ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.110 by root, Tue Dec 25 07:05:45 2007 UTC vs.
Revision 1.217 by root, Mon Nov 17 03:37:08 2008 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
		11	// a single header file is required
11	#include <ev.h>	12	#include <ev.h>
12		13
		14	#include <stdio.h> // for puts
		15
		16	// every watcher type has its own typedef'd struct
		17	// with the name ev_TYPE
13	ev_io stdin_watcher;	18	ev_io stdin_watcher;
14	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
15		20
16	/* called when data readable on stdin */	21	// all watcher callbacks have a similar signature
		22	// this callback is called when data is readable on stdin
17	static void	23	static void
18	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
19	{	25	{
20	/* puts ("stdin ready"); */	26	puts ("stdin ready");
21	ev_io_stop (EV_A_ w); /* just a syntax example */	27	// for one-shot events, one must manually stop the watcher
22	ev_unloop (EV_A_ EVUNLOOP_ALL); /* leave all loop calls */	28	// with its corresponding stop function.
		29	ev_io_stop (EV_A_ w);
		30
		31	// this causes all nested ev_loop's to stop iterating
		32	ev_unloop (EV_A_ EVUNLOOP_ALL);
23	}	33	}
24		34
		35	// another callback, this time for a time-out
25	static void	36	static void
26	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
27	{	38	{
28	/* puts ("timeout"); */	39	puts ("timeout");
29	ev_unloop (EV_A_ EVUNLOOP_ONE); /* leave one loop call */	40	// this causes the innermost ev_loop to stop iterating
		41	ev_unloop (EV_A_ EVUNLOOP_ONE);
30	}	42	}
31		43
32	int	44	int
33	main (void)	45	main (void)
34	{	46	{
		47	// use the default event loop unless you have special needs
35	struct ev_loop *loop = ev_default_loop (0);	48	ev_loop *loop = ev_default_loop (0);
36		49
37	/* initialise an io watcher, then start it */	50	// initialise an io watcher, then start it
		51	// this one will watch for stdin to become readable
38	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);	52	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
39	ev_io_start (loop, &stdin_watcher);	53	ev_io_start (loop, &stdin_watcher);
40		54
		55	// initialise a timer watcher, then start it
41	/* simple non-repeating 5.5 second timeout */	56	// simple non-repeating 5.5 second timeout
42	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);	57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
43	ev_timer_start (loop, &timeout_watcher);	58	ev_timer_start (loop, &timeout_watcher);
44		59
45	/* loop till timeout or data ready */	60	// now wait for events to arrive
46	ev_loop (loop, 0);	61	ev_loop (loop, 0);
47		62
		63	// unloop was called, so exit
48	return 0;	64	return 0;
49	}	65	}
50		66
51	=head1 DESCRIPTION	67	=head1 DESCRIPTION
52		68
53	The newest version of this document is also available as a html-formatted	69	The newest version of this document is also available as an html-formatted
54	web page you might find easier to navigate when reading it for the first	70	web page you might find easier to navigate when reading it for the first
55	time: L<http://cvs.schmorp.de/libev/ev.html>.	71	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
56		72
57	Libev is an event loop: you register interest in certain events (such as a	73	Libev is an event loop: you register interest in certain events (such as a
58	file descriptor being readable or a timeout occurring), and it will manage	74	file descriptor being readable or a timeout occurring), and it will manage
59	these event sources and provide your program with events.	75	these event sources and provide your program with events.
60		76
…		…
84	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	100	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
85	for example).	101	for example).
86		102
87	=head2 CONVENTIONS	103	=head2 CONVENTIONS
88		104
89	Libev is very configurable. In this manual the default configuration will	105	Libev is very configurable. In this manual the default (and most common)
90	be described, which supports multiple event loops. For more info about	106	configuration will be described, which supports multiple event loops. For
91	various configuration options please have a look at B<EMBED> section in	107	more info about various configuration options please have a look at
92	this manual. If libev was configured without support for multiple event	108	B<EMBED> section in this manual. If libev was configured without support
93	loops, then all functions taking an initial argument of name C<loop>	109	for multiple event loops, then all functions taking an initial argument of
94	(which is always of type C<struct ev_loop *>) will not have this argument.	110	name C<loop> (which is always of type C<ev_loop *>) will not have
		111	this argument.
95		112
96	=head2 TIME REPRESENTATION	113	=head2 TIME REPRESENTATION
97		114
98	Libev represents time as a single floating point number, representing the	115	Libev represents time as a single floating point number, representing the
99	(fractional) number of seconds since the (POSIX) epoch (somewhere near	116	(fractional) number of seconds since the (POSIX) epoch (somewhere near
100	the beginning of 1970, details are complicated, don't ask). This type is	117	the beginning of 1970, details are complicated, don't ask). This type is
101	called C<ev_tstamp>, which is what you should use too. It usually aliases	118	called C<ev_tstamp>, which is what you should use too. It usually aliases
102	to the C<double> type in C, and when you need to do any calculations on	119	to the C<double> type in C, and when you need to do any calculations on
103	it, you should treat it as some floatingpoint value. Unlike the name	120	it, you should treat it as some floating point value. Unlike the name
104	component C<stamp> might indicate, it is also used for time differences	121	component C<stamp> might indicate, it is also used for time differences
105	throughout libev.	122	throughout libev.
		123
		124	=head1 ERROR HANDLING
		125
		126	Libev knows three classes of errors: operating system errors, usage errors
		127	and internal errors (bugs).
		128
		129	When libev catches an operating system error it cannot handle (for example
		130	a system call indicating a condition libev cannot fix), it calls the callback
		131	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		132	abort. The default is to print a diagnostic message and to call C<abort
		133	()>.
		134
		135	When libev detects a usage error such as a negative timer interval, then
		136	it will print a diagnostic message and abort (via the C<assert> mechanism,
		137	so C<NDEBUG> will disable this checking): these are programming errors in
		138	the libev caller and need to be fixed there.
		139
		140	Libev also has a few internal error-checking C<assert>ions, and also has
		141	extensive consistency checking code. These do not trigger under normal
		142	circumstances, as they indicate either a bug in libev or worse.
		143
106		144
107	=head1 GLOBAL FUNCTIONS	145	=head1 GLOBAL FUNCTIONS
108		146
109	These functions can be called anytime, even before initialising the	147	These functions can be called anytime, even before initialising the
110	library in any way.	148	library in any way.
…		…
119		157
120	=item ev_sleep (ev_tstamp interval)	158	=item ev_sleep (ev_tstamp interval)
121		159
122	Sleep for the given interval: The current thread will be blocked until	160	Sleep for the given interval: The current thread will be blocked until
123	either it is interrupted or the given time interval has passed. Basically	161	either it is interrupted or the given time interval has passed. Basically
124	this is a subsecond-resolution C<sleep ()>.	162	this is a sub-second-resolution C<sleep ()>.
125		163
126	=item int ev_version_major ()	164	=item int ev_version_major ()
127		165
128	=item int ev_version_minor ()	166	=item int ev_version_minor ()
129		167
…		…
142	not a problem.	180	not a problem.
143		181
144	Example: Make sure we haven't accidentally been linked against the wrong	182	Example: Make sure we haven't accidentally been linked against the wrong
145	version.	183	version.
146		184
147	assert (("libev version mismatch",	185	assert (("libev version mismatch",
148	ev_version_major () == EV_VERSION_MAJOR	186	ev_version_major () == EV_VERSION_MAJOR
149	&& ev_version_minor () >= EV_VERSION_MINOR));	187	&& ev_version_minor () >= EV_VERSION_MINOR));
150		188
151	=item unsigned int ev_supported_backends ()	189	=item unsigned int ev_supported_backends ()
152		190
153	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	191	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
154	value) compiled into this binary of libev (independent of their	192	value) compiled into this binary of libev (independent of their
…		…
156	a description of the set values.	194	a description of the set values.
157		195
158	Example: make sure we have the epoll method, because yeah this is cool and	196	Example: make sure we have the epoll method, because yeah this is cool and
159	a must have and can we have a torrent of it please!!!11	197	a must have and can we have a torrent of it please!!!11
160		198
161	assert (("sorry, no epoll, no sex",	199	assert (("sorry, no epoll, no sex",
162	ev_supported_backends () & EVBACKEND_EPOLL));	200	ev_supported_backends () & EVBACKEND_EPOLL));
163		201
164	=item unsigned int ev_recommended_backends ()	202	=item unsigned int ev_recommended_backends ()
165		203
166	Return the set of all backends compiled into this binary of libev and also	204	Return the set of all backends compiled into this binary of libev and also
167	recommended for this platform. This set is often smaller than the one	205	recommended for this platform. This set is often smaller than the one
168	returned by C<ev_supported_backends>, as for example kqueue is broken on	206	returned by C<ev_supported_backends>, as for example kqueue is broken on
169	most BSDs and will not be autodetected unless you explicitly request it	207	most BSDs and will not be auto-detected unless you explicitly request it
170	(assuming you know what you are doing). This is the set of backends that	208	(assuming you know what you are doing). This is the set of backends that
171	libev will probe for if you specify no backends explicitly.	209	libev will probe for if you specify no backends explicitly.
172		210
173	=item unsigned int ev_embeddable_backends ()	211	=item unsigned int ev_embeddable_backends ()
174		212
…		…
178	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	216	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
179	recommended ones.	217	recommended ones.
180		218
181	See the description of C<ev_embed> watchers for more info.	219	See the description of C<ev_embed> watchers for more info.
182		220
183	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	221	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
184		222
185	Sets the allocation function to use (the prototype is similar - the	223	Sets the allocation function to use (the prototype is similar - the
186	semantics is identical - to the realloc C function). It is used to	224	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
187	allocate and free memory (no surprises here). If it returns zero when	225	used to allocate and free memory (no surprises here). If it returns zero
188	memory needs to be allocated, the library might abort or take some	226	when memory needs to be allocated (C<size != 0>), the library might abort
189	potentially destructive action. The default is your system realloc	227	or take some potentially destructive action.
190	function.	228
		229	Since some systems (at least OpenBSD and Darwin) fail to implement
		230	correct C<realloc> semantics, libev will use a wrapper around the system
		231	C<realloc> and C<free> functions by default.
191		232
192	You could override this function in high-availability programs to, say,	233	You could override this function in high-availability programs to, say,
193	free some memory if it cannot allocate memory, to use a special allocator,	234	free some memory if it cannot allocate memory, to use a special allocator,
194	or even to sleep a while and retry until some memory is available.	235	or even to sleep a while and retry until some memory is available.
195		236
196	Example: Replace the libev allocator with one that waits a bit and then	237	Example: Replace the libev allocator with one that waits a bit and then
197	retries).	238	retries (example requires a standards-compliant C<realloc>).
198		239
199	static void *	240	static void *
200	persistent_realloc (void *ptr, size_t size)	241	persistent_realloc (void *ptr, size_t size)
201	{	242	{
202	for (;;)	243	for (;;)
…		…
211	}	252	}
212		253
213	...	254	...
214	ev_set_allocator (persistent_realloc);	255	ev_set_allocator (persistent_realloc);
215		256
216	=item ev_set_syserr_cb (void (*cb)(const char *msg));	257	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
217		258
218	Set the callback function to call on a retryable syscall error (such	259	Set the callback function to call on a retryable system call error (such
219	as failed select, poll, epoll_wait). The message is a printable string	260	as failed select, poll, epoll_wait). The message is a printable string
220	indicating the system call or subsystem causing the problem. If this	261	indicating the system call or subsystem causing the problem. If this
221	callback is set, then libev will expect it to remedy the sitution, no	262	callback is set, then libev will expect it to remedy the situation, no
222	matter what, when it returns. That is, libev will generally retry the	263	matter what, when it returns. That is, libev will generally retry the
223	requested operation, or, if the condition doesn't go away, do bad stuff	264	requested operation, or, if the condition doesn't go away, do bad stuff
224	(such as abort).	265	(such as abort).
225		266
226	Example: This is basically the same thing that libev does internally, too.	267	Example: This is basically the same thing that libev does internally, too.
…		…
237		278
238	=back	279	=back
239		280
240	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	281	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
241		282
242	An event loop is described by a C<struct ev_loop *>. The library knows two	283	An event loop is described by a C<struct ev_loop *> (the C<struct>
243	types of such loops, the I<default> loop, which supports signals and child	284	is I<not> optional in this case, as there is also an C<ev_loop>
244	events, and dynamically created loops which do not.	285	I<function>).
245		286
246	If you use threads, a common model is to run the default event loop	287	The library knows two types of such loops, the I<default> loop, which
247	in your main thread (or in a separate thread) and for each thread you	288	supports signals and child events, and dynamically created loops which do
248	create, you also create another event loop. Libev itself does no locking	289	not.
249	whatsoever, so if you mix calls to the same event loop in different
250	threads, make sure you lock (this is usually a bad idea, though, even if
251	done correctly, because it's hideous and inefficient).
252		290
253	=over 4	291	=over 4
254		292
255	=item struct ev_loop *ev_default_loop (unsigned int flags)	293	=item struct ev_loop *ev_default_loop (unsigned int flags)
256		294
…		…
260	flags. If that is troubling you, check C<ev_backend ()> afterwards).	298	flags. If that is troubling you, check C<ev_backend ()> afterwards).
261		299
262	If you don't know what event loop to use, use the one returned from this	300	If you don't know what event loop to use, use the one returned from this
263	function.	301	function.
264		302
		303	Note that this function is I<not> thread-safe, so if you want to use it
		304	from multiple threads, you have to lock (note also that this is unlikely,
		305	as loops cannot be shared easily between threads anyway).
		306
		307	The default loop is the only loop that can handle C<ev_signal> and
		308	C<ev_child> watchers, and to do this, it always registers a handler
		309	for C<SIGCHLD>. If this is a problem for your application you can either
		310	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		311	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		312	C<ev_default_init>.
		313
265	The flags argument can be used to specify special behaviour or specific	314	The flags argument can be used to specify special behaviour or specific
266	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	315	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
267		316
268	The following flags are supported:	317	The following flags are supported:
269		318
…		…
274	The default flags value. Use this if you have no clue (it's the right	323	The default flags value. Use this if you have no clue (it's the right
275	thing, believe me).	324	thing, believe me).
276		325
277	=item C<EVFLAG_NOENV>	326	=item C<EVFLAG_NOENV>
278		327
279	If this flag bit is ored into the flag value (or the program runs setuid	328	If this flag bit is or'ed into the flag value (or the program runs setuid
280	or setgid) then libev will I<not> look at the environment variable	329	or setgid) then libev will I<not> look at the environment variable
281	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	330	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
282	override the flags completely if it is found in the environment. This is	331	override the flags completely if it is found in the environment. This is
283	useful to try out specific backends to test their performance, or to work	332	useful to try out specific backends to test their performance, or to work
284	around bugs.	333	around bugs.
…		…
290	enabling this flag.	339	enabling this flag.
291		340
292	This works by calling C<getpid ()> on every iteration of the loop,	341	This works by calling C<getpid ()> on every iteration of the loop,
293	and thus this might slow down your event loop if you do a lot of loop	342	and thus this might slow down your event loop if you do a lot of loop
294	iterations and little real work, but is usually not noticeable (on my	343	iterations and little real work, but is usually not noticeable (on my
295	Linux system for example, C<getpid> is actually a simple 5-insn sequence	344	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
296	without a syscall and thus I<very> fast, but my Linux system also has	345	without a system call and thus I<very> fast, but my GNU/Linux system also has
297	C<pthread_atfork> which is even faster).	346	C<pthread_atfork> which is even faster).
298		347
299	The big advantage of this flag is that you can forget about fork (and	348	The big advantage of this flag is that you can forget about fork (and
300	forget about forgetting to tell libev about forking) when you use this	349	forget about forgetting to tell libev about forking) when you use this
301	flag.	350	flag.
302		351
303	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>	352	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
304	environment variable.	353	environment variable.
305		354
306	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	355	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
307		356
308	This is your standard select(2) backend. Not I<completely> standard, as	357	This is your standard select(2) backend. Not I<completely> standard, as
…		…
310	but if that fails, expect a fairly low limit on the number of fds when	359	but if that fails, expect a fairly low limit on the number of fds when
311	using this backend. It doesn't scale too well (O(highest_fd)), but its	360	using this backend. It doesn't scale too well (O(highest_fd)), but its
312	usually the fastest backend for a low number of (low-numbered :) fds.	361	usually the fastest backend for a low number of (low-numbered :) fds.
313		362
314	To get good performance out of this backend you need a high amount of	363	To get good performance out of this backend you need a high amount of
315	parallelity (most of the file descriptors should be busy). If you are	364	parallelism (most of the file descriptors should be busy). If you are
316	writing a server, you should C<accept ()> in a loop to accept as many	365	writing a server, you should C<accept ()> in a loop to accept as many
317	connections as possible during one iteration. You might also want to have	366	connections as possible during one iteration. You might also want to have
318	a look at C<ev_set_io_collect_interval ()> to increase the amount of	367	a look at C<ev_set_io_collect_interval ()> to increase the amount of
319	readyness notifications you get per iteration.	368	readiness notifications you get per iteration.
		369
		370	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		371	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		372	C<exceptfds> set on that platform).
320		373
321	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	374	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
322		375
323	And this is your standard poll(2) backend. It's more complicated	376	And this is your standard poll(2) backend. It's more complicated
324	than select, but handles sparse fds better and has no artificial	377	than select, but handles sparse fds better and has no artificial
325	limit on the number of fds you can use (except it will slow down	378	limit on the number of fds you can use (except it will slow down
326	considerably with a lot of inactive fds). It scales similarly to select,	379	considerably with a lot of inactive fds). It scales similarly to select,
327	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	380	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
328	performance tips.	381	performance tips.
329		382
		383	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		384	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		385
330	=item C<EVBACKEND_EPOLL> (value 4, Linux)	386	=item C<EVBACKEND_EPOLL> (value 4, Linux)
331		387
332	For few fds, this backend is a bit little slower than poll and select,	388	For few fds, this backend is a bit little slower than poll and select,
333	but it scales phenomenally better. While poll and select usually scale	389	but it scales phenomenally better. While poll and select usually scale
334	like O(total_fds) where n is the total number of fds (or the highest fd),	390	like O(total_fds) where n is the total number of fds (or the highest fd),
335	epoll scales either O(1) or O(active_fds). The epoll design has a number	391	epoll scales either O(1) or O(active_fds).
336	of shortcomings, such as silently dropping events in some hard-to-detect	392
337	cases and rewiring a syscall per fd change, no fork support and bad	393	The epoll mechanism deserves honorable mention as the most misdesigned
338	support for dup.	394	of the more advanced event mechanisms: mere annoyances include silently
		395	dropping file descriptors, requiring a system call per change per file
		396	descriptor (and unnecessary guessing of parameters), problems with dup and
		397	so on. The biggest issue is fork races, however - if a program forks then
		398	I<both> parent and child process have to recreate the epoll set, which can
		399	take considerable time (one syscall per file descriptor) and is of course
		400	hard to detect.
		401
		402	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		403	of course I<doesn't>, and epoll just loves to report events for totally
		404	I<different> file descriptors (even already closed ones, so one cannot
		405	even remove them from the set) than registered in the set (especially
		406	on SMP systems). Libev tries to counter these spurious notifications by
		407	employing an additional generation counter and comparing that against the
		408	events to filter out spurious ones, recreating the set when required.
339		409
340	While stopping, setting and starting an I/O watcher in the same iteration	410	While stopping, setting and starting an I/O watcher in the same iteration
341	will result in some caching, there is still a syscall per such incident	411	will result in some caching, there is still a system call per such
342	(because the fd could point to a different file description now), so its	412	incident (because the same I<file descriptor> could point to a different
343	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	413	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
344	very well if you register events for both fds.	414	file descriptors might not work very well if you register events for both
345		415	file descriptors.
346	Please note that epoll sometimes generates spurious notifications, so you
347	need to use non-blocking I/O or other means to avoid blocking when no data
348	(or space) is available.
349		416
350	Best performance from this backend is achieved by not unregistering all	417	Best performance from this backend is achieved by not unregistering all
351	watchers for a file descriptor until it has been closed, if possible, i.e.	418	watchers for a file descriptor until it has been closed, if possible,
352	keep at least one watcher active per fd at all times.	419	i.e. keep at least one watcher active per fd at all times. Stopping and
		420	starting a watcher (without re-setting it) also usually doesn't cause
		421	extra overhead. A fork can both result in spurious notifications as well
		422	as in libev having to destroy and recreate the epoll object, which can
		423	take considerable time and thus should be avoided.
353		424
		425	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		426	faster than epoll for maybe up to a hundred file descriptors, depending on
		427	the usage. So sad.
		428
354	While nominally embeddeble in other event loops, this feature is broken in	429	While nominally embeddable in other event loops, this feature is broken in
355	all kernel versions tested so far.	430	all kernel versions tested so far.
		431
		432	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		433	C<EVBACKEND_POLL>.
356		434
357	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	435	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
358		436
359	Kqueue deserves special mention, as at the time of this writing, it	437	Kqueue deserves special mention, as at the time of this writing, it
360	was broken on all BSDs except NetBSD (usually it doesn't work reliably	438	was broken on all BSDs except NetBSD (usually it doesn't work reliably
361	with anything but sockets and pipes, except on Darwin, where of course	439	with anything but sockets and pipes, except on Darwin, where of course
362	it's completely useless). For this reason it's not being "autodetected"	440	it's completely useless). Unlike epoll, however, whose brokenness
		441	is by design, these kqueue bugs can (and eventually will) be fixed
		442	without API changes to existing programs. For this reason it's not being
363	unless you explicitly specify it explicitly in the flags (i.e. using	443	"auto-detected" unless you explicitly specify it in the flags (i.e. using
364	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	444	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
365	system like NetBSD.	445	system like NetBSD.
366		446
367	You still can embed kqueue into a normal poll or select backend and use it	447	You still can embed kqueue into a normal poll or select backend and use it
368	only for sockets (after having made sure that sockets work with kqueue on	448	only for sockets (after having made sure that sockets work with kqueue on
369	the target platform). See C<ev_embed> watchers for more info.	449	the target platform). See C<ev_embed> watchers for more info.
370		450
371	It scales in the same way as the epoll backend, but the interface to the	451	It scales in the same way as the epoll backend, but the interface to the
372	kernel is more efficient (which says nothing about its actual speed, of	452	kernel is more efficient (which says nothing about its actual speed, of
373	course). While stopping, setting and starting an I/O watcher does never	453	course). While stopping, setting and starting an I/O watcher does never
374	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to	454	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
375	two event changes per incident, support for C<fork ()> is very bad and it	455	two event changes per incident. Support for C<fork ()> is very bad (but
376	drops fds silently in similarly hard-to-detect cases.	456	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		457	cases
377		458
378	This backend usually performs well under most conditions.	459	This backend usually performs well under most conditions.
379		460
380	While nominally embeddable in other event loops, this doesn't work	461	While nominally embeddable in other event loops, this doesn't work
381	everywhere, so you might need to test for this. And since it is broken	462	everywhere, so you might need to test for this. And since it is broken
382	almost everywhere, you should only use it when you have a lot of sockets	463	almost everywhere, you should only use it when you have a lot of sockets
383	(for which it usually works), by embedding it into another event loop	464	(for which it usually works), by embedding it into another event loop
384	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	465	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,
385	sockets.	466	using it only for sockets.
		467
		468	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		469	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		470	C<NOTE_EOF>.
386		471
387	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	472	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
388		473
389	This is not implemented yet (and might never be, unless you send me an	474	This is not implemented yet (and might never be, unless you send me an
390	implementation). According to reports, C</dev/poll> only supports sockets	475	implementation). According to reports, C</dev/poll> only supports sockets
…		…
394	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	479	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
395		480
396	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	481	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
397	it's really slow, but it still scales very well (O(active_fds)).	482	it's really slow, but it still scales very well (O(active_fds)).
