1616	{
		1617	ev_timer_stop (loop, &tw);
		1618
		1619	for (int i = 0; i < nfd; ++i)
		1620	{
		1621	// set the relevant poll flags
		1622	// could also call adns_processreadable etc. here
		1623	struct pollfd *fd = fds + i;
		1624	int revents = ev_clear_pending (iow + i);
		1625	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1626	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1627
		1628	// now stop the watcher
		1629	ev_io_stop (loop, iow + i);
		1630	}
		1631
		1632	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1633	}
		1634
		1635	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		1636	in the prepare watcher and would dispose of the check watcher.
		1637
		1638	Method 3: If the module to be embedded supports explicit event
		1639	notification (adns does), you can also make use of the actual watcher
		1640	callbacks, and only destroy/create the watchers in the prepare watcher.
		1641
		1642	static void
		1643	timer_cb (EV_P_ ev_timer *w, int revents)
		1644	{
		1645	adns_state ads = (adns_state)w->data;
		1646	update_now (EV_A);
		1647
		1648	adns_processtimeouts (ads, &tv_now);
		1649	}
		1650
		1651	static void
		1652	io_cb (EV_P_ ev_io *w, int revents)
		1653	{
		1654	adns_state ads = (adns_state)w->data;
		1655	update_now (EV_A);
		1656
		1657	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		1658	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		1659	}
		1660
		1661	// do not ever call adns_afterpoll
		1662
		1663	Method 4: Do not use a prepare or check watcher because the module you
		1664	want to embed is too inflexible to support it. Instead, youc na override
		1665	their poll function. The drawback with this solution is that the main
		1666	loop is now no longer controllable by EV. The C<Glib::EV> module does
		1667	this.
		1668
		1669	static gint
		1670	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		1671	{
		1672	int got_events = 0;
		1673
		1674	for (n = 0; n < nfds; ++n)
		1675	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		1676
		1677	if (timeout >= 0)
		1678	// create/start timer
		1679
		1680	// poll
		1681	ev_loop (EV_A_ 0);
		1682
		1683	// stop timer again
		1684	if (timeout >= 0)
		1685	ev_timer_stop (EV_A_ &to);
		1686
		1687	// stop io watchers again - their callbacks should have set
		1688	for (n = 0; n < nfds; ++n)
		1689	ev_io_stop (EV_A_ iow [n]);
		1690
		1691	return got_events;
		1692	}
		1693
		1694
		1695	=head2 C<ev_embed> - when one backend isn't enough...
		1696
		1697	This is a rather advanced watcher type that lets you embed one event loop
		1698	into another (currently only C<ev_io> events are supported in the embedded
		1699	loop, other types of watchers might be handled in a delayed or incorrect
		1700	fashion and must not be used).
		1701
		1702	There are primarily two reasons you would want that: work around bugs and
		1703	prioritise I/O.
		1704
		1705	As an example for a bug workaround, the kqueue backend might only support
		1706	sockets on some platform, so it is unusable as generic backend, but you
		1707	still want to make use of it because you have many sockets and it scales
		1708	so nicely. In this case, you would create a kqueue-based loop and embed it
		1709	into your default loop (which might use e.g. poll). Overall operation will
		1710	be a bit slower because first libev has to poll and then call kevent, but
		1711	at least you can use both at what they are best.
		1712
		1713	As for prioritising I/O: rarely you have the case where some fds have
		1714	to be watched and handled very quickly (with low latency), and even
		1715	priorities and idle watchers might have too much overhead. In this case
		1716	you would put all the high priority stuff in one loop and all the rest in
		1717	a second one, and embed the second one in the first.
		1718
		1719	As long as the watcher is active, the callback will be invoked every time
		1720	there might be events pending in the embedded loop. The callback must then
		1721	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
		1722	their callbacks (you could also start an idle watcher to give the embedded
		1723	loop strictly lower priority for example). You can also set the callback
		1724	to C<0>, in which case the embed watcher will automatically execute the
		1725	embedded loop sweep.
