ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.30 by root, Fri Nov 23 04:36:03 2007 UTC vs.
Revision 1.68 by root, Fri Dec 7 18:09:43 2007 UTC

…		…
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
		8
		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	}
8		50
9	=head1 DESCRIPTION	51	=head1 DESCRIPTION
10		52
11	Libev is an event loop: you register interest in certain events (such as a	53	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	54	file descriptor being readable or a timeout occuring), and it will manage
…		…
21	details of the event, and then hand it over to libev by I<starting> the	63	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	64	watcher.
23		65
24	=head1 FEATURES	66	=head1 FEATURES
25		67
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	68	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	69	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	70	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
29	events (related to SIGCHLD), and event watchers dealing with the event	71	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	72	with customised rescheduling (C<ev_periodic>), synchronous signals
		73	(C<ev_signal>), process status change events (C<ev_child>), and event
		74	watchers dealing with the event loop mechanism itself (C<ev_idle>,
		75	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as
		76	file watchers (C<ev_stat>) and even limited support for fork events
		77	(C<ev_fork>).
		78
		79	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	80	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	81	for example).
33		82
34	=head1 CONVENTIONS	83	=head1 CONVENTIONS
35		84
36	Libev is very configurable. In this manual the default configuration	85	Libev is very configurable. In this manual the default configuration will
37	will be described, which supports multiple event loops. For more info	86	be described, which supports multiple event loops. For more info about
38	about various configuration options please have a look at the file	87	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	88	this manual. If libev was configured without support for multiple event
40	support for multiple event loops, then all functions taking an initial	89	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 *>)	90	(which is always of type C<struct ev_loop *>) will not have this argument.
42	will not have this argument.
43		91
44	=head1 TIME REPRESENTATION	92	=head1 TIME REPRESENTATION
45		93
46	Libev represents time as a single floating point number, representing the	94	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	95	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	96	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	97	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the double type in C.	98	to the C<double> type in C, and when you need to do any calculations on
		99	it, you should treat it as such.
51		100
52	=head1 GLOBAL FUNCTIONS	101	=head1 GLOBAL FUNCTIONS
53		102
54	These functions can be called anytime, even before initialising the	103	These functions can be called anytime, even before initialising the
55	library in any way.	104	library in any way.
…		…
75	Usually, it's a good idea to terminate if the major versions mismatch,	124	Usually, it's a good idea to terminate if the major versions mismatch,
76	as this indicates an incompatible change. Minor versions are usually	125	as this indicates an incompatible change. Minor versions are usually
77	compatible to older versions, so a larger minor version alone is usually	126	compatible to older versions, so a larger minor version alone is usually
78	not a problem.	127	not a problem.
79		128
		129	Example: Make sure we haven't accidentally been linked against the wrong
		130	version.
		131
		132	assert (("libev version mismatch",
		133	ev_version_major () == EV_VERSION_MAJOR
		134	&& ev_version_minor () >= EV_VERSION_MINOR));
		135
		136	=item unsigned int ev_supported_backends ()
		137
		138	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		139	value) compiled into this binary of libev (independent of their
		140	availability on the system you are running on). See C<ev_default_loop> for
		141	a description of the set values.
		142
		143	Example: make sure we have the epoll method, because yeah this is cool and
		144	a must have and can we have a torrent of it please!!!11
		145
		146	assert (("sorry, no epoll, no sex",
		147	ev_supported_backends () & EVBACKEND_EPOLL));
		148
		149	=item unsigned int ev_recommended_backends ()
		150
		151	Return the set of all backends compiled into this binary of libev and also
		152	recommended for this platform. This set is often smaller than the one
		153	returned by C<ev_supported_backends>, as for example kqueue is broken on
		154	most BSDs and will not be autodetected unless you explicitly request it
		155	(assuming you know what you are doing). This is the set of backends that
		156	libev will probe for if you specify no backends explicitly.
		157
		158	=item unsigned int ev_embeddable_backends ()
		159
		160	Returns the set of backends that are embeddable in other event loops. This
		161	is the theoretical, all-platform, value. To find which backends
		162	might be supported on the current system, you would need to look at
		163	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
		164	recommended ones.
		165
		166	See the description of C<ev_embed> watchers for more info.
		167
80	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	168	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
81		169
82	Sets the allocation function to use (the prototype is similar to the	170	Sets the allocation function to use (the prototype is similar - the
83	realloc C function, the semantics are identical). It is used to allocate	171	semantics is identical - to the realloc C function). It is used to
84	and free memory (no surprises here). If it returns zero when memory	172	allocate and free memory (no surprises here). If it returns zero when
85	needs to be allocated, the library might abort or take some potentially	173	memory needs to be allocated, the library might abort or take some
86	destructive action. The default is your system realloc function.	174	potentially destructive action. The default is your system realloc
		175	function.
87		176
88	You could override this function in high-availability programs to, say,	177	You could override this function in high-availability programs to, say,
89	free some memory if it cannot allocate memory, to use a special allocator,	178	free some memory if it cannot allocate memory, to use a special allocator,
90	or even to sleep a while and retry until some memory is available.	179	or even to sleep a while and retry until some memory is available.
		180
		181	Example: Replace the libev allocator with one that waits a bit and then
		182	retries).
		183
		184	static void *
		185	persistent_realloc (void *ptr, size_t size)
		186	{
		187	for (;;)
		188	{
		189	void *newptr = realloc (ptr, size);
		190
		191	if (newptr)
		192	return newptr;
		193
		194	sleep (60);
		195	}
		196	}
		197
		198	...
		199	ev_set_allocator (persistent_realloc);
91		200
92	=item ev_set_syserr_cb (void (*cb)(const char *msg));	201	=item ev_set_syserr_cb (void (*cb)(const char *msg));
93		202
94	Set the callback function to call on a retryable syscall error (such	203	Set the callback function to call on a retryable syscall error (such
95	as failed select, poll, epoll_wait). The message is a printable string	204	as failed select, poll, epoll_wait). The message is a printable string
…		…
97	callback is set, then libev will expect it to remedy the sitution, no	206	callback is set, then libev will expect it to remedy the sitution, no
98	matter what, when it returns. That is, libev will generally retry the	207	matter what, when it returns. That is, libev will generally retry the
99	requested operation, or, if the condition doesn't go away, do bad stuff	208	requested operation, or, if the condition doesn't go away, do bad stuff
100	(such as abort).	209	(such as abort).
101		210
		211	Example: This is basically the same thing that libev does internally, too.
		212
		213	static void
		214	fatal_error (const char *msg)
		215	{
		216	perror (msg);
		217	abort ();
		218	}
		219
		220	...
		221	ev_set_syserr_cb (fatal_error);
		222
102	=back	223	=back
103		224
104	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	225	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
105		226
106	An event loop is described by a C<struct ev_loop *>. The library knows two	227	An event loop is described by a C<struct ev_loop *>. The library knows two
…		…
119	=item struct ev_loop *ev_default_loop (unsigned int flags)	240	=item struct ev_loop *ev_default_loop (unsigned int flags)
120		241
121	This will initialise the default event loop if it hasn't been initialised	242	This will initialise the default event loop if it hasn't been initialised
122	yet and return it. If the default loop could not be initialised, returns	243	yet and return it. If the default loop could not be initialised, returns
123	false. If it already was initialised it simply returns it (and ignores the	244	false. If it already was initialised it simply returns it (and ignores the
124	flags).	245	flags. If that is troubling you, check C<ev_backend ()> afterwards).
125		246
126	If you don't know what event loop to use, use the one returned from this	247	If you don't know what event loop to use, use the one returned from this
127	function.	248	function.
128		249
129	The flags argument can be used to specify special behaviour or specific	250	The flags argument can be used to specify special behaviour or specific
130	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	251	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
131		252
132	It supports the following flags:	253	The following flags are supported:
133		254
134	=over 4	255	=over 4
135		256
136	=item C<EVFLAG_AUTO>	257	=item C<EVFLAG_AUTO>
137		258
…		…
145	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	266	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
146	override the flags completely if it is found in the environment. This is	267	override the flags completely if it is found in the environment. This is
147	useful to try out specific backends to test their performance, or to work	268	useful to try out specific backends to test their performance, or to work
148	around bugs.	269	around bugs.
149		270
		271	=item C<EVFLAG_FORKCHECK>
		272
		273	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		274	a fork, you can also make libev check for a fork in each iteration by
		275	enabling this flag.
		276
		277	This works by calling C<getpid ()> on every iteration of the loop,
		278	and thus this might slow down your event loop if you do a lot of loop
		279	iterations and little real work, but is usually not noticeable (on my
		280	Linux system for example, C<getpid> is actually a simple 5-insn sequence
		281	without a syscall and thus I<very> fast, but my Linux system also has
		282	C<pthread_atfork> which is even faster).
		283
		284	The big advantage of this flag is that you can forget about fork (and
		285	forget about forgetting to tell libev about forking) when you use this
		286	flag.
		287
		288	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		289	environment variable.
		290
150	=item C<EVMETHOD_SELECT> (value 1, portable select backend)	291	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
151		292
152	This is your standard select(2) backend. Not I<completely> standard, as	293	This is your standard select(2) backend. Not I<completely> standard, as
153	libev tries to roll its own fd_set with no limits on the number of fds,	294	libev tries to roll its own fd_set with no limits on the number of fds,
154	but if that fails, expect a fairly low limit on the number of fds when	295	but if that fails, expect a fairly low limit on the number of fds when
155	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	296	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
156	the fastest backend for a low number of fds.	297	the fastest backend for a low number of fds.
157		298
158	=item C<EVMETHOD_POLL> (value 2, poll backend, available everywhere except on windows)	299	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
159		300
160	And this is your standard poll(2) backend. It's more complicated than	301	And this is your standard poll(2) backend. It's more complicated than
161	select, but handles sparse fds better and has no artificial limit on the	302	select, but handles sparse fds better and has no artificial limit on the
162	number of fds you can use (except it will slow down considerably with a	303	number of fds you can use (except it will slow down considerably with a
163	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	304	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
164		305
165	=item C<EVMETHOD_EPOLL> (value 4, Linux)	306	=item C<EVBACKEND_EPOLL> (value 4, Linux)
166		307
167	For few fds, this backend is a bit little slower than poll and select,	308	For few fds, this backend is a bit little slower than poll and select,
168	but it scales phenomenally better. While poll and select usually scale like	309	but it scales phenomenally better. While poll and select usually scale like
169	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	310	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
170	either O(1) or O(active_fds).	311	either O(1) or O(active_fds).
