ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.21 by root, Mon Nov 12 09:07:10 2007 UTC vs.
Revision 1.77 by root, Sat Dec 8 22:11:14 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.
56		109
57	=over 4	110	=over 4
58		111
59	=item ev_tstamp ev_time ()	112	=item ev_tstamp ev_time ()
60		113
61	Returns the current time as libev would use it.	114	Returns the current time as libev would use it. Please note that the
		115	C<ev_now> function is usually faster and also often returns the timestamp
		116	you actually want to know.
62		117
63	=item int ev_version_major ()	118	=item int ev_version_major ()
64		119
65	=item int ev_version_minor ()	120	=item int ev_version_minor ()
66		121
…		…
73	Usually, it's a good idea to terminate if the major versions mismatch,	128	Usually, it's a good idea to terminate if the major versions mismatch,
74	as this indicates an incompatible change. Minor versions are usually	129	as this indicates an incompatible change. Minor versions are usually
75	compatible to older versions, so a larger minor version alone is usually	130	compatible to older versions, so a larger minor version alone is usually
76	not a problem.	131	not a problem.
77		132
		133	Example: Make sure we haven't accidentally been linked against the wrong
		134	version.
		135
		136	assert (("libev version mismatch",
		137	ev_version_major () == EV_VERSION_MAJOR
		138	&& ev_version_minor () >= EV_VERSION_MINOR));
		139
		140	=item unsigned int ev_supported_backends ()
		141
		142	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		143	value) compiled into this binary of libev (independent of their
		144	availability on the system you are running on). See C<ev_default_loop> for
		145	a description of the set values.
		146
		147	Example: make sure we have the epoll method, because yeah this is cool and
		148	a must have and can we have a torrent of it please!!!11
		149
		150	assert (("sorry, no epoll, no sex",
		151	ev_supported_backends () & EVBACKEND_EPOLL));
		152
		153	=item unsigned int ev_recommended_backends ()
		154
		155	Return the set of all backends compiled into this binary of libev and also
		156	recommended for this platform. This set is often smaller than the one
		157	returned by C<ev_supported_backends>, as for example kqueue is broken on
		158	most BSDs and will not be autodetected unless you explicitly request it
		159	(assuming you know what you are doing). This is the set of backends that
		160	libev will probe for if you specify no backends explicitly.
		161
		162	=item unsigned int ev_embeddable_backends ()
		163
		164	Returns the set of backends that are embeddable in other event loops. This
		165	is the theoretical, all-platform, value. To find which backends
		166	might be supported on the current system, you would need to look at
		167	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
		168	recommended ones.
		169
		170	See the description of C<ev_embed> watchers for more info.
		171
78	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	172	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
79		173
80	Sets the allocation function to use (the prototype is similar to the	174	Sets the allocation function to use (the prototype is similar - the
81	realloc C function, the semantics are identical). It is used to allocate	175	semantics is identical - to the realloc C function). It is used to
82	and free memory (no surprises here). If it returns zero when memory	176	allocate and free memory (no surprises here). If it returns zero when
83	needs to be allocated, the library might abort or take some potentially	177	memory needs to be allocated, the library might abort or take some
84	destructive action. The default is your system realloc function.	178	potentially destructive action. The default is your system realloc
		179	function.
85		180
86	You could override this function in high-availability programs to, say,	181	You could override this function in high-availability programs to, say,
87	free some memory if it cannot allocate memory, to use a special allocator,	182	free some memory if it cannot allocate memory, to use a special allocator,
88	or even to sleep a while and retry until some memory is available.	183	or even to sleep a while and retry until some memory is available.
		184
		185	Example: Replace the libev allocator with one that waits a bit and then
		186	retries).
		187
		188	static void *
		189	persistent_realloc (void *ptr, size_t size)
		190	{
		191	for (;;)
		192	{
		193	void *newptr = realloc (ptr, size);
		194
		195	if (newptr)
		196	return newptr;
		197
		198	sleep (60);
		199	}
		200	}
		201
		202	...
		203	ev_set_allocator (persistent_realloc);
89		204
90	=item ev_set_syserr_cb (void (*cb)(const char *msg));	205	=item ev_set_syserr_cb (void (*cb)(const char *msg));
91		206
92	Set the callback function to call on a retryable syscall error (such	207	Set the callback function to call on a retryable syscall error (such
93	as failed select, poll, epoll_wait). The message is a printable string	208	as failed select, poll, epoll_wait). The message is a printable string
…		…
95	callback is set, then libev will expect it to remedy the sitution, no	210	callback is set, then libev will expect it to remedy the sitution, no
96	matter what, when it returns. That is, libev will generally retry the	211	matter what, when it returns. That is, libev will generally retry the
97	requested operation, or, if the condition doesn't go away, do bad stuff	212	requested operation, or, if the condition doesn't go away, do bad stuff
98	(such as abort).	213	(such as abort).
99		214
		215	Example: This is basically the same thing that libev does internally, too.
		216
		217	static void
		218	fatal_error (const char *msg)
		219	{
		220	perror (msg);
		221	abort ();
		222	}
		223
		224	...
		225	ev_set_syserr_cb (fatal_error);
		226
100	=back	227	=back
101		228
102	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	229	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
103		230
104	An event loop is described by a C<struct ev_loop *>. The library knows two	231	An event loop is described by a C<struct ev_loop *>. The library knows two
…		…
117	=item struct ev_loop *ev_default_loop (unsigned int flags)	244	=item struct ev_loop *ev_default_loop (unsigned int flags)
118		245
119	This will initialise the default event loop if it hasn't been initialised	246	This will initialise the default event loop if it hasn't been initialised
120	yet and return it. If the default loop could not be initialised, returns	247	yet and return it. If the default loop could not be initialised, returns
121	false. If it already was initialised it simply returns it (and ignores the	248	false. If it already was initialised it simply returns it (and ignores the
122	flags).	249	flags. If that is troubling you, check C<ev_backend ()> afterwards).
123		250
124	If you don't know what event loop to use, use the one returned from this	251	If you don't know what event loop to use, use the one returned from this
125	function.	252	function.
126		253
127	The flags argument can be used to specify special behaviour or specific	254	The flags argument can be used to specify special behaviour or specific
128	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	255	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
129		256
130	It supports the following flags:	257	The following flags are supported:
131		258
132	=over 4	259	=over 4
133		260
134	=item C<EVFLAG_AUTO>	261	=item C<EVFLAG_AUTO>
135		262
…		…
143	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	270	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
144	override the flags completely if it is found in the environment. This is	271	override the flags completely if it is found in the environment. This is
145	useful to try out specific backends to test their performance, or to work	272	useful to try out specific backends to test their performance, or to work
146	around bugs.	273	around bugs.
147		274
		275	=item C<EVFLAG_FORKCHECK>
		276
		277	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		278	a fork, you can also make libev check for a fork in each iteration by
		279	enabling this flag.
		280
		281	This works by calling C<getpid ()> on every iteration of the loop,
		282	and thus this might slow down your event loop if you do a lot of loop
		283	iterations and little real work, but is usually not noticeable (on my
		284	Linux system for example, C<getpid> is actually a simple 5-insn sequence
		285	without a syscall and thus I<very> fast, but my Linux system also has
		286	C<pthread_atfork> which is even faster).
		287
		288	The big advantage of this flag is that you can forget about fork (and
		289	forget about forgetting to tell libev about forking) when you use this
		290	flag.
		291
		292	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		293	environment variable.
		294
148	=item C<EVMETHOD_SELECT> (portable select backend)	295	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
149		296
		297	This is your standard select(2) backend. Not I<completely> standard, as
		298	libev tries to roll its own fd_set with no limits on the number of fds,
		299	but if that fails, expect a fairly low limit on the number of fds when
		300	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
		301	the fastest backend for a low number of fds.
		302
150	=item C<EVMETHOD_POLL> (poll backend, available everywhere except on windows)	303	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
151		304
152	=item C<EVMETHOD_EPOLL> (linux only)	305	And this is your standard poll(2) backend. It's more complicated than
		306	select, but handles sparse fds better and has no artificial limit on the
		307	number of fds you can use (except it will slow down considerably with a
		308	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
153		309
154	=item C<EVMETHOD_KQUEUE> (some bsds only)	310	=item C<EVBACKEND_EPOLL> (value 4, Linux)
155		311
156	=item C<EVMETHOD_DEVPOLL> (solaris 8 only)	312	For few fds, this backend is a bit little slower than poll and select,
		313	but it scales phenomenally better. While poll and select usually scale like
		314	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
		315	either O(1) or O(active_fds).
157		316
158	=item C<EVMETHOD_PORT> (solaris 10 only)	317	While stopping and starting an I/O watcher in the same iteration will
		318	result in some caching, there is still a syscall per such incident
		319	(because the fd could point to a different file description now), so its
		320	best to avoid that. Also, dup()ed file descriptors might not work very
		321	well if you register events for both fds.
		322
		323	Please note that epoll sometimes generates spurious notifications, so you
		324	need to use non-blocking I/O or other means to avoid blocking when no data
		325	(or space) is available.
		326
		327	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
		328
		329	Kqueue deserves special mention, as at the time of this writing, it
		330	was broken on all BSDs except NetBSD (usually it doesn't work with
		331	anything but sockets and pipes, except on Darwin, where of course its
		332	completely useless). For this reason its not being "autodetected"
		333	unless you explicitly specify it explicitly in the flags (i.e. using
		334	C<EVBACKEND_KQUEUE>).
		335
		336	It scales in the same way as the epoll backend, but the interface to the
		337	kernel is more efficient (which says nothing about its actual speed, of
		338	course). While starting and stopping an I/O watcher does not cause an
		339	extra syscall as with epoll, it still adds up to four event changes per
		340	incident, so its best to avoid that.
		341
		342	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
		343
		344	This is not implemented yet (and might never be).
		345
		346	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
		347
		348	This uses the Solaris 10 port mechanism. As with everything on Solaris,
		349	it's really slow, but it still scales very well (O(active_fds)).
		350
		351	Please note that solaris ports can result in a lot of spurious
		352	notifications, so you need to use non-blocking I/O or other means to avoid
		353	blocking when no data (or space) is available.
		354
		355	=item C<EVBACKEND_ALL>
		356
		357	Try all backends (even potentially broken ones that wouldn't be tried
		358	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
		359	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
		360
		361	=back
159		362
160	If one or more of these are ored into the flags value, then only these	363	If one or more of these are ored into the flags value, then only these
161	backends will be tried (in the reverse order as given here). If one are	364	backends will be tried (in the reverse order as given here). If none are
162	specified, any backend will do.	365	specified, most compiled-in backend will be tried, usually in reverse
		366	order of their flag values :)
163		367
164	=back	368	The most typical usage is like this:
		369
		370	if (!ev_default_loop (0))
		371	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		372
		373	Restrict libev to the select and poll backends, and do not allow
		374	environment settings to be taken into account:
		375
		376	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		377
		378	Use whatever libev has to offer, but make sure that kqueue is used if
		379	available (warning, breaks stuff, best use only with your own private
		380	event loop and only if you know the OS supports your types of fds):
		381
		382	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
165		383
166	=item struct ev_loop *ev_loop_new (unsigned int flags)	384	=item struct ev_loop *ev_loop_new (unsigned int flags)
167		385
168	Similar to C<ev_default_loop>, but always creates a new event loop that is	386	Similar to C<ev_default_loop>, but always creates a new event loop that is
169	always distinct from the default loop. Unlike the default loop, it cannot	387	always distinct from the default loop. Unlike the default loop, it cannot
170	handle signal and child watchers, and attempts to do so will be greeted by	388	handle signal and child watchers, and attempts to do so will be greeted by
171	undefined behaviour (or a failed assertion if assertions are enabled).	389	undefined behaviour (or a failed assertion if assertions are enabled).