		1726
		1727	As long as the watcher is started it will automatically handle events. The
		1728	callback will be invoked whenever some events have been handled. You can
		1729	set the callback to C<0> to avoid having to specify one if you are not
		1730	interested in that.
		1731
		1732	Also, there have not currently been made special provisions for forking:
		1733	when you fork, you not only have to call C<ev_loop_fork> on both loops,
		1734	but you will also have to stop and restart any C<ev_embed> watchers
		1735	yourself.
		1736
		1737	Unfortunately, not all backends are embeddable, only the ones returned by
		1738	C<ev_embeddable_backends> are, which, unfortunately, does not include any
		1739	portable one.
		1740
		1741	So when you want to use this feature you will always have to be prepared
		1742	that you cannot get an embeddable loop. The recommended way to get around
		1743	this is to have a separate variables for your embeddable loop, try to
		1744	create it, and if that fails, use the normal loop for everything:
		1745
		1746	struct ev_loop *loop_hi = ev_default_init (0);
		1747	struct ev_loop *loop_lo = 0;
		1748	struct ev_embed embed;
		1749
		1750	// see if there is a chance of getting one that works
		1751	// (remember that a flags value of 0 means autodetection)
		1752	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		1753	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		1754	: 0;
		1755
		1756	// if we got one, then embed it, otherwise default to loop_hi
		1757	if (loop_lo)
		1758	{
		1759	ev_embed_init (&embed, 0, loop_lo);
		1760	ev_embed_start (loop_hi, &embed);
		1761	}
		1762	else
		1763	loop_lo = loop_hi;
		1764
		1765	=head3 Watcher-Specific Functions and Data Members
		1766
		1767	=over 4
		1768
		1769	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		1770
		1771	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		1772
		1773	Configures the watcher to embed the given loop, which must be
		1774	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		1775	invoked automatically, otherwise it is the responsibility of the callback
		1776	to invoke it (it will continue to be called until the sweep has been done,
		1777	if you do not want thta, you need to temporarily stop the embed watcher).
		1778
		1779	=item ev_embed_sweep (loop, ev_embed *)
		1780
		1781	Make a single, non-blocking sweep over the embedded loop. This works
		1782	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		1783	apropriate way for embedded loops.
		1784
		1785	=item struct ev_loop *loop [read-only]
		1786
		1787	The embedded event loop.
		1788
		1789	=back
		1790
		1791
		1792	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		1793
		1794	Fork watchers are called when a C<fork ()> was detected (usually because
		1795	whoever is a good citizen cared to tell libev about it by calling
		1796	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		1797	event loop blocks next and before C<ev_check> watchers are being called,
		1798	and only in the child after the fork. If whoever good citizen calling
		1799	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		1800	handlers will be invoked, too, of course.
		1801
		1802	=head3 Watcher-Specific Functions and Data Members
		1803
		1804	=over 4
		1805
		1806	=item ev_fork_init (ev_signal *, callback)
		1807
		1808	Initialises and configures the fork watcher - it has no parameters of any
		1809	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		1810	believe me.
		1811
		1812	=back
		1813
739		1814
740	=head1 OTHER FUNCTIONS	1815	=head1 OTHER FUNCTIONS
741		1816
742	There are some other functions of possible interest. Described. Here. Now.	1817	There are some other functions of possible interest. Described. Here. Now.
743		1818
…		…
773	/* stdin might have data for us, joy! */;	1848	/* stdin might have data for us, joy! */;
774	}	1849	}
775		1850
776	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	1851	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
777		1852
778	=item ev_feed_event (loop, watcher, int events)	1853	=item ev_feed_event (ev_loop *, watcher *, int revents)
779		1854
780	Feeds the given event set into the event loop, as if the specified event	1855	Feeds the given event set into the event loop, as if the specified event
781	had happened for the specified watcher (which must be a pointer to an	1856	had happened for the specified watcher (which must be a pointer to an
782	initialised but not necessarily started event watcher).	1857	initialised but not necessarily started event watcher).