…		…
173	result in some caching, there is still a syscall per such incident	314	result in some caching, there is still a syscall per such incident
174	(because the fd could point to a different file description now), so its	315	(because the fd could point to a different file description now), so its
175	best to avoid that. Also, dup()ed file descriptors might not work very	316	best to avoid that. Also, dup()ed file descriptors might not work very
176	well if you register events for both fds.	317	well if you register events for both fds.
177		318
		319	Please note that epoll sometimes generates spurious notifications, so you
		320	need to use non-blocking I/O or other means to avoid blocking when no data
		321	(or space) is available.
		322
178	=item C<EVMETHOD_KQUEUE> (value 8, most BSD clones)	323	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
179		324
180	Kqueue deserves special mention, as at the time of this writing, it	325	Kqueue deserves special mention, as at the time of this writing, it
181	was broken on all BSDs except NetBSD (usually it doesn't work with	326	was broken on all BSDs except NetBSD (usually it doesn't work with
182	anything but sockets and pipes, except on Darwin, where of course its	327	anything but sockets and pipes, except on Darwin, where of course its
183	completely useless). For this reason its not being "autodetected" unless	328	completely useless). For this reason its not being "autodetected"
184	you explicitly specify the flags (i.e. you don't use EVFLAG_AUTO).	329	unless you explicitly specify it explicitly in the flags (i.e. using
		330	C<EVBACKEND_KQUEUE>).
185		331
186	It scales in the same way as the epoll backend, but the interface to the	332	It scales in the same way as the epoll backend, but the interface to the
187	kernel is more efficient (which says nothing about its actual speed, of	333	kernel is more efficient (which says nothing about its actual speed, of
188	course). While starting and stopping an I/O watcher does not cause an	334	course). While starting and stopping an I/O watcher does not cause an
189	extra syscall as with epoll, it still adds up to four event changes per	335	extra syscall as with epoll, it still adds up to four event changes per
190	incident, so its best to avoid that.	336	incident, so its best to avoid that.
191		337
192	=item C<EVMETHOD_DEVPOLL> (value 16, Solaris 8)	338	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
193		339
194	This is not implemented yet (and might never be).	340	This is not implemented yet (and might never be).
195		341
196	=item C<EVMETHOD_PORT> (value 32, Solaris 10)	342	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
197		343
198	This uses the Solaris 10 port mechanism. As with everything on Solaris,	344	This uses the Solaris 10 port mechanism. As with everything on Solaris,
199	it's really slow, but it still scales very well (O(active_fds)).	345	it's really slow, but it still scales very well (O(active_fds)).
200		346
		347	Please note that solaris ports can result in a lot of spurious
		348	notifications, so you need to use non-blocking I/O or other means to avoid
		349	blocking when no data (or space) is available.
		350
201	=item C<EVMETHOD_ALL>	351	=item C<EVBACKEND_ALL>
202		352
203	Try all backends (even potentially broken ones that wouldn't be tried	353	Try all backends (even potentially broken ones that wouldn't be tried
204	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	354	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
205	C<EVMETHOD_ALL & ~EVMETHOD_KQUEUE>.	355	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
206		356
207	=back	357	=back
208		358
209	If one or more of these are ored into the flags value, then only these	359	If one or more of these are ored into the flags value, then only these
210	backends will be tried (in the reverse order as given here). If none are	360	backends will be tried (in the reverse order as given here). If none are
211	specified, most compiled-in backend will be tried, usually in reverse	361	specified, most compiled-in backend will be tried, usually in reverse
212	order of their flag values :)	362	order of their flag values :)
213		363
		364	The most typical usage is like this:
		365
		366	if (!ev_default_loop (0))
		367	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		368
		369	Restrict libev to the select and poll backends, and do not allow
		370	environment settings to be taken into account:
		371
		372	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		373
		374	Use whatever libev has to offer, but make sure that kqueue is used if
		375	available (warning, breaks stuff, best use only with your own private
		376	event loop and only if you know the OS supports your types of fds):
		377
		378	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
		379
214	=item struct ev_loop *ev_loop_new (unsigned int flags)	380	=item struct ev_loop *ev_loop_new (unsigned int flags)
215		381
216	Similar to C<ev_default_loop>, but always creates a new event loop that is	382	Similar to C<ev_default_loop>, but always creates a new event loop that is
217	always distinct from the default loop. Unlike the default loop, it cannot	383	always distinct from the default loop. Unlike the default loop, it cannot
218	handle signal and child watchers, and attempts to do so will be greeted by	384	handle signal and child watchers, and attempts to do so will be greeted by
219	undefined behaviour (or a failed assertion if assertions are enabled).	385	undefined behaviour (or a failed assertion if assertions are enabled).
220		386
		387	Example: Try to create a event loop that uses epoll and nothing else.
		388
		389	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
		390	if (!epoller)
		391	fatal ("no epoll found here, maybe it hides under your chair");
		392
221	=item ev_default_destroy ()	393	=item ev_default_destroy ()
222		394
223	Destroys the default loop again (frees all memory and kernel state	395	Destroys the default loop again (frees all memory and kernel state
224	etc.). This stops all registered event watchers (by not touching them in	396	etc.). None of the active event watchers will be stopped in the normal
225	any way whatsoever, although you cannot rely on this :).	397	sense, so e.g. C<ev_is_active> might still return true. It is your
		398	responsibility to either stop all watchers cleanly yoursef I<before>
		399	calling this function, or cope with the fact afterwards (which is usually
		400	the easiest thing, youc na just ignore the watchers and/or C<free ()> them
		401	for example).
226		402
227	=item ev_loop_destroy (loop)	403	=item ev_loop_destroy (loop)
228		404
229	Like C<ev_default_destroy>, but destroys an event loop created by an	405	Like C<ev_default_destroy>, but destroys an event loop created by an
230	earlier call to C<ev_loop_new>.	406	earlier call to C<ev_loop_new>.
…		…
244	it just in case after a fork. To make this easy, the function will fit in	420	it just in case after a fork. To make this easy, the function will fit in
245	quite nicely into a call to C<pthread_atfork>:	421	quite nicely into a call to C<pthread_atfork>:
246		422
247	pthread_atfork (0, 0, ev_default_fork);	423	pthread_atfork (0, 0, ev_default_fork);
248		424
		425	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
		426	without calling this function, so if you force one of those backends you
		427	do not need to care.
		428
249	=item ev_loop_fork (loop)	429	=item ev_loop_fork (loop)
250		430
251	Like C<ev_default_fork>, but acts on an event loop created by	431	Like C<ev_default_fork>, but acts on an event loop created by
252	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	432	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
253	after fork, and how you do this is entirely your own problem.	433	after fork, and how you do this is entirely your own problem.
254		434
		435	=item unsigned int ev_loop_count (loop)
		436
		437	Returns the count of loop iterations for the loop, which is identical to
		438	the number of times libev did poll for new events. It starts at C<0> and
		439	happily wraps around with enough iterations.
		440
		441	This value can sometimes be useful as a generation counter of sorts (it
		442	"ticks" the number of loop iterations), as it roughly corresponds with
		443	C<ev_prepare> and C<ev_check> calls.
		444
255	=item unsigned int ev_method (loop)	445	=item unsigned int ev_backend (loop)
256		446
257	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	447	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
258	use.	448	use.
259		449
260	=item ev_tstamp ev_now (loop)	450	=item ev_tstamp ev_now (loop)
261		451
262	Returns the current "event loop time", which is the time the event loop	452	Returns the current "event loop time", which is the time the event loop
263	got events and started processing them. This timestamp does not change	453	received events and started processing them. This timestamp does not
264	as long as callbacks are being processed, and this is also the base time	454	change as long as callbacks are being processed, and this is also the base
265	used for relative timers. You can treat it as the timestamp of the event	455	time used for relative timers. You can treat it as the timestamp of the
266	occuring (or more correctly, the mainloop finding out about it).	456	event occuring (or more correctly, libev finding out about it).
267		457
268	=item ev_loop (loop, int flags)	458	=item ev_loop (loop, int flags)
269		459
270	Finally, this is it, the event handler. This function usually is called	460	Finally, this is it, the event handler. This function usually is called
271	after you initialised all your watchers and you want to start handling	461	after you initialised all your watchers and you want to start handling
272	events.	462	events.
273		463
274	If the flags argument is specified as 0, it will not return until either	464	If the flags argument is specified as C<0>, it will not return until
275	no event watchers are active anymore or C<ev_unloop> was called.	465	either no event watchers are active anymore or C<ev_unloop> was called.
		466
		467	Please note that an explicit C<ev_unloop> is usually better than
		468	relying on all watchers to be stopped when deciding when a program has
		469	finished (especially in interactive programs), but having a program that
		470	automatically loops as long as it has to and no longer by virtue of
		471	relying on its watchers stopping correctly is a thing of beauty.
276		472
277	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	473	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
278	those events and any outstanding ones, but will not block your process in	474	those events and any outstanding ones, but will not block your process in
279	case there are no events and will return after one iteration of the loop.	475	case there are no events and will return after one iteration of the loop.
280		476
281	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	477	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
282	neccessary) and will handle those and any outstanding ones. It will block	478	neccessary) and will handle those and any outstanding ones. It will block
283	your process until at least one new event arrives, and will return after	479	your process until at least one new event arrives, and will return after
284	one iteration of the loop.	480	one iteration of the loop. This is useful if you are waiting for some
		481	external event in conjunction with something not expressible using other
		482	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
		483	usually a better approach for this kind of thing.
285		484
286	This flags value could be used to implement alternative looping
287	constructs, but the C<prepare> and C<check> watchers provide a better and
288	more generic mechanism.
