ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.131 by root, Tue Feb 19 17:09:28 2008 UTC vs.
Revision 1.259 by root, Sun Jul 19 01:36:34 2009 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	struct 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 ABOUT THIS DOCUMENT
52		68
		69	This document documents the libev software package.
		70
53	The newest version of this document is also available as a html-formatted	71	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	72	web page you might find easier to navigate when reading it for the first
55	time: L<http://cvs.schmorp.de/libev/ev.html>.	73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 ABOUT LIBEV
56		84
57	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
58	file descriptor being readable or a timeout occurring), and it will manage	86	file descriptor being readable or a timeout occurring), and it will manage
59	these event sources and provide your program with events.	87	these event sources and provide your program with events.
60		88
…		…
84	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	112	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
85	for example).	113	for example).
86		114
87	=head2 CONVENTIONS	115	=head2 CONVENTIONS
88		116
89	Libev is very configurable. In this manual the default configuration will	117	Libev is very configurable. In this manual the default (and most common)
90	be described, which supports multiple event loops. For more info about	118	configuration will be described, which supports multiple event loops. For
91	various configuration options please have a look at B<EMBED> section in	119	more info about various configuration options please have a look at
92	this manual. If libev was configured without support for multiple event	120	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>	121	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.	122	name C<loop> (which is always of type C<ev_loop *>) will not have
		123	this argument.
95		124
96	=head2 TIME REPRESENTATION	125	=head2 TIME REPRESENTATION
97		126
98	Libev represents time as a single floating point number, representing the	127	Libev represents time as a single floating point number, representing
99	(fractional) number of seconds since the (POSIX) epoch (somewhere near	128	the (fractional) number of seconds since the (POSIX) epoch (somewhere
100	the beginning of 1970, details are complicated, don't ask). This type is	129	near the beginning of 1970, details are complicated, don't ask). This
101	called C<ev_tstamp>, which is what you should use too. It usually aliases	130	type is called C<ev_tstamp>, which is what you should use too. It usually
102	to the C<double> type in C, and when you need to do any calculations on	131	aliases to the C<double> type in C. When you need to do any calculations
103	it, you should treat it as some floatingpoint value. Unlike the name	132	on 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	133	component C<stamp> might indicate, it is also used for time differences
105	throughout libev.	134	throughout libev.
		135
		136	=head1 ERROR HANDLING
		137
		138	Libev knows three classes of errors: operating system errors, usage errors
		139	and internal errors (bugs).
		140
		141	When libev catches an operating system error it cannot handle (for example
		142	a system call indicating a condition libev cannot fix), it calls the callback
		143	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		144	abort. The default is to print a diagnostic message and to call C<abort
		145	()>.
		146
		147	When libev detects a usage error such as a negative timer interval, then
		148	it will print a diagnostic message and abort (via the C<assert> mechanism,
		149	so C<NDEBUG> will disable this checking): these are programming errors in
		150	the libev caller and need to be fixed there.
		151
		152	Libev also has a few internal error-checking C<assert>ions, and also has
		153	extensive consistency checking code. These do not trigger under normal
		154	circumstances, as they indicate either a bug in libev or worse.
		155
106		156
107	=head1 GLOBAL FUNCTIONS	157	=head1 GLOBAL FUNCTIONS
108		158
109	These functions can be called anytime, even before initialising the	159	These functions can be called anytime, even before initialising the
110	library in any way.	160	library in any way.
…		…
119		169
120	=item ev_sleep (ev_tstamp interval)	170	=item ev_sleep (ev_tstamp interval)
121		171
122	Sleep for the given interval: The current thread will be blocked until	172	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	173	either it is interrupted or the given time interval has passed. Basically
124	this is a subsecond-resolution C<sleep ()>.	174	this is a sub-second-resolution C<sleep ()>.
125		175
126	=item int ev_version_major ()	176	=item int ev_version_major ()
127		177
128	=item int ev_version_minor ()	178	=item int ev_version_minor ()
129		179
…		…
142	not a problem.	192	not a problem.
143		193
144	Example: Make sure we haven't accidentally been linked against the wrong	194	Example: Make sure we haven't accidentally been linked against the wrong
145	version.	195	version.