783		1858
784	=item ev_feed_fd_event (loop, int fd, int revents)	1859	=item ev_feed_fd_event (ev_loop *, int fd, int revents)
785		1860
786	Feed an event on the given fd, as if a file descriptor backend detected	1861	Feed an event on the given fd, as if a file descriptor backend detected
787	the given events it.	1862	the given events it.
788		1863
789	=item ev_feed_signal_event (loop, int signum)	1864	=item ev_feed_signal_event (ev_loop *loop, int signum)
790		1865
791	Feed an event as if the given signal occured (loop must be the default loop!).	1866	Feed an event as if the given signal occured (C<loop> must be the default
		1867	loop!).
792		1868
793	=back	1869	=back
		1870
794		1871
795	=head1 LIBEVENT EMULATION	1872	=head1 LIBEVENT EMULATION
796		1873
797	Libev offers a compatibility emulation layer for libevent. It cannot	1874	Libev offers a compatibility emulation layer for libevent. It cannot
798	emulate the internals of libevent, so here are some usage hints:	1875	emulate the internals of libevent, so here are some usage hints:
…		…
819		1896
820	=back	1897	=back
821		1898
822	=head1 C++ SUPPORT	1899	=head1 C++ SUPPORT
823		1900
824	TBD.	1901	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		1902	you to use some convinience methods to start/stop watchers and also change
		1903	the callback model to a model using method callbacks on objects.
		1904
		1905	To use it,
		1906
		1907	#include <ev++.h>
		1908
		1909	This automatically includes F<ev.h> and puts all of its definitions (many
		1910	of them macros) into the global namespace. All C++ specific things are
		1911	put into the C<ev> namespace. It should support all the same embedding
		1912	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
		1913
		1914	Care has been taken to keep the overhead low. The only data member the C++
		1915	classes add (compared to plain C-style watchers) is the event loop pointer
		1916	that the watcher is associated with (or no additional members at all if
		1917	you disable C<EV_MULTIPLICITY> when embedding libev).
		1918
		1919	Currently, functions, and static and non-static member functions can be
		1920	used as callbacks. Other types should be easy to add as long as they only
		1921	need one additional pointer for context. If you need support for other
		1922	types of functors please contact the author (preferably after implementing
		1923	it).
		1924
		1925	Here is a list of things available in the C<ev> namespace:
		1926
		1927	=over 4
		1928
		1929	=item C<ev::READ>, C<ev::WRITE> etc.
		1930
		1931	These are just enum values with the same values as the C<EV_READ> etc.
		1932	macros from F<ev.h>.
		1933
		1934	=item C<ev::tstamp>, C<ev::now>
		1935
		1936	Aliases to the same types/functions as with the C<ev_> prefix.
		1937
		1938	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		1939
		1940	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		1941	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		1942	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		1943	defines by many implementations.
		1944
		1945	All of those classes have these methods:
		1946
		1947	=over 4
		1948
		1949	=item ev::TYPE::TYPE ()
		1950
		1951	=item ev::TYPE::TYPE (struct ev_loop *)
		1952
		1953	=item ev::TYPE::~TYPE
		1954
		1955	The constructor (optionally) takes an event loop to associate the watcher
		1956	with. If it is omitted, it will use C<EV_DEFAULT>.
		1957
		1958	The constructor calls C<ev_init> for you, which means you have to call the
		1959	C<set> method before starting it.
		1960
		1961	It will not set a callback, however: You have to call the templated C<set>
		1962	method to set a callback before you can start the watcher.
		1963
		1964	(The reason why you have to use a method is a limitation in C++ which does
		1965	not allow explicit template arguments for constructors).
		1966
		1967	The destructor automatically stops the watcher if it is active.
		1968
		1969	=item w->set<class, &class::method> (object *)
		1970
		1971	This method sets the callback method to call. The method has to have a
		1972	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		1973	first argument and the C<revents> as second. The object must be given as
		1974	parameter and is stored in the C<data> member of the watcher.