398		483
399	Please note that solaris event ports can deliver a lot of spurious	484	Please note that Solaris event ports can deliver a lot of spurious
400	notifications, so you need to use non-blocking I/O or other means to avoid	485	notifications, so you need to use non-blocking I/O or other means to avoid
401	blocking when no data (or space) is available.	486	blocking when no data (or space) is available.
402		487
403	While this backend scales well, it requires one system call per active	488	While this backend scales well, it requires one system call per active
404	file descriptor per loop iteration. For small and medium numbers of file	489	file descriptor per loop iteration. For small and medium numbers of file
405	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	490	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
406	might perform better.	491	might perform better.
407		492
		493	On the positive side, with the exception of the spurious readiness
		494	notifications, this backend actually performed fully to specification
		495	in all tests and is fully embeddable, which is a rare feat among the
		496	OS-specific backends (I vastly prefer correctness over speed hacks).
		497
		498	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		499	C<EVBACKEND_POLL>.
		500
408	=item C<EVBACKEND_ALL>	501	=item C<EVBACKEND_ALL>
409		502
410	Try all backends (even potentially broken ones that wouldn't be tried	503	Try all backends (even potentially broken ones that wouldn't be tried
411	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	504	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
412	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	505	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
413		506
414	It is definitely not recommended to use this flag.	507	It is definitely not recommended to use this flag.
415		508
416	=back	509	=back
417		510
418	If one or more of these are ored into the flags value, then only these	511	If one or more of these are or'ed into the flags value, then only these
419	backends will be tried (in the reverse order as given here). If none are	512	backends will be tried (in the reverse order as listed here). If none are
420	specified, most compiled-in backend will be tried, usually in reverse	513	specified, all backends in C<ev_recommended_backends ()> will be tried.
421	order of their flag values :)
422		514
423	The most typical usage is like this:	515	Example: This is the most typical usage.
424		516
425	if (!ev_default_loop (0))	517	if (!ev_default_loop (0))
426	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	518	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
427		519
428	Restrict libev to the select and poll backends, and do not allow	520	Example: Restrict libev to the select and poll backends, and do not allow
429	environment settings to be taken into account:	521	environment settings to be taken into account:
430		522
431	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	523	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
432		524
433	Use whatever libev has to offer, but make sure that kqueue is used if	525	Example: Use whatever libev has to offer, but make sure that kqueue is
434	available (warning, breaks stuff, best use only with your own private	526	used if available (warning, breaks stuff, best use only with your own
435	event loop and only if you know the OS supports your types of fds):	527	private event loop and only if you know the OS supports your types of
		528	fds):
436		529
437	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	530	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
438		531
439	=item struct ev_loop *ev_loop_new (unsigned int flags)	532	=item struct ev_loop *ev_loop_new (unsigned int flags)
440		533
441	Similar to C<ev_default_loop>, but always creates a new event loop that is	534	Similar to C<ev_default_loop>, but always creates a new event loop that is
442	always distinct from the default loop. Unlike the default loop, it cannot	535	always distinct from the default loop. Unlike the default loop, it cannot
443	handle signal and child watchers, and attempts to do so will be greeted by	536	handle signal and child watchers, and attempts to do so will be greeted by
444	undefined behaviour (or a failed assertion if assertions are enabled).	537	undefined behaviour (or a failed assertion if assertions are enabled).
445		538
		539	Note that this function I<is> thread-safe, and the recommended way to use
		540	libev with threads is indeed to create one loop per thread, and using the
		541	default loop in the "main" or "initial" thread.
		542
446	Example: Try to create a event loop that uses epoll and nothing else.	543	Example: Try to create a event loop that uses epoll and nothing else.
447		544
448	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	545	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
449	if (!epoller)	546	if (!epoller)
450	fatal ("no epoll found here, maybe it hides under your chair");	547	fatal ("no epoll found here, maybe it hides under your chair");
451		548
452	=item ev_default_destroy ()	549	=item ev_default_destroy ()
453		550
454	Destroys the default loop again (frees all memory and kernel state	551	Destroys the default loop again (frees all memory and kernel state
455	etc.). None of the active event watchers will be stopped in the normal	552	etc.). None of the active event watchers will be stopped in the normal
456	sense, so e.g. C<ev_is_active> might still return true. It is your	553	sense, so e.g. C<ev_is_active> might still return true. It is your
457	responsibility to either stop all watchers cleanly yoursef I<before>	554	responsibility to either stop all watchers cleanly yourself I<before>
458	calling this function, or cope with the fact afterwards (which is usually	555	calling this function, or cope with the fact afterwards (which is usually
459	the easiest thing, you can just ignore the watchers and/or C<free ()> them	556	the easiest thing, you can just ignore the watchers and/or C<free ()> them
460	for example).	557	for example).
461		558
462	Note that certain global state, such as signal state, will not be freed by	559	Note that certain global state, such as signal state (and installed signal
463	this function, and related watchers (such as signal and child watchers)	560	handlers), will not be freed by this function, and related watchers (such
464	would need to be stopped manually.	561	as signal and child watchers) would need to be stopped manually.
465		562
466	In general it is not advisable to call this function except in the	563	In general it is not advisable to call this function except in the
467	rare occasion where you really need to free e.g. the signal handling	564	rare occasion where you really need to free e.g. the signal handling
468	pipe fds. If you need dynamically allocated loops it is better to use	565	pipe fds. If you need dynamically allocated loops it is better to use
469	C<ev_loop_new> and C<ev_loop_destroy>).	566	C<ev_loop_new> and C<ev_loop_destroy>).
…		…
473	Like C<ev_default_destroy>, but destroys an event loop created by an	570	Like C<ev_default_destroy>, but destroys an event loop created by an
474	earlier call to C<ev_loop_new>.	571	earlier call to C<ev_loop_new>.
475		572
476	=item ev_default_fork ()	573	=item ev_default_fork ()
477		574
		575	This function sets a flag that causes subsequent C<ev_loop> iterations
478	This function reinitialises the kernel state for backends that have	576	to reinitialise the kernel state for backends that have one. Despite the
479	one. Despite the name, you can call it anytime, but it makes most sense	577	name, you can call it anytime, but it makes most sense after forking, in
480	after forking, in either the parent or child process (or both, but that	578	the child process (or both child and parent, but that again makes little
481	again makes little sense).	579	sense). You I<must> call it in the child before using any of the libev
		580	functions, and it will only take effect at the next C<ev_loop> iteration.
482		581
483	You I<must> call this function in the child process after forking if and	582	On the other hand, you only need to call this function in the child
484	only if you want to use the event library in both processes. If you just	583	process if and only if you want to use the event library in the child. If
485	fork+exec, you don't have to call it.	584	you just fork+exec, you don't have to call it at all.
486		585
487	The function itself is quite fast and it's usually not a problem to call	586	The function itself is quite fast and it's usually not a problem to call
488	it just in case after a fork. To make this easy, the function will fit in	587	it just in case after a fork. To make this easy, the function will fit in
489	quite nicely into a call to C<pthread_atfork>:	588	quite nicely into a call to C<pthread_atfork>:
490		589
491	pthread_atfork (0, 0, ev_default_fork);	590	pthread_atfork (0, 0, ev_default_fork);
492		591
493	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
494	without calling this function, so if you force one of those backends you
495	do not need to care.
496
497	=item ev_loop_fork (loop)	592	=item ev_loop_fork (loop)
498		593
499	Like C<ev_default_fork>, but acts on an event loop created by	594	Like C<ev_default_fork>, but acts on an event loop created by
500	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	595	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
501	after fork, and how you do this is entirely your own problem.	596	after fork that you want to re-use in the child, and how you do this is
		597	entirely your own problem.
		598
		599	=item int ev_is_default_loop (loop)
		600
		601	Returns true when the given loop is, in fact, the default loop, and false
		602	otherwise.
502		603
503	=item unsigned int ev_loop_count (loop)	604	=item unsigned int ev_loop_count (loop)
504		605
505	Returns the count of loop iterations for the loop, which is identical to	606	Returns the count of loop iterations for the loop, which is identical to
506	the number of times libev did poll for new events. It starts at C<0> and	607	the number of times libev did poll for new events. It starts at C<0> and
…		…
521	received events and started processing them. This timestamp does not	622	received events and started processing them. This timestamp does not
522	change as long as callbacks are being processed, and this is also the base	623	change as long as callbacks are being processed, and this is also the base
523	time used for relative timers. You can treat it as the timestamp of the	624	time used for relative timers. You can treat it as the timestamp of the
524	event occurring (or more correctly, libev finding out about it).	625	event occurring (or more correctly, libev finding out about it).
525		626
		627	=item ev_now_update (loop)
		628
		629	Establishes the current time by querying the kernel, updating the time
		630	returned by C<ev_now ()> in the progress. This is a costly operation and
		631	is usually done automatically within C<ev_loop ()>.
		632
		633	This function is rarely useful, but when some event callback runs for a
		634	very long time without entering the event loop, updating libev's idea of
		635	the current time is a good idea.
		636
		637	See also "The special problem of time updates" in the C<ev_timer> section.
		638
526	=item ev_loop (loop, int flags)	639	=item ev_loop (loop, int flags)
527		640
528	Finally, this is it, the event handler. This function usually is called	641	Finally, this is it, the event handler. This function usually is called
529	after you initialised all your watchers and you want to start handling	642	after you initialised all your watchers and you want to start handling
530	events.	643	events.
…		…
532	If the flags argument is specified as C<0>, it will not return until	645	If the flags argument is specified as C<0>, it will not return until
533	either no event watchers are active anymore or C<ev_unloop> was called.	646	either no event watchers are active anymore or C<ev_unloop> was called.
534		647
535	Please note that an explicit C<ev_unloop> is usually better than	648	Please note that an explicit C<ev_unloop> is usually better than
536	relying on all watchers to be stopped when deciding when a program has	649	relying on all watchers to be stopped when deciding when a program has
537	finished (especially in interactive programs), but having a program that	650	finished (especially in interactive programs), but having a program
538	automatically loops as long as it has to and no longer by virtue of	651	that automatically loops as long as it has to and no longer by virtue
539	relying on its watchers stopping correctly is a thing of beauty.	652	of relying on its watchers stopping correctly, that is truly a thing of
		653	beauty.
540		654
541	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	655	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
542	those events and any outstanding ones, but will not block your process in	656	those events and any already outstanding ones, but will not block your
543	case there are no events and will return after one iteration of the loop.	657	process in case there are no events and will return after one iteration of
		658	the loop.
544		659
545	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	660	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
546	neccessary) and will handle those and any outstanding ones. It will block	661	necessary) and will handle those and any already outstanding ones. It
547	your process until at least one new event arrives, and will return after	662	will block your process until at least one new event arrives (which could
548	one iteration of the loop. This is useful if you are waiting for some	663	be an event internal to libev itself, so there is no guarantee that a
549	external event in conjunction with something not expressible using other	664	user-registered callback will be called), and will return after one
		665	iteration of the loop.
		666
		667	This is useful if you are waiting for some external event in conjunction
		668	with something not expressible using other libev watchers (i.e. "roll your
550	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	669	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
551	usually a better approach for this kind of thing.	670	usually a better approach for this kind of thing.
552		671
553	Here are the gory details of what C<ev_loop> does:	672	Here are the gory details of what C<ev_loop> does:
554		673
555	- Before the first iteration, call any pending watchers.	674	- Before the first iteration, call any pending watchers.
556	* If there are no active watchers (reference count is zero), return.	675	* If EVFLAG_FORKCHECK was used, check for a fork.
557	- Queue all prepare watchers and then call all outstanding watchers.	676	- If a fork was detected (by any means), queue and call all fork watchers.
		677	- Queue and call all prepare watchers.
558	- If we have been forked, recreate the kernel state.	678	- If we have been forked, detach and recreate the kernel state
		679	as to not disturb the other process.
559	- Update the kernel state with all outstanding changes.	680	- Update the kernel state with all outstanding changes.
560	- Update the "event loop time".	681	- Update the "event loop time" (ev_now ()).
561	- Calculate for how long to block.	682	- Calculate for how long to sleep or block, if at all
		683	(active idle watchers, EVLOOP_NONBLOCK or not having
		684	any active watchers at all will result in not sleeping).
		685	- Sleep if the I/O and timer collect interval say so.
562	- Block the process, waiting for any events.	686	- Block the process, waiting for any events.
563	- Queue all outstanding I/O (fd) events.	687	- Queue all outstanding I/O (fd) events.
564	- Update the "event loop time" and do time jump handling.	688	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
565	- Queue all outstanding timers.	689	- Queue all expired timers.
566	- Queue all outstanding periodics.	690	- Queue all expired periodics.
567	- If no events are pending now, queue all idle watchers.	691	- Unless any events are pending now, queue all idle watchers.
568	- Queue all check watchers.	692	- Queue all check watchers.
569	- Call all queued watchers in reverse order (i.e. check watchers first).	693	- Call all queued watchers in reverse order (i.e. check watchers first).
570	Signals and child watchers are implemented as I/O watchers, and will	694	Signals and child watchers are implemented as I/O watchers, and will
571	be handled here by queueing them when their watcher gets executed.	695	be handled here by queueing them when their watcher gets executed.
572	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	696	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
573	were used, return, otherwise continue with step *.	697	were used, or there are no active watchers, return, otherwise
		698	continue with step *.
574		699
575	Example: Queue some jobs and then loop until no events are outsanding	700	Example: Queue some jobs and then loop until no events are outstanding
576	anymore.	701	anymore.
577		702
578	... queue jobs here, make sure they register event watchers as long	703	... queue jobs here, make sure they register event watchers as long
579	... as they still have work to do (even an idle watcher will do..)	704	... as they still have work to do (even an idle watcher will do..)
580	ev_loop (my_loop, 0);	705	ev_loop (my_loop, 0);
581	... jobs done. yeah!	706	... jobs done or somebody called unloop. yeah!
582		707
583	=item ev_unloop (loop, how)	708	=item ev_unloop (loop, how)
584		709
585	Can be used to make a call to C<ev_loop> return early (but only after it	710	Can be used to make a call to C<ev_loop> return early (but only after it
586	has processed all outstanding events). The C<how> argument must be either	711	has processed all outstanding events). The C<how> argument must be either
587	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	712	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
588	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	713	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
589		714
		715	This "unloop state" will be cleared when entering C<ev_loop> again.
		716
		717	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		718
590	=item ev_ref (loop)	719	=item ev_ref (loop)
591		720
592	=item ev_unref (loop)	721	=item ev_unref (loop)
593		722
594	Ref/unref can be used to add or remove a reference count on the event	723	Ref/unref can be used to add or remove a reference count on the event
595	loop: Every watcher keeps one reference, and as long as the reference	724	loop: Every watcher keeps one reference, and as long as the reference
596	count is nonzero, C<ev_loop> will not return on its own. If you have	725	count is nonzero, C<ev_loop> will not return on its own.
		726
597	a watcher you never unregister that should not keep C<ev_loop> from	727	If you have a watcher you never unregister that should not keep C<ev_loop>
598	returning, ev_unref() after starting, and ev_ref() before stopping it. For	728	from returning, call ev_unref() after starting, and ev_ref() before
		729	stopping it.
		730
599	example, libev itself uses this for its internal signal pipe: It is not	731	As an example, libev itself uses this for its internal signal pipe: It is
600	visible to the libev user and should not keep C<ev_loop> from exiting if	732	not visible to the libev user and should not keep C<ev_loop> from exiting
601	no event watchers registered by it are active. It is also an excellent	733	if no event watchers registered by it are active. It is also an excellent
602	way to do this for generic recurring timers or from within third-party	734	way to do this for generic recurring timers or from within third-party
603	libraries. Just remember to I<unref after start> and I<ref before stop>.	735	libraries. Just remember to I<unref after start> and I<ref before stop>
		736	(but only if the watcher wasn't active before, or was active before,
		737	respectively).
604		738
605	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	739	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
606	running when nothing else is active.	740	running when nothing else is active.
607		741
608	struct ev_signal exitsig;	742	ev_signal exitsig;
609	ev_signal_init (&exitsig, sig_cb, SIGINT);	743	ev_signal_init (&exitsig, sig_cb, SIGINT);
610	ev_signal_start (loop, &exitsig);	744	ev_signal_start (loop, &exitsig);
611	evf_unref (loop);	745	evf_unref (loop);
612		746
613	Example: For some weird reason, unregister the above signal handler again.	747	Example: For some weird reason, unregister the above signal handler again.
614		748
615	ev_ref (loop);	749	ev_ref (loop);
616	ev_signal_stop (loop, &exitsig);	750	ev_signal_stop (loop, &exitsig);
617		751
618	=item ev_set_io_collect_interval (loop, ev_tstamp interval)	752	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
619		753
620	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)	754	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
621		755
622	These advanced functions influence the time that libev will spend waiting	756	These advanced functions influence the time that libev will spend waiting
623	for events. Both are by default C<0>, meaning that libev will try to	757	for events. Both time intervals are by default C<0>, meaning that libev
624	invoke timer/periodic callbacks and I/O callbacks with minimum latency.	758	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		759	latency.
625		760
626	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	761	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
627	allows libev to delay invocation of I/O and timer/periodic callbacks to	762	allows libev to delay invocation of I/O and timer/periodic callbacks
628	increase efficiency of loop iterations.	763	to increase efficiency of loop iterations (or to increase power-saving
		764	opportunities).
629		765
630	The background is that sometimes your program runs just fast enough to	766	The idea is that sometimes your program runs just fast enough to handle
631	handle one (or very few) event(s) per loop iteration. While this makes	767	one (or very few) event(s) per loop iteration. While this makes the
632	the program responsive, it also wastes a lot of CPU time to poll for new	768	program responsive, it also wastes a lot of CPU time to poll for new
633	events, especially with backends like C<select ()> which have a high	769	events, especially with backends like C<select ()> which have a high
634	overhead for the actual polling but can deliver many events at once.	770	overhead for the actual polling but can deliver many events at once.
635		771
636	By setting a higher I<io collect interval> you allow libev to spend more	772	By setting a higher I<io collect interval> you allow libev to spend more
637	time collecting I/O events, so you can handle more events per iteration,	773	time collecting I/O events, so you can handle more events per iteration,
…		…
639	C<ev_timer>) will be not affected. Setting this to a non-null value will	775	C<ev_timer>) will be not affected. Setting this to a non-null value will
640	introduce an additional C<ev_sleep ()> call into most loop iterations.	776	introduce an additional C<ev_sleep ()> call into most loop iterations.
641		777
642	Likewise, by setting a higher I<timeout collect interval> you allow libev	778	Likewise, by setting a higher I<timeout collect interval> you allow libev
643	to spend more time collecting timeouts, at the expense of increased	779	to spend more time collecting timeouts, at the expense of increased
644	latency (the watcher callback will be called later). C<ev_io> watchers	780	latency/jitter/inexactness (the watcher callback will be called
645	will not be affected. Setting this to a non-null value will not introduce	781	later). C<ev_io> watchers will not be affected. Setting this to a non-null
646	any overhead in libev.	782	value will not introduce any overhead in libev.
647		783
648	Many (busy) programs can usually benefit by setting the io collect	784	Many (busy) programs can usually benefit by setting the I/O collect
649	interval to a value near C<0.1> or so, which is often enough for	785	interval to a value near C<0.1> or so, which is often enough for
650	interactive servers (of course not for games), likewise for timeouts. It	786	interactive servers (of course not for games), likewise for timeouts. It
651	usually doesn't make much sense to set it to a lower value than C<0.01>,	787	usually doesn't make much sense to set it to a lower value than C<0.01>,
652	as this approsaches the timing granularity of most systems.	788	as this approaches the timing granularity of most systems.
		789
		790	Setting the I<timeout collect interval> can improve the opportunity for
		791	saving power, as the program will "bundle" timer callback invocations that
		792	are "near" in time together, by delaying some, thus reducing the number of
		793	times the process sleeps and wakes up again. Another useful technique to
		794	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		795	they fire on, say, one-second boundaries only.
		796
		797	=item ev_loop_verify (loop)
		798
		799	This function only does something when C<EV_VERIFY> support has been
		800	compiled in, which is the default for non-minimal builds. It tries to go
		801	through all internal structures and checks them for validity. If anything
		802	is found to be inconsistent, it will print an error message to standard
		803	error and call C<abort ()>.
		804
		805	This can be used to catch bugs inside libev itself: under normal
		806	circumstances, this function will never abort as of course libev keeps its
		807	data structures consistent.
653		808
654	=back	809	=back
655		810
656		811
657	=head1 ANATOMY OF A WATCHER	812	=head1 ANATOMY OF A WATCHER
		813
		814	In the following description, uppercase C<TYPE> in names stands for the
		815	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		816	watchers and C<ev_io_start> for I/O watchers.
658		817
659	A watcher is a structure that you create and register to record your	818	A watcher is a structure that you create and register to record your
660	interest in some event. For instance, if you want to wait for STDIN to	819	interest in some event. For instance, if you want to wait for STDIN to
661	become readable, you would create an C<ev_io> watcher for that:	820	become readable, you would create an C<ev_io> watcher for that:
662		821
663	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	822	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
664	{	823	{
665	ev_io_stop (w);	824	ev_io_stop (w);
666	ev_unloop (loop, EVUNLOOP_ALL);	825	ev_unloop (loop, EVUNLOOP_ALL);
667	}	826	}
668		827
669	struct ev_loop *loop = ev_default_loop (0);	828	struct ev_loop *loop = ev_default_loop (0);
		829
670	struct ev_io stdin_watcher;	830	ev_io stdin_watcher;
		831
671	ev_init (&stdin_watcher, my_cb);	832	ev_init (&stdin_watcher, my_cb);
672	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	833	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
673	ev_io_start (loop, &stdin_watcher);	834	ev_io_start (loop, &stdin_watcher);
		835
674	ev_loop (loop, 0);	836	ev_loop (loop, 0);
675		837
676	As you can see, you are responsible for allocating the memory for your	838	As you can see, you are responsible for allocating the memory for your
677	watcher structures (and it is usually a bad idea to do this on the stack,	839	watcher structures (and it is I<usually> a bad idea to do this on the
678	although this can sometimes be quite valid).	840	stack).
		841
		842	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		843	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
679		844
680	Each watcher structure must be initialised by a call to C<ev_init	845	Each watcher structure must be initialised by a call to C<ev_init
681	(watcher *, callback)>, which expects a callback to be provided. This	846	(watcher *, callback)>, which expects a callback to be provided. This
682	callback gets invoked each time the event occurs (or, in the case of io	847	callback gets invoked each time the event occurs (or, in the case of I/O
683	watchers, each time the event loop detects that the file descriptor given	848	watchers, each time the event loop detects that the file descriptor given
684	is readable and/or writable).	849	is readable and/or writable).
685		850
686	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	851	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
687	with arguments specific to this watcher type. There is also a macro	852	macro to configure it, with arguments specific to the watcher type. There
688	to combine initialisation and setting in one call: C<< ev_<type>_init	853	is also a macro to combine initialisation and setting in one call: C<<
689	(watcher *, callback, ...) >>.	854	ev_TYPE_init (watcher *, callback, ...) >>.
690		855
691	To make the watcher actually watch out for events, you have to start it	856	To make the watcher actually watch out for events, you have to start it
692	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	857	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
693	*) >>), and you can stop watching for events at any time by calling the	858	*) >>), and you can stop watching for events at any time by calling the
694	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	859	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
695		860
696	As long as your watcher is active (has been started but not stopped) you	861	As long as your watcher is active (has been started but not stopped) you
697	must not touch the values stored in it. Most specifically you must never	862	must not touch the values stored in it. Most specifically you must never
698	reinitialise it or call its C<set> macro.	863	reinitialise it or call its C<ev_TYPE_set> macro.
699		864
700	Each and every callback receives the event loop pointer as first, the	865	Each and every callback receives the event loop pointer as first, the
701	registered watcher structure as second, and a bitset of received events as	866	registered watcher structure as second, and a bitset of received events as
702	third argument.	867	third argument.