172		390
		391	Example: Try to create a event loop that uses epoll and nothing else.
		392
		393	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
		394	if (!epoller)
		395	fatal ("no epoll found here, maybe it hides under your chair");
		396
173	=item ev_default_destroy ()	397	=item ev_default_destroy ()
174		398
175	Destroys the default loop again (frees all memory and kernel state	399	Destroys the default loop again (frees all memory and kernel state
176	etc.). This stops all registered event watchers (by not touching them in	400	etc.). None of the active event watchers will be stopped in the normal
177	any way whatsoever, although you cannot rely on this :).	401	sense, so e.g. C<ev_is_active> might still return true. It is your
		402	responsibility to either stop all watchers cleanly yoursef I<before>
		403	calling this function, or cope with the fact afterwards (which is usually
		404	the easiest thing, youc na just ignore the watchers and/or C<free ()> them
		405	for example).
178		406
179	=item ev_loop_destroy (loop)	407	=item ev_loop_destroy (loop)
180		408
181	Like C<ev_default_destroy>, but destroys an event loop created by an	409	Like C<ev_default_destroy>, but destroys an event loop created by an
182	earlier call to C<ev_loop_new>.	410	earlier call to C<ev_loop_new>.
…		…
186	This function reinitialises the kernel state for backends that have	414	This function reinitialises the kernel state for backends that have
187	one. Despite the name, you can call it anytime, but it makes most sense	415	one. Despite the name, you can call it anytime, but it makes most sense
188	after forking, in either the parent or child process (or both, but that	416	after forking, in either the parent or child process (or both, but that
189	again makes little sense).	417	again makes little sense).
190		418
191	You I<must> call this function after forking if and only if you want to	419	You I<must> call this function in the child process after forking if and
192	use the event library in both processes. If you just fork+exec, you don't	420	only if you want to use the event library in both processes. If you just
193	have to call it.	421	fork+exec, you don't have to call it.
194		422
195	The function itself is quite fast and it's usually not a problem to call	423	The function itself is quite fast and it's usually not a problem to call
196	it just in case after a fork. To make this easy, the function will fit in	424	it just in case after a fork. To make this easy, the function will fit in
197	quite nicely into a call to C<pthread_atfork>:	425	quite nicely into a call to C<pthread_atfork>:
198		426
199	pthread_atfork (0, 0, ev_default_fork);	427	pthread_atfork (0, 0, ev_default_fork);
200		428
		429	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
		430	without calling this function, so if you force one of those backends you
		431	do not need to care.
		432
201	=item ev_loop_fork (loop)	433	=item ev_loop_fork (loop)
202		434
203	Like C<ev_default_fork>, but acts on an event loop created by	435	Like C<ev_default_fork>, but acts on an event loop created by
204	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	436	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
205	after fork, and how you do this is entirely your own problem.	437	after fork, and how you do this is entirely your own problem.
206		438
		439	=item unsigned int ev_loop_count (loop)
		440
		441	Returns the count of loop iterations for the loop, which is identical to
		442	the number of times libev did poll for new events. It starts at C<0> and
		443	happily wraps around with enough iterations.
		444
		445	This value can sometimes be useful as a generation counter of sorts (it
		446	"ticks" the number of loop iterations), as it roughly corresponds with
		447	C<ev_prepare> and C<ev_check> calls.
		448
207	=item unsigned int ev_method (loop)	449	=item unsigned int ev_backend (loop)
208		450
209	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	451	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
210	use.	452	use.
211		453
212	=item ev_tstamp ev_now (loop)	454	=item ev_tstamp ev_now (loop)
213		455
214	Returns the current "event loop time", which is the time the event loop	456	Returns the current "event loop time", which is the time the event loop
215	got events and started processing them. This timestamp does not change	457	received events and started processing them. This timestamp does not
216	as long as callbacks are being processed, and this is also the base time	458	change as long as callbacks are being processed, and this is also the base
217	used for relative timers. You can treat it as the timestamp of the event	459	time used for relative timers. You can treat it as the timestamp of the
218	occuring (or more correctly, the mainloop finding out about it).	460	event occuring (or more correctly, libev finding out about it).
219		461
220	=item ev_loop (loop, int flags)	462	=item ev_loop (loop, int flags)
221		463
222	Finally, this is it, the event handler. This function usually is called	464	Finally, this is it, the event handler. This function usually is called
223	after you initialised all your watchers and you want to start handling	465	after you initialised all your watchers and you want to start handling
224	events.	466	events.
225		467
226	If the flags argument is specified as 0, it will not return until either	468	If the flags argument is specified as C<0>, it will not return until
227	no event watchers are active anymore or C<ev_unloop> was called.	469	either no event watchers are active anymore or C<ev_unloop> was called.
		470
		471	Please note that an explicit C<ev_unloop> is usually better than
		472	relying on all watchers to be stopped when deciding when a program has
		473	finished (especially in interactive programs), but having a program that
		474	automatically loops as long as it has to and no longer by virtue of
		475	relying on its watchers stopping correctly is a thing of beauty.
228		476
229	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	477	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
230	those events and any outstanding ones, but will not block your process in	478	those events and any outstanding ones, but will not block your process in
231	case there are no events and will return after one iteration of the loop.	479	case there are no events and will return after one iteration of the loop.
232		480
233	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	481	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
234	neccessary) and will handle those and any outstanding ones. It will block	482	neccessary) and will handle those and any outstanding ones. It will block
235	your process until at least one new event arrives, and will return after	483	your process until at least one new event arrives, and will return after
236	one iteration of the loop.	484	one iteration of the loop. This is useful if you are waiting for some
		485	external event in conjunction with something not expressible using other
		486	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
		487	usually a better approach for this kind of thing.
237		488
238	This flags value could be used to implement alternative looping	489	Here are the gory details of what C<ev_loop> does:
239	constructs, but the C<prepare> and C<check> watchers provide a better and	490
240	more generic mechanism.	491	- Before the first iteration, call any pending watchers.
		492	* If there are no active watchers (reference count is zero), return.
		493	- Queue all prepare watchers and then call all outstanding watchers.
		494	- If we have been forked, recreate the kernel state.
		495	- Update the kernel state with all outstanding changes.
		496	- Update the "event loop time".
		497	- Calculate for how long to block.
		498	- Block the process, waiting for any events.
		499	- Queue all outstanding I/O (fd) events.
		500	- Update the "event loop time" and do time jump handling.
		501	- Queue all outstanding timers.
		502	- Queue all outstanding periodics.
		503	- If no events are pending now, queue all idle watchers.
		504	- Queue all check watchers.
		505	- Call all queued watchers in reverse order (i.e. check watchers first).
		506	Signals and child watchers are implemented as I/O watchers, and will
		507	be handled here by queueing them when their watcher gets executed.
		508	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
		509	were used, return, otherwise continue with step *.
		510
		511	Example: Queue some jobs and then loop until no events are outsanding
		512	anymore.
		513
		514	... queue jobs here, make sure they register event watchers as long
		515	... as they still have work to do (even an idle watcher will do..)
		516	ev_loop (my_loop, 0);
		517	... jobs done. yeah!
241		518
242	=item ev_unloop (loop, how)	519	=item ev_unloop (loop, how)
243		520
244	Can be used to make a call to C<ev_loop> return early (but only after it	521	Can be used to make a call to C<ev_loop> return early (but only after it
245	has processed all outstanding events). The C<how> argument must be either	522	has processed all outstanding events). The C<how> argument must be either
246	C<EVUNLOOP_ONCE>, which will make the innermost C<ev_loop> call return, or	523	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
247	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	524	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
248		525
249	=item ev_ref (loop)	526	=item ev_ref (loop)
250		527
251	=item ev_unref (loop)	528	=item ev_unref (loop)
…		…
259	visible to the libev user and should not keep C<ev_loop> from exiting if	536	visible to the libev user and should not keep C<ev_loop> from exiting if
260	no event watchers registered by it are active. It is also an excellent	537	no event watchers registered by it are active. It is also an excellent
261	way to do this for generic recurring timers or from within third-party	538	way to do this for generic recurring timers or from within third-party
262	libraries. Just remember to I<unref after start> and I<ref before stop>.	539	libraries. Just remember to I<unref after start> and I<ref before stop>.
263		540
		541	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
		542	running when nothing else is active.
		543
		544	struct ev_signal exitsig;
		545	ev_signal_init (&exitsig, sig_cb, SIGINT);
		546	ev_signal_start (loop, &exitsig);
		547	evf_unref (loop);
		548
		549	Example: For some weird reason, unregister the above signal handler again.
		550
		551	ev_ref (loop);
		552	ev_signal_stop (loop, &exitsig);
		553
264	=back	554	=back
		555
265		556
266	=head1 ANATOMY OF A WATCHER	557	=head1 ANATOMY OF A WATCHER
267		558
268	A watcher is a structure that you create and register to record your	559	A watcher is a structure that you create and register to record your
269	interest in some event. For instance, if you want to wait for STDIN to	560	interest in some event. For instance, if you want to wait for STDIN to
…		…
302	*) >>), and you can stop watching for events at any time by calling the	593	*) >>), and you can stop watching for events at any time by calling the
303	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	594	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.
304		595
305	As long as your watcher is active (has been started but not stopped) you	596	As long as your watcher is active (has been started but not stopped) you
306	must not touch the values stored in it. Most specifically you must never	597	must not touch the values stored in it. Most specifically you must never
307	reinitialise it or call its set method.	598	reinitialise it or call its C<set> macro.
308
309	You can check whether an event is active by calling the C<ev_is_active
310	(watcher *)> macro. To see whether an event is outstanding (but the
311	callback for it has not been called yet) you can use the C<ev_is_pending
312	(watcher *)> macro.
313		599
314	Each and every callback receives the event loop pointer as first, the	600	Each and every callback receives the event loop pointer as first, the
315	registered watcher structure as second, and a bitset of received events as	601	registered watcher structure as second, and a bitset of received events as
316	third argument.	602	third argument.
317		603
…		…
341	The signal specified in the C<ev_signal> watcher has been received by a thread.	627	The signal specified in the C<ev_signal> watcher has been received by a thread.
342		628
343	=item C<EV_CHILD>	629	=item C<EV_CHILD>
344		630
345	The pid specified in the C<ev_child> watcher has received a status change.	631	The pid specified in the C<ev_child> watcher has received a status change.
		632
		633	=item C<EV_STAT>
		634
		635	The path specified in the C<ev_stat> watcher changed its attributes somehow.
346		636
347	=item C<EV_IDLE>	637	=item C<EV_IDLE>
348		638
349	The C<ev_idle> watcher has determined that you have nothing better to do.	639	The C<ev_idle> watcher has determined that you have nothing better to do.