289
290	Here are the gory details of what ev_loop does:	485	Here are the gory details of what C<ev_loop> does:
291		486
292	1. If there are no active watchers (reference count is zero), return.	487	* If there are no active watchers (reference count is zero), return.
293	2. Queue and immediately call all prepare watchers.	488	- Queue prepare watchers and then call all outstanding watchers.
294	3. If we have been forked, recreate the kernel state.	489	- If we have been forked, recreate the kernel state.
295	4. Update the kernel state with all outstanding changes.	490	- Update the kernel state with all outstanding changes.
296	5. Update the "event loop time".	491	- Update the "event loop time".
297	6. Calculate for how long to block.	492	- Calculate for how long to block.
298	7. Block the process, waiting for events.	493	- Block the process, waiting for any events.
		494	- Queue all outstanding I/O (fd) events.
299	8. Update the "event loop time" and do time jump handling.	495	- Update the "event loop time" and do time jump handling.
300	9. Queue all outstanding timers.	496	- Queue all outstanding timers.
301	10. Queue all outstanding periodics.	497	- Queue all outstanding periodics.
302	11. If no events are pending now, queue all idle watchers.	498	- If no events are pending now, queue all idle watchers.
303	12. Queue all check watchers.	499	- Queue all check watchers.
304	13. Call all queued watchers in reverse order (i.e. check watchers first).	500	- Call all queued watchers in reverse order (i.e. check watchers first).
		501	Signals and child watchers are implemented as I/O watchers, and will
		502	be handled here by queueing them when their watcher gets executed.
305	14. If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	503	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
306	was used, return, otherwise continue with step #1.	504	were used, return, otherwise continue with step *.
		505
		506	Example: Queue some jobs and then loop until no events are outsanding
		507	anymore.
		508
		509	... queue jobs here, make sure they register event watchers as long
		510	... as they still have work to do (even an idle watcher will do..)
		511	ev_loop (my_loop, 0);
		512	... jobs done. yeah!
307		513
308	=item ev_unloop (loop, how)	514	=item ev_unloop (loop, how)
309		515
310	Can be used to make a call to C<ev_loop> return early (but only after it	516	Can be used to make a call to C<ev_loop> return early (but only after it
311	has processed all outstanding events). The C<how> argument must be either	517	has processed all outstanding events). The C<how> argument must be either
…		…
325	visible to the libev user and should not keep C<ev_loop> from exiting if	531	visible to the libev user and should not keep C<ev_loop> from exiting if
326	no event watchers registered by it are active. It is also an excellent	532	no event watchers registered by it are active. It is also an excellent
327	way to do this for generic recurring timers or from within third-party	533	way to do this for generic recurring timers or from within third-party
328	libraries. Just remember to I<unref after start> and I<ref before stop>.	534	libraries. Just remember to I<unref after start> and I<ref before stop>.
329		535
		536	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
		537	running when nothing else is active.
		538
		539	struct ev_signal exitsig;
		540	ev_signal_init (&exitsig, sig_cb, SIGINT);
		541	ev_signal_start (loop, &exitsig);
		542	evf_unref (loop);
		543
		544	Example: For some weird reason, unregister the above signal handler again.
		545
		546	ev_ref (loop);
		547	ev_signal_stop (loop, &exitsig);
		548
330	=back	549	=back
		550
331		551
332	=head1 ANATOMY OF A WATCHER	552	=head1 ANATOMY OF A WATCHER
333		553
334	A watcher is a structure that you create and register to record your	554	A watcher is a structure that you create and register to record your
335	interest in some event. For instance, if you want to wait for STDIN to	555	interest in some event. For instance, if you want to wait for STDIN to
…		…
368	*) >>), and you can stop watching for events at any time by calling the	588	*) >>), and you can stop watching for events at any time by calling the
369	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	589	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.
370		590
371	As long as your watcher is active (has been started but not stopped) you	591	As long as your watcher is active (has been started but not stopped) you
372	must not touch the values stored in it. Most specifically you must never	592	must not touch the values stored in it. Most specifically you must never
373	reinitialise it or call its set method.	593	reinitialise it or call its C<set> macro.
374
375	You can check whether an event is active by calling the C<ev_is_active
376	(watcher *)> macro. To see whether an event is outstanding (but the
377	callback for it has not been called yet) you can use the C<ev_is_pending
378	(watcher *)> macro.
379		594
380	Each and every callback receives the event loop pointer as first, the	595	Each and every callback receives the event loop pointer as first, the
381	registered watcher structure as second, and a bitset of received events as	596	registered watcher structure as second, and a bitset of received events as
382	third argument.	597	third argument.
383		598
…		…
407	The signal specified in the C<ev_signal> watcher has been received by a thread.	622	The signal specified in the C<ev_signal> watcher has been received by a thread.
408		623
409	=item C<EV_CHILD>	624	=item C<EV_CHILD>
410		625
411	The pid specified in the C<ev_child> watcher has received a status change.	626	The pid specified in the C<ev_child> watcher has received a status change.
		627
		628	=item C<EV_STAT>
		629
		630	The path specified in the C<ev_stat> watcher changed its attributes somehow.
412		631
413	=item C<EV_IDLE>	632	=item C<EV_IDLE>
414		633
415	The C<ev_idle> watcher has determined that you have nothing better to do.	634	The C<ev_idle> watcher has determined that you have nothing better to do.
416		635
…		…
424	received events. Callbacks of both watcher types can start and stop as	643	received events. Callbacks of both watcher types can start and stop as
425	many watchers as they want, and all of them will be taken into account	644	many watchers as they want, and all of them will be taken into account
426	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	645	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
427	C<ev_loop> from blocking).	646	C<ev_loop> from blocking).
428		647
		648	=item C<EV_EMBED>
		649
		650	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		651
		652	=item C<EV_FORK>
		653
		654	The event loop has been resumed in the child process after fork (see
		655	C<ev_fork>).
		656
429	=item C<EV_ERROR>	657	=item C<EV_ERROR>
430		658
431	An unspecified error has occured, the watcher has been stopped. This might	659	An unspecified error has occured, the watcher has been stopped. This might
432	happen because the watcher could not be properly started because libev	660	happen because the watcher could not be properly started because libev
433	ran out of memory, a file descriptor was found to be closed or any other	661	ran out of memory, a file descriptor was found to be closed or any other
…		…
439	your callbacks is well-written it can just attempt the operation and cope	667	your callbacks is well-written it can just attempt the operation and cope
440	with the error from read() or write(). This will not work in multithreaded	668	with the error from read() or write(). This will not work in multithreaded
441	programs, though, so beware.	669	programs, though, so beware.
442		670
443	=back	671	=back
		672
		673	=head2 GENERIC WATCHER FUNCTIONS
		674
		675	In the following description, C<TYPE> stands for the watcher type,
		676	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
		677
		678	=over 4
		679
		680	=item C<ev_init> (ev_TYPE *watcher, callback)
		681
		682	This macro initialises the generic portion of a watcher. The contents
		683	of the watcher object can be arbitrary (so C<malloc> will do). Only
		684	the generic parts of the watcher are initialised, you I<need> to call
		685	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		686	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		687	which rolls both calls into one.
		688
		689	You can reinitialise a watcher at any time as long as it has been stopped
		690	(or never started) and there are no pending events outstanding.
		691
		692	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,
		693	int revents)>.
		694
		695	=item C<ev_TYPE_set> (ev_TYPE *, [args])
		696
		697	This macro initialises the type-specific parts of a watcher. You need to
		698	call C<ev_init> at least once before you call this macro, but you can
		699	call C<ev_TYPE_set> any number of times. You must not, however, call this
		700	macro on a watcher that is active (it can be pending, however, which is a
		701	difference to the C<ev_init> macro).
		702
		703	Although some watcher types do not have type-specific arguments
		704	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		705
		706	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		707
		708	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		709	calls into a single call. This is the most convinient method to initialise
		710	a watcher. The same limitations apply, of course.
		711
		712	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
		713
		714	Starts (activates) the given watcher. Only active watchers will receive
		715	events. If the watcher is already active nothing will happen.
		716
		717	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
		718
		719	Stops the given watcher again (if active) and clears the pending
		720	status. It is possible that stopped watchers are pending (for example,
		721	non-repeating timers are being stopped when they become pending), but
		722	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If
		723	you want to free or reuse the memory used by the watcher it is therefore a
		724	good idea to always call its C<ev_TYPE_stop> function.
		725
		726	=item bool ev_is_active (ev_TYPE *watcher)
		727
		728	Returns a true value iff the watcher is active (i.e. it has been started
		729	and not yet been stopped). As long as a watcher is active you must not modify
		730	it.
		731
		732	=item bool ev_is_pending (ev_TYPE *watcher)
		733
		734	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		735	events but its callback has not yet been invoked). As long as a watcher
		736	is pending (but not active) you must not call an init function on it (but
		737	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to
		738	libev (e.g. you cnanot C<free ()> it).
		739
		740	=item callback ev_cb (ev_TYPE *watcher)
		741
		742	Returns the callback currently set on the watcher.
		743
		744	=item ev_cb_set (ev_TYPE *watcher, callback)
		745
		746	Change the callback. You can change the callback at virtually any time
		747	(modulo threads).
		748
		749	=item ev_set_priority (ev_TYPE *watcher, priority)
		750
		751	=item int ev_priority (ev_TYPE *watcher)
		752
		753	Set and query the priority of the watcher. The priority is a small
		754	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		755	(default: C<-2>). Pending watchers with higher priority will be invoked
		756	before watchers with lower priority, but priority will not keep watchers
		757	from being executed (except for C<ev_idle> watchers).
		758
		759	This means that priorities are I<only> used for ordering callback
		760	invocation after new events have been received. This is useful, for
		761	example, to reduce latency after idling, or more often, to bind two
		762	watchers on the same event and make sure one is called first.
		763
		764	If you need to suppress invocation when higher priority events are pending
		765	you need to look at C<ev_idle> watchers, which provide this functionality.
		766
		767	The default priority used by watchers when no priority has been set is
		768	always C<0>, which is supposed to not be too high and not be too low :).