703		868
…		…
757	=item C<EV_FORK>	922	=item C<EV_FORK>
758		923
759	The event loop has been resumed in the child process after fork (see	924	The event loop has been resumed in the child process after fork (see
760	C<ev_fork>).	925	C<ev_fork>).
761		926
		927	=item C<EV_ASYNC>
		928
		929	The given async watcher has been asynchronously notified (see C<ev_async>).
		930
762	=item C<EV_ERROR>	931	=item C<EV_ERROR>
763		932
764	An unspecified error has occured, the watcher has been stopped. This might	933	An unspecified error has occurred, the watcher has been stopped. This might
765	happen because the watcher could not be properly started because libev	934	happen because the watcher could not be properly started because libev
766	ran out of memory, a file descriptor was found to be closed or any other	935	ran out of memory, a file descriptor was found to be closed or any other
		936	problem. Libev considers these application bugs.
		937
767	problem. You best act on it by reporting the problem and somehow coping	938	You best act on it by reporting the problem and somehow coping with the
768	with the watcher being stopped.	939	watcher being stopped. Note that well-written programs should not receive
		940	an error ever, so when your watcher receives it, this usually indicates a
		941	bug in your program.
769		942
770	Libev will usually signal a few "dummy" events together with an error,	943	Libev will usually signal a few "dummy" events together with an error, for
771	for example it might indicate that a fd is readable or writable, and if	944	example it might indicate that a fd is readable or writable, and if your
772	your callbacks is well-written it can just attempt the operation and cope	945	callbacks is well-written it can just attempt the operation and cope with
773	with the error from read() or write(). This will not work in multithreaded	946	the error from read() or write(). This will not work in multi-threaded
774	programs, though, so beware.	947	programs, though, as the fd could already be closed and reused for another
		948	thing, so beware.
775		949
776	=back	950	=back
777		951
778	=head2 GENERIC WATCHER FUNCTIONS	952	=head2 GENERIC WATCHER FUNCTIONS
779
780	In the following description, C<TYPE> stands for the watcher type,
781	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
782		953
783	=over 4	954	=over 4
784		955
785	=item C<ev_init> (ev_TYPE *watcher, callback)	956	=item C<ev_init> (ev_TYPE *watcher, callback)
786		957
…		…
792	which rolls both calls into one.	963	which rolls both calls into one.
793		964
794	You can reinitialise a watcher at any time as long as it has been stopped	965	You can reinitialise a watcher at any time as long as it has been stopped
795	(or never started) and there are no pending events outstanding.	966	(or never started) and there are no pending events outstanding.
796		967
797	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	968	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
798	int revents)>.	969	int revents)>.
		970
		971	Example: Initialise an C<ev_io> watcher in two steps.
		972
		973	ev_io w;
		974	ev_init (&w, my_cb);
		975	ev_io_set (&w, STDIN_FILENO, EV_READ);
799		976
800	=item C<ev_TYPE_set> (ev_TYPE *, [args])	977	=item C<ev_TYPE_set> (ev_TYPE *, [args])
801		978
802	This macro initialises the type-specific parts of a watcher. You need to	979	This macro initialises the type-specific parts of a watcher. You need to
803	call C<ev_init> at least once before you call this macro, but you can	980	call C<ev_init> at least once before you call this macro, but you can
…		…
806	difference to the C<ev_init> macro).	983	difference to the C<ev_init> macro).
807		984
808	Although some watcher types do not have type-specific arguments	985	Although some watcher types do not have type-specific arguments
809	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	986	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
810		987
		988	See C<ev_init>, above, for an example.
		989
811	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	990	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
812		991
813	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	992	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
814	calls into a single call. This is the most convinient method to initialise	993	calls into a single call. This is the most convenient method to initialise
815	a watcher. The same limitations apply, of course.	994	a watcher. The same limitations apply, of course.
		995
		996	Example: Initialise and set an C<ev_io> watcher in one step.
		997
		998	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
816		999
817	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1000	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
818		1001
819	Starts (activates) the given watcher. Only active watchers will receive	1002	Starts (activates) the given watcher. Only active watchers will receive
820	events. If the watcher is already active nothing will happen.	1003	events. If the watcher is already active nothing will happen.
821		1004
		1005	Example: Start the C<ev_io> watcher that is being abused as example in this
		1006	whole section.
		1007
		1008	ev_io_start (EV_DEFAULT_UC, &w);
		1009
822	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1010	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
823		1011
824	Stops the given watcher again (if active) and clears the pending	1012	Stops the given watcher if active, and clears the pending status (whether
		1013	the watcher was active or not).
		1014
825	status. It is possible that stopped watchers are pending (for example,	1015	It is possible that stopped watchers are pending - for example,
826	non-repeating timers are being stopped when they become pending), but	1016	non-repeating timers are being stopped when they become pending - but
827	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1017	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
828	you want to free or reuse the memory used by the watcher it is therefore a	1018	pending. If you want to free or reuse the memory used by the watcher it is
829	good idea to always call its C<ev_TYPE_stop> function.	1019	therefore a good idea to always call its C<ev_TYPE_stop> function.
830		1020
831	=item bool ev_is_active (ev_TYPE *watcher)	1021	=item bool ev_is_active (ev_TYPE *watcher)
832		1022
833	Returns a true value iff the watcher is active (i.e. it has been started	1023	Returns a true value iff the watcher is active (i.e. it has been started
834	and not yet been stopped). As long as a watcher is active you must not modify	1024	and not yet been stopped). As long as a watcher is active you must not modify
…		…
876	The default priority used by watchers when no priority has been set is	1066	The default priority used by watchers when no priority has been set is
877	always C<0>, which is supposed to not be too high and not be too low :).	1067	always C<0>, which is supposed to not be too high and not be too low :).
878		1068
879	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1069	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
880	fine, as long as you do not mind that the priority value you query might	1070	fine, as long as you do not mind that the priority value you query might
881	or might not have been adjusted to be within valid range.	1071	or might not have been clamped to the valid range.
882		1072
883	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1073	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
884		1074
885	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1075	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
886	C<loop> nor C<revents> need to be valid as long as the watcher callback	1076	C<loop> nor C<revents> need to be valid as long as the watcher callback
887	can deal with that fact.	1077	can deal with that fact, as both are simply passed through to the
		1078	callback.
888		1079
889	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1080	=item int ev_clear_pending (loop, ev_TYPE *watcher)
890		1081
891	If the watcher is pending, this function returns clears its pending status	1082	If the watcher is pending, this function clears its pending status and
892	and returns its C<revents> bitset (as if its callback was invoked). If the	1083	returns its C<revents> bitset (as if its callback was invoked). If the
893	watcher isn't pending it does nothing and returns C<0>.	1084	watcher isn't pending it does nothing and returns C<0>.
894		1085
		1086	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1087	callback to be invoked, which can be accomplished with this function.
		1088
895	=back	1089	=back
896		1090
897		1091
898	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1092	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
899		1093
900	Each watcher has, by default, a member C<void *data> that you can change	1094	Each watcher has, by default, a member C<void *data> that you can change
901	and read at any time, libev will completely ignore it. This can be used	1095	and read at any time: libev will completely ignore it. This can be used
902	to associate arbitrary data with your watcher. If you need more data and	1096	to associate arbitrary data with your watcher. If you need more data and
903	don't want to allocate memory and store a pointer to it in that data	1097	don't want to allocate memory and store a pointer to it in that data
904	member, you can also "subclass" the watcher type and provide your own	1098	member, you can also "subclass" the watcher type and provide your own
905	data:	1099	data:
906		1100
907	struct my_io	1101	struct my_io
908	{	1102	{
909	struct ev_io io;	1103	ev_io io;
910	int otherfd;	1104	int otherfd;
911	void *somedata;	1105	void *somedata;
912	struct whatever *mostinteresting;	1106	struct whatever *mostinteresting;
913	}	1107	};
		1108
		1109	...
		1110	struct my_io w;
		1111	ev_io_init (&w.io, my_cb, fd, EV_READ);
914		1112
915	And since your callback will be called with a pointer to the watcher, you	1113	And since your callback will be called with a pointer to the watcher, you
916	can cast it back to your own type:	1114	can cast it back to your own type:
917		1115
918	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1116	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
919	{	1117	{
920	struct my_io *w = (struct my_io *)w_;	1118	struct my_io *w = (struct my_io *)w_;
921	...	1119	...
922	}	1120	}
923		1121
924	More interesting and less C-conformant ways of casting your callback type	1122	More interesting and less C-conformant ways of casting your callback type
925	instead have been omitted.	1123	instead have been omitted.
926		1124
927	Another common scenario is having some data structure with multiple	1125	Another common scenario is to use some data structure with multiple
928	watchers:	1126	embedded watchers:
929		1127
930	struct my_biggy	1128	struct my_biggy
931	{	1129	{
932	int some_data;	1130	int some_data;
933	ev_timer t1;	1131	ev_timer t1;
934	ev_timer t2;	1132	ev_timer t2;
935	}	1133	}
936		1134
937	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1135	In this case getting the pointer to C<my_biggy> is a bit more
938	you need to use C<offsetof>:	1136	complicated: Either you store the address of your C<my_biggy> struct
		1137	in the C<data> member of the watcher (for woozies), or you need to use
		1138	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1139	programmers):
939		1140
940	#include <stddef.h>	1141	#include <stddef.h>
941		1142
942	static void	1143	static void
943	t1_cb (EV_P_ struct ev_timer *w, int revents)	1144	t1_cb (EV_P_ ev_timer *w, int revents)
944	{	1145	{
945	struct my_biggy big = (struct my_biggy *	1146	struct my_biggy big = (struct my_biggy *
946	(((char *)w) - offsetof (struct my_biggy, t1));	1147	(((char *)w) - offsetof (struct my_biggy, t1));
947	}	1148	}
948		1149
949	static void	1150	static void
950	t2_cb (EV_P_ struct ev_timer *w, int revents)	1151	t2_cb (EV_P_ ev_timer *w, int revents)
951	{	1152	{
952	struct my_biggy big = (struct my_biggy *	1153	struct my_biggy big = (struct my_biggy *
953	(((char *)w) - offsetof (struct my_biggy, t2));	1154	(((char *)w) - offsetof (struct my_biggy, t2));
954	}	1155	}
955		1156
956		1157
957	=head1 WATCHER TYPES	1158	=head1 WATCHER TYPES
958		1159
959	This section describes each watcher in detail, but will not repeat	1160	This section describes each watcher in detail, but will not repeat
…		…
983	In general you can register as many read and/or write event watchers per	1184	In general you can register as many read and/or write event watchers per
984	fd as you want (as long as you don't confuse yourself). Setting all file	1185	fd as you want (as long as you don't confuse yourself). Setting all file
985	descriptors to non-blocking mode is also usually a good idea (but not	1186	descriptors to non-blocking mode is also usually a good idea (but not
986	required if you know what you are doing).	1187	required if you know what you are doing).
987		1188
988	If you must do this, then force the use of a known-to-be-good backend	1189	If you cannot use non-blocking mode, then force the use of a
989	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1190	known-to-be-good backend (at the time of this writing, this includes only
990	C<EVBACKEND_POLL>).	1191	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>).
991		1192
992	Another thing you have to watch out for is that it is quite easy to	1193	Another thing you have to watch out for is that it is quite easy to
993	receive "spurious" readyness notifications, that is your callback might	1194	receive "spurious" readiness notifications, that is your callback might
994	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1195	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
995	because there is no data. Not only are some backends known to create a	1196	because there is no data. Not only are some backends known to create a
996	lot of those (for example solaris ports), it is very easy to get into	1197	lot of those (for example Solaris ports), it is very easy to get into
997	this situation even with a relatively standard program structure. Thus	1198	this situation even with a relatively standard program structure. Thus
998	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1199	it is best to always use non-blocking I/O: An extra C<read>(2) returning
999	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1200	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1000		1201
1001	If you cannot run the fd in non-blocking mode (for example you should not	1202	If you cannot run the fd in non-blocking mode (for example you should
1002	play around with an Xlib connection), then you have to seperately re-test	1203	not play around with an Xlib connection), then you have to separately
1003	whether a file descriptor is really ready with a known-to-be good interface	1204	re-test whether a file descriptor is really ready with a known-to-be good
1004	such as poll (fortunately in our Xlib example, Xlib already does this on	1205	interface such as poll (fortunately in our Xlib example, Xlib already
1005	its own, so its quite safe to use).	1206	does this on its own, so its quite safe to use). Some people additionally
		1207	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1208	indefinitely.
		1209
		1210	But really, best use non-blocking mode.
1006		1211
1007	=head3 The special problem of disappearing file descriptors	1212	=head3 The special problem of disappearing file descriptors
1008		1213
1009	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1214	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1010	descriptor (either by calling C<close> explicitly or by any other means,	1215	descriptor (either due to calling C<close> explicitly or any other means,
1011	such as C<dup>). The reason is that you register interest in some file	1216	such as C<dup2>). The reason is that you register interest in some file
1012	descriptor, but when it goes away, the operating system will silently drop	1217	descriptor, but when it goes away, the operating system will silently drop
1013	this interest. If another file descriptor with the same number then is	1218	this interest. If another file descriptor with the same number then is
1014	registered with libev, there is no efficient way to see that this is, in	1219	registered with libev, there is no efficient way to see that this is, in
1015	fact, a different file descriptor.	1220	fact, a different file descriptor.
1016		1221
…		…
1045	To support fork in your programs, you either have to call	1250	To support fork in your programs, you either have to call
1046	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1251	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
1047	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1252	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1048	C<EVBACKEND_POLL>.	1253	C<EVBACKEND_POLL>.
1049		1254
		1255	=head3 The special problem of SIGPIPE
		1256
		1257	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1258	when writing to a pipe whose other end has been closed, your program gets
		1259	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1260	this is sensible behaviour, for daemons, this is usually undesirable.
		1261
		1262	So when you encounter spurious, unexplained daemon exits, make sure you
		1263	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1264	somewhere, as that would have given you a big clue).
		1265
1050		1266
1051	=head3 Watcher-Specific Functions	1267	=head3 Watcher-Specific Functions
1052		1268
1053	=over 4	1269	=over 4
1054		1270
1055	=item ev_io_init (ev_io *, callback, int fd, int events)	1271	=item ev_io_init (ev_io *, callback, int fd, int events)
1056		1272
1057	=item ev_io_set (ev_io *, int fd, int events)	1273	=item ev_io_set (ev_io *, int fd, int events)
1058		1274
1059	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1275	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1060	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or	1276	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1061	C<EV_READ | EV_WRITE> to receive the given events.	1277	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1062		1278
1063	=item int fd [read-only]	1279	=item int fd [read-only]
1064		1280
1065	The file descriptor being watched.	1281	The file descriptor being watched.
1066		1282
1067	=item int events [read-only]	1283	=item int events [read-only]
1068		1284
1069	The events being watched.	1285	The events being watched.
1070		1286
1071	=back	1287	=back
		1288
		1289	=head3 Examples
1072		1290
1073	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1291	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1074	readable, but only once. Since it is likely line-buffered, you could	1292	readable, but only once. Since it is likely line-buffered, you could
1075	attempt to read a whole line in the callback.	1293	attempt to read a whole line in the callback.
1076		1294
1077	static void	1295	static void
1078	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1296	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1079	{	1297	{
1080	ev_io_stop (loop, w);	1298	ev_io_stop (loop, w);
1081	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1299	.. read from stdin here (or from w->fd) and handle any I/O errors
1082	}	1300	}
1083		1301
1084	...	1302	...
1085	struct ev_loop *loop = ev_default_init (0);	1303	struct ev_loop *loop = ev_default_init (0);
1086	struct ev_io stdin_readable;	1304	ev_io stdin_readable;
1087	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1305	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1088	ev_io_start (loop, &stdin_readable);	1306	ev_io_start (loop, &stdin_readable);
1089	ev_loop (loop, 0);	1307	ev_loop (loop, 0);
1090		1308
1091		1309
1092	=head2 C<ev_timer> - relative and optionally repeating timeouts	1310	=head2 C<ev_timer> - relative and optionally repeating timeouts
1093		1311
1094	Timer watchers are simple relative timers that generate an event after a	1312	Timer watchers are simple relative timers that generate an event after a
1095	given time, and optionally repeating in regular intervals after that.	1313	given time, and optionally repeating in regular intervals after that.
1096		1314
1097	The timers are based on real time, that is, if you register an event that	1315	The timers are based on real time, that is, if you register an event that
1098	times out after an hour and you reset your system clock to last years	1316	times out after an hour and you reset your system clock to January last
1099	time, it will still time out after (roughly) and hour. "Roughly" because	1317	year, it will still time out after (roughly) one hour. "Roughly" because
1100	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1318	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1101	monotonic clock option helps a lot here).	1319	monotonic clock option helps a lot here).
		1320
		1321	The callback is guaranteed to be invoked only I<after> its timeout has
		1322	passed, but if multiple timers become ready during the same loop iteration
		1323	then order of execution is undefined.
		1324
		1325	=head3 Be smart about timeouts
		1326
		1327	Many real-world problems involve some kind of timeout, usually for error
		1328	recovery. A typical example is an HTTP request - if the other side hangs,
		1329	you want to raise some error after a while.
		1330
		1331	What follows are some ways to handle this problem, from obvious and
		1332	inefficient to smart and efficient.
		1333
		1334	In the following, a 60 second activity timeout is assumed - a timeout that
		1335	gets reset to 60 seconds each time there is activity (e.g. each time some
		1336	data or other life sign was received).
		1337
		1338	=over 4
		1339
		1340	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1341
		1342	This is the most obvious, but not the most simple way: In the beginning,
		1343	start the watcher:
		1344
		1345	ev_timer_init (timer, callback, 60., 0.);
		1346	ev_timer_start (loop, timer);
		1347
		1348	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1349	and start it again:
		1350
		1351	ev_timer_stop (loop, timer);
		1352	ev_timer_set (timer, 60., 0.);
		1353	ev_timer_start (loop, timer);
		1354
		1355	This is relatively simple to implement, but means that each time there is
		1356	some activity, libev will first have to remove the timer from its internal
		1357	data structure and then add it again. Libev tries to be fast, but it's
		1358	still not a constant-time operation.
		1359
		1360	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1361
		1362	This is the easiest way, and involves using C<ev_timer_again> instead of
		1363	C<ev_timer_start>.
		1364
		1365	To implement this, configure an C<ev_timer> with a C<repeat> value
		1366	of C<60> and then call C<ev_timer_again> at start and each time you
		1367	successfully read or write some data. If you go into an idle state where
		1368	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1369	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1370
		1371	That means you can ignore both the C<ev_timer_start> function and the
		1372	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1373	member and C<ev_timer_again>.
		1374
		1375	At start:
		1376
		1377	ev_timer_init (timer, callback);
		1378	timer->repeat = 60.;
		1379	ev_timer_again (loop, timer);
		1380
		1381	Each time there is some activity:
		1382
		1383	ev_timer_again (loop, timer);
		1384
		1385	It is even possible to change the time-out on the fly, regardless of
		1386	whether the watcher is active or not:
		1387
		1388	timer->repeat = 30.;
		1389	ev_timer_again (loop, timer);
		1390
		1391	This is slightly more efficient then stopping/starting the timer each time
		1392	you want to modify its timeout value, as libev does not have to completely
		1393	remove and re-insert the timer from/into its internal data structure.
		1394
		1395	It is, however, even simpler than the "obvious" way to do it.
		1396
		1397	=item 3. Let the timer time out, but then re-arm it as required.
		1398
		1399	This method is more tricky, but usually most efficient: Most timeouts are
		1400	relatively long compared to the intervals between other activity - in
		1401	our example, within 60 seconds, there are usually many I/O events with
		1402	associated activity resets.
		1403
		1404	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1405	but remember the time of last activity, and check for a real timeout only
		1406	within the callback:
		1407
		1408	ev_tstamp last_activity; // time of last activity
		1409
		1410	static void
		1411	callback (EV_P_ ev_timer *w, int revents)
		1412	{
		1413	ev_tstamp now = ev_now (EV_A);
		1414	ev_tstamp timeout = last_activity + 60.;
		1415
		1416	// if last_activity + 60. is older than now, we did time out
		1417	if (timeout < now)
		1418	{
		1419	// timeout occured, take action
		1420	}
		1421	else
		1422	{
		1423	// callback was invoked, but there was some activity, re-arm
		1424	// the watcher to fire in last_activity + 60, which is
		1425	// guaranteed to be in the future, so "again" is positive:
		1426	w->repeat = timeout - now;
		1427	ev_timer_again (EV_A_ w);
		1428	}
		1429	}
		1430
		1431	To summarise the callback: first calculate the real timeout (defined
		1432	as "60 seconds after the last activity"), then check if that time has
		1433	been reached, which means something I<did>, in fact, time out. Otherwise
		1434	the callback was invoked too early (C<timeout> is in the future), so
		1435	re-schedule the timer to fire at that future time, to see if maybe we have
		1436	a timeout then.
		1437
		1438	Note how C<ev_timer_again> is used, taking advantage of the
		1439	C<ev_timer_again> optimisation when the timer is already running.
		1440
		1441	This scheme causes more callback invocations (about one every 60 seconds
		1442	minus half the average time between activity), but virtually no calls to
		1443	libev to change the timeout.
		1444
		1445	To start the timer, simply initialise the watcher and set C<last_activity>
		1446	to the current time (meaning we just have some activity :), then call the
		1447	callback, which will "do the right thing" and start the timer:
		1448
		1449	ev_timer_init (timer, callback);
		1450	last_activity = ev_now (loop);
		1451	callback (loop, timer, EV_TIMEOUT);
		1452
		1453	And when there is some activity, simply store the current time in
		1454	C<last_activity>, no libev calls at all:
		1455
		1456	last_actiivty = ev_now (loop);
		1457
		1458	This technique is slightly more complex, but in most cases where the
		1459	time-out is unlikely to be triggered, much more efficient.
		1460
		1461	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1462	callback :) - just change the timeout and invoke the callback, which will
		1463	fix things for you.
		1464
		1465	=item 4. Wee, just use a double-linked list for your timeouts.
		1466
		1467	If there is not one request, but many thousands (millions...), all
		1468	employing some kind of timeout with the same timeout value, then one can
		1469	do even better:
		1470
		1471	When starting the timeout, calculate the timeout value and put the timeout
		1472	at the I<end> of the list.
		1473
		1474	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1475	the list is expected to fire (for example, using the technique #3).
		1476
		1477	When there is some activity, remove the timer from the list, recalculate
		1478	the timeout, append it to the end of the list again, and make sure to
		1479	update the C<ev_timer> if it was taken from the beginning of the list.
		1480
		1481	This way, one can manage an unlimited number of timeouts in O(1) time for
		1482	starting, stopping and updating the timers, at the expense of a major
		1483	complication, and having to use a constant timeout. The constant timeout
		1484	ensures that the list stays sorted.
		1485
		1486	=back
		1487
		1488	So which method the best?
		1489
		1490	Method #2 is a simple no-brain-required solution that is adequate in most
		1491	situations. Method #3 requires a bit more thinking, but handles many cases
		1492	better, and isn't very complicated either. In most case, choosing either
		1493	one is fine, with #3 being better in typical situations.
		1494
		1495	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1496	rather complicated, but extremely efficient, something that really pays
		1497	off after the first million or so of active timers, i.e. it's usually
		1498	overkill :)
		1499
		1500	=head3 The special problem of time updates
		1501
		1502	Establishing the current time is a costly operation (it usually takes at
		1503	least two system calls): EV therefore updates its idea of the current
		1504	time only before and after C<ev_loop> collects new events, which causes a
		1505	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1506	lots of events in one iteration.