146		196
147	assert (("libev version mismatch",	197	assert (("libev version mismatch",
148	ev_version_major () == EV_VERSION_MAJOR	198	ev_version_major () == EV_VERSION_MAJOR
149	&& ev_version_minor () >= EV_VERSION_MINOR));	199	&& ev_version_minor () >= EV_VERSION_MINOR));
150		200
151	=item unsigned int ev_supported_backends ()	201	=item unsigned int ev_supported_backends ()
152		202
153	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	203	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
154	value) compiled into this binary of libev (independent of their	204	value) compiled into this binary of libev (independent of their
…		…
156	a description of the set values.	206	a description of the set values.
157		207
158	Example: make sure we have the epoll method, because yeah this is cool and	208	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	209	a must have and can we have a torrent of it please!!!11
160		210
161	assert (("sorry, no epoll, no sex",	211	assert (("sorry, no epoll, no sex",
162	ev_supported_backends () & EVBACKEND_EPOLL));	212	ev_supported_backends () & EVBACKEND_EPOLL));
163		213
164	=item unsigned int ev_recommended_backends ()	214	=item unsigned int ev_recommended_backends ()
165		215
166	Return the set of all backends compiled into this binary of libev and also	216	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	217	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	218	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	219	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	220	(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.	221	libev will probe for if you specify no backends explicitly.
172		222
173	=item unsigned int ev_embeddable_backends ()	223	=item unsigned int ev_embeddable_backends ()
174		224
…		…
178	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	228	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
179	recommended ones.	229	recommended ones.
180		230
181	See the description of C<ev_embed> watchers for more info.	231	See the description of C<ev_embed> watchers for more info.
182		232
183	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	233	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
184		234
185	Sets the allocation function to use (the prototype is similar - the	235	Sets the allocation function to use (the prototype is similar - the
186	semantics is identical - to the realloc C function). It is used to	236	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	237	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	238	when memory needs to be allocated (C<size != 0>), the library might abort
189	potentially destructive action. The default is your system realloc	239	or take some potentially destructive action.
190	function.	240
		241	Since some systems (at least OpenBSD and Darwin) fail to implement
		242	correct C<realloc> semantics, libev will use a wrapper around the system
		243	C<realloc> and C<free> functions by default.
191		244
192	You could override this function in high-availability programs to, say,	245	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,	246	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.	247	or even to sleep a while and retry until some memory is available.
195		248
196	Example: Replace the libev allocator with one that waits a bit and then	249	Example: Replace the libev allocator with one that waits a bit and then
197	retries).	250	retries (example requires a standards-compliant C<realloc>).
198		251
199	static void *	252	static void *
200	persistent_realloc (void *ptr, size_t size)	253	persistent_realloc (void *ptr, size_t size)
201	{	254	{
202	for (;;)	255	for (;;)
…		…
211	}	264	}
212		265
213	...	266	...
214	ev_set_allocator (persistent_realloc);	267	ev_set_allocator (persistent_realloc);
215		268
216	=item ev_set_syserr_cb (void (*cb)(const char *msg));	269	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
217		270
218	Set the callback function to call on a retryable syscall error (such	271	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	272	as failed select, poll, epoll_wait). The message is a printable string
220	indicating the system call or subsystem causing the problem. If this	273	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	274	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	275	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	276	requested operation, or, if the condition doesn't go away, do bad stuff
224	(such as abort).	277	(such as abort).
225		278
226	Example: This is basically the same thing that libev does internally, too.	279	Example: This is basically the same thing that libev does internally, too.
…		…
237		290
238	=back	291	=back
239		292
240	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	293	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
241		294
242	An event loop is described by a C<struct ev_loop *>. The library knows two	295	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	296	is I<not> optional in this case, as there is also an C<ev_loop>
244	events, and dynamically created loops which do not.	297	I<function>).
245		298
246	If you use threads, a common model is to run the default event loop	299	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	300	supports signals and child events, and dynamically created loops which do
248	create, you also create another event loop. Libev itself does no locking	301	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		302
253	=over 4	303	=over 4
254		304
255	=item struct ev_loop *ev_default_loop (unsigned int flags)	305	=item struct ev_loop *ev_default_loop (unsigned int flags)
256		306
…		…
260	flags. If that is troubling you, check C<ev_backend ()> afterwards).	310	flags. If that is troubling you, check C<ev_backend ()> afterwards).
261		311
262	If you don't know what event loop to use, use the one returned from this	312	If you don't know what event loop to use, use the one returned from this
263	function.	313	function.