		769
		770	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		771	fine, as long as you do not mind that the priority value you query might
		772	or might not have been adjusted to be within valid range.
		773
		774	=back
		775
444		776
445	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	777	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
446		778
447	Each watcher has, by default, a member C<void *data> that you can change	779	Each watcher has, by default, a member C<void *data> that you can change
448	and read at any time, libev will completely ignore it. This can be used	780	and read at any time, libev will completely ignore it. This can be used
…		…
466	{	798	{
467	struct my_io *w = (struct my_io *)w_;	799	struct my_io *w = (struct my_io *)w_;
468	...	800	...
469	}	801	}
470		802
471	More interesting and less C-conformant ways of catsing your callback type	803	More interesting and less C-conformant ways of casting your callback type
472	have been omitted....	804	instead have been omitted.
		805
		806	Another common scenario is having some data structure with multiple
		807	watchers:
		808
		809	struct my_biggy
		810	{
		811	int some_data;
		812	ev_timer t1;
		813	ev_timer t2;
		814	}
		815
		816	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		817	you need to use C<offsetof>:
		818
		819	#include <stddef.h>
		820
		821	static void
		822	t1_cb (EV_P_ struct ev_timer *w, int revents)
		823	{
		824	struct my_biggy big = (struct my_biggy *
		825	(((char *)w) - offsetof (struct my_biggy, t1));
		826	}
		827
		828	static void
		829	t2_cb (EV_P_ struct ev_timer *w, int revents)
		830	{
		831	struct my_biggy big = (struct my_biggy *
		832	(((char *)w) - offsetof (struct my_biggy, t2));
		833	}
473		834
474		835
475	=head1 WATCHER TYPES	836	=head1 WATCHER TYPES
476		837
477	This section describes each watcher in detail, but will not repeat	838	This section describes each watcher in detail, but will not repeat
478	information given in the last section.	839	information given in the last section. Any initialisation/set macros,
		840	functions and members specific to the watcher type are explained.
479		841
		842	Members are additionally marked with either I<[read-only]>, meaning that,
		843	while the watcher is active, you can look at the member and expect some
		844	sensible content, but you must not modify it (you can modify it while the
		845	watcher is stopped to your hearts content), or I<[read-write]>, which
		846	means you can expect it to have some sensible content while the watcher
		847	is active, but you can also modify it. Modifying it may not do something
		848	sensible or take immediate effect (or do anything at all), but libev will
		849	not crash or malfunction in any way.
		850
		851
480	=head2 C<ev_io> - is this file descriptor readable or writable	852	=head2 C<ev_io> - is this file descriptor readable or writable?
481		853
482	I/O watchers check whether a file descriptor is readable or writable	854	I/O watchers check whether a file descriptor is readable or writable
483	in each iteration of the event loop (This behaviour is called	855	in each iteration of the event loop, or, more precisely, when reading
484	level-triggering because you keep receiving events as long as the	856	would not block the process and writing would at least be able to write
485	condition persists. Remember you can stop the watcher if you don't want to	857	some data. This behaviour is called level-triggering because you keep
486	act on the event and neither want to receive future events).	858	receiving events as long as the condition persists. Remember you can stop
		859	the watcher if you don't want to act on the event and neither want to
		860	receive future events.
487		861
488	In general you can register as many read and/or write event watchers per	862	In general you can register as many read and/or write event watchers per
489	fd as you want (as long as you don't confuse yourself). Setting all file	863	fd as you want (as long as you don't confuse yourself). Setting all file
490	descriptors to non-blocking mode is also usually a good idea (but not	864	descriptors to non-blocking mode is also usually a good idea (but not
491	required if you know what you are doing).	865	required if you know what you are doing).
492		866
493	You have to be careful with dup'ed file descriptors, though. Some backends	867	You have to be careful with dup'ed file descriptors, though. Some backends
494	(the linux epoll backend is a notable example) cannot handle dup'ed file	868	(the linux epoll backend is a notable example) cannot handle dup'ed file
495	descriptors correctly if you register interest in two or more fds pointing	869	descriptors correctly if you register interest in two or more fds pointing
496	to the same underlying file/socket etc. description (that is, they share	870	to the same underlying file/socket/etc. description (that is, they share
497	the same underlying "file open").	871	the same underlying "file open").
498		872
499	If you must do this, then force the use of a known-to-be-good backend	873	If you must do this, then force the use of a known-to-be-good backend
500	(at the time of this writing, this includes only EVMETHOD_SELECT and	874	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
501	EVMETHOD_POLL).	875	C<EVBACKEND_POLL>).
		876
		877	Another thing you have to watch out for is that it is quite easy to
		878	receive "spurious" readyness notifications, that is your callback might
		879	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		880	because there is no data. Not only are some backends known to create a
		881	lot of those (for example solaris ports), it is very easy to get into
		882	this situation even with a relatively standard program structure. Thus
		883	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		884	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		885
		886	If you cannot run the fd in non-blocking mode (for example you should not
		887	play around with an Xlib connection), then you have to seperately re-test
		888	whether a file descriptor is really ready with a known-to-be good interface
		889	such as poll (fortunately in our Xlib example, Xlib already does this on
		890	its own, so its quite safe to use).
502		891
503	=over 4	892	=over 4
504		893
505	=item ev_io_init (ev_io *, callback, int fd, int events)	894	=item ev_io_init (ev_io *, callback, int fd, int events)
506		895
507	=item ev_io_set (ev_io *, int fd, int events)	896	=item ev_io_set (ev_io *, int fd, int events)
508		897
509	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	898	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
510	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	899	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or
511	EV_WRITE> to receive the given events.	900	C<EV_READ | EV_WRITE> to receive the given events.
		901
		902	=item int fd [read-only]
		903
		904	The file descriptor being watched.
		905
		906	=item int events [read-only]
		907
		908	The events being watched.
512		909
513	=back	910	=back
514		911
		912	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
		913	readable, but only once. Since it is likely line-buffered, you could
		914	attempt to read a whole line in the callback.
		915
		916	static void
		917	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		918	{
		919	ev_io_stop (loop, w);
		920	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		921	}
		922
		923	...
		924	struct ev_loop *loop = ev_default_init (0);
		925	struct ev_io stdin_readable;
		926	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		927	ev_io_start (loop, &stdin_readable);
		928	ev_loop (loop, 0);
		929
		930
515	=head2 C<ev_timer> - relative and optionally recurring timeouts	931	=head2 C<ev_timer> - relative and optionally repeating timeouts
516		932
517	Timer watchers are simple relative timers that generate an event after a	933	Timer watchers are simple relative timers that generate an event after a
518	given time, and optionally repeating in regular intervals after that.	934	given time, and optionally repeating in regular intervals after that.
519		935
520	The timers are based on real time, that is, if you register an event that	936	The timers are based on real time, that is, if you register an event that
…		…
555	=item ev_timer_again (loop)	971	=item ev_timer_again (loop)
556		972
557	This will act as if the timer timed out and restart it again if it is	973	This will act as if the timer timed out and restart it again if it is
558	repeating. The exact semantics are:	974	repeating. The exact semantics are:
559		975
		976	If the timer is pending, its pending status is cleared.
		977
560	If the timer is started but nonrepeating, stop it.	978	If the timer is started but nonrepeating, stop it (as if it timed out).
561		979
562	If the timer is repeating, either start it if necessary (with the repeat	980	If the timer is repeating, either start it if necessary (with the
563	value), or reset the running timer to the repeat value.	981	C<repeat> value), or reset the running timer to the C<repeat> value.
564		982
565	This sounds a bit complicated, but here is a useful and typical	983	This sounds a bit complicated, but here is a useful and typical
566	example: Imagine you have a tcp connection and you want a so-called idle	984	example: Imagine you have a tcp connection and you want a so-called idle
567	timeout, that is, you want to be called when there have been, say, 60	985	timeout, that is, you want to be called when there have been, say, 60
568	seconds of inactivity on the socket. The easiest way to do this is to	986	seconds of inactivity on the socket. The easiest way to do this is to
569	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	987	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
570	time you successfully read or write some data. If you go into an idle	988	C<ev_timer_again> each time you successfully read or write some data. If
571	state where you do not expect data to travel on the socket, you can stop	989	you go into an idle state where you do not expect data to travel on the
572	the timer, and again will automatically restart it if need be.	990	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
		991	automatically restart it if need be.
		992
		993	That means you can ignore the C<after> value and C<ev_timer_start>
		994	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
		995
		996	ev_timer_init (timer, callback, 0., 5.);
		997	ev_timer_again (loop, timer);
		998	...
		999	timer->again = 17.;
		1000	ev_timer_again (loop, timer);
		1001	...
		1002	timer->again = 10.;
		1003	ev_timer_again (loop, timer);
		1004
		1005	This is more slightly efficient then stopping/starting the timer each time
		1006	you want to modify its timeout value.
		1007
		1008	=item ev_tstamp repeat [read-write]
		1009
		1010	The current C<repeat> value. Will be used each time the watcher times out
		1011	or C<ev_timer_again> is called and determines the next timeout (if any),
		1012	which is also when any modifications are taken into account.
573		1013
574	=back	1014	=back
575		1015
		1016	Example: Create a timer that fires after 60 seconds.
		1017
		1018	static void
		1019	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1020	{
		1021	.. one minute over, w is actually stopped right here
		1022	}
		1023
		1024	struct ev_timer mytimer;
		1025	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		1026	ev_timer_start (loop, &mytimer);
		1027
		1028	Example: Create a timeout timer that times out after 10 seconds of
		1029	inactivity.
		1030
		1031	static void
		1032	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		1033	{
		1034	.. ten seconds without any activity
		1035	}
		1036
		1037	struct ev_timer mytimer;
		1038	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		1039	ev_timer_again (&mytimer); /* start timer */
		1040	ev_loop (loop, 0);
		1041
		1042	// and in some piece of code that gets executed on any "activity":
		1043	// reset the timeout to start ticking again at 10 seconds
		1044	ev_timer_again (&mytimer);
		1045
		1046
576	=head2 C<ev_periodic> - to cron or not to cron	1047	=head2 C<ev_periodic> - to cron or not to cron?