1102		1507
1103	The relative timeouts are calculated relative to the C<ev_now ()>	1508	The relative timeouts are calculated relative to the C<ev_now ()>
1104	time. This is usually the right thing as this timestamp refers to the time	1509	time. This is usually the right thing as this timestamp refers to the time
1105	of the event triggering whatever timeout you are modifying/starting. If	1510	of the event triggering whatever timeout you are modifying/starting. If
1106	you suspect event processing to be delayed and you I<need> to base the timeout	1511	you suspect event processing to be delayed and you I<need> to base the
1107	on the current time, use something like this to adjust for this:	1512	timeout on the current time, use something like this to adjust for this:
1108		1513
1109	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1514	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1110		1515
1111	The callback is guarenteed to be invoked only when its timeout has passed,	1516	If the event loop is suspended for a long time, you can also force an
1112	but if multiple timers become ready during the same loop iteration then	1517	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1113	order of execution is undefined.	1518	()>.
1114		1519
1115	=head3 Watcher-Specific Functions and Data Members	1520	=head3 Watcher-Specific Functions and Data Members
1116		1521
1117	=over 4	1522	=over 4
1118		1523
1119	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1524	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1120		1525
1121	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1526	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1122		1527
1123	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1528	Configure the timer to trigger after C<after> seconds. If C<repeat>
1124	C<0.>, then it will automatically be stopped. If it is positive, then the	1529	is C<0.>, then it will automatically be stopped once the timeout is
1125	timer will automatically be configured to trigger again C<repeat> seconds	1530	reached. If it is positive, then the timer will automatically be
1126	later, again, and again, until stopped manually.	1531	configured to trigger again C<repeat> seconds later, again, and again,
		1532	until stopped manually.
1127		1533
1128	The timer itself will do a best-effort at avoiding drift, that is, if you	1534	The timer itself will do a best-effort at avoiding drift, that is, if
1129	configure a timer to trigger every 10 seconds, then it will trigger at	1535	you configure a timer to trigger every 10 seconds, then it will normally
1130	exactly 10 second intervals. If, however, your program cannot keep up with	1536	trigger at exactly 10 second intervals. If, however, your program cannot
1131	the timer (because it takes longer than those 10 seconds to do stuff) the	1537	keep up with the timer (because it takes longer than those 10 seconds to
1132	timer will not fire more than once per event loop iteration.	1538	do stuff) the timer will not fire more than once per event loop iteration.
1133		1539
1134	=item ev_timer_again (loop)	1540	=item ev_timer_again (loop, ev_timer *)
1135		1541
1136	This will act as if the timer timed out and restart it again if it is	1542	This will act as if the timer timed out and restart it again if it is
1137	repeating. The exact semantics are:	1543	repeating. The exact semantics are:
1138		1544
1139	If the timer is pending, its pending status is cleared.	1545	If the timer is pending, its pending status is cleared.
1140		1546
1141	If the timer is started but nonrepeating, stop it (as if it timed out).	1547	If the timer is started but non-repeating, stop it (as if it timed out).
1142		1548
1143	If the timer is repeating, either start it if necessary (with the	1549	If the timer is repeating, either start it if necessary (with the
1144	C<repeat> value), or reset the running timer to the C<repeat> value.	1550	C<repeat> value), or reset the running timer to the C<repeat> value.
1145		1551
1146	This sounds a bit complicated, but here is a useful and typical	1552	This sounds a bit complicated, see "Be smart about timeouts", above, for a
1147	example: Imagine you have a tcp connection and you want a so-called idle	1553	usage example.
1148	timeout, that is, you want to be called when there have been, say, 60
1149	seconds of inactivity on the socket. The easiest way to do this is to
1150	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1151	C<ev_timer_again> each time you successfully read or write some data. If
1152	you go into an idle state where you do not expect data to travel on the
1153	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1154	automatically restart it if need be.
1155
1156	That means you can ignore the C<after> value and C<ev_timer_start>
1157	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1158
1159	ev_timer_init (timer, callback, 0., 5.);
1160	ev_timer_again (loop, timer);
1161	...
1162	timer->again = 17.;
1163	ev_timer_again (loop, timer);
1164	...
1165	timer->again = 10.;
1166	ev_timer_again (loop, timer);
1167
1168	This is more slightly efficient then stopping/starting the timer each time
1169	you want to modify its timeout value.
1170		1554
1171	=item ev_tstamp repeat [read-write]	1555	=item ev_tstamp repeat [read-write]
1172		1556
1173	The current C<repeat> value. Will be used each time the watcher times out	1557	The current C<repeat> value. Will be used each time the watcher times out
1174	or C<ev_timer_again> is called and determines the next timeout (if any),	1558	or C<ev_timer_again> is called, and determines the next timeout (if any),
1175	which is also when any modifications are taken into account.	1559	which is also when any modifications are taken into account.
1176		1560
1177	=back	1561	=back
1178		1562
		1563	=head3 Examples
		1564
1179	Example: Create a timer that fires after 60 seconds.	1565	Example: Create a timer that fires after 60 seconds.
1180		1566
1181	static void	1567	static void
1182	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1568	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1183	{	1569	{
1184	.. one minute over, w is actually stopped right here	1570	.. one minute over, w is actually stopped right here
1185	}	1571	}
1186		1572
1187	struct ev_timer mytimer;	1573	ev_timer mytimer;
1188	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1574	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1189	ev_timer_start (loop, &mytimer);	1575	ev_timer_start (loop, &mytimer);
1190		1576
1191	Example: Create a timeout timer that times out after 10 seconds of	1577	Example: Create a timeout timer that times out after 10 seconds of
1192	inactivity.	1578	inactivity.
1193		1579
1194	static void	1580	static void
1195	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1581	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1196	{	1582	{
1197	.. ten seconds without any activity	1583	.. ten seconds without any activity
1198	}	1584	}
1199		1585
1200	struct ev_timer mytimer;	1586	ev_timer mytimer;
1201	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1587	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1202	ev_timer_again (&mytimer); /* start timer */	1588	ev_timer_again (&mytimer); /* start timer */
1203	ev_loop (loop, 0);	1589	ev_loop (loop, 0);
1204		1590
1205	// and in some piece of code that gets executed on any "activity":	1591	// and in some piece of code that gets executed on any "activity":
1206	// reset the timeout to start ticking again at 10 seconds	1592	// reset the timeout to start ticking again at 10 seconds
1207	ev_timer_again (&mytimer);	1593	ev_timer_again (&mytimer);
1208		1594
1209		1595
1210	=head2 C<ev_periodic> - to cron or not to cron?	1596	=head2 C<ev_periodic> - to cron or not to cron?
1211		1597
1212	Periodic watchers are also timers of a kind, but they are very versatile	1598	Periodic watchers are also timers of a kind, but they are very versatile
1213	(and unfortunately a bit complex).	1599	(and unfortunately a bit complex).
1214		1600
1215	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1601	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
1216	but on wallclock time (absolute time). You can tell a periodic watcher	1602	but on wall clock time (absolute time). You can tell a periodic watcher
1217	to trigger "at" some specific point in time. For example, if you tell a	1603	to trigger after some specific point in time. For example, if you tell a
1218	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1604	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()
1219	+ 10.>) and then reset your system clock to the last year, then it will	1605	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1606	clock to January of the previous year, then it will take more than year
1220	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1607	to trigger the event (unlike an C<ev_timer>, which would still trigger
1221	roughly 10 seconds later).	1608	roughly 10 seconds later as it uses a relative timeout).
1222		1609
1223	They can also be used to implement vastly more complex timers, such as	1610	C<ev_periodic>s can also be used to implement vastly more complex timers,
1224	triggering an event on each midnight, local time or other, complicated,	1611	such as triggering an event on each "midnight, local time", or other
1225	rules.	1612	complicated rules.
1226		1613
1227	As with timers, the callback is guarenteed to be invoked only when the	1614	As with timers, the callback is guaranteed to be invoked only when the
1228	time (C<at>) has been passed, but if multiple periodic timers become ready	1615	time (C<at>) has passed, but if multiple periodic timers become ready
1229	during the same loop iteration then order of execution is undefined.	1616	during the same loop iteration, then order of execution is undefined.
1230		1617
1231	=head3 Watcher-Specific Functions and Data Members	1618	=head3 Watcher-Specific Functions and Data Members
1232		1619
1233	=over 4	1620	=over 4
1234		1621
1235	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1622	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
1236		1623
1237	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1624	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
1238		1625
1239	Lots of arguments, lets sort it out... There are basically three modes of	1626	Lots of arguments, lets sort it out... There are basically three modes of
1240	operation, and we will explain them from simplest to complex:	1627	operation, and we will explain them from simplest to most complex:
1241		1628
1242	=over 4	1629	=over 4
1243		1630
1244	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1631	=item * absolute timer (at = time, interval = reschedule_cb = 0)
1245		1632
1246	In this configuration the watcher triggers an event at the wallclock time	1633	In this configuration the watcher triggers an event after the wall clock
1247	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1634	time C<at> has passed. It will not repeat and will not adjust when a time
1248	that is, if it is to be run at January 1st 2011 then it will run when the	1635	jump occurs, that is, if it is to be run at January 1st 2011 then it will
1249	system time reaches or surpasses this time.	1636	only run when the system clock reaches or surpasses this time.
1250		1637
1251	=item * non-repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1638	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
1252		1639
1253	In this mode the watcher will always be scheduled to time out at the next	1640	In this mode the watcher will always be scheduled to time out at the next
1254	C<at + N * interval> time (for some integer N, which can also be negative)	1641	C<at + N * interval> time (for some integer N, which can also be negative)
1255	and then repeat, regardless of any time jumps.	1642	and then repeat, regardless of any time jumps.
1256		1643
1257	This can be used to create timers that do not drift with respect to system	1644	This can be used to create timers that do not drift with respect to the
1258	time:	1645	system clock, for example, here is a C<ev_periodic> that triggers each
		1646	hour, on the hour:
1259		1647
1260	ev_periodic_set (&periodic, 0., 3600., 0);	1648	ev_periodic_set (&periodic, 0., 3600., 0);
1261		1649
1262	This doesn't mean there will always be 3600 seconds in between triggers,	1650	This doesn't mean there will always be 3600 seconds in between triggers,
1263	but only that the the callback will be called when the system time shows a	1651	but only that the callback will be called when the system time shows a
1264	full hour (UTC), or more correctly, when the system time is evenly divisible	1652	full hour (UTC), or more correctly, when the system time is evenly divisible
1265	by 3600.	1653	by 3600.
1266		1654
1267	Another way to think about it (for the mathematically inclined) is that	1655	Another way to think about it (for the mathematically inclined) is that
1268	C<ev_periodic> will try to run the callback in this mode at the next possible	1656	C<ev_periodic> will try to run the callback in this mode at the next possible
1269	time where C<time = at (mod interval)>, regardless of any time jumps.	1657	time where C<time = at (mod interval)>, regardless of any time jumps.
1270		1658
1271	For numerical stability it is preferable that the C<at> value is near	1659	For numerical stability it is preferable that the C<at> value is near
1272	C<ev_now ()> (the current time), but there is no range requirement for	1660	C<ev_now ()> (the current time), but there is no range requirement for
1273	this value.	1661	this value, and in fact is often specified as zero.
		1662
		1663	Note also that there is an upper limit to how often a timer can fire (CPU
		1664	speed for example), so if C<interval> is very small then timing stability
		1665	will of course deteriorate. Libev itself tries to be exact to be about one
		1666	millisecond (if the OS supports it and the machine is fast enough).
1274		1667
1275	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1668	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1276		1669
1277	In this mode the values for C<interval> and C<at> are both being	1670	In this mode the values for C<interval> and C<at> are both being
1278	ignored. Instead, each time the periodic watcher gets scheduled, the	1671	ignored. Instead, each time the periodic watcher gets scheduled, the
1279	reschedule callback will be called with the watcher as first, and the	1672	reschedule callback will be called with the watcher as first, and the
1280	current time as second argument.	1673	current time as second argument.
1281		1674
1282	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1675	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
1283	ever, or make any event loop modifications>. If you need to stop it,	1676	ever, or make ANY event loop modifications whatsoever>.
1284	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
1285	starting an C<ev_prepare> watcher, which is legal).
1286		1677
		1678	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1679	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1680	only event loop modification you are allowed to do).
		1681
1287	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1682	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1288	ev_tstamp now)>, e.g.:	1683	*w, ev_tstamp now)>, e.g.:
1289		1684
		1685	static ev_tstamp
1290	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1686	my_rescheduler (ev_periodic *w, ev_tstamp now)
1291	{	1687	{
1292	return now + 60.;	1688	return now + 60.;
1293	}	1689	}
1294		1690
1295	It must return the next time to trigger, based on the passed time value	1691	It must return the next time to trigger, based on the passed time value
1296	(that is, the lowest time value larger than to the second argument). It	1692	(that is, the lowest time value larger than to the second argument). It
1297	will usually be called just before the callback will be triggered, but	1693	will usually be called just before the callback will be triggered, but
1298	might be called at other times, too.	1694	might be called at other times, too.
1299		1695
1300	NOTE: I<< This callback must always return a time that is later than the	1696	NOTE: I<< This callback must always return a time that is higher than or
1301	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1697	equal to the passed C<now> value >>.
1302		1698
1303	This can be used to create very complex timers, such as a timer that	1699	This can be used to create very complex timers, such as a timer that
1304	triggers on each midnight, local time. To do this, you would calculate the	1700	triggers on "next midnight, local time". To do this, you would calculate the
1305	next midnight after C<now> and return the timestamp value for this. How	1701	next midnight after C<now> and return the timestamp value for this. How
1306	you do this is, again, up to you (but it is not trivial, which is the main	1702	you do this is, again, up to you (but it is not trivial, which is the main
1307	reason I omitted it as an example).	1703	reason I omitted it as an example).
1308		1704
1309	=back	1705	=back
…		…
1313	Simply stops and restarts the periodic watcher again. This is only useful	1709	Simply stops and restarts the periodic watcher again. This is only useful
1314	when you changed some parameters or the reschedule callback would return	1710	when you changed some parameters or the reschedule callback would return
1315	a different time than the last time it was called (e.g. in a crond like	1711	a different time than the last time it was called (e.g. in a crond like
1316	program when the crontabs have changed).	1712	program when the crontabs have changed).
1317		1713
		1714	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1715
		1716	When active, returns the absolute time that the watcher is supposed to
		1717	trigger next.
		1718
1318	=item ev_tstamp offset [read-write]	1719	=item ev_tstamp offset [read-write]
1319		1720
1320	When repeating, this contains the offset value, otherwise this is the	1721	When repeating, this contains the offset value, otherwise this is the
1321	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	1722	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
1322		1723
…		…
1327		1728
1328	The current interval value. Can be modified any time, but changes only	1729	The current interval value. Can be modified any time, but changes only
1329	take effect when the periodic timer fires or C<ev_periodic_again> is being	1730	take effect when the periodic timer fires or C<ev_periodic_again> is being
1330	called.	1731	called.
1331		1732
1332	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	1733	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1333		1734
1334	The current reschedule callback, or C<0>, if this functionality is	1735	The current reschedule callback, or C<0>, if this functionality is
1335	switched off. Can be changed any time, but changes only take effect when	1736	switched off. Can be changed any time, but changes only take effect when
1336	the periodic timer fires or C<ev_periodic_again> is being called.	1737	the periodic timer fires or C<ev_periodic_again> is being called.
1337		1738
1338	=item ev_tstamp at [read-only]
1339
1340	When active, contains the absolute time that the watcher is supposed to
1341	trigger next.
1342
1343	=back	1739	=back
1344		1740
		1741	=head3 Examples
		1742
1345	Example: Call a callback every hour, or, more precisely, whenever the	1743	Example: Call a callback every hour, or, more precisely, whenever the
1346	system clock is divisible by 3600. The callback invocation times have	1744	system time is divisible by 3600. The callback invocation times have
1347	potentially a lot of jittering, but good long-term stability.	1745	potentially a lot of jitter, but good long-term stability.
1348		1746
1349	static void	1747	static void
1350	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1748	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1351	{	1749	{
1352	... its now a full hour (UTC, or TAI or whatever your clock follows)	1750	... its now a full hour (UTC, or TAI or whatever your clock follows)
1353	}	1751	}
1354		1752
1355	struct ev_periodic hourly_tick;	1753	ev_periodic hourly_tick;
1356	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1754	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1357	ev_periodic_start (loop, &hourly_tick);	1755	ev_periodic_start (loop, &hourly_tick);
1358		1756
1359	Example: The same as above, but use a reschedule callback to do it:	1757	Example: The same as above, but use a reschedule callback to do it:
1360		1758
1361	#include <math.h>	1759	#include <math.h>
1362		1760
1363	static ev_tstamp	1761	static ev_tstamp
1364	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1762	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1365	{	1763	{
1366	return fmod (now, 3600.) + 3600.;	1764	return now + (3600. - fmod (now, 3600.));
1367	}	1765	}
1368		1766
1369	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1767	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1370		1768
1371	Example: Call a callback every hour, starting now:	1769	Example: Call a callback every hour, starting now:
1372		1770
1373	struct ev_periodic hourly_tick;	1771	ev_periodic hourly_tick;
1374	ev_periodic_init (&hourly_tick, clock_cb,	1772	ev_periodic_init (&hourly_tick, clock_cb,
1375	fmod (ev_now (loop), 3600.), 3600., 0);	1773	fmod (ev_now (loop), 3600.), 3600., 0);
1376	ev_periodic_start (loop, &hourly_tick);	1774	ev_periodic_start (loop, &hourly_tick);
1377		1775
1378		1776
1379	=head2 C<ev_signal> - signal me when a signal gets signalled!	1777	=head2 C<ev_signal> - signal me when a signal gets signalled!
1380		1778
1381	Signal watchers will trigger an event when the process receives a specific	1779	Signal watchers will trigger an event when the process receives a specific
1382	signal one or more times. Even though signals are very asynchronous, libev	1780	signal one or more times. Even though signals are very asynchronous, libev
1383	will try it's best to deliver signals synchronously, i.e. as part of the	1781	will try it's best to deliver signals synchronously, i.e. as part of the
1384	normal event processing, like any other event.	1782	normal event processing, like any other event.
1385		1783
		1784	If you want signals asynchronously, just use C<sigaction> as you would
		1785	do without libev and forget about sharing the signal. You can even use
		1786	C<ev_async> from a signal handler to synchronously wake up an event loop.
		1787
1386	You can configure as many watchers as you like per signal. Only when the	1788	You can configure as many watchers as you like per signal. Only when the
1387	first watcher gets started will libev actually register a signal watcher	1789	first watcher gets started will libev actually register a signal handler
1388	with the kernel (thus it coexists with your own signal handlers as long	1790	with the kernel (thus it coexists with your own signal handlers as long as
1389	as you don't register any with libev). Similarly, when the last signal	1791	you don't register any with libev for the same signal). Similarly, when
1390	watcher for a signal is stopped libev will reset the signal handler to	1792	the last signal watcher for a signal is stopped, libev will reset the
1391	SIG_DFL (regardless of what it was set to before).	1793	signal handler to SIG_DFL (regardless of what it was set to before).
		1794
		1795	If possible and supported, libev will install its handlers with
		1796	C<SA_RESTART> behaviour enabled, so system calls should not be unduly
		1797	interrupted. If you have a problem with system calls getting interrupted by
		1798	signals you can block all signals in an C<ev_check> watcher and unblock
		1799	them in an C<ev_prepare> watcher.
1392		1800
1393	=head3 Watcher-Specific Functions and Data Members	1801	=head3 Watcher-Specific Functions and Data Members
1394		1802
1395	=over 4	1803	=over 4
1396		1804
…		…
1405		1813
1406	The signal the watcher watches out for.	1814	The signal the watcher watches out for.
1407		1815
1408	=back	1816	=back
1409		1817
		1818	=head3 Examples
		1819
		1820	Example: Try to exit cleanly on SIGINT.
		1821
		1822	static void
		1823	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		1824	{
		1825	ev_unloop (loop, EVUNLOOP_ALL);
		1826	}
		1827
		1828	ev_signal signal_watcher;
		1829	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1830	ev_signal_start (loop, &signal_watcher);
		1831
1410		1832
1411	=head2 C<ev_child> - watch out for process status changes	1833	=head2 C<ev_child> - watch out for process status changes
1412		1834
1413	Child watchers trigger when your process receives a SIGCHLD in response to	1835	Child watchers trigger when your process receives a SIGCHLD in response to
1414	some child status changes (most typically when a child of yours dies).	1836	some child status changes (most typically when a child of yours dies or
		1837	exits). It is permissible to install a child watcher I<after> the child
		1838	has been forked (which implies it might have already exited), as long
		1839	as the event loop isn't entered (or is continued from a watcher), i.e.,
		1840	forking and then immediately registering a watcher for the child is fine,
		1841	but forking and registering a watcher a few event loop iterations later is
		1842	not.
		1843
		1844	Only the default event loop is capable of handling signals, and therefore
		1845	you can only register child watchers in the default event loop.
		1846
		1847	=head3 Process Interaction
		1848
		1849	Libev grabs C<SIGCHLD> as soon as the default event loop is
		1850	initialised. This is necessary to guarantee proper behaviour even if
		1851	the first child watcher is started after the child exits. The occurrence
		1852	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		1853	synchronously as part of the event loop processing. Libev always reaps all
		1854	children, even ones not watched.
		1855
		1856	=head3 Overriding the Built-In Processing
		1857
		1858	Libev offers no special support for overriding the built-in child
		1859	processing, but if your application collides with libev's default child
		1860	handler, you can override it easily by installing your own handler for
		1861	C<SIGCHLD> after initialising the default loop, and making sure the
		1862	default loop never gets destroyed. You are encouraged, however, to use an
		1863	event-based approach to child reaping and thus use libev's support for
		1864	that, so other libev users can use C<ev_child> watchers freely.
		1865
		1866	=head3 Stopping the Child Watcher
		1867
		1868	Currently, the child watcher never gets stopped, even when the
		1869	child terminates, so normally one needs to stop the watcher in the
		1870	callback. Future versions of libev might stop the watcher automatically
		1871	when a child exit is detected.
1415		1872
1416	=head3 Watcher-Specific Functions and Data Members	1873	=head3 Watcher-Specific Functions and Data Members
1417		1874
1418	=over 4	1875	=over 4
1419		1876
1420	=item ev_child_init (ev_child *, callback, int pid)	1877	=item ev_child_init (ev_child *, callback, int pid, int trace)
1421		1878
1422	=item ev_child_set (ev_child *, int pid)	1879	=item ev_child_set (ev_child *, int pid, int trace)
1423		1880
1424	Configures the watcher to wait for status changes of process C<pid> (or	1881	Configures the watcher to wait for status changes of process C<pid> (or
1425	I<any> process if C<pid> is specified as C<0>). The callback can look	1882	I<any> process if C<pid> is specified as C<0>). The callback can look
1426	at the C<rstatus> member of the C<ev_child> watcher structure to see	1883	at the C<rstatus> member of the C<ev_child> watcher structure to see
1427	the status word (use the macros from C<sys/wait.h> and see your systems	1884	the status word (use the macros from C<sys/wait.h> and see your systems
1428	C<waitpid> documentation). The C<rpid> member contains the pid of the	1885	C<waitpid> documentation). The C<rpid> member contains the pid of the
1429	process causing the status change.	1886	process causing the status change. C<trace> must be either C<0> (only
		1887	activate the watcher when the process terminates) or C<1> (additionally
		1888	activate the watcher when the process is stopped or continued).
1430		1889
1431	=item int pid [read-only]	1890	=item int pid [read-only]
1432		1891
1433	The process id this watcher watches out for, or C<0>, meaning any process id.	1892	The process id this watcher watches out for, or C<0>, meaning any process id.
1434		1893
…		…
1441	The process exit/trace status caused by C<rpid> (see your systems	1900	The process exit/trace status caused by C<rpid> (see your systems
1442	C<waitpid> and C<sys/wait.h> documentation for details).	1901	C<waitpid> and C<sys/wait.h> documentation for details).