264		314
		315	Note that this function is I<not> thread-safe, so if you want to use it
		316	from multiple threads, you have to lock (note also that this is unlikely,
		317	as loops cannot be shared easily between threads anyway).
		318
265	The default loop is the only loop that can handle C<ev_signal> and	319	The default loop is the only loop that can handle C<ev_signal> and
266	C<ev_child> watchers, and to do this, it always registers a handler	320	C<ev_child> watchers, and to do this, it always registers a handler
267	for C<SIGCHLD>. If this is a problem for your app you can either	321	for C<SIGCHLD>. If this is a problem for your application you can either
268	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	322	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
269	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	323	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
270	C<ev_default_init>.	324	C<ev_default_init>.
271		325
272	The flags argument can be used to specify special behaviour or specific	326	The flags argument can be used to specify special behaviour or specific
…		…
281	The default flags value. Use this if you have no clue (it's the right	335	The default flags value. Use this if you have no clue (it's the right
282	thing, believe me).	336	thing, believe me).
283		337
284	=item C<EVFLAG_NOENV>	338	=item C<EVFLAG_NOENV>
285		339
286	If this flag bit is ored into the flag value (or the program runs setuid	340	If this flag bit is or'ed into the flag value (or the program runs setuid
287	or setgid) then libev will I<not> look at the environment variable	341	or setgid) then libev will I<not> look at the environment variable
288	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	342	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
289	override the flags completely if it is found in the environment. This is	343	override the flags completely if it is found in the environment. This is
290	useful to try out specific backends to test their performance, or to work	344	useful to try out specific backends to test their performance, or to work
291	around bugs.	345	around bugs.
…		…
297	enabling this flag.	351	enabling this flag.
298		352
299	This works by calling C<getpid ()> on every iteration of the loop,	353	This works by calling C<getpid ()> on every iteration of the loop,
300	and thus this might slow down your event loop if you do a lot of loop	354	and thus this might slow down your event loop if you do a lot of loop
301	iterations and little real work, but is usually not noticeable (on my	355	iterations and little real work, but is usually not noticeable (on my
302	Linux system for example, C<getpid> is actually a simple 5-insn sequence	356	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
303	without a syscall and thus I<very> fast, but my Linux system also has	357	without a system call and thus I<very> fast, but my GNU/Linux system also has
304	C<pthread_atfork> which is even faster).	358	C<pthread_atfork> which is even faster).
305		359
306	The big advantage of this flag is that you can forget about fork (and	360	The big advantage of this flag is that you can forget about fork (and
307	forget about forgetting to tell libev about forking) when you use this	361	forget about forgetting to tell libev about forking) when you use this
308	flag.	362	flag.
309		363
310	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>	364	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
311	environment variable.	365	environment variable.
312		366
313	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	367	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
314		368
315	This is your standard select(2) backend. Not I<completely> standard, as	369	This is your standard select(2) backend. Not I<completely> standard, as
…		…
317	but if that fails, expect a fairly low limit on the number of fds when	371	but if that fails, expect a fairly low limit on the number of fds when
318	using this backend. It doesn't scale too well (O(highest_fd)), but its	372	using this backend. It doesn't scale too well (O(highest_fd)), but its
319	usually the fastest backend for a low number of (low-numbered :) fds.	373	usually the fastest backend for a low number of (low-numbered :) fds.
320		374
321	To get good performance out of this backend you need a high amount of	375	To get good performance out of this backend you need a high amount of
322	parallelity (most of the file descriptors should be busy). If you are	376	parallelism (most of the file descriptors should be busy). If you are
323	writing a server, you should C<accept ()> in a loop to accept as many	377	writing a server, you should C<accept ()> in a loop to accept as many
324	connections as possible during one iteration. You might also want to have	378	connections as possible during one iteration. You might also want to have
325	a look at C<ev_set_io_collect_interval ()> to increase the amount of	379	a look at C<ev_set_io_collect_interval ()> to increase the amount of
326	readyness notifications you get per iteration.	380	readiness notifications you get per iteration.
		381
		382	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		383	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		384	C<exceptfds> set on that platform).
327		385
328	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	386	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
329		387
330	And this is your standard poll(2) backend. It's more complicated	388	And this is your standard poll(2) backend. It's more complicated
331	than select, but handles sparse fds better and has no artificial	389	than select, but handles sparse fds better and has no artificial
332	limit on the number of fds you can use (except it will slow down	390	limit on the number of fds you can use (except it will slow down
333	considerably with a lot of inactive fds). It scales similarly to select,	391	considerably with a lot of inactive fds). It scales similarly to select,
334	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	392	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
335	performance tips.	393	performance tips.