350		640
…		…
358	received events. Callbacks of both watcher types can start and stop as	648	received events. Callbacks of both watcher types can start and stop as
359	many watchers as they want, and all of them will be taken into account	649	many watchers as they want, and all of them will be taken into account
360	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	650	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
361	C<ev_loop> from blocking).	651	C<ev_loop> from blocking).
362		652
		653	=item C<EV_EMBED>
		654
		655	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		656
		657	=item C<EV_FORK>
		658
		659	The event loop has been resumed in the child process after fork (see
		660	C<ev_fork>).
		661
363	=item C<EV_ERROR>	662	=item C<EV_ERROR>
364		663
365	An unspecified error has occured, the watcher has been stopped. This might	664	An unspecified error has occured, the watcher has been stopped. This might
366	happen because the watcher could not be properly started because libev	665	happen because the watcher could not be properly started because libev
367	ran out of memory, a file descriptor was found to be closed or any other	666	ran out of memory, a file descriptor was found to be closed or any other
…		…
373	your callbacks is well-written it can just attempt the operation and cope	672	your callbacks is well-written it can just attempt the operation and cope
374	with the error from read() or write(). This will not work in multithreaded	673	with the error from read() or write(). This will not work in multithreaded
375	programs, though, so beware.	674	programs, though, so beware.
376		675
377	=back	676	=back
		677
		678	=head2 GENERIC WATCHER FUNCTIONS
		679
		680	In the following description, C<TYPE> stands for the watcher type,
		681	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
		682
		683	=over 4
		684
		685	=item C<ev_init> (ev_TYPE *watcher, callback)
		686
		687	This macro initialises the generic portion of a watcher. The contents
		688	of the watcher object can be arbitrary (so C<malloc> will do). Only
		689	the generic parts of the watcher are initialised, you I<need> to call
		690	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		691	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		692	which rolls both calls into one.
		693
		694	You can reinitialise a watcher at any time as long as it has been stopped
		695	(or never started) and there are no pending events outstanding.
		696
		697	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,
		698	int revents)>.
		699
		700	=item C<ev_TYPE_set> (ev_TYPE *, [args])
		701
		702	This macro initialises the type-specific parts of a watcher. You need to
		703	call C<ev_init> at least once before you call this macro, but you can
		704	call C<ev_TYPE_set> any number of times. You must not, however, call this
		705	macro on a watcher that is active (it can be pending, however, which is a
		706	difference to the C<ev_init> macro).
		707
		708	Although some watcher types do not have type-specific arguments
		709	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		710
		711	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		712
		713	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		714	calls into a single call. This is the most convinient method to initialise
		715	a watcher. The same limitations apply, of course.
		716
		717	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
		718
		719	Starts (activates) the given watcher. Only active watchers will receive
		720	events. If the watcher is already active nothing will happen.
		721
		722	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
		723
		724	Stops the given watcher again (if active) and clears the pending
		725	status. It is possible that stopped watchers are pending (for example,
		726	non-repeating timers are being stopped when they become pending), but
		727	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If
		728	you want to free or reuse the memory used by the watcher it is therefore a
		729	good idea to always call its C<ev_TYPE_stop> function.
		730
		731	=item bool ev_is_active (ev_TYPE *watcher)
		732
		733	Returns a true value iff the watcher is active (i.e. it has been started
		734	and not yet been stopped). As long as a watcher is active you must not modify
		735	it.
		736
		737	=item bool ev_is_pending (ev_TYPE *watcher)
		738
		739	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		740	events but its callback has not yet been invoked). As long as a watcher
		741	is pending (but not active) you must not call an init function on it (but
		742	C<ev_TYPE_set> is safe), you must not change its priority, and you must
		743	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		744	it).
		745
		746	=item callback ev_cb (ev_TYPE *watcher)
		747
		748	Returns the callback currently set on the watcher.
		749
		750	=item ev_cb_set (ev_TYPE *watcher, callback)
		751
		752	Change the callback. You can change the callback at virtually any time
		753	(modulo threads).
		754
		755	=item ev_set_priority (ev_TYPE *watcher, priority)
		756
		757	=item int ev_priority (ev_TYPE *watcher)
		758
		759	Set and query the priority of the watcher. The priority is a small
		760	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		761	(default: C<-2>). Pending watchers with higher priority will be invoked
		762	before watchers with lower priority, but priority will not keep watchers
		763	from being executed (except for C<ev_idle> watchers).
		764
		765	This means that priorities are I<only> used for ordering callback
		766	invocation after new events have been received. This is useful, for
		767	example, to reduce latency after idling, or more often, to bind two
		768	watchers on the same event and make sure one is called first.
		769
		770	If you need to suppress invocation when higher priority events are pending
		771	you need to look at C<ev_idle> watchers, which provide this functionality.
		772
		773	You I<must not> change the priority of a watcher as long as it is active or
		774	pending.
		775
		776	The default priority used by watchers when no priority has been set is
		777	always C<0>, which is supposed to not be too high and not be too low :).
		778
		779	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		780	fine, as long as you do not mind that the priority value you query might
		781	or might not have been adjusted to be within valid range.
		782
		783	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		784
		785	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		786	C<loop> nor C<revents> need to be valid as long as the watcher callback
		787	can deal with that fact.
		788
		789	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		790
		791	If the watcher is pending, this function returns clears its pending status
		792	and returns its C<revents> bitset (as if its callback was invoked). If the
		793	watcher isn't pending it does nothing and returns C<0>.
		794
		795	=back
		796
378		797
379	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	798	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
380		799
381	Each watcher has, by default, a member C<void *data> that you can change	800	Each watcher has, by default, a member C<void *data> that you can change
382	and read at any time, libev will completely ignore it. This can be used	801	and read at any time, libev will completely ignore it. This can be used
…		…
400	{	819	{
401	struct my_io *w = (struct my_io *)w_;	820	struct my_io *w = (struct my_io *)w_;
402	...	821	...
403	}	822	}
404		823
405	More interesting and less C-conformant ways of catsing your callback type	824	More interesting and less C-conformant ways of casting your callback type
406	have been omitted....	825	instead have been omitted.
		826
		827	Another common scenario is having some data structure with multiple
		828	watchers:
		829
		830	struct my_biggy
		831	{
		832	int some_data;
		833	ev_timer t1;
		834	ev_timer t2;
		835	}
		836
		837	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		838	you need to use C<offsetof>:
		839
		840	#include <stddef.h>
		841
		842	static void
		843	t1_cb (EV_P_ struct ev_timer *w, int revents)
		844	{
		845	struct my_biggy big = (struct my_biggy *
		846	(((char *)w) - offsetof (struct my_biggy, t1));
		847	}
		848
		849	static void
		850	t2_cb (EV_P_ struct ev_timer *w, int revents)
		851	{
		852	struct my_biggy big = (struct my_biggy *
		853	(((char *)w) - offsetof (struct my_biggy, t2));
		854	}
407		855
408		856
409	=head1 WATCHER TYPES	857	=head1 WATCHER TYPES
410		858
411	This section describes each watcher in detail, but will not repeat	859	This section describes each watcher in detail, but will not repeat
412	information given in the last section.	860	information given in the last section. Any initialisation/set macros,
		861	functions and members specific to the watcher type are explained.
413		862
		863	Members are additionally marked with either I<[read-only]>, meaning that,
		864	while the watcher is active, you can look at the member and expect some
		865	sensible content, but you must not modify it (you can modify it while the
		866	watcher is stopped to your hearts content), or I<[read-write]>, which
		867	means you can expect it to have some sensible content while the watcher
		868	is active, but you can also modify it. Modifying it may not do something
		869	sensible or take immediate effect (or do anything at all), but libev will
		870	not crash or malfunction in any way.
		871
		872
414	=head2 C<ev_io> - is this file descriptor readable or writable	873	=head2 C<ev_io> - is this file descriptor readable or writable?
415		874
416	I/O watchers check whether a file descriptor is readable or writable	875	I/O watchers check whether a file descriptor is readable or writable
417	in each iteration of the event loop (This behaviour is called	876	in each iteration of the event loop, or, more precisely, when reading
418	level-triggering because you keep receiving events as long as the	877	would not block the process and writing would at least be able to write
419	condition persists. Remember you can stop the watcher if you don't want to	878	some data. This behaviour is called level-triggering because you keep
420	act on the event and neither want to receive future events).	879	receiving events as long as the condition persists. Remember you can stop
		880	the watcher if you don't want to act on the event and neither want to
		881	receive future events.
421		882
422	In general you can register as many read and/or write event watchers oer	883	In general you can register as many read and/or write event watchers per
423	fd as you want (as long as you don't confuse yourself). Setting all file	884	fd as you want (as long as you don't confuse yourself). Setting all file
424	descriptors to non-blocking mode is also usually a good idea (but not	885	descriptors to non-blocking mode is also usually a good idea (but not
425	required if you know what you are doing).	886	required if you know what you are doing).
426		887
427	You have to be careful with dup'ed file descriptors, though. Some backends	888	You have to be careful with dup'ed file descriptors, though. Some backends
428	(the linux epoll backend is a notable example) cannot handle dup'ed file	889	(the linux epoll backend is a notable example) cannot handle dup'ed file
429	descriptors correctly if you register interest in two or more fds pointing	890	descriptors correctly if you register interest in two or more fds pointing
430	to the same file/socket etc. description.	891	to the same underlying file/socket/etc. description (that is, they share
		892	the same underlying "file open").
431		893
432	If you must do this, then force the use of a known-to-be-good backend	894	If you must do this, then force the use of a known-to-be-good backend
433	(at the time of this writing, this includes only EVMETHOD_SELECT and	895	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
434	EVMETHOD_POLL).	896	C<EVBACKEND_POLL>).
		897
		898	Another thing you have to watch out for is that it is quite easy to
		899	receive "spurious" readyness notifications, that is your callback might
		900	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		901	because there is no data. Not only are some backends known to create a
		902	lot of those (for example solaris ports), it is very easy to get into
		903	this situation even with a relatively standard program structure. Thus
		904	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		905	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		906
		907	If you cannot run the fd in non-blocking mode (for example you should not
		908	play around with an Xlib connection), then you have to seperately re-test
		909	whether a file descriptor is really ready with a known-to-be good interface
		910	such as poll (fortunately in our Xlib example, Xlib already does this on
		911	its own, so its quite safe to use).
435		912
436	=over 4	913	=over 4
437		914
438	=item ev_io_init (ev_io *, callback, int fd, int events)	915	=item ev_io_init (ev_io *, callback, int fd, int events)
439		916
440	=item ev_io_set (ev_io *, int fd, int events)	917	=item ev_io_set (ev_io *, int fd, int events)
441		918
442	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	919	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
443	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	920	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or
444	EV_WRITE> to receive the given events.	921	C<EV_READ | EV_WRITE> to receive the given events.
		922
		923	=item int fd [read-only]
		924
		925	The file descriptor being watched.
		926
		927	=item int events [read-only]
		928
		929	The events being watched.
445		930
446	=back	931	=back
447		932
		933	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
		934	readable, but only once. Since it is likely line-buffered, you could
		935	attempt to read a whole line in the callback.
		936
		937	static void
		938	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		939	{
		940	ev_io_stop (loop, w);
		941	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		942	}
		943
		944	...