1443		1902
1444	=back	1903	=back
1445		1904
1446	Example: Try to exit cleanly on SIGINT and SIGTERM.	1905	=head3 Examples
1447		1906
		1907	Example: C<fork()> a new process and install a child handler to wait for
		1908	its completion.
		1909
		1910	ev_child cw;
		1911
1448	static void	1912	static void
1449	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	1913	child_cb (EV_P_ ev_child *w, int revents)
1450	{	1914	{
1451	ev_unloop (loop, EVUNLOOP_ALL);	1915	ev_child_stop (EV_A_ w);
		1916	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1452	}	1917	}
1453		1918
1454	struct ev_signal signal_watcher;	1919	pid_t pid = fork ();
1455	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	1920
1456	ev_signal_start (loop, &sigint_cb);	1921	if (pid < 0)
		1922	// error
		1923	else if (pid == 0)
		1924	{
		1925	// the forked child executes here
		1926	exit (1);
		1927	}
		1928	else
		1929	{
		1930	ev_child_init (&cw, child_cb, pid, 0);
		1931	ev_child_start (EV_DEFAULT_ &cw);
		1932	}
1457		1933
1458		1934
1459	=head2 C<ev_stat> - did the file attributes just change?	1935	=head2 C<ev_stat> - did the file attributes just change?
1460		1936
1461	This watches a filesystem path for attribute changes. That is, it calls	1937	This watches a file system path for attribute changes. That is, it calls
1462	C<stat> regularly (or when the OS says it changed) and sees if it changed	1938	C<stat> on that path in regular intervals (or when the OS says it changed)
1463	compared to the last time, invoking the callback if it did.	1939	and sees if it changed compared to the last time, invoking the callback if
		1940	it did.
1464		1941
1465	The path does not need to exist: changing from "path exists" to "path does	1942	The path does not need to exist: changing from "path exists" to "path does
1466	not exist" is a status change like any other. The condition "path does	1943	not exist" is a status change like any other. The condition "path does not
1467	not exist" is signified by the C<st_nlink> field being zero (which is	1944	exist" (or more correctly "path cannot be stat'ed") is signified by the
1468	otherwise always forced to be at least one) and all the other fields of	1945	C<st_nlink> field being zero (which is otherwise always forced to be at
1469	the stat buffer having unspecified contents.	1946	least one) and all the other fields of the stat buffer having unspecified
		1947	contents.
1470		1948
1471	The path I<should> be absolute and I<must not> end in a slash. If it is	1949	The path I<must not> end in a slash or contain special components such as
		1950	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1472	relative and your working directory changes, the behaviour is undefined.	1951	your working directory changes, then the behaviour is undefined.
1473		1952
1474	Since there is no standard to do this, the portable implementation simply	1953	Since there is no portable change notification interface available, the
1475	calls C<stat (2)> regularly on the path to see if it changed somehow. You	1954	portable implementation simply calls C<stat(2)> regularly on the path
1476	can specify a recommended polling interval for this case. If you specify	1955	to see if it changed somehow. You can specify a recommended polling
1477	a polling interval of C<0> (highly recommended!) then a I<suitable,	1956	interval for this case. If you specify a polling interval of C<0> (highly
1478	unspecified default> value will be used (which you can expect to be around	1957	recommended!) then a I<suitable, unspecified default> value will be used
1479	five seconds, although this might change dynamically). Libev will also	1958	(which you can expect to be around five seconds, although this might
1480	impose a minimum interval which is currently around C<0.1>, but thats	1959	change dynamically). Libev will also impose a minimum interval which is
1481	usually overkill.	1960	currently around C<0.1>, but that's usually overkill.
1482		1961
1483	This watcher type is not meant for massive numbers of stat watchers,	1962	This watcher type is not meant for massive numbers of stat watchers,
1484	as even with OS-supported change notifications, this can be	1963	as even with OS-supported change notifications, this can be
1485	resource-intensive.	1964	resource-intensive.
1486		1965
1487	At the time of this writing, only the Linux inotify interface is	1966	At the time of this writing, the only OS-specific interface implemented
1488	implemented (implementing kqueue support is left as an exercise for the	1967	is the Linux inotify interface (implementing kqueue support is left as an
1489	reader). Inotify will be used to give hints only and should not change the	1968	exercise for the reader. Note, however, that the author sees no way of
1490	semantics of C<ev_stat> watchers, which means that libev sometimes needs	1969	implementing C<ev_stat> semantics with kqueue, except as a hint).
1491	to fall back to regular polling again even with inotify, but changes are
1492	usually detected immediately, and if the file exists there will be no
1493	polling.
1494		1970
1495	=head3 Inotify	1971	=head3 ABI Issues (Largefile Support)
1496		1972
		1973	Libev by default (unless the user overrides this) uses the default
		1974	compilation environment, which means that on systems with large file
		1975	support disabled by default, you get the 32 bit version of the stat
		1976	structure. When using the library from programs that change the ABI to
		1977	use 64 bit file offsets the programs will fail. In that case you have to
		1978	compile libev with the same flags to get binary compatibility. This is
		1979	obviously the case with any flags that change the ABI, but the problem is
		1980	most noticeably displayed with ev_stat and large file support.
		1981
		1982	The solution for this is to lobby your distribution maker to make large
		1983	file interfaces available by default (as e.g. FreeBSD does) and not
		1984	optional. Libev cannot simply switch on large file support because it has
		1985	to exchange stat structures with application programs compiled using the
		1986	default compilation environment.
		1987
		1988	=head3 Inotify and Kqueue
		1989
1497	When C<inotify (7)> support has been compiled into libev (generally only	1990	When C<inotify (7)> support has been compiled into libev and present at
1498	available on Linux) and present at runtime, it will be used to speed up	1991	runtime, it will be used to speed up change detection where possible. The
1499	change detection where possible. The inotify descriptor will be created lazily	1992	inotify descriptor will be created lazily when the first C<ev_stat>
1500	when the first C<ev_stat> watcher is being started.	1993	watcher is being started.
1501		1994
1502	Inotify presense does not change the semantics of C<ev_stat> watchers	1995	Inotify presence does not change the semantics of C<ev_stat> watchers
1503	except that changes might be detected earlier, and in some cases, to avoid	1996	except that changes might be detected earlier, and in some cases, to avoid
1504	making regular C<stat> calls. Even in the presense of inotify support	1997	making regular C<stat> calls. Even in the presence of inotify support
1505	there are many cases where libev has to resort to regular C<stat> polling.	1998	there are many cases where libev has to resort to regular C<stat> polling,
		1999	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2000	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2001	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2002	xfs are fully working) libev usually gets away without polling.
1506		2003
1507	(There is no support for kqueue, as apparently it cannot be used to	2004	There is no support for kqueue, as apparently it cannot be used to
1508	implement this functionality, due to the requirement of having a file	2005	implement this functionality, due to the requirement of having a file
1509	descriptor open on the object at all times).	2006	descriptor open on the object at all times, and detecting renames, unlinks
		2007	etc. is difficult.
		2008
		2009	=head3 C<stat ()> is a synchronous operation
		2010
		2011	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2012	the process. The exception are C<ev_stat> watchers - those call C<stat
		2013	()>, which is a synchronous operation.
		2014
		2015	For local paths, this usually doesn't matter: unless the system is very
		2016	busy or the intervals between stat's are large, a stat call will be fast,
		2017	as the path data is suually in memory already (except when starting the
		2018	watcher).
		2019
		2020	For networked file systems, calling C<stat ()> can block an indefinite
		2021	time due to network issues, and even under good conditions, a stat call
		2022	often takes multiple milliseconds.
		2023
		2024	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2025	paths, although this is fully supported by libev.
1510		2026
1511	=head3 The special problem of stat time resolution	2027	=head3 The special problem of stat time resolution
1512		2028
1513	The C<stat ()> syscall only supports full-second resolution portably, and	2029	The C<stat ()> system call only supports full-second resolution portably,
1514	even on systems where the resolution is higher, many filesystems still	2030	and even on systems where the resolution is higher, most file systems
1515	only support whole seconds.	2031	still only support whole seconds.
1516		2032
1517	That means that, if the time is the only thing that changes, you might	2033	That means that, if the time is the only thing that changes, you can
1518	miss updates: on the first update, C<ev_stat> detects a change and calls	2034	easily miss updates: on the first update, C<ev_stat> detects a change and
1519	your callback, which does something. When there is another update within	2035	calls your callback, which does something. When there is another update
1520	the same second, C<ev_stat> will be unable to detect it.	2036	within the same second, C<ev_stat> will be unable to detect unless the
		2037	stat data does change in other ways (e.g. file size).
1521		2038
1522	The solution to this is to delay acting on a change for a second (or till	2039	The solution to this is to delay acting on a change for slightly more
1523	the next second boundary), using a roughly one-second delay C<ev_timer>	2040	than a second (or till slightly after the next full second boundary), using
1524	(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>	2041	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1525	is added to work around small timing inconsistencies of some operating	2042	ev_timer_again (loop, w)>).
1526	systems.	2043
		2044	The C<.02> offset is added to work around small timing inconsistencies
		2045	of some operating systems (where the second counter of the current time
		2046	might be be delayed. One such system is the Linux kernel, where a call to
		2047	C<gettimeofday> might return a timestamp with a full second later than
		2048	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2049	update file times then there will be a small window where the kernel uses
		2050	the previous second to update file times but libev might already execute
		2051	the timer callback).
1527		2052
1528	=head3 Watcher-Specific Functions and Data Members	2053	=head3 Watcher-Specific Functions and Data Members
1529		2054
1530	=over 4	2055	=over 4
1531		2056
…		…
1537	C<path>. The C<interval> is a hint on how quickly a change is expected to	2062	C<path>. The C<interval> is a hint on how quickly a change is expected to
1538	be detected and should normally be specified as C<0> to let libev choose	2063	be detected and should normally be specified as C<0> to let libev choose
1539	a suitable value. The memory pointed to by C<path> must point to the same	2064	a suitable value. The memory pointed to by C<path> must point to the same
1540	path for as long as the watcher is active.	2065	path for as long as the watcher is active.
1541		2066
1542	The callback will be receive C<EV_STAT> when a change was detected,	2067	The callback will receive an C<EV_STAT> event when a change was detected,
1543	relative to the attributes at the time the watcher was started (or the	2068	relative to the attributes at the time the watcher was started (or the
1544	last change was detected).	2069	last change was detected).
1545		2070
1546	=item ev_stat_stat (ev_stat *)	2071	=item ev_stat_stat (loop, ev_stat *)
1547		2072
1548	Updates the stat buffer immediately with new values. If you change the	2073	Updates the stat buffer immediately with new values. If you change the
1549	watched path in your callback, you could call this fucntion to avoid	2074	watched path in your callback, you could call this function to avoid
1550	detecting this change (while introducing a race condition). Can also be	2075	detecting this change (while introducing a race condition if you are not
1551	useful simply to find out the new values.	2076	the only one changing the path). Can also be useful simply to find out the
		2077	new values.
1552		2078
1553	=item ev_statdata attr [read-only]	2079	=item ev_statdata attr [read-only]
1554		2080
1555	The most-recently detected attributes of the file. Although the type is of	2081	The most-recently detected attributes of the file. Although the type is
1556	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types	2082	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
1557	suitable for your system. If the C<st_nlink> member is C<0>, then there	2083	suitable for your system, but you can only rely on the POSIX-standardised
		2084	members to be present. If the C<st_nlink> member is C<0>, then there was
1558	was some error while C<stat>ing the file.	2085	some error while C<stat>ing the file.
1559		2086
1560	=item ev_statdata prev [read-only]	2087	=item ev_statdata prev [read-only]
1561		2088
1562	The previous attributes of the file. The callback gets invoked whenever	2089	The previous attributes of the file. The callback gets invoked whenever
1563	C<prev> != C<attr>.	2090	C<prev> != C<attr>, or, more precisely, one or more of these members
		2091	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2092	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1564		2093
1565	=item ev_tstamp interval [read-only]	2094	=item ev_tstamp interval [read-only]
1566		2095
1567	The specified interval.	2096	The specified interval.
1568		2097
1569	=item const char *path [read-only]	2098	=item const char *path [read-only]
1570		2099
1571	The filesystem path that is being watched.	2100	The file system path that is being watched.
1572		2101
1573	=back	2102	=back
1574		2103
1575	=head3 Examples	2104	=head3 Examples
1576		2105
1577	Example: Watch C</etc/passwd> for attribute changes.	2106	Example: Watch C</etc/passwd> for attribute changes.
1578		2107
1579	static void	2108	static void
1580	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	2109	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1581	{	2110	{
1582	/* /etc/passwd changed in some way */	2111	/* /etc/passwd changed in some way */
1583	if (w->attr.st_nlink)	2112	if (w->attr.st_nlink)
1584	{	2113	{
1585	printf ("passwd current size %ld\n", (long)w->attr.st_size);	2114	printf ("passwd current size %ld\n", (long)w->attr.st_size);
1586	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);	2115	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
1587	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);	2116	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
1588	}	2117	}
1589	else	2118	else
1590	/* you shalt not abuse printf for puts */	2119	/* you shalt not abuse printf for puts */
1591	puts ("wow, /etc/passwd is not there, expect problems. "	2120	puts ("wow, /etc/passwd is not there, expect problems. "
1592	"if this is windows, they already arrived\n");	2121	"if this is windows, they already arrived\n");
1593	}	2122	}
1594		2123
1595	...	2124	...
1596	ev_stat passwd;	2125	ev_stat passwd;
1597		2126
1598	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);	2127	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1599	ev_stat_start (loop, &passwd);	2128	ev_stat_start (loop, &passwd);
1600		2129
1601	Example: Like above, but additionally use a one-second delay so we do not	2130	Example: Like above, but additionally use a one-second delay so we do not
1602	miss updates (however, frequent updates will delay processing, too, so	2131	miss updates (however, frequent updates will delay processing, too, so
1603	one might do the work both on C<ev_stat> callback invocation I<and> on	2132	one might do the work both on C<ev_stat> callback invocation I<and> on
1604	C<ev_timer> callback invocation).	2133	C<ev_timer> callback invocation).
1605		2134
1606	static ev_stat passwd;	2135	static ev_stat passwd;
1607	static ev_timer timer;	2136	static ev_timer timer;
1608		2137
1609	static void	2138	static void
1610	timer_cb (EV_P_ ev_timer *w, int revents)	2139	timer_cb (EV_P_ ev_timer *w, int revents)
1611	{	2140	{
1612	ev_timer_stop (EV_A_ w);	2141	ev_timer_stop (EV_A_ w);
1613		2142
1614	/* now it's one second after the most recent passwd change */	2143	/* now it's one second after the most recent passwd change */
1615	}	2144	}
1616		2145
1617	static void	2146	static void
1618	stat_cb (EV_P_ ev_stat *w, int revents)	2147	stat_cb (EV_P_ ev_stat *w, int revents)
1619	{	2148	{
1620	/* reset the one-second timer */	2149	/* reset the one-second timer */
1621	ev_timer_again (EV_A_ &timer);	2150	ev_timer_again (EV_A_ &timer);
1622	}	2151	}
1623		2152
1624	...	2153	...
1625	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	2154	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1626	ev_stat_start (loop, &passwd);	2155	ev_stat_start (loop, &passwd);
1627	ev_timer_init (&timer, timer_cb, 0., 1.01);	2156	ev_timer_init (&timer, timer_cb, 0., 1.02);
1628		2157
1629		2158
1630	=head2 C<ev_idle> - when you've got nothing better to do...	2159	=head2 C<ev_idle> - when you've got nothing better to do...
1631		2160
1632	Idle watchers trigger events when no other events of the same or higher	2161	Idle watchers trigger events when no other events of the same or higher
1633	priority are pending (prepare, check and other idle watchers do not	2162	priority are pending (prepare, check and other idle watchers do not count
1634	count).	2163	as receiving "events").
1635		2164
1636	That is, as long as your process is busy handling sockets or timeouts	2165	That is, as long as your process is busy handling sockets or timeouts
1637	(or even signals, imagine) of the same or higher priority it will not be	2166	(or even signals, imagine) of the same or higher priority it will not be
1638	triggered. But when your process is idle (or only lower-priority watchers	2167	triggered. But when your process is idle (or only lower-priority watchers
1639	are pending), the idle watchers are being called once per event loop	2168	are pending), the idle watchers are being called once per event loop
…		…
1658	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2187	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1659	believe me.	2188	believe me.
1660		2189
1661	=back	2190	=back
1662		2191
		2192	=head3 Examples
		2193
1663	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2194	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1664	callback, free it. Also, use no error checking, as usual.	2195	callback, free it. Also, use no error checking, as usual.
1665		2196
1666	static void	2197	static void
1667	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2198	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1668	{	2199	{
1669	free (w);	2200	free (w);
1670	// now do something you wanted to do when the program has	2201	// now do something you wanted to do when the program has
1671	// no longer asnything immediate to do.	2202	// no longer anything immediate to do.
1672	}	2203	}
1673		2204
1674	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2205	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1675	ev_idle_init (idle_watcher, idle_cb);	2206	ev_idle_init (idle_watcher, idle_cb);
1676	ev_idle_start (loop, idle_cb);	2207	ev_idle_start (loop, idle_cb);
1677		2208
1678		2209
1679	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2210	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1680		2211
1681	Prepare and check watchers are usually (but not always) used in tandem:	2212	Prepare and check watchers are usually (but not always) used in pairs:
1682	prepare watchers get invoked before the process blocks and check watchers	2213	prepare watchers get invoked before the process blocks and check watchers
1683	afterwards.	2214	afterwards.
1684		2215
1685	You I<must not> call C<ev_loop> or similar functions that enter	2216	You I<must not> call C<ev_loop> or similar functions that enter
1686	the current event loop from either C<ev_prepare> or C<ev_check>	2217	the current event loop from either C<ev_prepare> or C<ev_check>
…		…
1689	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2220	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1690	C<ev_check> so if you have one watcher of each kind they will always be	2221	C<ev_check> so if you have one watcher of each kind they will always be
1691	called in pairs bracketing the blocking call.	2222	called in pairs bracketing the blocking call.
1692		2223
1693	Their main purpose is to integrate other event mechanisms into libev and	2224	Their main purpose is to integrate other event mechanisms into libev and
1694	their use is somewhat advanced. This could be used, for example, to track	2225	their use is somewhat advanced. They could be used, for example, to track
1695	variable changes, implement your own watchers, integrate net-snmp or a	2226	variable changes, implement your own watchers, integrate net-snmp or a
1696	coroutine library and lots more. They are also occasionally useful if	2227	coroutine library and lots more. They are also occasionally useful if
1697	you cache some data and want to flush it before blocking (for example,	2228	you cache some data and want to flush it before blocking (for example,
1698	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2229	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1699	watcher).	2230	watcher).
1700		2231
1701	This is done by examining in each prepare call which file descriptors need	2232	This is done by examining in each prepare call which file descriptors
1702	to be watched by the other library, registering C<ev_io> watchers for	2233	need to be watched by the other library, registering C<ev_io> watchers
1703	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2234	for them and starting an C<ev_timer> watcher for any timeouts (many
1704	provide just this functionality). Then, in the check watcher you check for	2235	libraries provide exactly this functionality). Then, in the check watcher,
1705	any events that occured (by checking the pending status of all watchers	2236	you check for any events that occurred (by checking the pending status
1706	and stopping them) and call back into the library. The I/O and timer	2237	of all watchers and stopping them) and call back into the library. The
1707	callbacks will never actually be called (but must be valid nevertheless,	2238	I/O and timer callbacks will never actually be called (but must be valid
1708	because you never know, you know?).	2239	nevertheless, because you never know, you know?).
1709		2240
1710	As another example, the Perl Coro module uses these hooks to integrate	2241	As another example, the Perl Coro module uses these hooks to integrate
1711	coroutines into libev programs, by yielding to other active coroutines	2242	coroutines into libev programs, by yielding to other active coroutines
1712	during each prepare and only letting the process block if no coroutines	2243	during each prepare and only letting the process block if no coroutines
1713	are ready to run (it's actually more complicated: it only runs coroutines	2244	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1716	loop from blocking if lower-priority coroutines are active, thus mapping	2247	loop from blocking if lower-priority coroutines are active, thus mapping
1717	low-priority coroutines to idle/background tasks).	2248	low-priority coroutines to idle/background tasks).
1718		2249
1719	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2250	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1720	priority, to ensure that they are being run before any other watchers	2251	priority, to ensure that they are being run before any other watchers
		2252	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2253
1721	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2254	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1722	too) should not activate ("feed") events into libev. While libev fully	2255	activate ("feed") events into libev. While libev fully supports this, they
1723	supports this, they will be called before other C<ev_check> watchers	2256	might get executed before other C<ev_check> watchers did their job. As
1724	did their job. As C<ev_check> watchers are often used to embed other	2257	C<ev_check> watchers are often used to embed other (non-libev) event
1725	(non-libev) event loops those other event loops might be in an unusable	2258	loops those other event loops might be in an unusable state until their
1726	state until their C<ev_check> watcher ran (always remind yourself to	2259	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1727	coexist peacefully with others).	2260	others).
1728		2261
1729	=head3 Watcher-Specific Functions and Data Members	2262	=head3 Watcher-Specific Functions and Data Members
1730		2263
1731	=over 4	2264	=over 4
1732		2265
…		…
1734		2267
1735	=item ev_check_init (ev_check *, callback)	2268	=item ev_check_init (ev_check *, callback)
1736		2269
1737	Initialises and configures the prepare or check watcher - they have no	2270	Initialises and configures the prepare or check watcher - they have no
1738	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2271	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1739	macros, but using them is utterly, utterly and completely pointless.	2272	macros, but using them is utterly, utterly, utterly and completely
		2273	pointless.
1740		2274
1741	=back	2275	=back
		2276
		2277	=head3 Examples
1742		2278
1743	There are a number of principal ways to embed other event loops or modules	2279	There are a number of principal ways to embed other event loops or modules
1744	into libev. Here are some ideas on how to include libadns into libev	2280	into libev. Here are some ideas on how to include libadns into libev
1745	(there is a Perl module named C<EV::ADNS> that does this, which you could	2281	(there is a Perl module named C<EV::ADNS> that does this, which you could
1746	use for an actually working example. Another Perl module named C<EV::Glib>	2282	use as a working example. Another Perl module named C<EV::Glib> embeds a
1747	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV	2283	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
1748	into the Glib event loop).	2284	Glib event loop).
1749		2285
1750	Method 1: Add IO watchers and a timeout watcher in a prepare handler,	2286	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1751	and in a check watcher, destroy them and call into libadns. What follows	2287	and in a check watcher, destroy them and call into libadns. What follows
1752	is pseudo-code only of course. This requires you to either use a low	2288	is pseudo-code only of course. This requires you to either use a low
1753	priority for the check watcher or use C<ev_clear_pending> explicitly, as	2289	priority for the check watcher or use C<ev_clear_pending> explicitly, as
1754	the callbacks for the IO/timeout watchers might not have been called yet.	2290	the callbacks for the IO/timeout watchers might not have been called yet.
1755		2291
1756	static ev_io iow [nfd];	2292	static ev_io iow [nfd];
1757	static ev_timer tw;	2293	static ev_timer tw;
1758		2294
1759	static void	2295	static void
1760	io_cb (ev_loop *loop, ev_io *w, int revents)	2296	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1761	{	2297	{
1762	}	2298	}
1763		2299
1764	// create io watchers for each fd and a timer before blocking	2300	// create io watchers for each fd and a timer before blocking
1765	static void	2301	static void
1766	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2302	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1767	{	2303	{
1768	int timeout = 3600000;	2304	int timeout = 3600000;
1769	struct pollfd fds [nfd];	2305	struct pollfd fds [nfd];
1770	// actual code will need to loop here and realloc etc.	2306	// actual code will need to loop here and realloc etc.