		1975
		1976	This method synthesizes efficient thunking code to call your method from
		1977	the C callback that libev requires. If your compiler can inline your
		1978	callback (i.e. it is visible to it at the place of the C<set> call and
		1979	your compiler is good :), then the method will be fully inlined into the
		1980	thunking function, making it as fast as a direct C callback.
		1981
		1982	Example: simple class declaration and watcher initialisation
		1983
		1984	struct myclass
		1985	{
		1986	void io_cb (ev::io &w, int revents) { }
		1987	}
		1988
		1989	myclass obj;
		1990	ev::io iow;
		1991	iow.set <myclass, &myclass::io_cb> (&obj);
		1992
		1993	=item w->set<function> (void *data = 0)
		1994
		1995	Also sets a callback, but uses a static method or plain function as
		1996	callback. The optional C<data> argument will be stored in the watcher's
		1997	C<data> member and is free for you to use.
		1998
		1999	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2000
		2001	See the method-C<set> above for more details.
		2002
		2003	Example:
		2004
		2005	static void io_cb (ev::io &w, int revents) { }
		2006	iow.set <io_cb> ();
		2007
		2008	=item w->set (struct ev_loop *)
		2009
		2010	Associates a different C<struct ev_loop> with this watcher. You can only
		2011	do this when the watcher is inactive (and not pending either).
		2012
		2013	=item w->set ([args])
		2014
		2015	Basically the same as C<ev_TYPE_set>, with the same args. Must be
		2016	called at least once. Unlike the C counterpart, an active watcher gets
		2017	automatically stopped and restarted when reconfiguring it with this
		2018	method.
		2019
		2020	=item w->start ()
		2021
		2022	Starts the watcher. Note that there is no C<loop> argument, as the
		2023	constructor already stores the event loop.
		2024
		2025	=item w->stop ()
		2026
		2027	Stops the watcher if it is active. Again, no C<loop> argument.
		2028
		2029	=item w->again () (C<ev::timer>, C<ev::periodic> only)
		2030
		2031	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		2032	C<ev_TYPE_again> function.
		2033
		2034	=item w->sweep () (C<ev::embed> only)
		2035
		2036	Invokes C<ev_embed_sweep>.
		2037
		2038	=item w->update () (C<ev::stat> only)
		2039
		2040	Invokes C<ev_stat_stat>.
		2041
		2042	=back
		2043
		2044	=back
		2045
		2046	Example: Define a class with an IO and idle watcher, start one of them in
		2047	the constructor.
		2048
		2049	class myclass
		2050	{
		2051	ev_io io; void io_cb (ev::io &w, int revents);
		2052	ev_idle idle void idle_cb (ev::idle &w, int revents);
		2053
		2054	myclass ();
		2055	}
		2056
		2057	myclass::myclass (int fd)
		2058	{
		2059	io .set <myclass, &myclass::io_cb > (this);
		2060	idle.set <myclass, &myclass::idle_cb> (this);
		2061
		2062	io.start (fd, ev::READ);
		2063	}
		2064
		2065
		2066	=head1 MACRO MAGIC
		2067
		2068	Libev can be compiled with a variety of options, the most fundamantal
		2069	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		2070	functions and callbacks have an initial C<struct ev_loop *> argument.
		2071
		2072	To make it easier to write programs that cope with either variant, the
		2073	following macros are defined:
		2074
		2075	=over 4
		2076
		2077	=item C<EV_A>, C<EV_A_>
		2078
		2079	This provides the loop I<argument> for functions, if one is required ("ev
		2080	loop argument"). The C<EV_A> form is used when this is the sole argument,
		2081	C<EV_A_> is used when other arguments are following. Example:
		2082
		2083	ev_unref (EV_A);
		2084	ev_timer_add (EV_A_ watcher);
		2085	ev_loop (EV_A_ 0);
		2086
		2087	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		2088	which is often provided by the following macro.