		945	struct ev_loop *loop = ev_default_init (0);
		946	struct ev_io stdin_readable;
		947	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		948	ev_io_start (loop, &stdin_readable);
		949	ev_loop (loop, 0);
		950
		951
448	=head2 C<ev_timer> - relative and optionally recurring timeouts	952	=head2 C<ev_timer> - relative and optionally repeating timeouts
449		953
450	Timer watchers are simple relative timers that generate an event after a	954	Timer watchers are simple relative timers that generate an event after a
451	given time, and optionally repeating in regular intervals after that.	955	given time, and optionally repeating in regular intervals after that.
452		956
453	The timers are based on real time, that is, if you register an event that	957	The timers are based on real time, that is, if you register an event that
454	times out after an hour and youreset your system clock to last years	958	times out after an hour and you reset your system clock to last years
455	time, it will still time out after (roughly) and hour. "Roughly" because	959	time, it will still time out after (roughly) and hour. "Roughly" because
456	detecting time jumps is hard, and soem inaccuracies are unavoidable (the	960	detecting time jumps is hard, and some inaccuracies are unavoidable (the
457	monotonic clock option helps a lot here).	961	monotonic clock option helps a lot here).
458		962
459	The relative timeouts are calculated relative to the C<ev_now ()>	963	The relative timeouts are calculated relative to the C<ev_now ()>
460	time. This is usually the right thing as this timestamp refers to the time	964	time. This is usually the right thing as this timestamp refers to the time
461	of the event triggering whatever timeout you are modifying/starting. If	965	of the event triggering whatever timeout you are modifying/starting. If
462	you suspect event processing to be delayed and you *need* to base the timeout	966	you suspect event processing to be delayed and you I<need> to base the timeout
463	ion the current time, use something like this to adjust for this:	967	on the current time, use something like this to adjust for this:
464		968
465	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	969	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
		970
		971	The callback is guarenteed to be invoked only when its timeout has passed,
		972	but if multiple timers become ready during the same loop iteration then
		973	order of execution is undefined.
466		974
467	=over 4	975	=over 4
468		976
469	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	977	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
470		978
…		…
476	later, again, and again, until stopped manually.	984	later, again, and again, until stopped manually.
477		985
478	The timer itself will do a best-effort at avoiding drift, that is, if you	986	The timer itself will do a best-effort at avoiding drift, that is, if you
479	configure a timer to trigger every 10 seconds, then it will trigger at	987	configure a timer to trigger every 10 seconds, then it will trigger at
480	exactly 10 second intervals. If, however, your program cannot keep up with	988	exactly 10 second intervals. If, however, your program cannot keep up with
481	the timer (ecause it takes longer than those 10 seconds to do stuff) the	989	the timer (because it takes longer than those 10 seconds to do stuff) the
482	timer will not fire more than once per event loop iteration.	990	timer will not fire more than once per event loop iteration.
483		991
484	=item ev_timer_again (loop)	992	=item ev_timer_again (loop)
485		993
486	This will act as if the timer timed out and restart it again if it is	994	This will act as if the timer timed out and restart it again if it is
487	repeating. The exact semantics are:	995	repeating. The exact semantics are:
488		996
		997	If the timer is pending, its pending status is cleared.
		998
489	If the timer is started but nonrepeating, stop it.	999	If the timer is started but nonrepeating, stop it (as if it timed out).
490		1000
491	If the timer is repeating, either start it if necessary (with the repeat	1001	If the timer is repeating, either start it if necessary (with the
492	value), or reset the running timer to the repeat value.	1002	C<repeat> value), or reset the running timer to the C<repeat> value.
493		1003
494	This sounds a bit complicated, but here is a useful and typical	1004	This sounds a bit complicated, but here is a useful and typical
495	example: Imagine you have a tcp connection and you want a so-called idle	1005	example: Imagine you have a tcp connection and you want a so-called idle
496	timeout, that is, you want to be called when there have been, say, 60	1006	timeout, that is, you want to be called when there have been, say, 60
497	seconds of inactivity on the socket. The easiest way to do this is to	1007	seconds of inactivity on the socket. The easiest way to do this is to
498	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	1008	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
499	time you successfully read or write some data. If you go into an idle	1009	C<ev_timer_again> each time you successfully read or write some data. If
500	state where you do not expect data to travel on the socket, you can stop	1010	you go into an idle state where you do not expect data to travel on the
501	the timer, and again will automatically restart it if need be.	1011	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
		1012	automatically restart it if need be.
		1013
		1014	That means you can ignore the C<after> value and C<ev_timer_start>
		1015	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
		1016
		1017	ev_timer_init (timer, callback, 0., 5.);
		1018	ev_timer_again (loop, timer);
		1019	...
		1020	timer->again = 17.;
		1021	ev_timer_again (loop, timer);
		1022	...
		1023	timer->again = 10.;
		1024	ev_timer_again (loop, timer);
		1025
		1026	This is more slightly efficient then stopping/starting the timer each time
		1027	you want to modify its timeout value.
		1028
		1029	=item ev_tstamp repeat [read-write]
		1030
		1031	The current C<repeat> value. Will be used each time the watcher times out
		1032	or C<ev_timer_again> is called and determines the next timeout (if any),
		1033	which is also when any modifications are taken into account.
502		1034
503	=back	1035	=back
504		1036
		1037	Example: Create a timer that fires after 60 seconds.
		1038
		1039	static void
		1040	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1041	{
		1042	.. one minute over, w is actually stopped right here
		1043	}
		1044
		1045	struct ev_timer mytimer;
		1046	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		1047	ev_timer_start (loop, &mytimer);
		1048
		1049	Example: Create a timeout timer that times out after 10 seconds of
		1050	inactivity.
		1051
		1052	static void
		1053	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1054	{
		1055	.. ten seconds without any activity
		1056	}
		1057
		1058	struct ev_timer mytimer;
		1059	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		1060	ev_timer_again (&mytimer); /* start timer */
		1061	ev_loop (loop, 0);
		1062
		1063	// and in some piece of code that gets executed on any "activity":
		1064	// reset the timeout to start ticking again at 10 seconds
		1065	ev_timer_again (&mytimer);
		1066
		1067
505	=head2 C<ev_periodic> - to cron or not to cron	1068	=head2 C<ev_periodic> - to cron or not to cron?
506		1069
507	Periodic watchers are also timers of a kind, but they are very versatile	1070	Periodic watchers are also timers of a kind, but they are very versatile
508	(and unfortunately a bit complex).	1071	(and unfortunately a bit complex).
509		1072
510	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1073	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
511	but on wallclock time (absolute time). You can tell a periodic watcher	1074	but on wallclock time (absolute time). You can tell a periodic watcher
512	to trigger "at" some specific point in time. For example, if you tell a	1075	to trigger "at" some specific point in time. For example, if you tell a
513	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	1076	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
514	+ 10.>) and then reset your system clock to the last year, then it will	1077	+ 10.>) and then reset your system clock to the last year, then it will
515	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1078	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
516	roughly 10 seconds later and of course not if you reset your system time	1079	roughly 10 seconds later and of course not if you reset your system time
517	again).	1080	again).
518		1081
519	They can also be used to implement vastly more complex timers, such as	1082	They can also be used to implement vastly more complex timers, such as
520	triggering an event on eahc midnight, local time.	1083	triggering an event on eahc midnight, local time.
521		1084
		1085	As with timers, the callback is guarenteed to be invoked only when the
		1086	time (C<at>) has been passed, but if multiple periodic timers become ready
		1087	during the same loop iteration then order of execution is undefined.
		1088
522	=over 4	1089	=over 4
523		1090
524	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1091	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
525		1092
526	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1093	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
527		1094
528	Lots of arguments, lets sort it out... There are basically three modes of	1095	Lots of arguments, lets sort it out... There are basically three modes of
529	operation, and we will explain them from simplest to complex:	1096	operation, and we will explain them from simplest to complex:
530
531		1097
532	=over 4	1098	=over 4
533		1099
534	=item * absolute timer (interval = reschedule_cb = 0)	1100	=item * absolute timer (interval = reschedule_cb = 0)
535		1101
…		…
599	Simply stops and restarts the periodic watcher again. This is only useful	1165	Simply stops and restarts the periodic watcher again. This is only useful
600	when you changed some parameters or the reschedule callback would return	1166	when you changed some parameters or the reschedule callback would return
601	a different time than the last time it was called (e.g. in a crond like	1167	a different time than the last time it was called (e.g. in a crond like
602	program when the crontabs have changed).	1168	program when the crontabs have changed).
603		1169
		1170	=item ev_tstamp interval [read-write]
		1171
		1172	The current interval value. Can be modified any time, but changes only
		1173	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1174	called.
		1175
		1176	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]
		1177
		1178	The current reschedule callback, or C<0>, if this functionality is
		1179	switched off. Can be changed any time, but changes only take effect when
		1180	the periodic timer fires or C<ev_periodic_again> is being called.
		1181
604	=back	1182	=back
605		1183
		1184	Example: Call a callback every hour, or, more precisely, whenever the
		1185	system clock is divisible by 3600. The callback invocation times have
		1186	potentially a lot of jittering, but good long-term stability.
		1187
		1188	static void
		1189	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		1190	{
		1191	... its now a full hour (UTC, or TAI or whatever your clock follows)
		1192	}
		1193
		1194	struct ev_periodic hourly_tick;
		1195	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		1196	ev_periodic_start (loop, &hourly_tick);
		1197
		1198	Example: The same as above, but use a reschedule callback to do it:
		1199
		1200	#include <math.h>
		1201
		1202	static ev_tstamp
		1203	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		1204	{
		1205	return fmod (now, 3600.) + 3600.;
		1206	}
		1207
		1208	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		1209
		1210	Example: Call a callback every hour, starting now:
		1211
		1212	struct ev_periodic hourly_tick;
		1213	ev_periodic_init (&hourly_tick, clock_cb,
		1214	fmod (ev_now (loop), 3600.), 3600., 0);
		1215	ev_periodic_start (loop, &hourly_tick);
		1216
		1217
606	=head2 C<ev_signal> - signal me when a signal gets signalled	1218	=head2 C<ev_signal> - signal me when a signal gets signalled!
607		1219
608	Signal watchers will trigger an event when the process receives a specific	1220	Signal watchers will trigger an event when the process receives a specific
609	signal one or more times. Even though signals are very asynchronous, libev	1221	signal one or more times. Even though signals are very asynchronous, libev
610	will try it's best to deliver signals synchronously, i.e. as part of the	1222	will try it's best to deliver signals synchronously, i.e. as part of the
611	normal event processing, like any other event.	1223	normal event processing, like any other event.
…		…
624	=item ev_signal_set (ev_signal *, int signum)	1236	=item ev_signal_set (ev_signal *, int signum)
625		1237
626	Configures the watcher to trigger on the given signal number (usually one	1238	Configures the watcher to trigger on the given signal number (usually one
627	of the C<SIGxxx> constants).	1239	of the C<SIGxxx> constants).