1771	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2307	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1772		2308
1773	/* the callback is illegal, but won't be called as we stop during check */	2309	/* the callback is illegal, but won't be called as we stop during check */
1774	ev_timer_init (&tw, 0, timeout * 1e-3);	2310	ev_timer_init (&tw, 0, timeout * 1e-3);
1775	ev_timer_start (loop, &tw);	2311	ev_timer_start (loop, &tw);
1776		2312
1777	// create one ev_io per pollfd	2313	// create one ev_io per pollfd
1778	for (int i = 0; i < nfd; ++i)	2314	for (int i = 0; i < nfd; ++i)
1779	{	2315	{
1780	ev_io_init (iow + i, io_cb, fds [i].fd,	2316	ev_io_init (iow + i, io_cb, fds [i].fd,
1781	((fds [i].events & POLLIN ? EV_READ : 0)	2317	((fds [i].events & POLLIN ? EV_READ : 0)
1782	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	2318	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1783		2319
1784	fds [i].revents = 0;	2320	fds [i].revents = 0;
1785	ev_io_start (loop, iow + i);	2321	ev_io_start (loop, iow + i);
1786	}	2322	}
1787	}	2323	}
1788		2324
1789	// stop all watchers after blocking	2325	// stop all watchers after blocking
1790	static void	2326	static void
1791	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2327	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
1792	{	2328	{
1793	ev_timer_stop (loop, &tw);	2329	ev_timer_stop (loop, &tw);
1794		2330
1795	for (int i = 0; i < nfd; ++i)	2331	for (int i = 0; i < nfd; ++i)
1796	{	2332	{
1797	// set the relevant poll flags	2333	// set the relevant poll flags
1798	// could also call adns_processreadable etc. here	2334	// could also call adns_processreadable etc. here
1799	struct pollfd *fd = fds + i;	2335	struct pollfd *fd = fds + i;
1800	int revents = ev_clear_pending (iow + i);	2336	int revents = ev_clear_pending (iow + i);
1801	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;	2337	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1802	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;	2338	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1803		2339
1804	// now stop the watcher	2340	// now stop the watcher
1805	ev_io_stop (loop, iow + i);	2341	ev_io_stop (loop, iow + i);
1806	}	2342	}
1807		2343
1808	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	2344	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1809	}	2345	}
1810		2346
1811	Method 2: This would be just like method 1, but you run C<adns_afterpoll>	2347	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
1812	in the prepare watcher and would dispose of the check watcher.	2348	in the prepare watcher and would dispose of the check watcher.
1813		2349
1814	Method 3: If the module to be embedded supports explicit event	2350	Method 3: If the module to be embedded supports explicit event
1815	notification (adns does), you can also make use of the actual watcher	2351	notification (libadns does), you can also make use of the actual watcher
1816	callbacks, and only destroy/create the watchers in the prepare watcher.	2352	callbacks, and only destroy/create the watchers in the prepare watcher.
1817		2353
1818	static void	2354	static void
1819	timer_cb (EV_P_ ev_timer *w, int revents)	2355	timer_cb (EV_P_ ev_timer *w, int revents)
1820	{	2356	{
1821	adns_state ads = (adns_state)w->data;	2357	adns_state ads = (adns_state)w->data;
1822	update_now (EV_A);	2358	update_now (EV_A);
1823		2359
1824	adns_processtimeouts (ads, &tv_now);	2360	adns_processtimeouts (ads, &tv_now);
1825	}	2361	}
1826		2362
1827	static void	2363	static void
1828	io_cb (EV_P_ ev_io *w, int revents)	2364	io_cb (EV_P_ ev_io *w, int revents)
1829	{	2365	{
1830	adns_state ads = (adns_state)w->data;	2366	adns_state ads = (adns_state)w->data;
1831	update_now (EV_A);	2367	update_now (EV_A);
1832		2368
1833	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);	2369	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
1834	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);	2370	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
1835	}	2371	}
1836		2372
1837	// do not ever call adns_afterpoll	2373	// do not ever call adns_afterpoll
1838		2374
1839	Method 4: Do not use a prepare or check watcher because the module you	2375	Method 4: Do not use a prepare or check watcher because the module you
1840	want to embed is too inflexible to support it. Instead, youc na override	2376	want to embed is not flexible enough to support it. Instead, you can
1841	their poll function. The drawback with this solution is that the main	2377	override their poll function. The drawback with this solution is that the
1842	loop is now no longer controllable by EV. The C<Glib::EV> module does	2378	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
1843	this.	2379	this approach, effectively embedding EV as a client into the horrible
		2380	libglib event loop.
1844		2381
1845	static gint	2382	static gint
1846	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2383	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
1847	{	2384	{
1848	int got_events = 0;	2385	int got_events = 0;
1849		2386
1850	for (n = 0; n < nfds; ++n)	2387	for (n = 0; n < nfds; ++n)
1851	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events	2388	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
1852		2389
1853	if (timeout >= 0)	2390	if (timeout >= 0)
1854	// create/start timer	2391	// create/start timer
1855		2392
1856	// poll	2393	// poll
1857	ev_loop (EV_A_ 0);	2394	ev_loop (EV_A_ 0);
1858		2395
1859	// stop timer again	2396	// stop timer again
1860	if (timeout >= 0)	2397	if (timeout >= 0)
1861	ev_timer_stop (EV_A_ &to);	2398	ev_timer_stop (EV_A_ &to);
1862		2399
1863	// stop io watchers again - their callbacks should have set	2400	// stop io watchers again - their callbacks should have set
1864	for (n = 0; n < nfds; ++n)	2401	for (n = 0; n < nfds; ++n)
1865	ev_io_stop (EV_A_ iow [n]);	2402	ev_io_stop (EV_A_ iow [n]);
1866		2403
1867	return got_events;	2404	return got_events;
1868	}	2405	}
1869		2406
1870		2407
1871	=head2 C<ev_embed> - when one backend isn't enough...	2408	=head2 C<ev_embed> - when one backend isn't enough...
1872		2409
1873	This is a rather advanced watcher type that lets you embed one event loop	2410	This is a rather advanced watcher type that lets you embed one event loop
…		…
1879	prioritise I/O.	2416	prioritise I/O.
1880		2417
1881	As an example for a bug workaround, the kqueue backend might only support	2418	As an example for a bug workaround, the kqueue backend might only support
1882	sockets on some platform, so it is unusable as generic backend, but you	2419	sockets on some platform, so it is unusable as generic backend, but you
1883	still want to make use of it because you have many sockets and it scales	2420	still want to make use of it because you have many sockets and it scales
1884	so nicely. In this case, you would create a kqueue-based loop and embed it	2421	so nicely. In this case, you would create a kqueue-based loop and embed
1885	into your default loop (which might use e.g. poll). Overall operation will	2422	it into your default loop (which might use e.g. poll). Overall operation
1886	be a bit slower because first libev has to poll and then call kevent, but	2423	will be a bit slower because first libev has to call C<poll> and then
1887	at least you can use both at what they are best.	2424	C<kevent>, but at least you can use both mechanisms for what they are
		2425	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1888		2426
1889	As for prioritising I/O: rarely you have the case where some fds have	2427	As for prioritising I/O: under rare circumstances you have the case where
1890	to be watched and handled very quickly (with low latency), and even	2428	some fds have to be watched and handled very quickly (with low latency),
1891	priorities and idle watchers might have too much overhead. In this case	2429	and even priorities and idle watchers might have too much overhead. In
1892	you would put all the high priority stuff in one loop and all the rest in	2430	this case you would put all the high priority stuff in one loop and all
1893	a second one, and embed the second one in the first.	2431	the rest in a second one, and embed the second one in the first.
1894		2432
1895	As long as the watcher is active, the callback will be invoked every time	2433	As long as the watcher is active, the callback will be invoked every time
1896	there might be events pending in the embedded loop. The callback must then	2434	there might be events pending in the embedded loop. The callback must then
1897	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2435	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
1898	their callbacks (you could also start an idle watcher to give the embedded	2436	their callbacks (you could also start an idle watcher to give the embedded
…		…
1906	interested in that.	2444	interested in that.
1907		2445
1908	Also, there have not currently been made special provisions for forking:	2446	Also, there have not currently been made special provisions for forking:
1909	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2447	when you fork, you not only have to call C<ev_loop_fork> on both loops,
1910	but you will also have to stop and restart any C<ev_embed> watchers	2448	but you will also have to stop and restart any C<ev_embed> watchers
1911	yourself.	2449	yourself - but you can use a fork watcher to handle this automatically,
		2450	and future versions of libev might do just that.
1912		2451
1913	Unfortunately, not all backends are embeddable, only the ones returned by	2452	Unfortunately, not all backends are embeddable: only the ones returned by
1914	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2453	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1915	portable one.	2454	portable one.
1916		2455
1917	So when you want to use this feature you will always have to be prepared	2456	So when you want to use this feature you will always have to be prepared
1918	that you cannot get an embeddable loop. The recommended way to get around	2457	that you cannot get an embeddable loop. The recommended way to get around
1919	this is to have a separate variables for your embeddable loop, try to	2458	this is to have a separate variables for your embeddable loop, try to
1920	create it, and if that fails, use the normal loop for everything:	2459	create it, and if that fails, use the normal loop for everything.
1921		2460
1922	struct ev_loop *loop_hi = ev_default_init (0);	2461	=head3 C<ev_embed> and fork
1923	struct ev_loop *loop_lo = 0;
1924	struct ev_embed embed;
1925
1926	// see if there is a chance of getting one that works
1927	// (remember that a flags value of 0 means autodetection)
1928	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1929	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1930	: 0;
1931		2462
1932	// if we got one, then embed it, otherwise default to loop_hi	2463	While the C<ev_embed> watcher is running, forks in the embedding loop will
1933	if (loop_lo)	2464	automatically be applied to the embedded loop as well, so no special
1934	{	2465	fork handling is required in that case. When the watcher is not running,
1935	ev_embed_init (&embed, 0, loop_lo);	2466	however, it is still the task of the libev user to call C<ev_loop_fork ()>
1936	ev_embed_start (loop_hi, &embed);	2467	as applicable.
1937	}
1938	else
1939	loop_lo = loop_hi;
1940		2468
1941	=head3 Watcher-Specific Functions and Data Members	2469	=head3 Watcher-Specific Functions and Data Members
1942		2470
1943	=over 4	2471	=over 4
1944		2472
…		…
1948		2476
1949	Configures the watcher to embed the given loop, which must be	2477	Configures the watcher to embed the given loop, which must be
1950	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	2478	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1951	invoked automatically, otherwise it is the responsibility of the callback	2479	invoked automatically, otherwise it is the responsibility of the callback
1952	to invoke it (it will continue to be called until the sweep has been done,	2480	to invoke it (it will continue to be called until the sweep has been done,
1953	if you do not want thta, you need to temporarily stop the embed watcher).	2481	if you do not want that, you need to temporarily stop the embed watcher).
1954		2482
1955	=item ev_embed_sweep (loop, ev_embed *)	2483	=item ev_embed_sweep (loop, ev_embed *)
1956		2484
1957	Make a single, non-blocking sweep over the embedded loop. This works	2485	Make a single, non-blocking sweep over the embedded loop. This works
1958	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	2486	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1959	apropriate way for embedded loops.	2487	appropriate way for embedded loops.
1960		2488
1961	=item struct ev_loop *other [read-only]	2489	=item struct ev_loop *other [read-only]
1962		2490
1963	The embedded event loop.	2491	The embedded event loop.
1964		2492
1965	=back	2493	=back
		2494
		2495	=head3 Examples
		2496
		2497	Example: Try to get an embeddable event loop and embed it into the default
		2498	event loop. If that is not possible, use the default loop. The default
		2499	loop is stored in C<loop_hi>, while the embeddable loop is stored in
		2500	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
		2501	used).
		2502
		2503	struct ev_loop *loop_hi = ev_default_init (0);
		2504	struct ev_loop *loop_lo = 0;
		2505	ev_embed embed;
		2506
		2507	// see if there is a chance of getting one that works
		2508	// (remember that a flags value of 0 means autodetection)
		2509	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		2510	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		2511	: 0;
		2512
		2513	// if we got one, then embed it, otherwise default to loop_hi
		2514	if (loop_lo)
		2515	{
		2516	ev_embed_init (&embed, 0, loop_lo);
		2517	ev_embed_start (loop_hi, &embed);
		2518	}
		2519	else
		2520	loop_lo = loop_hi;
		2521
		2522	Example: Check if kqueue is available but not recommended and create
		2523	a kqueue backend for use with sockets (which usually work with any
		2524	kqueue implementation). Store the kqueue/socket-only event loop in
		2525	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		2526
		2527	struct ev_loop *loop = ev_default_init (0);
		2528	struct ev_loop *loop_socket = 0;
		2529	ev_embed embed;
		2530
		2531	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2532	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2533	{
		2534	ev_embed_init (&embed, 0, loop_socket);
		2535	ev_embed_start (loop, &embed);
		2536	}
		2537
		2538	if (!loop_socket)
		2539	loop_socket = loop;
		2540
		2541	// now use loop_socket for all sockets, and loop for everything else
1966		2542
1967		2543
1968	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	2544	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
1969		2545
1970	Fork watchers are called when a C<fork ()> was detected (usually because	2546	Fork watchers are called when a C<fork ()> was detected (usually because
…		…
1986	believe me.	2562	believe me.
1987		2563
1988	=back	2564	=back
1989		2565
1990		2566
		2567	=head2 C<ev_async> - how to wake up another event loop
		2568
		2569	In general, you cannot use an C<ev_loop> from multiple threads or other
		2570	asynchronous sources such as signal handlers (as opposed to multiple event
		2571	loops - those are of course safe to use in different threads).
		2572
		2573	Sometimes, however, you need to wake up another event loop you do not
		2574	control, for example because it belongs to another thread. This is what
		2575	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		2576	can signal it by calling C<ev_async_send>, which is thread- and signal
		2577	safe.
		2578
		2579	This functionality is very similar to C<ev_signal> watchers, as signals,
		2580	too, are asynchronous in nature, and signals, too, will be compressed
		2581	(i.e. the number of callback invocations may be less than the number of
		2582	C<ev_async_sent> calls).
		2583
		2584	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		2585	just the default loop.
		2586
		2587	=head3 Queueing
		2588
		2589	C<ev_async> does not support queueing of data in any way. The reason
		2590	is that the author does not know of a simple (or any) algorithm for a
		2591	multiple-writer-single-reader queue that works in all cases and doesn't
		2592	need elaborate support such as pthreads.
		2593
		2594	That means that if you want to queue data, you have to provide your own
		2595	queue. But at least I can tell you how to implement locking around your
		2596	queue:
		2597
		2598	=over 4
		2599
		2600	=item queueing from a signal handler context
		2601
		2602	To implement race-free queueing, you simply add to the queue in the signal
		2603	handler but you block the signal handler in the watcher callback. Here is
		2604	an example that does that for some fictitious SIGUSR1 handler:
		2605
		2606	static ev_async mysig;
		2607
		2608	static void
		2609	sigusr1_handler (void)
		2610	{
		2611	sometype data;
		2612
		2613	// no locking etc.
		2614	queue_put (data);
		2615	ev_async_send (EV_DEFAULT_ &mysig);
		2616	}
		2617
		2618	static void
		2619	mysig_cb (EV_P_ ev_async *w, int revents)
		2620	{
		2621	sometype data;
		2622	sigset_t block, prev;
		2623
		2624	sigemptyset (&block);
		2625	sigaddset (&block, SIGUSR1);
		2626	sigprocmask (SIG_BLOCK, &block, &prev);
		2627
		2628	while (queue_get (&data))
		2629	process (data);
		2630
		2631	if (sigismember (&prev, SIGUSR1)
		2632	sigprocmask (SIG_UNBLOCK, &block, 0);
		2633	}
		2634
		2635	(Note: pthreads in theory requires you to use C<pthread_setmask>
		2636	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		2637	either...).
		2638
		2639	=item queueing from a thread context
		2640
		2641	The strategy for threads is different, as you cannot (easily) block
		2642	threads but you can easily preempt them, so to queue safely you need to
		2643	employ a traditional mutex lock, such as in this pthread example:
		2644
		2645	static ev_async mysig;
		2646	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		2647
		2648	static void
		2649	otherthread (void)
		2650	{
		2651	// only need to lock the actual queueing operation
		2652	pthread_mutex_lock (&mymutex);
		2653	queue_put (data);
		2654	pthread_mutex_unlock (&mymutex);
		2655
		2656	ev_async_send (EV_DEFAULT_ &mysig);
		2657	}
		2658
		2659	static void
		2660	mysig_cb (EV_P_ ev_async *w, int revents)
		2661	{
		2662	pthread_mutex_lock (&mymutex);
		2663
		2664	while (queue_get (&data))
		2665	process (data);
		2666
		2667	pthread_mutex_unlock (&mymutex);
		2668	}
		2669
		2670	=back
		2671
		2672
		2673	=head3 Watcher-Specific Functions and Data Members
		2674
		2675	=over 4
		2676
		2677	=item ev_async_init (ev_async *, callback)
		2678
		2679	Initialises and configures the async watcher - it has no parameters of any
		2680	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
		2681	trust me.
		2682
		2683	=item ev_async_send (loop, ev_async *)
		2684
		2685	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		2686	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		2687	C<ev_feed_event>, this call is safe to do from other threads, signal or
		2688	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
		2689	section below on what exactly this means).
		2690
		2691	This call incurs the overhead of a system call only once per loop iteration,
		2692	so while the overhead might be noticeable, it doesn't apply to repeated
		2693	calls to C<ev_async_send>.
		2694
		2695	=item bool = ev_async_pending (ev_async *)
		2696
		2697	Returns a non-zero value when C<ev_async_send> has been called on the
		2698	watcher but the event has not yet been processed (or even noted) by the
		2699	event loop.
		2700
		2701	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2702	the loop iterates next and checks for the watcher to have become active,
		2703	it will reset the flag again. C<ev_async_pending> can be used to very
		2704	quickly check whether invoking the loop might be a good idea.
		2705
		2706	Not that this does I<not> check whether the watcher itself is pending, only
		2707	whether it has been requested to make this watcher pending.
		2708
		2709	=back
		2710
		2711
1991	=head1 OTHER FUNCTIONS	2712	=head1 OTHER FUNCTIONS
1992		2713
1993	There are some other functions of possible interest. Described. Here. Now.	2714	There are some other functions of possible interest. Described. Here. Now.
1994		2715
1995	=over 4	2716	=over 4
1996		2717
1997	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	2718	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
1998		2719
1999	This function combines a simple timer and an I/O watcher, calls your	2720	This function combines a simple timer and an I/O watcher, calls your
2000	callback on whichever event happens first and automatically stop both	2721	callback on whichever event happens first and automatically stops both
2001	watchers. This is useful if you want to wait for a single event on an fd	2722	watchers. This is useful if you want to wait for a single event on an fd
2002	or timeout without having to allocate/configure/start/stop/free one or	2723	or timeout without having to allocate/configure/start/stop/free one or
2003	more watchers yourself.	2724	more watchers yourself.
2004		2725
2005	If C<fd> is less than 0, then no I/O watcher will be started and events	2726	If C<fd> is less than 0, then no I/O watcher will be started and the
2006	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	2727	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2007	C<events> set will be craeted and started.	2728	the given C<fd> and C<events> set will be created and started.
2008		2729
2009	If C<timeout> is less than 0, then no timeout watcher will be	2730	If C<timeout> is less than 0, then no timeout watcher will be
2010	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	2731	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2011	repeat = 0) will be started. While C<0> is a valid timeout, it is of	2732	repeat = 0) will be started. C<0> is a valid timeout.
2012	dubious value.
2013		2733
2014	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	2734	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2015	passed an C<revents> set like normal event callbacks (a combination of	2735	passed an C<revents> set like normal event callbacks (a combination of
2016	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	2736	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2017	value passed to C<ev_once>:	2737	value passed to C<ev_once>. Note that it is possible to receive I<both>
		2738	a timeout and an io event at the same time - you probably should give io
		2739	events precedence.
2018		2740
		2741	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		2742
2019	static void stdin_ready (int revents, void *arg)	2743	static void stdin_ready (int revents, void *arg)
2020	{	2744	{
2021	if (revents & EV_TIMEOUT)
2022	/* doh, nothing entered */;
2023	else if (revents & EV_READ)	2745	if (revents & EV_READ)
2024	/* stdin might have data for us, joy! */;	2746	/* stdin might have data for us, joy! */;
		2747	else if (revents & EV_TIMEOUT)
		2748	/* doh, nothing entered */;
2025	}	2749	}
2026		2750
2027	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	2751	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2028		2752
2029	=item ev_feed_event (ev_loop *, watcher *, int revents)	2753	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
2030		2754
2031	Feeds the given event set into the event loop, as if the specified event	2755	Feeds the given event set into the event loop, as if the specified event
2032	had happened for the specified watcher (which must be a pointer to an	2756	had happened for the specified watcher (which must be a pointer to an
2033	initialised but not necessarily started event watcher).	2757	initialised but not necessarily started event watcher).
2034		2758
2035	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	2759	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)
2036		2760
2037	Feed an event on the given fd, as if a file descriptor backend detected	2761	Feed an event on the given fd, as if a file descriptor backend detected
2038	the given events it.	2762	the given events it.
2039		2763
2040	=item ev_feed_signal_event (ev_loop *loop, int signum)	2764	=item ev_feed_signal_event (struct ev_loop *loop, int signum)
2041		2765
2042	Feed an event as if the given signal occured (C<loop> must be the default	2766	Feed an event as if the given signal occurred (C<loop> must be the default
2043	loop!).	2767	loop!).
2044		2768
2045	=back	2769	=back
2046		2770
2047		2771
…		…
2063		2787
2064	=item * Priorities are not currently supported. Initialising priorities	2788	=item * Priorities are not currently supported. Initialising priorities
2065	will fail and all watchers will have the same priority, even though there	2789	will fail and all watchers will have the same priority, even though there
2066	is an ev_pri field.	2790	is an ev_pri field.
2067		2791
		2792	=item * In libevent, the last base created gets the signals, in libev, the
		2793	first base created (== the default loop) gets the signals.
		2794
2068	=item * Other members are not supported.	2795	=item * Other members are not supported.
2069		2796
2070	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2797	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2071	to use the libev header file and library.	2798	to use the libev header file and library.
2072		2799
2073	=back	2800	=back
2074		2801
2075	=head1 C++ SUPPORT	2802	=head1 C++ SUPPORT
2076		2803
2077	Libev comes with some simplistic wrapper classes for C++ that mainly allow	2804	Libev comes with some simplistic wrapper classes for C++ that mainly allow
2078	you to use some convinience methods to start/stop watchers and also change	2805	you to use some convenience methods to start/stop watchers and also change
2079	the callback model to a model using method callbacks on objects.	2806	the callback model to a model using method callbacks on objects.
2080		2807
2081	To use it,	2808	To use it,
2082		2809
2083	#include <ev++.h>	2810	#include <ev++.h>
2084		2811
2085	This automatically includes F<ev.h> and puts all of its definitions (many	2812	This automatically includes F<ev.h> and puts all of its definitions (many
2086	of them macros) into the global namespace. All C++ specific things are	2813	of them macros) into the global namespace. All C++ specific things are
2087	put into the C<ev> namespace. It should support all the same embedding	2814	put into the C<ev> namespace. It should support all the same embedding
2088	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.	2815	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
…		…
2155	your compiler is good :), then the method will be fully inlined into the	2882	your compiler is good :), then the method will be fully inlined into the
2156	thunking function, making it as fast as a direct C callback.	2883	thunking function, making it as fast as a direct C callback.