		2089
		2090	=item C<EV_P>, C<EV_P_>
		2091
		2092	This provides the loop I<parameter> for functions, if one is required ("ev
		2093	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		2094	C<EV_P_> is used when other parameters are following. Example:
		2095
		2096	// this is how ev_unref is being declared
		2097	static void ev_unref (EV_P);
		2098
		2099	// this is how you can declare your typical callback
		2100	static void cb (EV_P_ ev_timer *w, int revents)
		2101
		2102	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		2103	suitable for use with C<EV_A>.
		2104
		2105	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		2106
		2107	Similar to the other two macros, this gives you the value of the default
		2108	loop, if multiple loops are supported ("ev loop default").
		2109
		2110	=back
		2111
		2112	Example: Declare and initialise a check watcher, utilising the above
		2113	macros so it will work regardless of whether multiple loops are supported
		2114	or not.
		2115
		2116	static void
		2117	check_cb (EV_P_ ev_timer *w, int revents)
		2118	{
		2119	ev_check_stop (EV_A_ w);
		2120	}
		2121
		2122	ev_check check;
		2123	ev_check_init (&check, check_cb);
		2124	ev_check_start (EV_DEFAULT_ &check);
		2125	ev_loop (EV_DEFAULT_ 0);
		2126
		2127	=head1 EMBEDDING
		2128
		2129	Libev can (and often is) directly embedded into host
		2130	applications. Examples of applications that embed it include the Deliantra
		2131	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		2132	and rxvt-unicode.
		2133
		2134	The goal is to enable you to just copy the neecssary files into your
		2135	source directory without having to change even a single line in them, so
		2136	you can easily upgrade by simply copying (or having a checked-out copy of
		2137	libev somewhere in your source tree).
		2138
		2139	=head2 FILESETS
		2140
		2141	Depending on what features you need you need to include one or more sets of files
		2142	in your app.
		2143
		2144	=head3 CORE EVENT LOOP
		2145
		2146	To include only the libev core (all the C<ev_*> functions), with manual
		2147	configuration (no autoconf):
		2148
		2149	#define EV_STANDALONE 1
		2150	#include "ev.c"
		2151
		2152	This will automatically include F<ev.h>, too, and should be done in a
		2153	single C source file only to provide the function implementations. To use
		2154	it, do the same for F<ev.h> in all files wishing to use this API (best
		2155	done by writing a wrapper around F<ev.h> that you can include instead and
		2156	where you can put other configuration options):
		2157
		2158	#define EV_STANDALONE 1
		2159	#include "ev.h"
		2160
		2161	Both header files and implementation files can be compiled with a C++
		2162	compiler (at least, thats a stated goal, and breakage will be treated
		2163	as a bug).
		2164
		2165	You need the following files in your source tree, or in a directory
		2166	in your include path (e.g. in libev/ when using -Ilibev):
		2167
		2168	ev.h
		2169	ev.c
		2170	ev_vars.h
		2171	ev_wrap.h
		2172
		2173	ev_win32.c required on win32 platforms only
		2174
		2175	ev_select.c only when select backend is enabled (which is enabled by default)
		2176	ev_poll.c only when poll backend is enabled (disabled by default)
		2177	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		2178	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		2179	ev_port.c only when the solaris port backend is enabled (disabled by default)
		2180
		2181	F<ev.c> includes the backend files directly when enabled, so you only need
		2182	to compile this single file.
		2183
		2184	=head3 LIBEVENT COMPATIBILITY API
		2185
		2186	To include the libevent compatibility API, also include:
		2187
		2188	#include "event.c"
		2189
		2190	in the file including F<ev.c>, and:
		2191
		2192	#include "event.h"
		2193
		2194	in the files that want to use the libevent API. This also includes F<ev.h>.
		2195
		2196	You need the following additional files for this:
		2197
		2198	event.h
		2199	event.c
		2200
		2201	=head3 AUTOCONF SUPPORT
		2202
		2203	Instead of using C<EV_STANDALONE=1> and providing your config in
		2204	whatever way you want, you can also C<m4_include([libev.m4])> in your
		2205	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		2206	include F<config.h> and configure itself accordingly.