628		1240
		1241	=item int signum [read-only]
		1242
		1243	The signal the watcher watches out for.
		1244
629	=back	1245	=back
630		1246
		1247
631	=head2 C<ev_child> - wait for pid status changes	1248	=head2 C<ev_child> - watch out for process status changes
632		1249
633	Child watchers trigger when your process receives a SIGCHLD in response to	1250	Child watchers trigger when your process receives a SIGCHLD in response to
634	some child status changes (most typically when a child of yours dies).	1251	some child status changes (most typically when a child of yours dies).
635		1252
636	=over 4	1253	=over 4
…		…
644	at the C<rstatus> member of the C<ev_child> watcher structure to see	1261	at the C<rstatus> member of the C<ev_child> watcher structure to see
645	the status word (use the macros from C<sys/wait.h> and see your systems	1262	the status word (use the macros from C<sys/wait.h> and see your systems
646	C<waitpid> documentation). The C<rpid> member contains the pid of the	1263	C<waitpid> documentation). The C<rpid> member contains the pid of the
647	process causing the status change.	1264	process causing the status change.
648		1265
		1266	=item int pid [read-only]
		1267
		1268	The process id this watcher watches out for, or C<0>, meaning any process id.
		1269
		1270	=item int rpid [read-write]
		1271
		1272	The process id that detected a status change.
		1273
		1274	=item int rstatus [read-write]
		1275
		1276	The process exit/trace status caused by C<rpid> (see your systems
		1277	C<waitpid> and C<sys/wait.h> documentation for details).
		1278
649	=back	1279	=back
650		1280
		1281	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1282
		1283	static void
		1284	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1285	{
		1286	ev_unloop (loop, EVUNLOOP_ALL);
		1287	}
		1288
		1289	struct ev_signal signal_watcher;
		1290	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1291	ev_signal_start (loop, &sigint_cb);
		1292
		1293
		1294	=head2 C<ev_stat> - did the file attributes just change?
		1295
		1296	This watches a filesystem path for attribute changes. That is, it calls
		1297	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1298	compared to the last time, invoking the callback if it did.
		1299
		1300	The path does not need to exist: changing from "path exists" to "path does
		1301	not exist" is a status change like any other. The condition "path does
		1302	not exist" is signified by the C<st_nlink> field being zero (which is
		1303	otherwise always forced to be at least one) and all the other fields of
		1304	the stat buffer having unspecified contents.
		1305
		1306	The path I<should> be absolute and I<must not> end in a slash. If it is
		1307	relative and your working directory changes, the behaviour is undefined.
		1308
		1309	Since there is no standard to do this, the portable implementation simply
		1310	calls C<stat (2)> regularly on the path to see if it changed somehow. You
		1311	can specify a recommended polling interval for this case. If you specify
		1312	a polling interval of C<0> (highly recommended!) then a I<suitable,
		1313	unspecified default> value will be used (which you can expect to be around
		1314	five seconds, although this might change dynamically). Libev will also
		1315	impose a minimum interval which is currently around C<0.1>, but thats
		1316	usually overkill.
		1317
		1318	This watcher type is not meant for massive numbers of stat watchers,
		1319	as even with OS-supported change notifications, this can be
		1320	resource-intensive.
		1321
		1322	At the time of this writing, only the Linux inotify interface is
		1323	implemented (implementing kqueue support is left as an exercise for the
		1324	reader). Inotify will be used to give hints only and should not change the
		1325	semantics of C<ev_stat> watchers, which means that libev sometimes needs
		1326	to fall back to regular polling again even with inotify, but changes are
		1327	usually detected immediately, and if the file exists there will be no
		1328	polling.
		1329
		1330	=over 4
		1331
		1332	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		1333
		1334	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		1335
		1336	Configures the watcher to wait for status changes of the given
		1337	C<path>. The C<interval> is a hint on how quickly a change is expected to
		1338	be detected and should normally be specified as C<0> to let libev choose
		1339	a suitable value. The memory pointed to by C<path> must point to the same
		1340	path for as long as the watcher is active.
		1341
		1342	The callback will be receive C<EV_STAT> when a change was detected,
		1343	relative to the attributes at the time the watcher was started (or the
		1344	last change was detected).
		1345
		1346	=item ev_stat_stat (ev_stat *)
		1347
		1348	Updates the stat buffer immediately with new values. If you change the
		1349	watched path in your callback, you could call this fucntion to avoid
		1350	detecting this change (while introducing a race condition). Can also be
		1351	useful simply to find out the new values.
		1352
		1353	=item ev_statdata attr [read-only]
		1354
		1355	The most-recently detected attributes of the file. Although the type is of
		1356	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		1357	suitable for your system. If the C<st_nlink> member is C<0>, then there
		1358	was some error while C<stat>ing the file.
		1359
		1360	=item ev_statdata prev [read-only]
		1361
		1362	The previous attributes of the file. The callback gets invoked whenever
		1363	C<prev> != C<attr>.
		1364
		1365	=item ev_tstamp interval [read-only]
		1366
		1367	The specified interval.
		1368
		1369	=item const char *path [read-only]
		1370
		1371	The filesystem path that is being watched.
		1372
		1373	=back
		1374
		1375	Example: Watch C</etc/passwd> for attribute changes.
		1376
		1377	static void
		1378	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		1379	{
		1380	/* /etc/passwd changed in some way */
		1381	if (w->attr.st_nlink)
		1382	{
		1383	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1384	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1385	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1386	}
		1387	else
		1388	/* you shalt not abuse printf for puts */
		1389	puts ("wow, /etc/passwd is not there, expect problems. "
		1390	"if this is windows, they already arrived\n");
		1391	}
		1392
		1393	...
		1394	ev_stat passwd;
		1395
		1396	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
		1397	ev_stat_start (loop, &passwd);
		1398
		1399
651	=head2 C<ev_idle> - when you've got nothing better to do	1400	=head2 C<ev_idle> - when you've got nothing better to do...
652		1401
653	Idle watchers trigger events when there are no other events are pending	1402	Idle watchers trigger events when no other events of the same or higher
654	(prepare, check and other idle watchers do not count). That is, as long	1403	priority are pending (prepare, check and other idle watchers do not
655	as your process is busy handling sockets or timeouts (or even signals,	1404	count).
656	imagine) it will not be triggered. But when your process is idle all idle	1405
657	watchers are being called again and again, once per event loop iteration -	1406	That is, as long as your process is busy handling sockets or timeouts
		1407	(or even signals, imagine) of the same or higher priority it will not be
		1408	triggered. But when your process is idle (or only lower-priority watchers
		1409	are pending), the idle watchers are being called once per event loop
658	until stopped, that is, or your process receives more events and becomes	1410	iteration - until stopped, that is, or your process receives more events
659	busy.	1411	and becomes busy again with higher priority stuff.
660		1412
661	The most noteworthy effect is that as long as any idle watchers are	1413	The most noteworthy effect is that as long as any idle watchers are
662	active, the process will not block when waiting for new events.	1414	active, the process will not block when waiting for new events.
663		1415
664	Apart from keeping your process non-blocking (which is a useful	1416	Apart from keeping your process non-blocking (which is a useful
…		…
674	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1426	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
675	believe me.	1427	believe me.
676		1428
677	=back	1429	=back
678		1430
		1431	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
		1432	callback, free it. Also, use no error checking, as usual.
		1433
		1434	static void
		1435	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
		1436	{
		1437	free (w);
		1438	// now do something you wanted to do when the program has
		1439	// no longer asnything immediate to do.
		1440	}
		1441
		1442	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1443	ev_idle_init (idle_watcher, idle_cb);
		1444	ev_idle_start (loop, idle_cb);
		1445
		1446
679	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1447	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
680		1448
681	Prepare and check watchers are usually (but not always) used in tandem:	1449	Prepare and check watchers are usually (but not always) used in tandem:
682	prepare watchers get invoked before the process blocks and check watchers	1450	prepare watchers get invoked before the process blocks and check watchers
683	afterwards.	1451	afterwards.
684		1452
		1453	You I<must not> call C<ev_loop> or similar functions that enter
		1454	the current event loop from either C<ev_prepare> or C<ev_check>
		1455	watchers. Other loops than the current one are fine, however. The
		1456	rationale behind this is that you do not need to check for recursion in
		1457	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		1458	C<ev_check> so if you have one watcher of each kind they will always be
		1459	called in pairs bracketing the blocking call.
		1460
685	Their main purpose is to integrate other event mechanisms into libev. This	1461	Their main purpose is to integrate other event mechanisms into libev and
686	could be used, for example, to track variable changes, implement your own	1462	their use is somewhat advanced. This could be used, for example, to track
687	watchers, integrate net-snmp or a coroutine library and lots more.	1463	variable changes, implement your own watchers, integrate net-snmp or a
		1464	coroutine library and lots more. They are also occasionally useful if
		1465	you cache some data and want to flush it before blocking (for example,
		1466	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		1467	watcher).
688		1468
689	This is done by examining in each prepare call which file descriptors need	1469	This is done by examining in each prepare call which file descriptors need
690	to be watched by the other library, registering C<ev_io> watchers for	1470	to be watched by the other library, registering C<ev_io> watchers for
691	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1471	them and starting an C<ev_timer> watcher for any timeouts (many libraries
692	provide just this functionality). Then, in the check watcher you check for	1472	provide just this functionality). Then, in the check watcher you check for
…		…
702	with priority higher than or equal to the event loop and one coroutine	1482	with priority higher than or equal to the event loop and one coroutine
703	of lower priority, but only once, using idle watchers to keep the event	1483	of lower priority, but only once, using idle watchers to keep the event
704	loop from blocking if lower-priority coroutines are active, thus mapping	1484	loop from blocking if lower-priority coroutines are active, thus mapping
705	low-priority coroutines to idle/background tasks).	1485	low-priority coroutines to idle/background tasks).
706		1486
		1487	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1488	priority, to ensure that they are being run before any other watchers
		1489	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1490	too) should not activate ("feed") events into libev. While libev fully
		1491	supports this, they will be called before other C<ev_check> watchers did
		1492	their job. As C<ev_check> watchers are often used to embed other event
		1493	loops those other event loops might be in an unusable state until their
		1494	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		1495	others).
		1496
707	=over 4	1497	=over 4
708		1498
709	=item ev_prepare_init (ev_prepare *, callback)	1499	=item ev_prepare_init (ev_prepare *, callback)
710		1500
711	=item ev_check_init (ev_check *, callback)	1501	=item ev_check_init (ev_check *, callback)
…		…
714	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1504	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
715	macros, but using them is utterly, utterly and completely pointless.	1505	macros, but using them is utterly, utterly and completely pointless.
716		1506
717	=back	1507	=back
718		1508
		1509	There are a number of principal ways to embed other event loops or modules
		1510	into libev. Here are some ideas on how to include libadns into libev
		1511	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1512	use for an actually working example. Another Perl module named C<EV::Glib>
		1513	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV
		1514	into the Glib event loop).
		1515
		1516	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		1517	and in a check watcher, destroy them and call into libadns. What follows
		1518	is pseudo-code only of course. This requires you to either use a low
		1519	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1520	the callbacks for the IO/timeout watchers might not have been called yet.
		1521
		1522	static ev_io iow [nfd];
		1523	static ev_timer tw;
		1524
		1525	static void
		1526	io_cb (ev_loop *loop, ev_io *w, int revents)
		1527	{
		1528	}
		1529
		1530	// create io watchers for each fd and a timer before blocking
		1531	static void
		1532	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
		1533	{
		1534	int timeout = 3600000;
		1535	struct pollfd fds [nfd];
		1536	// actual code will need to loop here and realloc etc.