2157		2884
2158	Example: simple class declaration and watcher initialisation	2885	Example: simple class declaration and watcher initialisation
2159		2886
2160	struct myclass	2887	struct myclass
2161	{	2888	{
2162	void io_cb (ev::io &w, int revents) { }	2889	void io_cb (ev::io &w, int revents) { }
2163	}	2890	}
2164		2891
2165	myclass obj;	2892	myclass obj;
2166	ev::io iow;	2893	ev::io iow;
2167	iow.set <myclass, &myclass::io_cb> (&obj);	2894	iow.set <myclass, &myclass::io_cb> (&obj);
2168		2895
2169	=item w->set<function> (void *data = 0)	2896	=item w->set<function> (void *data = 0)
2170		2897
2171	Also sets a callback, but uses a static method or plain function as	2898	Also sets a callback, but uses a static method or plain function as
2172	callback. The optional C<data> argument will be stored in the watcher's	2899	callback. The optional C<data> argument will be stored in the watcher's
…		…
2174		2901
2175	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	2902	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2176		2903
2177	See the method-C<set> above for more details.	2904	See the method-C<set> above for more details.
2178		2905
2179	Example:	2906	Example: Use a plain function as callback.
2180		2907
2181	static void io_cb (ev::io &w, int revents) { }	2908	static void io_cb (ev::io &w, int revents) { }
2182	iow.set <io_cb> ();	2909	iow.set <io_cb> ();
2183		2910
2184	=item w->set (struct ev_loop *)	2911	=item w->set (struct ev_loop *)
2185		2912
2186	Associates a different C<struct ev_loop> with this watcher. You can only	2913	Associates a different C<struct ev_loop> with this watcher. You can only
2187	do this when the watcher is inactive (and not pending either).	2914	do this when the watcher is inactive (and not pending either).
2188		2915
2189	=item w->set ([args])	2916	=item w->set ([arguments])
2190		2917
2191	Basically the same as C<ev_TYPE_set>, with the same args. Must be	2918	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be
2192	called at least once. Unlike the C counterpart, an active watcher gets	2919	called at least once. Unlike the C counterpart, an active watcher gets
2193	automatically stopped and restarted when reconfiguring it with this	2920	automatically stopped and restarted when reconfiguring it with this
2194	method.	2921	method.
2195		2922
2196	=item w->start ()	2923	=item w->start ()
…		…
2220	=back	2947	=back
2221		2948
2222	Example: Define a class with an IO and idle watcher, start one of them in	2949	Example: Define a class with an IO and idle watcher, start one of them in
2223	the constructor.	2950	the constructor.
2224		2951
2225	class myclass	2952	class myclass
2226	{	2953	{
2227	ev_io io; void io_cb (ev::io &w, int revents);	2954	ev::io io ; void io_cb (ev::io &w, int revents);
2228	ev_idle idle void idle_cb (ev::idle &w, int revents);	2955	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2229		2956
2230	myclass ();	2957	myclass (int fd)
2231	}	2958	{
2232
2233	myclass::myclass (int fd)
2234	{
2235	io .set <myclass, &myclass::io_cb > (this);	2959	io .set <myclass, &myclass::io_cb > (this);
2236	idle.set <myclass, &myclass::idle_cb> (this);	2960	idle.set <myclass, &myclass::idle_cb> (this);
2237		2961
2238	io.start (fd, ev::READ);	2962	io.start (fd, ev::READ);
		2963	}
2239	}	2964	};
		2965
		2966
		2967	=head1 OTHER LANGUAGE BINDINGS
		2968
		2969	Libev does not offer other language bindings itself, but bindings for a
		2970	number of languages exist in the form of third-party packages. If you know
		2971	any interesting language binding in addition to the ones listed here, drop
		2972	me a note.
		2973
		2974	=over 4
		2975
		2976	=item Perl
		2977
		2978	The EV module implements the full libev API and is actually used to test
		2979	libev. EV is developed together with libev. Apart from the EV core module,
		2980	there are additional modules that implement libev-compatible interfaces
		2981	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		2982	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		2983	and C<EV::Glib>).
		2984
		2985	It can be found and installed via CPAN, its homepage is at
		2986	L<http://software.schmorp.de/pkg/EV>.
		2987
		2988	=item Python
		2989
		2990	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		2991	seems to be quite complete and well-documented. Note, however, that the
		2992	patch they require for libev is outright dangerous as it breaks the ABI
		2993	for everybody else, and therefore, should never be applied in an installed
		2994	libev (if python requires an incompatible ABI then it needs to embed
		2995	libev).
		2996
		2997	=item Ruby
		2998
		2999	Tony Arcieri has written a ruby extension that offers access to a subset
		3000	of the libev API and adds file handle abstractions, asynchronous DNS and
		3001	more on top of it. It can be found via gem servers. Its homepage is at
		3002	L<http://rev.rubyforge.org/>.
		3003
		3004	=item D
		3005
		3006	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		3007	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3008
		3009	=item Ocaml
		3010
		3011	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3012	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3013
		3014	=back
2240		3015
2241		3016
2242	=head1 MACRO MAGIC	3017	=head1 MACRO MAGIC
2243		3018
2244	Libev can be compiled with a variety of options, the most fundamantal	3019	Libev can be compiled with a variety of options, the most fundamental
2245	of which is C<EV_MULTIPLICITY>. This option determines whether (most)	3020	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
2246	functions and callbacks have an initial C<struct ev_loop *> argument.	3021	functions and callbacks have an initial C<struct ev_loop *> argument.
2247		3022
2248	To make it easier to write programs that cope with either variant, the	3023	To make it easier to write programs that cope with either variant, the
2249	following macros are defined:	3024	following macros are defined:
…		…
2254		3029
2255	This provides the loop I<argument> for functions, if one is required ("ev	3030	This provides the loop I<argument> for functions, if one is required ("ev
2256	loop argument"). The C<EV_A> form is used when this is the sole argument,	3031	loop argument"). The C<EV_A> form is used when this is the sole argument,
2257	C<EV_A_> is used when other arguments are following. Example:	3032	C<EV_A_> is used when other arguments are following. Example:
2258		3033
2259	ev_unref (EV_A);	3034	ev_unref (EV_A);
2260	ev_timer_add (EV_A_ watcher);	3035	ev_timer_add (EV_A_ watcher);
2261	ev_loop (EV_A_ 0);	3036	ev_loop (EV_A_ 0);
2262		3037
2263	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	3038	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
2264	which is often provided by the following macro.	3039	which is often provided by the following macro.
2265		3040
2266	=item C<EV_P>, C<EV_P_>	3041	=item C<EV_P>, C<EV_P_>
2267		3042
2268	This provides the loop I<parameter> for functions, if one is required ("ev	3043	This provides the loop I<parameter> for functions, if one is required ("ev
2269	loop parameter"). The C<EV_P> form is used when this is the sole parameter,	3044	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
2270	C<EV_P_> is used when other parameters are following. Example:	3045	C<EV_P_> is used when other parameters are following. Example:
2271		3046
2272	// this is how ev_unref is being declared	3047	// this is how ev_unref is being declared
2273	static void ev_unref (EV_P);	3048	static void ev_unref (EV_P);
2274		3049
2275	// this is how you can declare your typical callback	3050	// this is how you can declare your typical callback
2276	static void cb (EV_P_ ev_timer *w, int revents)	3051	static void cb (EV_P_ ev_timer *w, int revents)
2277		3052
2278	It declares a parameter C<loop> of type C<struct ev_loop *>, quite	3053	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
2279	suitable for use with C<EV_A>.	3054	suitable for use with C<EV_A>.
2280		3055
2281	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	3056	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
2282		3057
2283	Similar to the other two macros, this gives you the value of the default	3058	Similar to the other two macros, this gives you the value of the default
2284	loop, if multiple loops are supported ("ev loop default").	3059	loop, if multiple loops are supported ("ev loop default").
		3060
		3061	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		3062
		3063	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		3064	default loop has been initialised (C<UC> == unchecked). Their behaviour
		3065	is undefined when the default loop has not been initialised by a previous
		3066	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		3067
		3068	It is often prudent to use C<EV_DEFAULT> when initialising the first
		3069	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
2285		3070
2286	=back	3071	=back
2287		3072
2288	Example: Declare and initialise a check watcher, utilising the above	3073	Example: Declare and initialise a check watcher, utilising the above
2289	macros so it will work regardless of whether multiple loops are supported	3074	macros so it will work regardless of whether multiple loops are supported
2290	or not.	3075	or not.
2291		3076
2292	static void	3077	static void
2293	check_cb (EV_P_ ev_timer *w, int revents)	3078	check_cb (EV_P_ ev_timer *w, int revents)
2294	{	3079	{
2295	ev_check_stop (EV_A_ w);	3080	ev_check_stop (EV_A_ w);
2296	}	3081	}
2297		3082
2298	ev_check check;	3083	ev_check check;
2299	ev_check_init (&check, check_cb);	3084	ev_check_init (&check, check_cb);
2300	ev_check_start (EV_DEFAULT_ &check);	3085	ev_check_start (EV_DEFAULT_ &check);
2301	ev_loop (EV_DEFAULT_ 0);	3086	ev_loop (EV_DEFAULT_ 0);
2302		3087
2303	=head1 EMBEDDING	3088	=head1 EMBEDDING
2304		3089
2305	Libev can (and often is) directly embedded into host	3090	Libev can (and often is) directly embedded into host
2306	applications. Examples of applications that embed it include the Deliantra	3091	applications. Examples of applications that embed it include the Deliantra
…		…
2313	libev somewhere in your source tree).	3098	libev somewhere in your source tree).
2314		3099
2315	=head2 FILESETS	3100	=head2 FILESETS
2316		3101
2317	Depending on what features you need you need to include one or more sets of files	3102	Depending on what features you need you need to include one or more sets of files
2318	in your app.	3103	in your application.
2319		3104
2320	=head3 CORE EVENT LOOP	3105	=head3 CORE EVENT LOOP
2321		3106
2322	To include only the libev core (all the C<ev_*> functions), with manual	3107	To include only the libev core (all the C<ev_*> functions), with manual
2323	configuration (no autoconf):	3108	configuration (no autoconf):
2324		3109
2325	#define EV_STANDALONE 1	3110	#define EV_STANDALONE 1
2326	#include "ev.c"	3111	#include "ev.c"
2327		3112
2328	This will automatically include F<ev.h>, too, and should be done in a	3113	This will automatically include F<ev.h>, too, and should be done in a
2329	single C source file only to provide the function implementations. To use	3114	single C source file only to provide the function implementations. To use
2330	it, do the same for F<ev.h> in all files wishing to use this API (best	3115	it, do the same for F<ev.h> in all files wishing to use this API (best
2331	done by writing a wrapper around F<ev.h> that you can include instead and	3116	done by writing a wrapper around F<ev.h> that you can include instead and
2332	where you can put other configuration options):	3117	where you can put other configuration options):
2333		3118
2334	#define EV_STANDALONE 1	3119	#define EV_STANDALONE 1
2335	#include "ev.h"	3120	#include "ev.h"
2336		3121
2337	Both header files and implementation files can be compiled with a C++	3122	Both header files and implementation files can be compiled with a C++
2338	compiler (at least, thats a stated goal, and breakage will be treated	3123	compiler (at least, that's a stated goal, and breakage will be treated
2339	as a bug).	3124	as a bug).
2340		3125
2341	You need the following files in your source tree, or in a directory	3126	You need the following files in your source tree, or in a directory
2342	in your include path (e.g. in libev/ when using -Ilibev):	3127	in your include path (e.g. in libev/ when using -Ilibev):
2343		3128
2344	ev.h	3129	ev.h
2345	ev.c	3130	ev.c
2346	ev_vars.h	3131	ev_vars.h
2347	ev_wrap.h	3132	ev_wrap.h
2348		3133
2349	ev_win32.c required on win32 platforms only	3134	ev_win32.c required on win32 platforms only
2350		3135
2351	ev_select.c only when select backend is enabled (which is enabled by default)	3136	ev_select.c only when select backend is enabled (which is enabled by default)
2352	ev_poll.c only when poll backend is enabled (disabled by default)	3137	ev_poll.c only when poll backend is enabled (disabled by default)
2353	ev_epoll.c only when the epoll backend is enabled (disabled by default)	3138	ev_epoll.c only when the epoll backend is enabled (disabled by default)
2354	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	3139	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
2355	ev_port.c only when the solaris port backend is enabled (disabled by default)	3140	ev_port.c only when the solaris port backend is enabled (disabled by default)
2356		3141
2357	F<ev.c> includes the backend files directly when enabled, so you only need	3142	F<ev.c> includes the backend files directly when enabled, so you only need
2358	to compile this single file.	3143	to compile this single file.
2359		3144
2360	=head3 LIBEVENT COMPATIBILITY API	3145	=head3 LIBEVENT COMPATIBILITY API
2361		3146
2362	To include the libevent compatibility API, also include:	3147	To include the libevent compatibility API, also include:
2363		3148
2364	#include "event.c"	3149	#include "event.c"
2365		3150
2366	in the file including F<ev.c>, and:	3151	in the file including F<ev.c>, and:
2367		3152
2368	#include "event.h"	3153	#include "event.h"
2369		3154
2370	in the files that want to use the libevent API. This also includes F<ev.h>.	3155	in the files that want to use the libevent API. This also includes F<ev.h>.
2371		3156
2372	You need the following additional files for this:	3157	You need the following additional files for this:
2373		3158
2374	event.h	3159	event.h
2375	event.c	3160	event.c
2376		3161
2377	=head3 AUTOCONF SUPPORT	3162	=head3 AUTOCONF SUPPORT
2378		3163
2379	Instead of using C<EV_STANDALONE=1> and providing your config in	3164	Instead of using C<EV_STANDALONE=1> and providing your configuration in
2380	whatever way you want, you can also C<m4_include([libev.m4])> in your	3165	whatever way you want, you can also C<m4_include([libev.m4])> in your
2381	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then	3166	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
2382	include F<config.h> and configure itself accordingly.	3167	include F<config.h> and configure itself accordingly.
2383		3168
2384	For this of course you need the m4 file:	3169	For this of course you need the m4 file:
2385		3170
2386	libev.m4	3171	libev.m4
2387		3172
2388	=head2 PREPROCESSOR SYMBOLS/MACROS	3173	=head2 PREPROCESSOR SYMBOLS/MACROS
2389		3174
2390	Libev can be configured via a variety of preprocessor symbols you have to define	3175	Libev can be configured via a variety of preprocessor symbols you have to
2391	before including any of its files. The default is not to build for multiplicity	3176	define before including any of its files. The default in the absence of
2392	and only include the select backend.	3177	autoconf is documented for every option.
2393		3178
2394	=over 4	3179	=over 4
2395		3180
2396	=item EV_STANDALONE	3181	=item EV_STANDALONE
2397		3182
…		…
2402	F<event.h> that are not directly supported by the libev core alone.	3187	F<event.h> that are not directly supported by the libev core alone.
2403		3188
2404	=item EV_USE_MONOTONIC	3189	=item EV_USE_MONOTONIC
2405		3190
2406	If defined to be C<1>, libev will try to detect the availability of the	3191	If defined to be C<1>, libev will try to detect the availability of the
2407	monotonic clock option at both compiletime and runtime. Otherwise no use	3192	monotonic clock option at both compile time and runtime. Otherwise no use
2408	of the monotonic clock option will be attempted. If you enable this, you	3193	of the monotonic clock option will be attempted. If you enable this, you
2409	usually have to link against librt or something similar. Enabling it when	3194	usually have to link against librt or something similar. Enabling it when
2410	the functionality isn't available is safe, though, although you have	3195	the functionality isn't available is safe, though, although you have
2411	to make sure you link against any libraries where the C<clock_gettime>	3196	to make sure you link against any libraries where the C<clock_gettime>
2412	function is hiding in (often F<-lrt>).	3197	function is hiding in (often F<-lrt>).
2413		3198
2414	=item EV_USE_REALTIME	3199	=item EV_USE_REALTIME
2415		3200
2416	If defined to be C<1>, libev will try to detect the availability of the	3201	If defined to be C<1>, libev will try to detect the availability of the
2417	realtime clock option at compiletime (and assume its availability at	3202	real-time clock option at compile time (and assume its availability at
2418	runtime if successful). Otherwise no use of the realtime clock option will	3203	runtime if successful). Otherwise no use of the real-time clock option will
2419	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3204	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
2420	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3205	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
2421	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3206	note about libraries in the description of C<EV_USE_MONOTONIC>, though.
2422		3207
2423	=item EV_USE_NANOSLEEP	3208	=item EV_USE_NANOSLEEP
2424		3209
2425	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3210	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2426	and will use it for delays. Otherwise it will use C<select ()>.	3211	and will use it for delays. Otherwise it will use C<select ()>.
2427		3212
		3213	=item EV_USE_EVENTFD
		3214
		3215	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		3216	available and will probe for kernel support at runtime. This will improve
		3217	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		3218	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		3219	2.7 or newer, otherwise disabled.
		3220
2428	=item EV_USE_SELECT	3221	=item EV_USE_SELECT
2429		3222
2430	If undefined or defined to be C<1>, libev will compile in support for the	3223	If undefined or defined to be C<1>, libev will compile in support for the
2431	C<select>(2) backend. No attempt at autodetection will be done: if no	3224	C<select>(2) backend. No attempt at auto-detection will be done: if no
2432	other method takes over, select will be it. Otherwise the select backend	3225	other method takes over, select will be it. Otherwise the select backend
2433	will not be compiled in.	3226	will not be compiled in.
2434		3227
2435	=item EV_SELECT_USE_FD_SET	3228	=item EV_SELECT_USE_FD_SET
2436		3229
2437	If defined to C<1>, then the select backend will use the system C<fd_set>	3230	If defined to C<1>, then the select backend will use the system C<fd_set>
2438	structure. This is useful if libev doesn't compile due to a missing	3231	structure. This is useful if libev doesn't compile due to a missing
2439	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	3232	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on
2440	exotic systems. This usually limits the range of file descriptors to some	3233	exotic systems. This usually limits the range of file descriptors to some
2441	low limit such as 1024 or might have other limitations (winsocket only	3234	low limit such as 1024 or might have other limitations (winsocket only
2442	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3235	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
2443	influence the size of the C<fd_set> used.	3236	influence the size of the C<fd_set> used.
2444		3237
…		…
2450	be used is the winsock select). This means that it will call	3243	be used is the winsock select). This means that it will call
2451	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3244	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2452	it is assumed that all these functions actually work on fds, even	3245	it is assumed that all these functions actually work on fds, even
2453	on win32. Should not be defined on non-win32 platforms.	3246	on win32. Should not be defined on non-win32 platforms.
2454		3247
		3248	=item EV_FD_TO_WIN32_HANDLE
		3249
		3250	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		3251	file descriptors to socket handles. When not defining this symbol (the
		3252	default), then libev will call C<_get_osfhandle>, which is usually
		3253	correct. In some cases, programs use their own file descriptor management,
		3254	in which case they can provide this function to map fds to socket handles.
		3255
2455	=item EV_USE_POLL	3256	=item EV_USE_POLL
2456		3257
2457	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3258	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2458	backend. Otherwise it will be enabled on non-win32 platforms. It	3259	backend. Otherwise it will be enabled on non-win32 platforms. It
2459	takes precedence over select.	3260	takes precedence over select.
2460		3261
2461	=item EV_USE_EPOLL	3262	=item EV_USE_EPOLL
2462		3263
2463	If defined to be C<1>, libev will compile in support for the Linux	3264	If defined to be C<1>, libev will compile in support for the Linux
2464	C<epoll>(7) backend. Its availability will be detected at runtime,	3265	C<epoll>(7) backend. Its availability will be detected at runtime,
2465	otherwise another method will be used as fallback. This is the	3266	otherwise another method will be used as fallback. This is the preferred
2466	preferred backend for GNU/Linux systems.	3267	backend for GNU/Linux systems. If undefined, it will be enabled if the
		3268	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2467		3269
2468	=item EV_USE_KQUEUE	3270	=item EV_USE_KQUEUE
2469		3271
2470	If defined to be C<1>, libev will compile in support for the BSD style	3272	If defined to be C<1>, libev will compile in support for the BSD style
2471	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	3273	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
2484	otherwise another method will be used as fallback. This is the preferred	3286	otherwise another method will be used as fallback. This is the preferred
2485	backend for Solaris 10 systems.	3287	backend for Solaris 10 systems.
2486		3288
2487	=item EV_USE_DEVPOLL	3289	=item EV_USE_DEVPOLL
2488		3290
2489	reserved for future expansion, works like the USE symbols above.	3291	Reserved for future expansion, works like the USE symbols above.
2490		3292
2491	=item EV_USE_INOTIFY	3293	=item EV_USE_INOTIFY
2492		3294
2493	If defined to be C<1>, libev will compile in support for the Linux inotify	3295	If defined to be C<1>, libev will compile in support for the Linux inotify
2494	interface to speed up C<ev_stat> watchers. Its actual availability will	3296	interface to speed up C<ev_stat> watchers. Its actual availability will
2495	be detected at runtime.	3297	be detected at runtime. If undefined, it will be enabled if the headers
		3298	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		3299
		3300	=item EV_ATOMIC_T
		3301
		3302	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		3303	access is atomic with respect to other threads or signal contexts. No such
		3304	type is easily found in the C language, so you can provide your own type
		3305	that you know is safe for your purposes. It is used both for signal handler "locking"
		3306	as well as for signal and thread safety in C<ev_async> watchers.
		3307
		3308	In the absence of this define, libev will use C<sig_atomic_t volatile>
		3309	(from F<signal.h>), which is usually good enough on most platforms.
2496		3310
2497	=item EV_H	3311	=item EV_H
2498		3312
2499	The name of the F<ev.h> header file used to include it. The default if	3313	The name of the F<ev.h> header file used to include it. The default if
2500	undefined is C<"ev.h"> in F<event.h> and F<ev.c>. This can be used to	3314	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
2501	virtually rename the F<ev.h> header file in case of conflicts.	3315	used to virtually rename the F<ev.h> header file in case of conflicts.
2502		3316
2503	=item EV_CONFIG_H	3317	=item EV_CONFIG_H
2504		3318
2505	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	3319	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2506	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	3320	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2507	C<EV_H>, above.	3321	C<EV_H>, above.
2508		3322
2509	=item EV_EVENT_H	3323	=item EV_EVENT_H
2510		3324
2511	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	3325	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2512	of how the F<event.h> header can be found, the dfeault is C<"event.h">.	3326	of how the F<event.h> header can be found, the default is C<"event.h">.
2513		3327
2514	=item EV_PROTOTYPES	3328	=item EV_PROTOTYPES
2515		3329
2516	If defined to be C<0>, then F<ev.h> will not define any function	3330	If defined to be C<0>, then F<ev.h> will not define any function
2517	prototypes, but still define all the structs and other symbols. This is	3331	prototypes, but still define all the structs and other symbols. This is
…		…
2538	When doing priority-based operations, libev usually has to linearly search	3352	When doing priority-based operations, libev usually has to linearly search
2539	all the priorities, so having many of them (hundreds) uses a lot of space	3353	all the priorities, so having many of them (hundreds) uses a lot of space
2540	and time, so using the defaults of five priorities (-2 .. +2) is usually	3354	and time, so using the defaults of five priorities (-2 .. +2) is usually
2541	fine.	3355	fine.
2542		3356
2543	If your embedding app does not need any priorities, defining these both to	3357	If your embedding application does not need any priorities, defining these
2544	C<0> will save some memory and cpu.	3358	both to C<0> will save some memory and CPU.
2545		3359
2546	=item EV_PERIODIC_ENABLE	3360	=item EV_PERIODIC_ENABLE
2547		3361
2548	If undefined or defined to be C<1>, then periodic timers are supported. If	3362	If undefined or defined to be C<1>, then periodic timers are supported. If
2549	defined to be C<0>, then they are not. Disabling them saves a few kB of	3363	defined to be C<0>, then they are not. Disabling them saves a few kB of
…		…
2556	code.	3370	code.