		2207
		2208	For this of course you need the m4 file:
		2209
		2210	libev.m4
		2211
		2212	=head2 PREPROCESSOR SYMBOLS/MACROS
		2213
		2214	Libev can be configured via a variety of preprocessor symbols you have to define
		2215	before including any of its files. The default is not to build for multiplicity
		2216	and only include the select backend.
		2217
		2218	=over 4
		2219
		2220	=item EV_STANDALONE
		2221
		2222	Must always be C<1> if you do not use autoconf configuration, which
		2223	keeps libev from including F<config.h>, and it also defines dummy
		2224	implementations for some libevent functions (such as logging, which is not
		2225	supported). It will also not define any of the structs usually found in
		2226	F<event.h> that are not directly supported by the libev core alone.
		2227
		2228	=item EV_USE_MONOTONIC
		2229
		2230	If defined to be C<1>, libev will try to detect the availability of the
		2231	monotonic clock option at both compiletime and runtime. Otherwise no use
		2232	of the monotonic clock option will be attempted. If you enable this, you
		2233	usually have to link against librt or something similar. Enabling it when
		2234	the functionality isn't available is safe, though, althoguh you have
		2235	to make sure you link against any libraries where the C<clock_gettime>
		2236	function is hiding in (often F<-lrt>).
		2237
		2238	=item EV_USE_REALTIME
		2239
		2240	If defined to be C<1>, libev will try to detect the availability of the
		2241	realtime clock option at compiletime (and assume its availability at
		2242	runtime if successful). Otherwise no use of the realtime clock option will
		2243	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		2244	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries
		2245	in the description of C<EV_USE_MONOTONIC>, though.
		2246
		2247	=item EV_USE_SELECT
		2248
		2249	If undefined or defined to be C<1>, libev will compile in support for the
		2250	C<select>(2) backend. No attempt at autodetection will be done: if no
		2251	other method takes over, select will be it. Otherwise the select backend
		2252	will not be compiled in.
		2253
		2254	=item EV_SELECT_USE_FD_SET
		2255
		2256	If defined to C<1>, then the select backend will use the system C<fd_set>
		2257	structure. This is useful if libev doesn't compile due to a missing
		2258	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on
		2259	exotic systems. This usually limits the range of file descriptors to some
		2260	low limit such as 1024 or might have other limitations (winsocket only
		2261	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		2262	influence the size of the C<fd_set> used.
		2263
		2264	=item EV_SELECT_IS_WINSOCKET
		2265
		2266	When defined to C<1>, the select backend will assume that
		2267	select/socket/connect etc. don't understand file descriptors but
		2268	wants osf handles on win32 (this is the case when the select to
		2269	be used is the winsock select). This means that it will call
		2270	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		2271	it is assumed that all these functions actually work on fds, even
		2272	on win32. Should not be defined on non-win32 platforms.
		2273
		2274	=item EV_USE_POLL
		2275
		2276	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		2277	backend. Otherwise it will be enabled on non-win32 platforms. It
		2278	takes precedence over select.
		2279
		2280	=item EV_USE_EPOLL
		2281
		2282	If defined to be C<1>, libev will compile in support for the Linux
		2283	C<epoll>(7) backend. Its availability will be detected at runtime,
		2284	otherwise another method will be used as fallback. This is the
		2285	preferred backend for GNU/Linux systems.
		2286
		2287	=item EV_USE_KQUEUE
		2288
		2289	If defined to be C<1>, libev will compile in support for the BSD style
		2290	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		2291	otherwise another method will be used as fallback. This is the preferred
		2292	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		2293	supports some types of fds correctly (the only platform we found that
		2294	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		2295	not be used unless explicitly requested. The best way to use it is to find
		2296	out whether kqueue supports your type of fd properly and use an embedded
		2297	kqueue loop.
		2298
		2299	=item EV_USE_PORT
		2300
		2301	If defined to be C<1>, libev will compile in support for the Solaris
		2302	10 port style backend. Its availability will be detected at runtime,
		2303	otherwise another method will be used as fallback. This is the preferred
		2304	backend for Solaris 10 systems.