		1537	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1538
		1539	/* the callback is illegal, but won't be called as we stop during check */
		1540	ev_timer_init (&tw, 0, timeout * 1e-3);
		1541	ev_timer_start (loop, &tw);
		1542
		1543	// create one ev_io per pollfd
		1544	for (int i = 0; i < nfd; ++i)
		1545	{
		1546	ev_io_init (iow + i, io_cb, fds [i].fd,
		1547	((fds [i].events & POLLIN ? EV_READ : 0)
		1548	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1549
		1550	fds [i].revents = 0;
		1551	ev_io_start (loop, iow + i);
		1552	}
		1553	}
		1554
		1555	// stop all watchers after blocking
		1556	static void
		1557	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
		1558	{
		1559	ev_timer_stop (loop, &tw);
		1560
		1561	for (int i = 0; i < nfd; ++i)
		1562	{
		1563	// set the relevant poll flags
		1564	// could also call adns_processreadable etc. here
		1565	struct pollfd *fd = fds + i;
		1566	int revents = ev_clear_pending (iow + i);
		1567	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1568	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1569
		1570	// now stop the watcher
		1571	ev_io_stop (loop, iow + i);
		1572	}
		1573
		1574	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1575	}
		1576
		1577	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		1578	in the prepare watcher and would dispose of the check watcher.
		1579
		1580	Method 3: If the module to be embedded supports explicit event
		1581	notification (adns does), you can also make use of the actual watcher
		1582	callbacks, and only destroy/create the watchers in the prepare watcher.
		1583
		1584	static void
		1585	timer_cb (EV_P_ ev_timer *w, int revents)
		1586	{
		1587	adns_state ads = (adns_state)w->data;
		1588	update_now (EV_A);
		1589
		1590	adns_processtimeouts (ads, &tv_now);
		1591	}
		1592
		1593	static void
		1594	io_cb (EV_P_ ev_io *w, int revents)
		1595	{
		1596	adns_state ads = (adns_state)w->data;
		1597	update_now (EV_A);
		1598
		1599	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		1600	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		1601	}
		1602
		1603	// do not ever call adns_afterpoll
		1604
		1605	Method 4: Do not use a prepare or check watcher because the module you
		1606	want to embed is too inflexible to support it. Instead, youc na override
		1607	their poll function. The drawback with this solution is that the main
		1608	loop is now no longer controllable by EV. The C<Glib::EV> module does
		1609	this.
		1610
		1611	static gint
		1612	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		1613	{
		1614	int got_events = 0;
		1615
		1616	for (n = 0; n < nfds; ++n)
		1617	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		1618
		1619	if (timeout >= 0)
		1620	// create/start timer
		1621
		1622	// poll
		1623	ev_loop (EV_A_ 0);
		1624
		1625	// stop timer again
		1626	if (timeout >= 0)
		1627	ev_timer_stop (EV_A_ &to);
		1628
		1629	// stop io watchers again - their callbacks should have set
		1630	for (n = 0; n < nfds; ++n)
		1631	ev_io_stop (EV_A_ iow [n]);
		1632
		1633	return got_events;
		1634	}
		1635
		1636
		1637	=head2 C<ev_embed> - when one backend isn't enough...
		1638
		1639	This is a rather advanced watcher type that lets you embed one event loop
		1640	into another (currently only C<ev_io> events are supported in the embedded
		1641	loop, other types of watchers might be handled in a delayed or incorrect
		1642	fashion and must not be used).
		1643
		1644	There are primarily two reasons you would want that: work around bugs and
		1645	prioritise I/O.
		1646
		1647	As an example for a bug workaround, the kqueue backend might only support
		1648	sockets on some platform, so it is unusable as generic backend, but you
		1649	still want to make use of it because you have many sockets and it scales
		1650	so nicely. In this case, you would create a kqueue-based loop and embed it
		1651	into your default loop (which might use e.g. poll). Overall operation will
		1652	be a bit slower because first libev has to poll and then call kevent, but
		1653	at least you can use both at what they are best.
		1654
		1655	As for prioritising I/O: rarely you have the case where some fds have
		1656	to be watched and handled very quickly (with low latency), and even
		1657	priorities and idle watchers might have too much overhead. In this case
		1658	you would put all the high priority stuff in one loop and all the rest in
		1659	a second one, and embed the second one in the first.
		1660
		1661	As long as the watcher is active, the callback will be invoked every time
		1662	there might be events pending in the embedded loop. The callback must then
		1663	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
		1664	their callbacks (you could also start an idle watcher to give the embedded
		1665	loop strictly lower priority for example). You can also set the callback
		1666	to C<0>, in which case the embed watcher will automatically execute the
		1667	embedded loop sweep.
		1668
		1669	As long as the watcher is started it will automatically handle events. The
		1670	callback will be invoked whenever some events have been handled. You can
		1671	set the callback to C<0> to avoid having to specify one if you are not
		1672	interested in that.
		1673
		1674	Also, there have not currently been made special provisions for forking:
		1675	when you fork, you not only have to call C<ev_loop_fork> on both loops,
		1676	but you will also have to stop and restart any C<ev_embed> watchers
		1677	yourself.
		1678
		1679	Unfortunately, not all backends are embeddable, only the ones returned by
		1680	C<ev_embeddable_backends> are, which, unfortunately, does not include any
		1681	portable one.
		1682
		1683	So when you want to use this feature you will always have to be prepared
		1684	that you cannot get an embeddable loop. The recommended way to get around
		1685	this is to have a separate variables for your embeddable loop, try to
		1686	create it, and if that fails, use the normal loop for everything:
		1687
		1688	struct ev_loop *loop_hi = ev_default_init (0);
		1689	struct ev_loop *loop_lo = 0;
		1690	struct ev_embed embed;
		1691
		1692	// see if there is a chance of getting one that works
		1693	// (remember that a flags value of 0 means autodetection)
		1694	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		1695	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		1696	: 0;
		1697
		1698	// if we got one, then embed it, otherwise default to loop_hi
		1699	if (loop_lo)
		1700	{
		1701	ev_embed_init (&embed, 0, loop_lo);
		1702	ev_embed_start (loop_hi, &embed);
		1703	}
		1704	else
		1705	loop_lo = loop_hi;
		1706
		1707	=over 4
		1708
		1709	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		1710
		1711	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		1712
		1713	Configures the watcher to embed the given loop, which must be
		1714	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		1715	invoked automatically, otherwise it is the responsibility of the callback
		1716	to invoke it (it will continue to be called until the sweep has been done,
		1717	if you do not want thta, you need to temporarily stop the embed watcher).
		1718
		1719	=item ev_embed_sweep (loop, ev_embed *)
		1720
		1721	Make a single, non-blocking sweep over the embedded loop. This works
		1722	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		1723	apropriate way for embedded loops.
		1724
		1725	=item struct ev_loop *loop [read-only]
		1726
		1727	The embedded event loop.
		1728
		1729	=back
		1730
		1731
		1732	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		1733
		1734	Fork watchers are called when a C<fork ()> was detected (usually because
		1735	whoever is a good citizen cared to tell libev about it by calling
		1736	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		1737	event loop blocks next and before C<ev_check> watchers are being called,
		1738	and only in the child after the fork. If whoever good citizen calling
		1739	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		1740	handlers will be invoked, too, of course.
		1741
		1742	=over 4
		1743
		1744	=item ev_fork_init (ev_signal *, callback)
		1745
		1746	Initialises and configures the fork watcher - it has no parameters of any
		1747	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		1748	believe me.
		1749
		1750	=back
		1751
		1752
719	=head1 OTHER FUNCTIONS	1753	=head1 OTHER FUNCTIONS
720		1754
721	There are some other functions of possible interest. Described. Here. Now.	1755	There are some other functions of possible interest. Described. Here. Now.
722		1756
723	=over 4	1757	=over 4
…		…
725	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	1759	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
726		1760
727	This function combines a simple timer and an I/O watcher, calls your	1761	This function combines a simple timer and an I/O watcher, calls your
728	callback on whichever event happens first and automatically stop both	1762	callback on whichever event happens first and automatically stop both
729	watchers. This is useful if you want to wait for a single event on an fd	1763	watchers. This is useful if you want to wait for a single event on an fd
730	or timeout without havign to allocate/configure/start/stop/free one or	1764	or timeout without having to allocate/configure/start/stop/free one or
731	more watchers yourself.	1765	more watchers yourself.
732		1766
733	If C<fd> is less than 0, then no I/O watcher will be started and events	1767	If C<fd> is less than 0, then no I/O watcher will be started and events
734	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	1768	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and
735	C<events> set will be craeted and started.	1769	C<events> set will be craeted and started.
…		…
752	/* stdin might have data for us, joy! */;	1786	/* stdin might have data for us, joy! */;
753	}	1787	}
754		1788
755	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	1789	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
756		1790
757	=item ev_feed_event (loop, watcher, int events)	1791	=item ev_feed_event (ev_loop *, watcher *, int revents)
758		1792
759	Feeds the given event set into the event loop, as if the specified event	1793	Feeds the given event set into the event loop, as if the specified event
760	had happened for the specified watcher (which must be a pointer to an	1794	had happened for the specified watcher (which must be a pointer to an
761	initialised but not necessarily started event watcher).	1795	initialised but not necessarily started event watcher).
762		1796
763	=item ev_feed_fd_event (loop, int fd, int revents)	1797	=item ev_feed_fd_event (ev_loop *, int fd, int revents)
764		1798
765	Feed an event on the given fd, as if a file descriptor backend detected	1799	Feed an event on the given fd, as if a file descriptor backend detected
766	the given events it.	1800	the given events it.
767		1801
768	=item ev_feed_signal_event (loop, int signum)	1802	=item ev_feed_signal_event (ev_loop *loop, int signum)
769		1803
770	Feed an event as if the given signal occured (loop must be the default loop!).	1804	Feed an event as if the given signal occured (C<loop> must be the default
		1805	loop!).
771		1806
772	=back	1807	=back
773		1808
		1809
774	=head1 LIBEVENT EMULATION	1810	=head1 LIBEVENT EMULATION
775		1811
776	TBD.	1812	Libev offers a compatibility emulation layer for libevent. It cannot
		1813	emulate the internals of libevent, so here are some usage hints:
		1814
		1815	=over 4
		1816
		1817	=item * Use it by including <event.h>, as usual.
		1818
		1819	=item * The following members are fully supported: ev_base, ev_callback,
		1820	ev_arg, ev_fd, ev_res, ev_events.
		1821
		1822	=item * Avoid using ev_flags and the EVLIST_*-macros, while it is
		1823	maintained by libev, it does not work exactly the same way as in libevent (consider
		1824	it a private API).
		1825
		1826	=item * Priorities are not currently supported. Initialising priorities
		1827	will fail and all watchers will have the same priority, even though there
		1828	is an ev_pri field.
		1829
		1830	=item * Other members are not supported.
		1831
		1832	=item * The libev emulation is I<not> ABI compatible to libevent, you need
		1833	to use the libev header file and library.