2557		3371
2558	=item EV_EMBED_ENABLE	3372	=item EV_EMBED_ENABLE
2559		3373
2560	If undefined or defined to be C<1>, then embed watchers are supported. If	3374	If undefined or defined to be C<1>, then embed watchers are supported. If
2561	defined to be C<0>, then they are not.	3375	defined to be C<0>, then they are not. Embed watchers rely on most other
		3376	watcher types, which therefore must not be disabled.
2562		3377
2563	=item EV_STAT_ENABLE	3378	=item EV_STAT_ENABLE
2564		3379
2565	If undefined or defined to be C<1>, then stat watchers are supported. If	3380	If undefined or defined to be C<1>, then stat watchers are supported. If
2566	defined to be C<0>, then they are not.	3381	defined to be C<0>, then they are not.
…		…
2568	=item EV_FORK_ENABLE	3383	=item EV_FORK_ENABLE
2569		3384
2570	If undefined or defined to be C<1>, then fork watchers are supported. If	3385	If undefined or defined to be C<1>, then fork watchers are supported. If
2571	defined to be C<0>, then they are not.	3386	defined to be C<0>, then they are not.
2572		3387
		3388	=item EV_ASYNC_ENABLE
		3389
		3390	If undefined or defined to be C<1>, then async watchers are supported. If
		3391	defined to be C<0>, then they are not.
		3392
2573	=item EV_MINIMAL	3393	=item EV_MINIMAL
2574		3394
2575	If you need to shave off some kilobytes of code at the expense of some	3395	If you need to shave off some kilobytes of code at the expense of some
2576	speed, define this symbol to C<1>. Currently only used for gcc to override	3396	speed, define this symbol to C<1>. Currently this is used to override some
2577	some inlining decisions, saves roughly 30% codesize of amd64.	3397	inlining decisions, saves roughly 30% code size on amd64. It also selects a
		3398	much smaller 2-heap for timer management over the default 4-heap.
2578		3399
2579	=item EV_PID_HASHSIZE	3400	=item EV_PID_HASHSIZE
2580		3401
2581	C<ev_child> watchers use a small hash table to distribute workload by	3402	C<ev_child> watchers use a small hash table to distribute workload by
2582	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3403	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
2589	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	3410	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
2590	usually more than enough. If you need to manage thousands of C<ev_stat>	3411	usually more than enough. If you need to manage thousands of C<ev_stat>
2591	watchers you might want to increase this value (I<must> be a power of	3412	watchers you might want to increase this value (I<must> be a power of
2592	two).	3413	two).
2593		3414
		3415	=item EV_USE_4HEAP
		3416
		3417	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3418	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		3419	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		3420	faster performance with many (thousands) of watchers.
		3421
		3422	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3423	(disabled).
		3424
		3425	=item EV_HEAP_CACHE_AT
		3426
		3427	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3428	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		3429	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3430	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3431	but avoids random read accesses on heap changes. This improves performance
		3432	noticeably with many (hundreds) of watchers.
		3433
		3434	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3435	(disabled).
		3436
		3437	=item EV_VERIFY
		3438
		3439	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3440	be done: If set to C<0>, no internal verification code will be compiled
		3441	in. If set to C<1>, then verification code will be compiled in, but not
		3442	called. If set to C<2>, then the internal verification code will be
		3443	called once per loop, which can slow down libev. If set to C<3>, then the
		3444	verification code will be called very frequently, which will slow down
		3445	libev considerably.
		3446
		3447	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3448	C<0>.
		3449
2594	=item EV_COMMON	3450	=item EV_COMMON
2595		3451
2596	By default, all watchers have a C<void *data> member. By redefining	3452	By default, all watchers have a C<void *data> member. By redefining
2597	this macro to a something else you can include more and other types of	3453	this macro to a something else you can include more and other types of
2598	members. You have to define it each time you include one of the files,	3454	members. You have to define it each time you include one of the files,
2599	though, and it must be identical each time.	3455	though, and it must be identical each time.
2600		3456
2601	For example, the perl EV module uses something like this:	3457	For example, the perl EV module uses something like this:
2602		3458
2603	#define EV_COMMON \	3459	#define EV_COMMON \
2604	SV *self; /* contains this struct */ \	3460	SV *self; /* contains this struct */ \
2605	SV *cb_sv, *fh /* note no trailing ";" */	3461	SV *cb_sv, *fh /* note no trailing ";" */
2606		3462
2607	=item EV_CB_DECLARE (type)	3463	=item EV_CB_DECLARE (type)
2608		3464
2609	=item EV_CB_INVOKE (watcher, revents)	3465	=item EV_CB_INVOKE (watcher, revents)
2610		3466
…		…
2615	definition and a statement, respectively. See the F<ev.h> header file for	3471	definition and a statement, respectively. See the F<ev.h> header file for
2616	their default definitions. One possible use for overriding these is to	3472	their default definitions. One possible use for overriding these is to
2617	avoid the C<struct ev_loop *> as first argument in all cases, or to use	3473	avoid the C<struct ev_loop *> as first argument in all cases, or to use
2618	method calls instead of plain function calls in C++.	3474	method calls instead of plain function calls in C++.
2619		3475
		3476	=back
		3477
2620	=head2 EXPORTED API SYMBOLS	3478	=head2 EXPORTED API SYMBOLS
2621		3479
2622	If you need to re-export the API (e.g. via a dll) and you need a list of	3480	If you need to re-export the API (e.g. via a DLL) and you need a list of
2623	exported symbols, you can use the provided F<Symbol.*> files which list	3481	exported symbols, you can use the provided F<Symbol.*> files which list
2624	all public symbols, one per line:	3482	all public symbols, one per line:
2625		3483
2626	Symbols.ev for libev proper	3484	Symbols.ev for libev proper
2627	Symbols.event for the libevent emulation	3485	Symbols.event for the libevent emulation
2628		3486
2629	This can also be used to rename all public symbols to avoid clashes with	3487	This can also be used to rename all public symbols to avoid clashes with
2630	multiple versions of libev linked together (which is obviously bad in	3488	multiple versions of libev linked together (which is obviously bad in
2631	itself, but sometimes it is inconvinient to avoid this).	3489	itself, but sometimes it is inconvenient to avoid this).
2632		3490
2633	A sed command like this will create wrapper C<#define>'s that you need to	3491	A sed command like this will create wrapper C<#define>'s that you need to
2634	include before including F<ev.h>:	3492	include before including F<ev.h>:
2635		3493
2636	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h	3494	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
…		…
2653	file.	3511	file.
2654		3512
2655	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	3513	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2656	that everybody includes and which overrides some configure choices:	3514	that everybody includes and which overrides some configure choices:
2657		3515
2658	#define EV_MINIMAL 1	3516	#define EV_MINIMAL 1
2659	#define EV_USE_POLL 0	3517	#define EV_USE_POLL 0
2660	#define EV_MULTIPLICITY 0	3518	#define EV_MULTIPLICITY 0
2661	#define EV_PERIODIC_ENABLE 0	3519	#define EV_PERIODIC_ENABLE 0
2662	#define EV_STAT_ENABLE 0	3520	#define EV_STAT_ENABLE 0
2663	#define EV_FORK_ENABLE 0	3521	#define EV_FORK_ENABLE 0
2664	#define EV_CONFIG_H <config.h>	3522	#define EV_CONFIG_H <config.h>
2665	#define EV_MINPRI 0	3523	#define EV_MINPRI 0
2666	#define EV_MAXPRI 0	3524	#define EV_MAXPRI 0
2667		3525
2668	#include "ev++.h"	3526	#include "ev++.h"
2669		3527
2670	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	3528	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2671		3529
2672	#include "ev_cpp.h"	3530	#include "ev_cpp.h"
2673	#include "ev.c"	3531	#include "ev.c"
2674		3532
		3533	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
2675		3534
		3535	=head2 THREADS AND COROUTINES
		3536
		3537	=head3 THREADS
		3538
		3539	All libev functions are reentrant and thread-safe unless explicitly
		3540	documented otherwise, but libev implements no locking itself. This means
		3541	that you can use as many loops as you want in parallel, as long as there
		3542	are no concurrent calls into any libev function with the same loop
		3543	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		3544	of course): libev guarantees that different event loops share no data
		3545	structures that need any locking.
		3546
		3547	Or to put it differently: calls with different loop parameters can be done
		3548	concurrently from multiple threads, calls with the same loop parameter
		3549	must be done serially (but can be done from different threads, as long as
		3550	only one thread ever is inside a call at any point in time, e.g. by using
		3551	a mutex per loop).
		3552
		3553	Specifically to support threads (and signal handlers), libev implements
		3554	so-called C<ev_async> watchers, which allow some limited form of
		3555	concurrency on the same event loop, namely waking it up "from the
		3556	outside".
		3557
		3558	If you want to know which design (one loop, locking, or multiple loops
		3559	without or something else still) is best for your problem, then I cannot
		3560	help you, but here is some generic advice:
		3561
		3562	=over 4
		3563
		3564	=item * most applications have a main thread: use the default libev loop
		3565	in that thread, or create a separate thread running only the default loop.
		3566
		3567	This helps integrating other libraries or software modules that use libev
		3568	themselves and don't care/know about threading.
		3569
		3570	=item * one loop per thread is usually a good model.
		3571
		3572	Doing this is almost never wrong, sometimes a better-performance model
		3573	exists, but it is always a good start.
		3574
		3575	=item * other models exist, such as the leader/follower pattern, where one
		3576	loop is handed through multiple threads in a kind of round-robin fashion.
		3577
		3578	Choosing a model is hard - look around, learn, know that usually you can do
		3579	better than you currently do :-)
		3580
		3581	=item * often you need to talk to some other thread which blocks in the
		3582	event loop.
		3583
		3584	C<ev_async> watchers can be used to wake them up from other threads safely
		3585	(or from signal contexts...).
		3586
		3587	An example use would be to communicate signals or other events that only
		3588	work in the default loop by registering the signal watcher with the
		3589	default loop and triggering an C<ev_async> watcher from the default loop
		3590	watcher callback into the event loop interested in the signal.
		3591
		3592	=back
		3593
		3594	=head3 COROUTINES
		3595
		3596	Libev is very accommodating to coroutines ("cooperative threads"):
		3597	libev fully supports nesting calls to its functions from different
		3598	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3599	different coroutines, and switch freely between both coroutines running the
		3600	loop, as long as you don't confuse yourself). The only exception is that
		3601	you must not do this from C<ev_periodic> reschedule callbacks.
		3602
		3603	Care has been taken to ensure that libev does not keep local state inside
		3604	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		3605	they do not call any callbacks.
		3606
		3607	=head2 COMPILER WARNINGS
		3608
		3609	Depending on your compiler and compiler settings, you might get no or a
		3610	lot of warnings when compiling libev code. Some people are apparently
		3611	scared by this.
		3612
		3613	However, these are unavoidable for many reasons. For one, each compiler
		3614	has different warnings, and each user has different tastes regarding
		3615	warning options. "Warn-free" code therefore cannot be a goal except when
		3616	targeting a specific compiler and compiler-version.
		3617
		3618	Another reason is that some compiler warnings require elaborate
		3619	workarounds, or other changes to the code that make it less clear and less
		3620	maintainable.
		3621
		3622	And of course, some compiler warnings are just plain stupid, or simply
		3623	wrong (because they don't actually warn about the condition their message
		3624	seems to warn about). For example, certain older gcc versions had some
		3625	warnings that resulted an extreme number of false positives. These have
		3626	been fixed, but some people still insist on making code warn-free with
		3627	such buggy versions.
		3628
		3629	While libev is written to generate as few warnings as possible,
		3630	"warn-free" code is not a goal, and it is recommended not to build libev
		3631	with any compiler warnings enabled unless you are prepared to cope with
		3632	them (e.g. by ignoring them). Remember that warnings are just that:
		3633	warnings, not errors, or proof of bugs.
		3634
		3635
		3636	=head2 VALGRIND
		3637
		3638	Valgrind has a special section here because it is a popular tool that is
		3639	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		3640
		3641	If you think you found a bug (memory leak, uninitialised data access etc.)
		3642	in libev, then check twice: If valgrind reports something like:
		3643
		3644	==2274== definitely lost: 0 bytes in 0 blocks.
		3645	==2274== possibly lost: 0 bytes in 0 blocks.
		3646	==2274== still reachable: 256 bytes in 1 blocks.
		3647
		3648	Then there is no memory leak, just as memory accounted to global variables
		3649	is not a memleak - the memory is still being referenced, and didn't leak.
		3650
		3651	Similarly, under some circumstances, valgrind might report kernel bugs
		3652	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		3653	although an acceptable workaround has been found here), or it might be
		3654	confused.
		3655
		3656	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		3657	make it into some kind of religion.
		3658
		3659	If you are unsure about something, feel free to contact the mailing list
		3660	with the full valgrind report and an explanation on why you think this
		3661	is a bug in libev (best check the archives, too :). However, don't be
		3662	annoyed when you get a brisk "this is no bug" answer and take the chance
		3663	of learning how to interpret valgrind properly.
		3664
		3665	If you need, for some reason, empty reports from valgrind for your project
		3666	I suggest using suppression lists.
		3667
		3668
		3669	=head1 PORTABILITY NOTES
		3670
		3671	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		3672
		3673	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		3674	requires, and its I/O model is fundamentally incompatible with the POSIX
		3675	model. Libev still offers limited functionality on this platform in
		3676	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		3677	descriptors. This only applies when using Win32 natively, not when using
		3678	e.g. cygwin.
		3679
		3680	Lifting these limitations would basically require the full
		3681	re-implementation of the I/O system. If you are into these kinds of
		3682	things, then note that glib does exactly that for you in a very portable
		3683	way (note also that glib is the slowest event library known to man).
		3684
		3685	There is no supported compilation method available on windows except
		3686	embedding it into other applications.
		3687
		3688	Not a libev limitation but worth mentioning: windows apparently doesn't
		3689	accept large writes: instead of resulting in a partial write, windows will
		3690	either accept everything or return C<ENOBUFS> if the buffer is too large,
		3691	so make sure you only write small amounts into your sockets (less than a
		3692	megabyte seems safe, but this apparently depends on the amount of memory
		3693	available).
		3694
		3695	Due to the many, low, and arbitrary limits on the win32 platform and
		3696	the abysmal performance of winsockets, using a large number of sockets
		3697	is not recommended (and not reasonable). If your program needs to use
		3698	more than a hundred or so sockets, then likely it needs to use a totally
		3699	different implementation for windows, as libev offers the POSIX readiness
		3700	notification model, which cannot be implemented efficiently on windows
		3701	(Microsoft monopoly games).
		3702
		3703	A typical way to use libev under windows is to embed it (see the embedding
		3704	section for details) and use the following F<evwrap.h> header file instead
		3705	of F<ev.h>:
		3706
		3707	#define EV_STANDALONE /* keeps ev from requiring config.h */
		3708	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		3709
		3710	#include "ev.h"
		3711
		3712	And compile the following F<evwrap.c> file into your project (make sure
		3713	you do I<not> compile the F<ev.c> or any other embedded source files!):
		3714
		3715	#include "evwrap.h"
		3716	#include "ev.c"
		3717
		3718	=over 4
		3719
		3720	=item The winsocket select function
		3721
		3722	The winsocket C<select> function doesn't follow POSIX in that it
		3723	requires socket I<handles> and not socket I<file descriptors> (it is
		3724	also extremely buggy). This makes select very inefficient, and also
		3725	requires a mapping from file descriptors to socket handles (the Microsoft
		3726	C runtime provides the function C<_open_osfhandle> for this). See the
		3727	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		3728	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		3729
		3730	The configuration for a "naked" win32 using the Microsoft runtime
		3731	libraries and raw winsocket select is:
		3732
		3733	#define EV_USE_SELECT 1
		3734	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		3735
		3736	Note that winsockets handling of fd sets is O(n), so you can easily get a
		3737	complexity in the O(n²) range when using win32.
		3738
		3739	=item Limited number of file descriptors
		3740
		3741	Windows has numerous arbitrary (and low) limits on things.
		3742
		3743	Early versions of winsocket's select only supported waiting for a maximum
		3744	of C<64> handles (probably owning to the fact that all windows kernels
		3745	can only wait for C<64> things at the same time internally; Microsoft
		3746	recommends spawning a chain of threads and wait for 63 handles and the
		3747	previous thread in each. Great).
		3748
		3749	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		3750	to some high number (e.g. C<2048>) before compiling the winsocket select
		3751	call (which might be in libev or elsewhere, for example, perl does its own
		3752	select emulation on windows).
		3753
		3754	Another limit is the number of file descriptors in the Microsoft runtime
		3755	libraries, which by default is C<64> (there must be a hidden I<64> fetish
		3756	or something like this inside Microsoft). You can increase this by calling
		3757	C<_setmaxstdio>, which can increase this limit to C<2048> (another
		3758	arbitrary limit), but is broken in many versions of the Microsoft runtime
		3759	libraries.
		3760
		3761	This might get you to about C<512> or C<2048> sockets (depending on
		3762	windows version and/or the phase of the moon). To get more, you need to
		3763	wrap all I/O functions and provide your own fd management, but the cost of
		3764	calling select (O(n²)) will likely make this unworkable.
		3765
		3766	=back
		3767
		3768	=head2 PORTABILITY REQUIREMENTS
		3769
		3770	In addition to a working ISO-C implementation and of course the
		3771	backend-specific APIs, libev relies on a few additional extensions:
		3772
		3773	=over 4
		3774
		3775	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		3776	calling conventions regardless of C<ev_watcher_type *>.
		3777
		3778	Libev assumes not only that all watcher pointers have the same internal
		3779	structure (guaranteed by POSIX but not by ISO C for example), but it also
		3780	assumes that the same (machine) code can be used to call any watcher
		3781	callback: The watcher callbacks have different type signatures, but libev
		3782	calls them using an C<ev_watcher *> internally.
		3783
		3784	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3785
		3786	The type C<sig_atomic_t volatile> (or whatever is defined as
		3787	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		3788	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3789	believed to be sufficiently portable.
		3790
		3791	=item C<sigprocmask> must work in a threaded environment
		3792
		3793	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3794	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3795	pthread implementations will either allow C<sigprocmask> in the "main
		3796	thread" or will block signals process-wide, both behaviours would
		3797	be compatible with libev. Interaction between C<sigprocmask> and
		3798	C<pthread_sigmask> could complicate things, however.
		3799
		3800	The most portable way to handle signals is to block signals in all threads
		3801	except the initial one, and run the default loop in the initial thread as
		3802	well.
		3803
		3804	=item C<long> must be large enough for common memory allocation sizes
		3805
		3806	To improve portability and simplify its API, libev uses C<long> internally
		3807	instead of C<size_t> when allocating its data structures. On non-POSIX
		3808	systems (Microsoft...) this might be unexpectedly low, but is still at
		3809	least 31 bits everywhere, which is enough for hundreds of millions of
		3810	watchers.
		3811
		3812	=item C<double> must hold a time value in seconds with enough accuracy
		3813
		3814	The type C<double> is used to represent timestamps. It is required to
		3815	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3816	enough for at least into the year 4000. This requirement is fulfilled by
		3817	implementations implementing IEEE 754 (basically all existing ones).
		3818
		3819	=back
		3820
		3821	If you know of other additional requirements drop me a note.
		3822
		3823
2676	=head1 COMPLEXITIES	3824	=head1 ALGORITHMIC COMPLEXITIES
2677		3825
2678	In this section the complexities of (many of) the algorithms used inside	3826	In this section the complexities of (many of) the algorithms used inside
2679	libev will be explained. For complexity discussions about backends see the	3827	libev will be documented. For complexity discussions about backends see
2680	documentation for C<ev_default_init>.	3828	the documentation for C<ev_default_init>.
2681		3829
2682	All of the following are about amortised time: If an array needs to be	3830	All of the following are about amortised time: If an array needs to be
2683	extended, libev needs to realloc and move the whole array, but this	3831	extended, libev needs to realloc and move the whole array, but this
2684	happens asymptotically never with higher number of elements, so O(1) might	3832	happens asymptotically rarer with higher number of elements, so O(1) might
2685	mean it might do a lengthy realloc operation in rare cases, but on average	3833	mean that libev does a lengthy realloc operation in rare cases, but on
2686	it is much faster and asymptotically approaches constant time.	3834	average it is much faster and asymptotically approaches constant time.
2687		3835
2688	=over 4	3836	=over 4
2689		3837
2690	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	3838	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2691		3839
2692	This means that, when you have a watcher that triggers in one hour and	3840	This means that, when you have a watcher that triggers in one hour and
2693	there are 100 watchers that would trigger before that then inserting will	3841	there are 100 watchers that would trigger before that, then inserting will
2694	have to skip roughly seven (C<ld 100>) of these watchers.	3842	have to skip roughly seven (C<ld 100>) of these watchers.
2695		3843
2696	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	3844	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
2697		3845
2698	That means that changing a timer costs less than removing/adding them	3846	That means that changing a timer costs less than removing/adding them,
2699	as only the relative motion in the event queue has to be paid for.	3847	as only the relative motion in the event queue has to be paid for.
2700		3848
2701	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	3849	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
2702		3850
2703	These just add the watcher into an array or at the head of a list.	3851	These just add the watcher into an array or at the head of a list.
2704		3852
2705	=item Stopping check/prepare/idle watchers: O(1)	3853	=item Stopping check/prepare/idle/fork/async watchers: O(1)
2706		3854
2707	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	3855	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
2708		3856
2709	These watchers are stored in lists then need to be walked to find the	3857	These watchers are stored in lists, so they need to be walked to find the
2710	correct watcher to remove. The lists are usually short (you don't usually	3858	correct watcher to remove. The lists are usually short (you don't usually
2711	have many watchers waiting for the same fd or signal).	3859	have many watchers waiting for the same fd or signal: one is typical, two
		3860	is rare).
2712		3861
2713	=item Finding the next timer in each loop iteration: O(1)	3862	=item Finding the next timer in each loop iteration: O(1)
2714		3863
2715	By virtue of using a binary heap, the next timer is always found at the	3864	By virtue of using a binary or 4-heap, the next timer is always found at a
2716	beginning of the storage array.	3865	fixed position in the storage array.
2717		3866
2718	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	3867	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2719		3868
2720	A change means an I/O watcher gets started or stopped, which requires	3869	A change means an I/O watcher gets started or stopped, which requires
2721	libev to recalculate its status (and possibly tell the kernel, depending	3870	libev to recalculate its status (and possibly tell the kernel, depending
2722	on backend and wether C<ev_io_set> was used).	3871	on backend and whether C<ev_io_set> was used).
2723		3872
2724	=item Activating one watcher (putting it into the pending state): O(1)	3873	=item Activating one watcher (putting it into the pending state): O(1)
2725		3874
2726	=item Priority handling: O(number_of_priorities)	3875	=item Priority handling: O(number_of_priorities)
2727		3876
2728	Priorities are implemented by allocating some space for each	3877	Priorities are implemented by allocating some space for each
2729	priority. When doing priority-based operations, libev usually has to	3878	priority. When doing priority-based operations, libev usually has to
2730	linearly search all the priorities, but starting/stopping and activating	3879	linearly search all the priorities, but starting/stopping and activating
2731	watchers becomes O(1) w.r.t. prioritiy handling.	3880	watchers becomes O(1) with respect to priority handling.
		3881
		3882	=item Sending an ev_async: O(1)
		3883
		3884	=item Processing ev_async_send: O(number_of_async_watchers)
		3885
		3886	=item Processing signals: O(max_signal_number)
		3887
		3888	Sending involves a system call I<iff> there were no other C<ev_async_send>
		3889	calls in the current loop iteration. Checking for async and signal events
		3890	involves iterating over all running async watchers or all signal numbers.
2732		3891
2733	=back	3892	=back
2734		3893
2735		3894
2736	=head1 AUTHOR	3895	=head1 AUTHOR
2737		3896
2738	Marc Lehmann <libev@schmorp.de>.	3897	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
2739		3898

Diff Legend

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