		2305
		2306	=item EV_USE_DEVPOLL
		2307
		2308	reserved for future expansion, works like the USE symbols above.
		2309
		2310	=item EV_USE_INOTIFY
		2311
		2312	If defined to be C<1>, libev will compile in support for the Linux inotify
		2313	interface to speed up C<ev_stat> watchers. Its actual availability will
		2314	be detected at runtime.
		2315
		2316	=item EV_H
		2317
		2318	The name of the F<ev.h> header file used to include it. The default if
		2319	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This
		2320	can be used to virtually rename the F<ev.h> header file in case of conflicts.
		2321
		2322	=item EV_CONFIG_H
		2323
		2324	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		2325	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		2326	C<EV_H>, above.
		2327
		2328	=item EV_EVENT_H
		2329
		2330	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		2331	of how the F<event.h> header can be found.
		2332
		2333	=item EV_PROTOTYPES
		2334
		2335	If defined to be C<0>, then F<ev.h> will not define any function
		2336	prototypes, but still define all the structs and other symbols. This is
		2337	occasionally useful if you want to provide your own wrapper functions
		2338	around libev functions.
		2339
		2340	=item EV_MULTIPLICITY
		2341
		2342	If undefined or defined to C<1>, then all event-loop-specific functions
		2343	will have the C<struct ev_loop *> as first argument, and you can create
		2344	additional independent event loops. Otherwise there will be no support
		2345	for multiple event loops and there is no first event loop pointer
		2346	argument. Instead, all functions act on the single default loop.
		2347
		2348	=item EV_MINPRI
		2349
		2350	=item EV_MAXPRI
		2351
		2352	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2353	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2354	provide for more priorities by overriding those symbols (usually defined
		2355	to be C<-2> and C<2>, respectively).
		2356
		2357	When doing priority-based operations, libev usually has to linearly search
		2358	all the priorities, so having many of them (hundreds) uses a lot of space
		2359	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2360	fine.
		2361
		2362	If your embedding app does not need any priorities, defining these both to
		2363	C<0> will save some memory and cpu.
		2364
		2365	=item EV_PERIODIC_ENABLE
		2366
		2367	If undefined or defined to be C<1>, then periodic timers are supported. If
		2368	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2369	code.
		2370
		2371	=item EV_IDLE_ENABLE
		2372
		2373	If undefined or defined to be C<1>, then idle watchers are supported. If
		2374	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2375	code.
		2376
		2377	=item EV_EMBED_ENABLE
		2378
		2379	If undefined or defined to be C<1>, then embed watchers are supported. If
		2380	defined to be C<0>, then they are not.
		2381
		2382	=item EV_STAT_ENABLE
		2383
		2384	If undefined or defined to be C<1>, then stat watchers are supported. If
		2385	defined to be C<0>, then they are not.
		2386
		2387	=item EV_FORK_ENABLE
		2388
		2389	If undefined or defined to be C<1>, then fork watchers are supported. If
		2390	defined to be C<0>, then they are not.
		2391
		2392	=item EV_MINIMAL
		2393
		2394	If you need to shave off some kilobytes of code at the expense of some
		2395	speed, define this symbol to C<1>. Currently only used for gcc to override
		2396	some inlining decisions, saves roughly 30% codesize of amd64.
		2397
		2398	=item EV_PID_HASHSIZE
		2399
		2400	C<ev_child> watchers use a small hash table to distribute workload by
		2401	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		2402	than enough. If you need to manage thousands of children you might want to
		2403	increase this value (I<must> be a power of two).
		2404
		2405	=item EV_INOTIFY_HASHSIZE
		2406
		2407	C<ev_staz> watchers use a small hash table to distribute workload by
		2408	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		2409	usually more than enough. If you need to manage thousands of C<ev_stat>
		2410	watchers you might want to increase this value (I<must> be a power of
		2411	two).
		2412
		2413	=item EV_COMMON
		2414
		2415	By default, all watchers have a C<void *data> member. By redefining
		2416	this macro to a something else you can include more and other types of
		2417	members. You have to define it each time you include one of the files,
		2418	though, and it must be identical each time.