		1834
		1835	=back
777		1836
778	=head1 C++ SUPPORT	1837	=head1 C++ SUPPORT
779		1838
780	TBD.	1839	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		1840	you to use some convinience methods to start/stop watchers and also change
		1841	the callback model to a model using method callbacks on objects.
		1842
		1843	To use it,
		1844
		1845	#include <ev++.h>
		1846
		1847	This automatically includes F<ev.h> and puts all of its definitions (many
		1848	of them macros) into the global namespace. All C++ specific things are
		1849	put into the C<ev> namespace. It should support all the same embedding
		1850	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
		1851
		1852	Care has been taken to keep the overhead low. The only data member the C++
		1853	classes add (compared to plain C-style watchers) is the event loop pointer
		1854	that the watcher is associated with (or no additional members at all if
		1855	you disable C<EV_MULTIPLICITY> when embedding libev).
		1856
		1857	Currently, functions, and static and non-static member functions can be
		1858	used as callbacks. Other types should be easy to add as long as they only
		1859	need one additional pointer for context. If you need support for other
		1860	types of functors please contact the author (preferably after implementing
		1861	it).
		1862
		1863	Here is a list of things available in the C<ev> namespace:
		1864
		1865	=over 4
		1866
		1867	=item C<ev::READ>, C<ev::WRITE> etc.
		1868
		1869	These are just enum values with the same values as the C<EV_READ> etc.
		1870	macros from F<ev.h>.
		1871
		1872	=item C<ev::tstamp>, C<ev::now>
		1873
		1874	Aliases to the same types/functions as with the C<ev_> prefix.
		1875
		1876	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		1877
		1878	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		1879	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		1880	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		1881	defines by many implementations.
		1882
		1883	All of those classes have these methods:
		1884
		1885	=over 4
		1886
		1887	=item ev::TYPE::TYPE ()
		1888
		1889	=item ev::TYPE::TYPE (struct ev_loop *)
		1890
		1891	=item ev::TYPE::~TYPE
		1892
		1893	The constructor (optionally) takes an event loop to associate the watcher
		1894	with. If it is omitted, it will use C<EV_DEFAULT>.
		1895
		1896	The constructor calls C<ev_init> for you, which means you have to call the
		1897	C<set> method before starting it.
		1898
		1899	It will not set a callback, however: You have to call the templated C<set>
		1900	method to set a callback before you can start the watcher.
		1901
		1902	(The reason why you have to use a method is a limitation in C++ which does
		1903	not allow explicit template arguments for constructors).
		1904
		1905	The destructor automatically stops the watcher if it is active.
		1906
		1907	=item w->set<class, &class::method> (object *)
		1908
		1909	This method sets the callback method to call. The method has to have a
		1910	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		1911	first argument and the C<revents> as second. The object must be given as
		1912	parameter and is stored in the C<data> member of the watcher.
		1913
		1914	This method synthesizes efficient thunking code to call your method from
		1915	the C callback that libev requires. If your compiler can inline your
		1916	callback (i.e. it is visible to it at the place of the C<set> call and
		1917	your compiler is good :), then the method will be fully inlined into the
		1918	thunking function, making it as fast as a direct C callback.
		1919
		1920	Example: simple class declaration and watcher initialisation
		1921
		1922	struct myclass
		1923	{
		1924	void io_cb (ev::io &w, int revents) { }
		1925	}
		1926
		1927	myclass obj;
		1928	ev::io iow;
		1929	iow.set <myclass, &myclass::io_cb> (&obj);
		1930
		1931	=item w->set<function> (void *data = 0)
		1932
		1933	Also sets a callback, but uses a static method or plain function as
		1934	callback. The optional C<data> argument will be stored in the watcher's
		1935	C<data> member and is free for you to use.
		1936
		1937	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		1938
		1939	See the method-C<set> above for more details.
		1940
		1941	Example:
		1942
		1943	static void io_cb (ev::io &w, int revents) { }
		1944	iow.set <io_cb> ();
		1945
		1946	=item w->set (struct ev_loop *)
		1947
		1948	Associates a different C<struct ev_loop> with this watcher. You can only
		1949	do this when the watcher is inactive (and not pending either).
		1950
		1951	=item w->set ([args])
		1952
		1953	Basically the same as C<ev_TYPE_set>, with the same args. Must be
		1954	called at least once. Unlike the C counterpart, an active watcher gets
		1955	automatically stopped and restarted when reconfiguring it with this
		1956	method.
		1957
		1958	=item w->start ()
		1959
		1960	Starts the watcher. Note that there is no C<loop> argument, as the
		1961	constructor already stores the event loop.
		1962
		1963	=item w->stop ()
		1964
		1965	Stops the watcher if it is active. Again, no C<loop> argument.
		1966
		1967	=item w->again () C<ev::timer>, C<ev::periodic> only
		1968
		1969	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		1970	C<ev_TYPE_again> function.
		1971
		1972	=item w->sweep () C<ev::embed> only
		1973
		1974	Invokes C<ev_embed_sweep>.
		1975
		1976	=item w->update () C<ev::stat> only
		1977
		1978	Invokes C<ev_stat_stat>.
		1979
		1980	=back
		1981
		1982	=back
		1983
		1984	Example: Define a class with an IO and idle watcher, start one of them in
		1985	the constructor.
		1986
		1987	class myclass
		1988	{
		1989	ev_io io; void io_cb (ev::io &w, int revents);
		1990	ev_idle idle void idle_cb (ev::idle &w, int revents);
		1991
		1992	myclass ();
		1993	}
		1994
		1995	myclass::myclass (int fd)
		1996	{
		1997	io .set <myclass, &myclass::io_cb > (this);
		1998	idle.set <myclass, &myclass::idle_cb> (this);
		1999
		2000	io.start (fd, ev::READ);
		2001	}
		2002
		2003
		2004	=head1 MACRO MAGIC
		2005
		2006	Libev can be compiled with a variety of options, the most fundemantal is
		2007	C<EV_MULTIPLICITY>. This option determines whether (most) functions and
		2008	callbacks have an initial C<struct ev_loop *> argument.
		2009
		2010	To make it easier to write programs that cope with either variant, the
		2011	following macros are defined:
		2012
		2013	=over 4
		2014
		2015	=item C<EV_A>, C<EV_A_>
		2016
		2017	This provides the loop I<argument> for functions, if one is required ("ev
		2018	loop argument"). The C<EV_A> form is used when this is the sole argument,
		2019	C<EV_A_> is used when other arguments are following. Example:
		2020
		2021	ev_unref (EV_A);
		2022	ev_timer_add (EV_A_ watcher);
		2023	ev_loop (EV_A_ 0);
		2024
		2025	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		2026	which is often provided by the following macro.
		2027
		2028	=item C<EV_P>, C<EV_P_>
		2029
		2030	This provides the loop I<parameter> for functions, if one is required ("ev
		2031	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		2032	C<EV_P_> is used when other parameters are following. Example:
		2033
		2034	// this is how ev_unref is being declared
		2035	static void ev_unref (EV_P);
		2036
		2037	// this is how you can declare your typical callback
		2038	static void cb (EV_P_ ev_timer *w, int revents)
		2039
		2040	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		2041	suitable for use with C<EV_A>.
		2042
		2043	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		2044
		2045	Similar to the other two macros, this gives you the value of the default
		2046	loop, if multiple loops are supported ("ev loop default").
		2047
		2048	=back
		2049
		2050	Example: Declare and initialise a check watcher, utilising the above
		2051	macros so it will work regardless of whether multiple loops are supported
		2052	or not.
		2053
		2054	static void
		2055	check_cb (EV_P_ ev_timer *w, int revents)
		2056	{
		2057	ev_check_stop (EV_A_ w);
		2058	}
		2059
		2060	ev_check check;
		2061	ev_check_init (&check, check_cb);
		2062	ev_check_start (EV_DEFAULT_ &check);
		2063	ev_loop (EV_DEFAULT_ 0);
		2064
		2065	=head1 EMBEDDING
		2066
		2067	Libev can (and often is) directly embedded into host
		2068	applications. Examples of applications that embed it include the Deliantra
		2069	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		2070	and rxvt-unicode.
		2071
		2072	The goal is to enable you to just copy the neecssary files into your
		2073	source directory without having to change even a single line in them, so
		2074	you can easily upgrade by simply copying (or having a checked-out copy of
		2075	libev somewhere in your source tree).
		2076
		2077	=head2 FILESETS
		2078
		2079	Depending on what features you need you need to include one or more sets of files
		2080	in your app.
		2081
		2082	=head3 CORE EVENT LOOP
		2083
		2084	To include only the libev core (all the C<ev_*> functions), with manual
		2085	configuration (no autoconf):
		2086
		2087	#define EV_STANDALONE 1
		2088	#include "ev.c"
		2089
		2090	This will automatically include F<ev.h>, too, and should be done in a
		2091	single C source file only to provide the function implementations. To use
		2092	it, do the same for F<ev.h> in all files wishing to use this API (best
		2093	done by writing a wrapper around F<ev.h> that you can include instead and
		2094	where you can put other configuration options):
		2095
		2096	#define EV_STANDALONE 1
		2097	#include "ev.h"
		2098
		2099	Both header files and implementation files can be compiled with a C++
		2100	compiler (at least, thats a stated goal, and breakage will be treated
		2101	as a bug).
		2102
		2103	You need the following files in your source tree, or in a directory
		2104	in your include path (e.g. in libev/ when using -Ilibev):
		2105
		2106	ev.h
		2107	ev.c
		2108	ev_vars.h
		2109	ev_wrap.h
		2110
		2111	ev_win32.c required on win32 platforms only
		2112
		2113	ev_select.c only when select backend is enabled (which is enabled by default)
		2114	ev_poll.c only when poll backend is enabled (disabled by default)
		2115	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		2116	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		2117	ev_port.c only when the solaris port backend is enabled (disabled by default)
		2118
		2119	F<ev.c> includes the backend files directly when enabled, so you only need
		2120	to compile this single file.
		2121
		2122	=head3 LIBEVENT COMPATIBILITY API
		2123
		2124	To include the libevent compatibility API, also include:
		2125
		2126	#include "event.c"
		2127
		2128	in the file including F<ev.c>, and:
		2129
		2130	#include "event.h"
		2131
		2132	in the files that want to use the libevent API. This also includes F<ev.h>.
		2133
		2134	You need the following additional files for this:
		2135
		2136	event.h
		2137	event.c
		2138
		2139	=head3 AUTOCONF SUPPORT
		2140
		2141	Instead of using C<EV_STANDALONE=1> and providing your config in
		2142	whatever way you want, you can also C<m4_include([libev.m4])> in your
		2143	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		2144	include F<config.h> and configure itself accordingly.
		2145
		2146	For this of course you need the m4 file:
		2147
		2148	libev.m4
		2149
		2150	=head2 PREPROCESSOR SYMBOLS/MACROS
		2151
		2152	Libev can be configured via a variety of preprocessor symbols you have to define
		2153	before including any of its files. The default is not to build for multiplicity
		2154	and only include the select backend.