577		1048
578	Periodic watchers are also timers of a kind, but they are very versatile	1049	Periodic watchers are also timers of a kind, but they are very versatile
579	(and unfortunately a bit complex).	1050	(and unfortunately a bit complex).
580		1051
581	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1052	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
582	but on wallclock time (absolute time). You can tell a periodic watcher	1053	but on wallclock time (absolute time). You can tell a periodic watcher
583	to trigger "at" some specific point in time. For example, if you tell a	1054	to trigger "at" some specific point in time. For example, if you tell a
584	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	1055	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
585	+ 10.>) and then reset your system clock to the last year, then it will	1056	+ 10.>) and then reset your system clock to the last year, then it will
586	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1057	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
587	roughly 10 seconds later and of course not if you reset your system time	1058	roughly 10 seconds later and of course not if you reset your system time
588	again).	1059	again).
589		1060
…		…
673	Simply stops and restarts the periodic watcher again. This is only useful	1144	Simply stops and restarts the periodic watcher again. This is only useful
674	when you changed some parameters or the reschedule callback would return	1145	when you changed some parameters or the reschedule callback would return
675	a different time than the last time it was called (e.g. in a crond like	1146	a different time than the last time it was called (e.g. in a crond like
676	program when the crontabs have changed).	1147	program when the crontabs have changed).
677		1148
		1149	=item ev_tstamp interval [read-write]
		1150
		1151	The current interval value. Can be modified any time, but changes only
		1152	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1153	called.
		1154
		1155	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]
		1156
		1157	The current reschedule callback, or C<0>, if this functionality is
		1158	switched off. Can be changed any time, but changes only take effect when
		1159	the periodic timer fires or C<ev_periodic_again> is being called.
		1160
678	=back	1161	=back
679		1162
		1163	Example: Call a callback every hour, or, more precisely, whenever the
		1164	system clock is divisible by 3600. The callback invocation times have
		1165	potentially a lot of jittering, but good long-term stability.
		1166
		1167	static void
		1168	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		1169	{
		1170	... its now a full hour (UTC, or TAI or whatever your clock follows)
		1171	}
		1172
		1173	struct ev_periodic hourly_tick;
		1174	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		1175	ev_periodic_start (loop, &hourly_tick);
		1176
		1177	Example: The same as above, but use a reschedule callback to do it:
		1178
		1179	#include <math.h>
		1180
		1181	static ev_tstamp
		1182	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		1183	{
		1184	return fmod (now, 3600.) + 3600.;
		1185	}
		1186
		1187	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		1188
		1189	Example: Call a callback every hour, starting now:
		1190
		1191	struct ev_periodic hourly_tick;
		1192	ev_periodic_init (&hourly_tick, clock_cb,
		1193	fmod (ev_now (loop), 3600.), 3600., 0);
		1194	ev_periodic_start (loop, &hourly_tick);
		1195
		1196
680	=head2 C<ev_signal> - signal me when a signal gets signalled	1197	=head2 C<ev_signal> - signal me when a signal gets signalled!
681		1198
682	Signal watchers will trigger an event when the process receives a specific	1199	Signal watchers will trigger an event when the process receives a specific
683	signal one or more times. Even though signals are very asynchronous, libev	1200	signal one or more times. Even though signals are very asynchronous, libev
684	will try it's best to deliver signals synchronously, i.e. as part of the	1201	will try it's best to deliver signals synchronously, i.e. as part of the
685	normal event processing, like any other event.	1202	normal event processing, like any other event.
…		…
698	=item ev_signal_set (ev_signal *, int signum)	1215	=item ev_signal_set (ev_signal *, int signum)
699		1216
700	Configures the watcher to trigger on the given signal number (usually one	1217	Configures the watcher to trigger on the given signal number (usually one
701	of the C<SIGxxx> constants).	1218	of the C<SIGxxx> constants).
702		1219
		1220	=item int signum [read-only]
		1221
		1222	The signal the watcher watches out for.
		1223
703	=back	1224	=back
704		1225
		1226
705	=head2 C<ev_child> - wait for pid status changes	1227	=head2 C<ev_child> - watch out for process status changes
706		1228
707	Child watchers trigger when your process receives a SIGCHLD in response to	1229	Child watchers trigger when your process receives a SIGCHLD in response to
708	some child status changes (most typically when a child of yours dies).	1230	some child status changes (most typically when a child of yours dies).
709		1231
710	=over 4	1232	=over 4
…		…
718	at the C<rstatus> member of the C<ev_child> watcher structure to see	1240	at the C<rstatus> member of the C<ev_child> watcher structure to see
719	the status word (use the macros from C<sys/wait.h> and see your systems	1241	the status word (use the macros from C<sys/wait.h> and see your systems
720	C<waitpid> documentation). The C<rpid> member contains the pid of the	1242	C<waitpid> documentation). The C<rpid> member contains the pid of the
721	process causing the status change.	1243	process causing the status change.
722		1244
		1245	=item int pid [read-only]
		1246
		1247	The process id this watcher watches out for, or C<0>, meaning any process id.
		1248
		1249	=item int rpid [read-write]
		1250
		1251	The process id that detected a status change.
		1252
		1253	=item int rstatus [read-write]
		1254
		1255	The process exit/trace status caused by C<rpid> (see your systems
		1256	C<waitpid> and C<sys/wait.h> documentation for details).
		1257
723	=back	1258	=back
724		1259
		1260	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1261
		1262	static void
		1263	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1264	{
		1265	ev_unloop (loop, EVUNLOOP_ALL);
		1266	}
		1267
		1268	struct ev_signal signal_watcher;
		1269	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1270	ev_signal_start (loop, &sigint_cb);
		1271
		1272
		1273	=head2 C<ev_stat> - did the file attributes just change?
		1274
		1275	This watches a filesystem path for attribute changes. That is, it calls
		1276	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1277	compared to the last time, invoking the callback if it did.
		1278
		1279	The path does not need to exist: changing from "path exists" to "path does
		1280	not exist" is a status change like any other. The condition "path does
		1281	not exist" is signified by the C<st_nlink> field being zero (which is
		1282	otherwise always forced to be at least one) and all the other fields of
		1283	the stat buffer having unspecified contents.
		1284
		1285	The path I<should> be absolute and I<must not> end in a slash. If it is
		1286	relative and your working directory changes, the behaviour is undefined.
		1287
		1288	Since there is no standard to do this, the portable implementation simply
		1289	calls C<stat (2)> regularly on the path to see if it changed somehow. You
		1290	can specify a recommended polling interval for this case. If you specify
		1291	a polling interval of C<0> (highly recommended!) then a I<suitable,
		1292	unspecified default> value will be used (which you can expect to be around
		1293	five seconds, although this might change dynamically). Libev will also
		1294	impose a minimum interval which is currently around C<0.1>, but thats
		1295	usually overkill.
		1296
		1297	This watcher type is not meant for massive numbers of stat watchers,
		1298	as even with OS-supported change notifications, this can be
		1299	resource-intensive.
		1300
		1301	At the time of this writing, only the Linux inotify interface is
		1302	implemented (implementing kqueue support is left as an exercise for the
		1303	reader). Inotify will be used to give hints only and should not change the
		1304	semantics of C<ev_stat> watchers, which means that libev sometimes needs
		1305	to fall back to regular polling again even with inotify, but changes are
		1306	usually detected immediately, and if the file exists there will be no
		1307	polling.
		1308
		1309	=over 4
		1310
		1311	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		1312
		1313	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		1314
		1315	Configures the watcher to wait for status changes of the given
		1316	C<path>. The C<interval> is a hint on how quickly a change is expected to
		1317	be detected and should normally be specified as C<0> to let libev choose
		1318	a suitable value. The memory pointed to by C<path> must point to the same
		1319	path for as long as the watcher is active.
		1320
		1321	The callback will be receive C<EV_STAT> when a change was detected,
		1322	relative to the attributes at the time the watcher was started (or the
		1323	last change was detected).
		1324
		1325	=item ev_stat_stat (ev_stat *)
		1326
		1327	Updates the stat buffer immediately with new values. If you change the
		1328	watched path in your callback, you could call this fucntion to avoid
		1329	detecting this change (while introducing a race condition). Can also be
		1330	useful simply to find out the new values.
		1331
		1332	=item ev_statdata attr [read-only]
		1333
		1334	The most-recently detected attributes of the file. Although the type is of
		1335	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		1336	suitable for your system. If the C<st_nlink> member is C<0>, then there
		1337	was some error while C<stat>ing the file.
		1338
		1339	=item ev_statdata prev [read-only]
		1340
		1341	The previous attributes of the file. The callback gets invoked whenever
		1342	C<prev> != C<attr>.
		1343
		1344	=item ev_tstamp interval [read-only]
		1345
		1346	The specified interval.
		1347
		1348	=item const char *path [read-only]
		1349
		1350	The filesystem path that is being watched.
		1351
		1352	=back
		1353
		1354	Example: Watch C</etc/passwd> for attribute changes.
		1355
		1356	static void
		1357	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		1358	{
		1359	/* /etc/passwd changed in some way */
		1360	if (w->attr.st_nlink)
		1361	{
		1362	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1363	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1364	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1365	}
		1366	else
		1367	/* you shalt not abuse printf for puts */
		1368	puts ("wow, /etc/passwd is not there, expect problems. "
		1369	"if this is windows, they already arrived\n");
		1370	}
		1371
		1372	...
		1373	ev_stat passwd;
		1374
		1375	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
		1376	ev_stat_start (loop, &passwd);
		1377
		1378
725	=head2 C<ev_idle> - when you've got nothing better to do	1379	=head2 C<ev_idle> - when you've got nothing better to do...
726		1380
727	Idle watchers trigger events when there are no other events are pending	1381	Idle watchers trigger events when no other events of the same or higher
728	(prepare, check and other idle watchers do not count). That is, as long	1382	priority are pending (prepare, check and other idle watchers do not
729	as your process is busy handling sockets or timeouts (or even signals,	1383	count).