		2419
		2420	For example, the perl EV module uses something like this:
		2421
		2422	#define EV_COMMON \
		2423	SV *self; /* contains this struct */ \
		2424	SV *cb_sv, *fh /* note no trailing ";" */
		2425
		2426	=item EV_CB_DECLARE (type)
		2427
		2428	=item EV_CB_INVOKE (watcher, revents)
		2429
		2430	=item ev_set_cb (ev, cb)
		2431
		2432	Can be used to change the callback member declaration in each watcher,
		2433	and the way callbacks are invoked and set. Must expand to a struct member
		2434	definition and a statement, respectively. See the F<ev.v> header file for
		2435	their default definitions. One possible use for overriding these is to
		2436	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		2437	method calls instead of plain function calls in C++.
		2438
		2439	=head2 EXAMPLES
		2440
		2441	For a real-world example of a program the includes libev
		2442	verbatim, you can have a look at the EV perl module
		2443	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		2444	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		2445	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		2446	will be compiled. It is pretty complex because it provides its own header
		2447	file.
		2448
		2449	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		2450	that everybody includes and which overrides some configure choices:
		2451
		2452	#define EV_MINIMAL 1
		2453	#define EV_USE_POLL 0
		2454	#define EV_MULTIPLICITY 0
		2455	#define EV_PERIODIC_ENABLE 0
		2456	#define EV_STAT_ENABLE 0
		2457	#define EV_FORK_ENABLE 0
		2458	#define EV_CONFIG_H <config.h>
		2459	#define EV_MINPRI 0
		2460	#define EV_MAXPRI 0
		2461
		2462	#include "ev++.h"
		2463
		2464	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		2465
		2466	#include "ev_cpp.h"
		2467	#include "ev.c"
		2468
		2469
		2470	=head1 COMPLEXITIES
		2471
		2472	In this section the complexities of (many of) the algorithms used inside
		2473	libev will be explained. For complexity discussions about backends see the
		2474	documentation for C<ev_default_init>.
		2475
		2476	All of the following are about amortised time: If an array needs to be
		2477	extended, libev needs to realloc and move the whole array, but this
		2478	happens asymptotically never with higher number of elements, so O(1) might
		2479	mean it might do a lengthy realloc operation in rare cases, but on average
		2480	it is much faster and asymptotically approaches constant time.
		2481
		2482	=over 4
		2483
		2484	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		2485
		2486	This means that, when you have a watcher that triggers in one hour and
		2487	there are 100 watchers that would trigger before that then inserting will
		2488	have to skip those 100 watchers.
		2489
		2490	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
		2491
		2492	That means that for changing a timer costs less than removing/adding them
		2493	as only the relative motion in the event queue has to be paid for.
		2494
		2495	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
		2496
		2497	These just add the watcher into an array or at the head of a list.
		2498	=item Stopping check/prepare/idle watchers: O(1)
		2499
		2500	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		2501
		2502	These watchers are stored in lists then need to be walked to find the
		2503	correct watcher to remove. The lists are usually short (you don't usually
		2504	have many watchers waiting for the same fd or signal).
		2505
		2506	=item Finding the next timer per loop iteration: O(1)
		2507
		2508	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		2509
		2510	A change means an I/O watcher gets started or stopped, which requires
		2511	libev to recalculate its status (and possibly tell the kernel).
		2512
		2513	=item Activating one watcher: O(1)
		2514
		2515	=item Priority handling: O(number_of_priorities)
		2516
		2517	Priorities are implemented by allocating some space for each
		2518	priority. When doing priority-based operations, libev usually has to
		2519	linearly search all the priorities.
		2520
		2521	=back
		2522
825		2523
826	=head1 AUTHOR	2524	=head1 AUTHOR
827		2525
828	Marc Lehmann <libev@schmorp.de>.	2526	Marc Lehmann <libev@schmorp.de>.
829		2527

Diff Legend

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