		2155
		2156	=over 4
		2157
		2158	=item EV_STANDALONE
		2159
		2160	Must always be C<1> if you do not use autoconf configuration, which
		2161	keeps libev from including F<config.h>, and it also defines dummy
		2162	implementations for some libevent functions (such as logging, which is not
		2163	supported). It will also not define any of the structs usually found in
		2164	F<event.h> that are not directly supported by the libev core alone.
		2165
		2166	=item EV_USE_MONOTONIC
		2167
		2168	If defined to be C<1>, libev will try to detect the availability of the
		2169	monotonic clock option at both compiletime and runtime. Otherwise no use
		2170	of the monotonic clock option will be attempted. If you enable this, you
		2171	usually have to link against librt or something similar. Enabling it when
		2172	the functionality isn't available is safe, though, althoguh you have
		2173	to make sure you link against any libraries where the C<clock_gettime>
		2174	function is hiding in (often F<-lrt>).
		2175
		2176	=item EV_USE_REALTIME
		2177
		2178	If defined to be C<1>, libev will try to detect the availability of the
		2179	realtime clock option at compiletime (and assume its availability at
		2180	runtime if successful). Otherwise no use of the realtime clock option will
		2181	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		2182	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries
		2183	in the description of C<EV_USE_MONOTONIC>, though.
		2184
		2185	=item EV_USE_SELECT
		2186
		2187	If undefined or defined to be C<1>, libev will compile in support for the
		2188	C<select>(2) backend. No attempt at autodetection will be done: if no
		2189	other method takes over, select will be it. Otherwise the select backend
		2190	will not be compiled in.
		2191
		2192	=item EV_SELECT_USE_FD_SET
		2193
		2194	If defined to C<1>, then the select backend will use the system C<fd_set>
		2195	structure. This is useful if libev doesn't compile due to a missing
		2196	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on
		2197	exotic systems. This usually limits the range of file descriptors to some
		2198	low limit such as 1024 or might have other limitations (winsocket only
		2199	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		2200	influence the size of the C<fd_set> used.
		2201
		2202	=item EV_SELECT_IS_WINSOCKET
		2203
		2204	When defined to C<1>, the select backend will assume that
		2205	select/socket/connect etc. don't understand file descriptors but
		2206	wants osf handles on win32 (this is the case when the select to
		2207	be used is the winsock select). This means that it will call
		2208	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		2209	it is assumed that all these functions actually work on fds, even
		2210	on win32. Should not be defined on non-win32 platforms.
		2211
		2212	=item EV_USE_POLL
		2213
		2214	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		2215	backend. Otherwise it will be enabled on non-win32 platforms. It
		2216	takes precedence over select.
		2217
		2218	=item EV_USE_EPOLL
		2219
		2220	If defined to be C<1>, libev will compile in support for the Linux
		2221	C<epoll>(7) backend. Its availability will be detected at runtime,
		2222	otherwise another method will be used as fallback. This is the
		2223	preferred backend for GNU/Linux systems.
		2224
		2225	=item EV_USE_KQUEUE
		2226
		2227	If defined to be C<1>, libev will compile in support for the BSD style
		2228	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		2229	otherwise another method will be used as fallback. This is the preferred
		2230	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		2231	supports some types of fds correctly (the only platform we found that
		2232	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		2233	not be used unless explicitly requested. The best way to use it is to find
		2234	out whether kqueue supports your type of fd properly and use an embedded
		2235	kqueue loop.
		2236
		2237	=item EV_USE_PORT
		2238
		2239	If defined to be C<1>, libev will compile in support for the Solaris
		2240	10 port style backend. Its availability will be detected at runtime,
		2241	otherwise another method will be used as fallback. This is the preferred
		2242	backend for Solaris 10 systems.
		2243
		2244	=item EV_USE_DEVPOLL
		2245
		2246	reserved for future expansion, works like the USE symbols above.
		2247
		2248	=item EV_USE_INOTIFY
		2249
		2250	If defined to be C<1>, libev will compile in support for the Linux inotify
		2251	interface to speed up C<ev_stat> watchers. Its actual availability will
		2252	be detected at runtime.
		2253
		2254	=item EV_H
		2255
		2256	The name of the F<ev.h> header file used to include it. The default if
		2257	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This
		2258	can be used to virtually rename the F<ev.h> header file in case of conflicts.
		2259
		2260	=item EV_CONFIG_H
		2261
		2262	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		2263	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		2264	C<EV_H>, above.
		2265
		2266	=item EV_EVENT_H
		2267
		2268	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		2269	of how the F<event.h> header can be found.
		2270
		2271	=item EV_PROTOTYPES
		2272
		2273	If defined to be C<0>, then F<ev.h> will not define any function
		2274	prototypes, but still define all the structs and other symbols. This is
		2275	occasionally useful if you want to provide your own wrapper functions
		2276	around libev functions.
		2277
		2278	=item EV_MULTIPLICITY
		2279
		2280	If undefined or defined to C<1>, then all event-loop-specific functions
		2281	will have the C<struct ev_loop *> as first argument, and you can create
		2282	additional independent event loops. Otherwise there will be no support
		2283	for multiple event loops and there is no first event loop pointer
		2284	argument. Instead, all functions act on the single default loop.
		2285
		2286	=item EV_MINPRI
		2287
		2288	=item EV_MAXPRI
		2289
		2290	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2291	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2292	provide for more priorities by overriding those symbols (usually defined
		2293	to be C<-2> and C<2>, respectively).
		2294
		2295	When doing priority-based operations, libev usually has to linearly search
		2296	all the priorities, so having many of them (hundreds) uses a lot of space
		2297	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2298	fine.
		2299
		2300	If your embedding app does not need any priorities, defining these both to
		2301	C<0> will save some memory and cpu.
		2302
		2303	=item EV_PERIODIC_ENABLE
		2304
		2305	If undefined or defined to be C<1>, then periodic timers are supported. If
		2306	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2307	code.
		2308
		2309	=item EV_IDLE_ENABLE
		2310
		2311	If undefined or defined to be C<1>, then idle watchers are supported. If
		2312	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2313	code.
		2314
		2315	=item EV_EMBED_ENABLE
		2316
		2317	If undefined or defined to be C<1>, then embed watchers are supported. If
		2318	defined to be C<0>, then they are not.
		2319
		2320	=item EV_STAT_ENABLE
		2321
		2322	If undefined or defined to be C<1>, then stat watchers are supported. If
		2323	defined to be C<0>, then they are not.
		2324
		2325	=item EV_FORK_ENABLE
		2326
		2327	If undefined or defined to be C<1>, then fork watchers are supported. If
		2328	defined to be C<0>, then they are not.
		2329
		2330	=item EV_MINIMAL
		2331
		2332	If you need to shave off some kilobytes of code at the expense of some
		2333	speed, define this symbol to C<1>. Currently only used for gcc to override
		2334	some inlining decisions, saves roughly 30% codesize of amd64.
		2335
		2336	=item EV_PID_HASHSIZE
		2337
		2338	C<ev_child> watchers use a small hash table to distribute workload by
		2339	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		2340	than enough. If you need to manage thousands of children you might want to
		2341	increase this value (I<must> be a power of two).
		2342
		2343	=item EV_INOTIFY_HASHSIZE
		2344
		2345	C<ev_staz> watchers use a small hash table to distribute workload by
		2346	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		2347	usually more than enough. If you need to manage thousands of C<ev_stat>
		2348	watchers you might want to increase this value (I<must> be a power of
		2349	two).
		2350
		2351	=item EV_COMMON
		2352
		2353	By default, all watchers have a C<void *data> member. By redefining
		2354	this macro to a something else you can include more and other types of
		2355	members. You have to define it each time you include one of the files,
		2356	though, and it must be identical each time.
		2357
		2358	For example, the perl EV module uses something like this:
		2359
		2360	#define EV_COMMON \
		2361	SV *self; /* contains this struct */ \
		2362	SV *cb_sv, *fh /* note no trailing ";" */
		2363
		2364	=item EV_CB_DECLARE (type)
		2365
		2366	=item EV_CB_INVOKE (watcher, revents)
		2367
		2368	=item ev_set_cb (ev, cb)
		2369
		2370	Can be used to change the callback member declaration in each watcher,
		2371	and the way callbacks are invoked and set. Must expand to a struct member
		2372	definition and a statement, respectively. See the F<ev.v> header file for
		2373	their default definitions. One possible use for overriding these is to
		2374	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		2375	method calls instead of plain function calls in C++.
		2376
		2377	=head2 EXAMPLES
		2378
		2379	For a real-world example of a program the includes libev
		2380	verbatim, you can have a look at the EV perl module
		2381	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		2382	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		2383	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		2384	will be compiled. It is pretty complex because it provides its own header
		2385	file.
		2386
		2387	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		2388	that everybody includes and which overrides some configure choices:
		2389
		2390	#define EV_MINIMAL 1
		2391	#define EV_USE_POLL 0
		2392	#define EV_MULTIPLICITY 0
		2393	#define EV_PERIODIC_ENABLE 0
		2394	#define EV_STAT_ENABLE 0
		2395	#define EV_FORK_ENABLE 0
		2396	#define EV_CONFIG_H <config.h>
		2397	#define EV_MINPRI 0
		2398	#define EV_MAXPRI 0
		2399
		2400	#include "ev++.h"
		2401
		2402	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		2403
		2404	#include "ev_cpp.h"
		2405	#include "ev.c"
		2406
		2407
		2408	=head1 COMPLEXITIES
		2409
		2410	In this section the complexities of (many of) the algorithms used inside
		2411	libev will be explained. For complexity discussions about backends see the
		2412	documentation for C<ev_default_init>.
		2413
		2414	All of the following are about amortised time: If an array needs to be
		2415	extended, libev needs to realloc and move the whole array, but this
		2416	happens asymptotically never with higher number of elements, so O(1) might
		2417	mean it might do a lengthy realloc operation in rare cases, but on average
		2418	it is much faster and asymptotically approaches constant time.
		2419
		2420	=over 4
		2421
		2422	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		2423
		2424	This means that, when you have a watcher that triggers in one hour and
		2425	there are 100 watchers that would trigger before that then inserting will
		2426	have to skip those 100 watchers.
		2427
		2428	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
		2429
		2430	That means that for changing a timer costs less than removing/adding them
		2431	as only the relative motion in the event queue has to be paid for.
		2432
		2433	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
		2434
		2435	These just add the watcher into an array or at the head of a list.
		2436	=item Stopping check/prepare/idle watchers: O(1)
		2437
		2438	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		2439
		2440	These watchers are stored in lists then need to be walked to find the
		2441	correct watcher to remove. The lists are usually short (you don't usually
		2442	have many watchers waiting for the same fd or signal).
		2443
		2444	=item Finding the next timer per loop iteration: O(1)
		2445
		2446	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		2447
		2448	A change means an I/O watcher gets started or stopped, which requires
		2449	libev to recalculate its status (and possibly tell the kernel).
		2450
		2451	=item Activating one watcher: O(1)
		2452
		2453	=item Priority handling: O(number_of_priorities)
		2454
		2455	Priorities are implemented by allocating some space for each
		2456	priority. When doing priority-based operations, libev usually has to
		2457	linearly search all the priorities.
		2458
		2459	=back
		2460
781		2461
782	=head1 AUTHOR	2462	=head1 AUTHOR
783		2463
784	Marc Lehmann <libev@schmorp.de>.	2464	Marc Lehmann <libev@schmorp.de>.
785		2465

Diff Legend

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