730	imagine) it will not be triggered. But when your process is idle all idle	1384
731	watchers are being called again and again, once per event loop iteration -	1385	That is, as long as your process is busy handling sockets or timeouts
		1386	(or even signals, imagine) of the same or higher priority it will not be
		1387	triggered. But when your process is idle (or only lower-priority watchers
		1388	are pending), the idle watchers are being called once per event loop
732	until stopped, that is, or your process receives more events and becomes	1389	iteration - until stopped, that is, or your process receives more events
733	busy.	1390	and becomes busy again with higher priority stuff.
734		1391
735	The most noteworthy effect is that as long as any idle watchers are	1392	The most noteworthy effect is that as long as any idle watchers are
736	active, the process will not block when waiting for new events.	1393	active, the process will not block when waiting for new events.
737		1394
738	Apart from keeping your process non-blocking (which is a useful	1395	Apart from keeping your process non-blocking (which is a useful
…		…
748	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1405	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
749	believe me.	1406	believe me.
750		1407
751	=back	1408	=back
752		1409
		1410	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
		1411	callback, free it. Also, use no error checking, as usual.
		1412
		1413	static void
		1414	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
		1415	{
		1416	free (w);
		1417	// now do something you wanted to do when the program has
		1418	// no longer asnything immediate to do.
		1419	}
		1420
		1421	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1422	ev_idle_init (idle_watcher, idle_cb);
		1423	ev_idle_start (loop, idle_cb);
		1424
		1425
753	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1426	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
754		1427
755	Prepare and check watchers are usually (but not always) used in tandem:	1428	Prepare and check watchers are usually (but not always) used in tandem:
756	prepare watchers get invoked before the process blocks and check watchers	1429	prepare watchers get invoked before the process blocks and check watchers
757	afterwards.	1430	afterwards.
758		1431
		1432	You I<must not> call C<ev_loop> or similar functions that enter
		1433	the current event loop from either C<ev_prepare> or C<ev_check>
		1434	watchers. Other loops than the current one are fine, however. The
		1435	rationale behind this is that you do not need to check for recursion in
		1436	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		1437	C<ev_check> so if you have one watcher of each kind they will always be
		1438	called in pairs bracketing the blocking call.
		1439
759	Their main purpose is to integrate other event mechanisms into libev. This	1440	Their main purpose is to integrate other event mechanisms into libev and
760	could be used, for example, to track variable changes, implement your own	1441	their use is somewhat advanced. This could be used, for example, to track
761	watchers, integrate net-snmp or a coroutine library and lots more.	1442	variable changes, implement your own watchers, integrate net-snmp or a
		1443	coroutine library and lots more. They are also occasionally useful if
		1444	you cache some data and want to flush it before blocking (for example,
		1445	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		1446	watcher).
762		1447
763	This is done by examining in each prepare call which file descriptors need	1448	This is done by examining in each prepare call which file descriptors need
764	to be watched by the other library, registering C<ev_io> watchers for	1449	to be watched by the other library, registering C<ev_io> watchers for
765	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1450	them and starting an C<ev_timer> watcher for any timeouts (many libraries
766	provide just this functionality). Then, in the check watcher you check for	1451	provide just this functionality). Then, in the check watcher you check for
…		…
788	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1473	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
789	macros, but using them is utterly, utterly and completely pointless.	1474	macros, but using them is utterly, utterly and completely pointless.
790		1475
791	=back	1476	=back
792		1477
		1478	Example: To include a library such as adns, you would add IO watchers
		1479	and a timeout watcher in a prepare handler, as required by libadns, and
		1480	in a check watcher, destroy them and call into libadns. What follows is
		1481	pseudo-code only of course:
		1482
		1483	static ev_io iow [nfd];
		1484	static ev_timer tw;
		1485
		1486	static void
		1487	io_cb (ev_loop *loop, ev_io *w, int revents)
		1488	{
		1489	// set the relevant poll flags
		1490	// could also call adns_processreadable etc. here
		1491	struct pollfd *fd = (struct pollfd *)w->data;
		1492	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1493	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1494	}
		1495
		1496	// create io watchers for each fd and a timer before blocking
		1497	static void
		1498	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
		1499	{
		1500	int timeout = 3600000;
		1501	struct pollfd fds [nfd];
		1502	// actual code will need to loop here and realloc etc.
		1503	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1504
		1505	/* the callback is illegal, but won't be called as we stop during check */
		1506	ev_timer_init (&tw, 0, timeout * 1e-3);
		1507	ev_timer_start (loop, &tw);
		1508
		1509	// create on ev_io per pollfd
		1510	for (int i = 0; i < nfd; ++i)
		1511	{
		1512	ev_io_init (iow + i, io_cb, fds [i].fd,
		1513	((fds [i].events & POLLIN ? EV_READ : 0)
		1514	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1515
		1516	fds [i].revents = 0;
		1517	iow [i].data = fds + i;
		1518	ev_io_start (loop, iow + i);
		1519	}
		1520	}
		1521
		1522	// stop all watchers after blocking
		1523	static void
		1524	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
		1525	{
		1526	ev_timer_stop (loop, &tw);
		1527
		1528	for (int i = 0; i < nfd; ++i)
		1529	ev_io_stop (loop, iow + i);
		1530
		1531	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1532	}
		1533
		1534
		1535	=head2 C<ev_embed> - when one backend isn't enough...
		1536
		1537	This is a rather advanced watcher type that lets you embed one event loop
		1538	into another (currently only C<ev_io> events are supported in the embedded
		1539	loop, other types of watchers might be handled in a delayed or incorrect
		1540	fashion and must not be used).
		1541
		1542	There are primarily two reasons you would want that: work around bugs and
		1543	prioritise I/O.
		1544
		1545	As an example for a bug workaround, the kqueue backend might only support
		1546	sockets on some platform, so it is unusable as generic backend, but you
		1547	still want to make use of it because you have many sockets and it scales
		1548	so nicely. In this case, you would create a kqueue-based loop and embed it
		1549	into your default loop (which might use e.g. poll). Overall operation will
		1550	be a bit slower because first libev has to poll and then call kevent, but
		1551	at least you can use both at what they are best.
		1552
		1553	As for prioritising I/O: rarely you have the case where some fds have
		1554	to be watched and handled very quickly (with low latency), and even
		1555	priorities and idle watchers might have too much overhead. In this case
		1556	you would put all the high priority stuff in one loop and all the rest in
		1557	a second one, and embed the second one in the first.
		1558
		1559	As long as the watcher is active, the callback will be invoked every time
		1560	there might be events pending in the embedded loop. The callback must then
		1561	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
		1562	their callbacks (you could also start an idle watcher to give the embedded
		1563	loop strictly lower priority for example). You can also set the callback
		1564	to C<0>, in which case the embed watcher will automatically execute the
		1565	embedded loop sweep.
		1566
		1567	As long as the watcher is started it will automatically handle events. The
		1568	callback will be invoked whenever some events have been handled. You can
		1569	set the callback to C<0> to avoid having to specify one if you are not
		1570	interested in that.
		1571
		1572	Also, there have not currently been made special provisions for forking:
		1573	when you fork, you not only have to call C<ev_loop_fork> on both loops,
		1574	but you will also have to stop and restart any C<ev_embed> watchers
		1575	yourself.
		1576
		1577	Unfortunately, not all backends are embeddable, only the ones returned by
		1578	C<ev_embeddable_backends> are, which, unfortunately, does not include any
		1579	portable one.
		1580
		1581	So when you want to use this feature you will always have to be prepared
		1582	that you cannot get an embeddable loop. The recommended way to get around
		1583	this is to have a separate variables for your embeddable loop, try to
		1584	create it, and if that fails, use the normal loop for everything:
		1585
		1586	struct ev_loop *loop_hi = ev_default_init (0);
		1587	struct ev_loop *loop_lo = 0;
		1588	struct ev_embed embed;
		1589
		1590	// see if there is a chance of getting one that works
		1591	// (remember that a flags value of 0 means autodetection)
		1592	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		1593	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		1594	: 0;
		1595
		1596	// if we got one, then embed it, otherwise default to loop_hi
		1597	if (loop_lo)
		1598	{
		1599	ev_embed_init (&embed, 0, loop_lo);
		1600	ev_embed_start (loop_hi, &embed);
		1601	}
		1602	else
		1603	loop_lo = loop_hi;
		1604
		1605	=over 4
		1606
		1607	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		1608
		1609	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		1610
		1611	Configures the watcher to embed the given loop, which must be
		1612	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		1613	invoked automatically, otherwise it is the responsibility of the callback
		1614	to invoke it (it will continue to be called until the sweep has been done,
		1615	if you do not want thta, you need to temporarily stop the embed watcher).
		1616
		1617	=item ev_embed_sweep (loop, ev_embed *)
		1618
		1619	Make a single, non-blocking sweep over the embedded loop. This works
		1620	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		1621	apropriate way for embedded loops.
		1622
		1623	=item struct ev_loop *loop [read-only]
		1624
		1625	The embedded event loop.
		1626
		1627	=back
		1628
		1629
		1630	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		1631
		1632	Fork watchers are called when a C<fork ()> was detected (usually because
		1633	whoever is a good citizen cared to tell libev about it by calling
		1634	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		1635	event loop blocks next and before C<ev_check> watchers are being called,
		1636	and only in the child after the fork. If whoever good citizen calling
		1637	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		1638	handlers will be invoked, too, of course.
		1639
		1640	=over 4
		1641
		1642	=item ev_fork_init (ev_signal *, callback)
		1643
		1644	Initialises and configures the fork watcher - it has no parameters of any
		1645	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		1646	believe me.
		1647
		1648	=back
		1649
		1650
793	=head1 OTHER FUNCTIONS	1651	=head1 OTHER FUNCTIONS
794		1652
795	There are some other functions of possible interest. Described. Here. Now.	1653	There are some other functions of possible interest. Described. Here. Now.
796		1654
797	=over 4	1655	=over 4
…		…
826	/* stdin might have data for us, joy! */;	1684	/* stdin might have data for us, joy! */;
827	}	1685	}
828		1686
829	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	1687	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
830		1688
831	=item ev_feed_event (loop, watcher, int events)	1689	=item ev_feed_event (ev_loop *, watcher *, int revents)
832		1690
833	Feeds the given event set into the event loop, as if the specified event	1691	Feeds the given event set into the event loop, as if the specified event
834	had happened for the specified watcher (which must be a pointer to an	1692	had happened for the specified watcher (which must be a pointer to an
835	initialised but not necessarily started event watcher).	1693	initialised but not necessarily started event watcher).
836		1694
837	=item ev_feed_fd_event (loop, int fd, int revents)	1695	=item ev_feed_fd_event (ev_loop *, int fd, int revents)
838		1696
839	Feed an event on the given fd, as if a file descriptor backend detected	1697	Feed an event on the given fd, as if a file descriptor backend detected
840	the given events it.	1698	the given events it.
841		1699
842	=item ev_feed_signal_event (loop, int signum)	1700	=item ev_feed_signal_event (ev_loop *loop, int signum)
843		1701
844	Feed an event as if the given signal occured (loop must be the default loop!).	1702	Feed an event as if the given signal occured (C<loop> must be the default
		1703	loop!).
845		1704
846	=back	1705	=back
		1706
847		1707
848	=head1 LIBEVENT EMULATION	1708	=head1 LIBEVENT EMULATION
849		1709
850	Libev offers a compatibility emulation layer for libevent. It cannot	1710	Libev offers a compatibility emulation layer for libevent. It cannot
851	emulate the internals of libevent, so here are some usage hints:	1711	emulate the internals of libevent, so here are some usage hints:
…		…
872		1732
873	=back	1733	=back
874		1734
875	=head1 C++ SUPPORT	1735	=head1 C++ SUPPORT
876		1736
877	TBD.	1737	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		1738	you to use some convinience methods to start/stop watchers and also change
		1739	the callback model to a model using method callbacks on objects.
		1740
		1741	To use it,
		1742
		1743	#include <ev++.h>
		1744
		1745	(it is not installed by default). This automatically includes F<ev.h>
		1746	and puts all of its definitions (many of them macros) into the global
		1747	namespace. All C++ specific things are put into the C<ev> namespace.
		1748
		1749	It should support all the same embedding options as F<ev.h>, most notably
		1750	C<EV_MULTIPLICITY>.
		1751
		1752	Here is a list of things available in the C<ev> namespace:
		1753
		1754	=over 4
		1755
		1756	=item C<ev::READ>, C<ev::WRITE> etc.
		1757
		1758	These are just enum values with the same values as the C<EV_READ> etc.
		1759	macros from F<ev.h>.
		1760
		1761	=item C<ev::tstamp>, C<ev::now>
		1762
		1763	Aliases to the same types/functions as with the C<ev_> prefix.
		1764
		1765	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		1766
		1767	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		1768	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		1769	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		1770	defines by many implementations.
		1771
		1772	All of those classes have these methods:
		1773
		1774	=over 4
		1775
		1776	=item ev::TYPE::TYPE (object *, object::method *)
		1777
		1778	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)
		1779
		1780	=item ev::TYPE::~TYPE
		1781
		1782	The constructor takes a pointer to an object and a method pointer to
		1783	the event handler callback to call in this class. The constructor calls
		1784	C<ev_init> for you, which means you have to call the C<set> method
		1785	before starting it. If you do not specify a loop then the constructor
		1786	automatically associates the default loop with this watcher.
		1787
		1788	The destructor automatically stops the watcher if it is active.
		1789
		1790	=item w->set (struct ev_loop *)
		1791
		1792	Associates a different C<struct ev_loop> with this watcher. You can only
		1793	do this when the watcher is inactive (and not pending either).
		1794
		1795	=item w->set ([args])
		1796
		1797	Basically the same as C<ev_TYPE_set>, with the same args. Must be
		1798	called at least once. Unlike the C counterpart, an active watcher gets
		1799	automatically stopped and restarted.
		1800
		1801	=item w->start ()
		1802
		1803	Starts the watcher. Note that there is no C<loop> argument as the
		1804	constructor already takes the loop.
		1805
		1806	=item w->stop ()
		1807
		1808	Stops the watcher if it is active. Again, no C<loop> argument.
		1809
		1810	=item w->again () C<ev::timer>, C<ev::periodic> only
		1811
		1812	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		1813	C<ev_TYPE_again> function.
		1814
		1815	=item w->sweep () C<ev::embed> only
		1816
		1817	Invokes C<ev_embed_sweep>.
		1818
		1819	=item w->update () C<ev::stat> only
		1820
		1821	Invokes C<ev_stat_stat>.
		1822
		1823	=back
		1824
		1825	=back
		1826
		1827	Example: Define a class with an IO and idle watcher, start one of them in
		1828	the constructor.
		1829
		1830	class myclass
		1831	{
		1832	ev_io io; void io_cb (ev::io &w, int revents);
		1833	ev_idle idle void idle_cb (ev::idle &w, int revents);
		1834
		1835	myclass ();
		1836	}
		1837
		1838	myclass::myclass (int fd)
		1839	: io (this, &myclass::io_cb),
		1840	idle (this, &myclass::idle_cb)
		1841	{
		1842	io.start (fd, ev::READ);
		1843	}
		1844
		1845
		1846	=head1 MACRO MAGIC
		1847
		1848	Libev can be compiled with a variety of options, the most fundemantal is
		1849	C<EV_MULTIPLICITY>. This option determines whether (most) functions and
		1850	callbacks have an initial C<struct ev_loop *> argument.
		1851
		1852	To make it easier to write programs that cope with either variant, the
		1853	following macros are defined:
		1854
		1855	=over 4
		1856
		1857	=item C<EV_A>, C<EV_A_>
		1858
		1859	This provides the loop I<argument> for functions, if one is required ("ev
		1860	loop argument"). The C<EV_A> form is used when this is the sole argument,
		1861	C<EV_A_> is used when other arguments are following. Example:
		1862
		1863	ev_unref (EV_A);
		1864	ev_timer_add (EV_A_ watcher);
		1865	ev_loop (EV_A_ 0);
		1866
		1867	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		1868	which is often provided by the following macro.
		1869
		1870	=item C<EV_P>, C<EV_P_>
		1871
		1872	This provides the loop I<parameter> for functions, if one is required ("ev
		1873	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		1874	C<EV_P_> is used when other parameters are following. Example:
		1875
		1876	// this is how ev_unref is being declared
		1877	static void ev_unref (EV_P);
		1878
		1879	// this is how you can declare your typical callback
		1880	static void cb (EV_P_ ev_timer *w, int revents)
		1881
		1882	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		1883	suitable for use with C<EV_A>.
		1884
		1885	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		1886
		1887	Similar to the other two macros, this gives you the value of the default
		1888	loop, if multiple loops are supported ("ev loop default").
		1889
		1890	=back
		1891
		1892	Example: Declare and initialise a check watcher, utilising the above
		1893	macros so it will work regardless of whether multiple loops are supported
		1894	or not.
		1895
		1896	static void
		1897	check_cb (EV_P_ ev_timer *w, int revents)
		1898	{
		1899	ev_check_stop (EV_A_ w);
		1900	}
		1901
		1902	ev_check check;
		1903	ev_check_init (&check, check_cb);
		1904	ev_check_start (EV_DEFAULT_ &check);
		1905	ev_loop (EV_DEFAULT_ 0);
		1906
		1907	=head1 EMBEDDING
		1908
		1909	Libev can (and often is) directly embedded into host
		1910	applications. Examples of applications that embed it include the Deliantra
		1911	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		1912	and rxvt-unicode.
		1913
		1914	The goal is to enable you to just copy the neecssary files into your
		1915	source directory without having to change even a single line in them, so
		1916	you can easily upgrade by simply copying (or having a checked-out copy of
		1917	libev somewhere in your source tree).
		1918
		1919	=head2 FILESETS
		1920
		1921	Depending on what features you need you need to include one or more sets of files
		1922	in your app.
		1923
		1924	=head3 CORE EVENT LOOP
		1925
		1926	To include only the libev core (all the C<ev_*> functions), with manual
		1927	configuration (no autoconf):
		1928
		1929	#define EV_STANDALONE 1
		1930	#include "ev.c"
		1931
		1932	This will automatically include F<ev.h>, too, and should be done in a
		1933	single C source file only to provide the function implementations. To use
		1934	it, do the same for F<ev.h> in all files wishing to use this API (best
		1935	done by writing a wrapper around F<ev.h> that you can include instead and
		1936	where you can put other configuration options):
		1937
		1938	#define EV_STANDALONE 1
		1939	#include "ev.h"
		1940
		1941	Both header files and implementation files can be compiled with a C++
		1942	compiler (at least, thats a stated goal, and breakage will be treated
		1943	as a bug).
		1944
		1945	You need the following files in your source tree, or in a directory
		1946	in your include path (e.g. in libev/ when using -Ilibev):
		1947
		1948	ev.h
		1949	ev.c
		1950	ev_vars.h
		1951	ev_wrap.h
		1952
		1953	ev_win32.c required on win32 platforms only
		1954
		1955	ev_select.c only when select backend is enabled (which is enabled by default)
		1956	ev_poll.c only when poll backend is enabled (disabled by default)
		1957	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		1958	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		1959	ev_port.c only when the solaris port backend is enabled (disabled by default)
		1960
		1961	F<ev.c> includes the backend files directly when enabled, so you only need
		1962	to compile this single file.
		1963
		1964	=head3 LIBEVENT COMPATIBILITY API
		1965
		1966	To include the libevent compatibility API, also include:
		1967
		1968	#include "event.c"
		1969
		1970	in the file including F<ev.c>, and:
		1971
		1972	#include "event.h"
		1973
		1974	in the files that want to use the libevent API. This also includes F<ev.h>.
		1975
		1976	You need the following additional files for this:
		1977
		1978	event.h
		1979	event.c
		1980
		1981	=head3 AUTOCONF SUPPORT
		1982
		1983	Instead of using C<EV_STANDALONE=1> and providing your config in
		1984	whatever way you want, you can also C<m4_include([libev.m4])> in your
		1985	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		1986	include F<config.h> and configure itself accordingly.
		1987
		1988	For this of course you need the m4 file:
		1989
		1990	libev.m4
		1991
		1992	=head2 PREPROCESSOR SYMBOLS/MACROS
		1993
		1994	Libev can be configured via a variety of preprocessor symbols you have to define
		1995	before including any of its files. The default is not to build for multiplicity
		1996	and only include the select backend.
		1997
		1998	=over 4
		1999
		2000	=item EV_STANDALONE
		2001
		2002	Must always be C<1> if you do not use autoconf configuration, which
		2003	keeps libev from including F<config.h>, and it also defines dummy
		2004	implementations for some libevent functions (such as logging, which is not
		2005	supported). It will also not define any of the structs usually found in
		2006	F<event.h> that are not directly supported by the libev core alone.
		2007
		2008	=item EV_USE_MONOTONIC
		2009
		2010	If defined to be C<1>, libev will try to detect the availability of the
		2011	monotonic clock option at both compiletime and runtime. Otherwise no use
		2012	of the monotonic clock option will be attempted. If you enable this, you
		2013	usually have to link against librt or something similar. Enabling it when
		2014	the functionality isn't available is safe, though, althoguh you have
		2015	to make sure you link against any libraries where the C<clock_gettime>
		2016	function is hiding in (often F<-lrt>).
		2017
		2018	=item EV_USE_REALTIME
		2019
		2020	If defined to be C<1>, libev will try to detect the availability of the
		2021	realtime clock option at compiletime (and assume its availability at
		2022	runtime if successful). Otherwise no use of the realtime clock option will
		2023	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		2024	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries
		2025	in the description of C<EV_USE_MONOTONIC>, though.
		2026
		2027	=item EV_USE_SELECT
		2028
		2029	If undefined or defined to be C<1>, libev will compile in support for the
		2030	C<select>(2) backend. No attempt at autodetection will be done: if no
		2031	other method takes over, select will be it. Otherwise the select backend
		2032	will not be compiled in.
		2033
		2034	=item EV_SELECT_USE_FD_SET
		2035
		2036	If defined to C<1>, then the select backend will use the system C<fd_set>
		2037	structure. This is useful if libev doesn't compile due to a missing
		2038	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on
		2039	exotic systems. This usually limits the range of file descriptors to some
		2040	low limit such as 1024 or might have other limitations (winsocket only
		2041	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		2042	influence the size of the C<fd_set> used.
		2043
		2044	=item EV_SELECT_IS_WINSOCKET
		2045
		2046	When defined to C<1>, the select backend will assume that
		2047	select/socket/connect etc. don't understand file descriptors but
		2048	wants osf handles on win32 (this is the case when the select to
		2049	be used is the winsock select). This means that it will call
		2050	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		2051	it is assumed that all these functions actually work on fds, even
		2052	on win32. Should not be defined on non-win32 platforms.
		2053
		2054	=item EV_USE_POLL
		2055
		2056	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		2057	backend. Otherwise it will be enabled on non-win32 platforms. It
		2058	takes precedence over select.
		2059
		2060	=item EV_USE_EPOLL
		2061
		2062	If defined to be C<1>, libev will compile in support for the Linux
		2063	C<epoll>(7) backend. Its availability will be detected at runtime,
		2064	otherwise another method will be used as fallback. This is the
		2065	preferred backend for GNU/Linux systems.
		2066
		2067	=item EV_USE_KQUEUE
		2068
		2069	If defined to be C<1>, libev will compile in support for the BSD style
		2070	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		2071	otherwise another method will be used as fallback. This is the preferred
		2072	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		2073	supports some types of fds correctly (the only platform we found that
		2074	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		2075	not be used unless explicitly requested. The best way to use it is to find
		2076	out whether kqueue supports your type of fd properly and use an embedded
		2077	kqueue loop.
		2078
		2079	=item EV_USE_PORT
		2080
		2081	If defined to be C<1>, libev will compile in support for the Solaris
		2082	10 port style backend. Its availability will be detected at runtime,
		2083	otherwise another method will be used as fallback. This is the preferred
		2084	backend for Solaris 10 systems.
		2085
		2086	=item EV_USE_DEVPOLL
		2087
		2088	reserved for future expansion, works like the USE symbols above.
		2089
		2090	=item EV_USE_INOTIFY
		2091
		2092	If defined to be C<1>, libev will compile in support for the Linux inotify
		2093	interface to speed up C<ev_stat> watchers. Its actual availability will
		2094	be detected at runtime.
		2095
		2096	=item EV_H
		2097
		2098	The name of the F<ev.h> header file used to include it. The default if
		2099	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This
		2100	can be used to virtually rename the F<ev.h> header file in case of conflicts.
		2101
		2102	=item EV_CONFIG_H
		2103
		2104	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		2105	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		2106	C<EV_H>, above.
		2107
		2108	=item EV_EVENT_H
		2109
		2110	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		2111	of how the F<event.h> header can be found.
		2112
		2113	=item EV_PROTOTYPES
		2114
		2115	If defined to be C<0>, then F<ev.h> will not define any function
		2116	prototypes, but still define all the structs and other symbols. This is
		2117	occasionally useful if you want to provide your own wrapper functions
		2118	around libev functions.
		2119
		2120	=item EV_MULTIPLICITY
		2121
		2122	If undefined or defined to C<1>, then all event-loop-specific functions
		2123	will have the C<struct ev_loop *> as first argument, and you can create
		2124	additional independent event loops. Otherwise there will be no support
		2125	for multiple event loops and there is no first event loop pointer
		2126	argument. Instead, all functions act on the single default loop.
		2127
		2128	=item EV_PERIODIC_ENABLE
		2129
		2130	If undefined or defined to be C<1>, then periodic timers are supported. If
		2131	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2132	code.
		2133
		2134	=item EV_IDLE_ENABLE
		2135
		2136	If undefined or defined to be C<1>, then idle watchers are supported. If
		2137	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2138	code.
		2139
		2140	=item EV_EMBED_ENABLE
		2141
		2142	If undefined or defined to be C<1>, then embed watchers are supported. If
		2143	defined to be C<0>, then they are not.
		2144
		2145	=item EV_STAT_ENABLE
		2146
		2147	If undefined or defined to be C<1>, then stat watchers are supported. If
		2148	defined to be C<0>, then they are not.
		2149
		2150	=item EV_FORK_ENABLE
		2151
		2152	If undefined or defined to be C<1>, then fork watchers are supported. If
		2153	defined to be C<0>, then they are not.
		2154
		2155	=item EV_MINIMAL
		2156
		2157	If you need to shave off some kilobytes of code at the expense of some
		2158	speed, define this symbol to C<1>. Currently only used for gcc to override
		2159	some inlining decisions, saves roughly 30% codesize of amd64.
		2160
		2161	=item EV_PID_HASHSIZE
		2162
		2163	C<ev_child> watchers use a small hash table to distribute workload by
		2164	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		2165	than enough. If you need to manage thousands of children you might want to
		2166	increase this value (I<must> be a power of two).
		2167
		2168	=item EV_INOTIFY_HASHSIZE
		2169
		2170	C<ev_staz> watchers use a small hash table to distribute workload by
		2171	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		2172	usually more than enough. If you need to manage thousands of C<ev_stat>
		2173	watchers you might want to increase this value (I<must> be a power of
		2174	two).
		2175
		2176	=item EV_COMMON
		2177
		2178	By default, all watchers have a C<void *data> member. By redefining
		2179	this macro to a something else you can include more and other types of
		2180	members. You have to define it each time you include one of the files,
		2181	though, and it must be identical each time.
		2182
		2183	For example, the perl EV module uses something like this:
		2184
		2185	#define EV_COMMON \
		2186	SV *self; /* contains this struct */ \
		2187	SV *cb_sv, *fh /* note no trailing ";" */
		2188
		2189	=item EV_CB_DECLARE (type)
		2190
		2191	=item EV_CB_INVOKE (watcher, revents)
		2192
		2193	=item ev_set_cb (ev, cb)
		2194
		2195	Can be used to change the callback member declaration in each watcher,
		2196	and the way callbacks are invoked and set. Must expand to a struct member
		2197	definition and a statement, respectively. See the F<ev.v> header file for
		2198	their default definitions. One possible use for overriding these is to
		2199	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		2200	method calls instead of plain function calls in C++.
		2201
		2202	=head2 EXAMPLES
		2203
		2204	For a real-world example of a program the includes libev
		2205	verbatim, you can have a look at the EV perl module
		2206	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		2207	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		2208	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		2209	will be compiled. It is pretty complex because it provides its own header
		2210	file.
		2211
		2212	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		2213	that everybody includes and which overrides some configure choices:
		2214
		2215	#define EV_MINIMAL 1
		2216	#define EV_USE_POLL 0
		2217	#define EV_MULTIPLICITY 0
		2218	#define EV_PERIODIC_ENABLE 0
		2219	#define EV_STAT_ENABLE 0
		2220	#define EV_FORK_ENABLE 0
		2221	#define EV_CONFIG_H <config.h>
		2222	#define EV_MINPRI 0
		2223	#define EV_MAXPRI 0
		2224
		2225	#include "ev++.h"
		2226
		2227	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		2228
		2229	#include "ev_cpp.h"
		2230	#include "ev.c"
		2231
		2232
		2233	=head1 COMPLEXITIES
		2234
		2235	In this section the complexities of (many of) the algorithms used inside
		2236	libev will be explained. For complexity discussions about backends see the
		2237	documentation for C<ev_default_init>.
		2238
		2239	=over 4
		2240
		2241	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		2242
		2243	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
		2244
		2245	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
		2246
		2247	=item Stopping check/prepare/idle watchers: O(1)
		2248
		2249	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		2250
		2251	=item Finding the next timer per loop iteration: O(1)
		2252
		2253	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		2254
		2255	=item Activating one watcher: O(1)
		2256
		2257	=back
		2258
878		2259
879	=head1 AUTHOR	2260	=head1 AUTHOR
880		2261
881	Marc Lehmann <libev@schmorp.de>.	2262	Marc Lehmann <libev@schmorp.de>.
882		2263

Diff Legend

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