336		394
		395	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		396	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		397
337	=item C<EVBACKEND_EPOLL> (value 4, Linux)	398	=item C<EVBACKEND_EPOLL> (value 4, Linux)
338		399
339	For few fds, this backend is a bit little slower than poll and select,	400	For few fds, this backend is a bit little slower than poll and select,
340	but it scales phenomenally better. While poll and select usually scale	401	but it scales phenomenally better. While poll and select usually scale
341	like O(total_fds) where n is the total number of fds (or the highest fd),	402	like O(total_fds) where n is the total number of fds (or the highest fd),
342	epoll scales either O(1) or O(active_fds). The epoll design has a number	403	epoll scales either O(1) or O(active_fds).
343	of shortcomings, such as silently dropping events in some hard-to-detect	404
344	cases and rewiring a syscall per fd change, no fork support and bad	405	The epoll mechanism deserves honorable mention as the most misdesigned
345	support for dup.	406	of the more advanced event mechanisms: mere annoyances include silently
		407	dropping file descriptors, requiring a system call per change per file
		408	descriptor (and unnecessary guessing of parameters), problems with dup and
		409	so on. The biggest issue is fork races, however - if a program forks then
		410	I<both> parent and child process have to recreate the epoll set, which can
		411	take considerable time (one syscall per file descriptor) and is of course
		412	hard to detect.
		413
		414	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		415	of course I<doesn't>, and epoll just loves to report events for totally
		416	I<different> file descriptors (even already closed ones, so one cannot
		417	even remove them from the set) than registered in the set (especially
		418	on SMP systems). Libev tries to counter these spurious notifications by
		419	employing an additional generation counter and comparing that against the
		420	events to filter out spurious ones, recreating the set when required.
346		421
347	While stopping, setting and starting an I/O watcher in the same iteration	422	While stopping, setting and starting an I/O watcher in the same iteration
348	will result in some caching, there is still a syscall per such incident	423	will result in some caching, there is still a system call per such
349	(because the fd could point to a different file description now), so its	424	incident (because the same I<file descriptor> could point to a different
350	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	425	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
351	very well if you register events for both fds.	426	file descriptors might not work very well if you register events for both
352		427	file descriptors.
353	Please note that epoll sometimes generates spurious notifications, so you
354	need to use non-blocking I/O or other means to avoid blocking when no data
355	(or space) is available.
356		428
357	Best performance from this backend is achieved by not unregistering all	429	Best performance from this backend is achieved by not unregistering all
358	watchers for a file descriptor until it has been closed, if possible, i.e.	430	watchers for a file descriptor until it has been closed, if possible,
359	keep at least one watcher active per fd at all times.	431	i.e. keep at least one watcher active per fd at all times. Stopping and
		432	starting a watcher (without re-setting it) also usually doesn't cause
		433	extra overhead. A fork can both result in spurious notifications as well
		434	as in libev having to destroy and recreate the epoll object, which can
		435	take considerable time and thus should be avoided.
360		436
		437	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		438	faster than epoll for maybe up to a hundred file descriptors, depending on
		439	the usage. So sad.
		440
361	While nominally embeddeble in other event loops, this feature is broken in	441	While nominally embeddable in other event loops, this feature is broken in
362	all kernel versions tested so far.	442	all kernel versions tested so far.
		443
		444	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		445	C<EVBACKEND_POLL>.
363		446
364	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	447	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
365		448
366	Kqueue deserves special mention, as at the time of this writing, it	449	Kqueue deserves special mention, as at the time of this writing, it
367	was broken on all BSDs except NetBSD (usually it doesn't work reliably	450	was broken on all BSDs except NetBSD (usually it doesn't work reliably
368	with anything but sockets and pipes, except on Darwin, where of course	451	with anything but sockets and pipes, except on Darwin, where of course
369	it's completely useless). For this reason it's not being "autodetected"	452	it's completely useless). Unlike epoll, however, whose brokenness
		453	is by design, these kqueue bugs can (and eventually will) be fixed
		454	without API changes to existing programs. For this reason it's not being
370	unless you explicitly specify it explicitly in the flags (i.e. using	455	"auto-detected" unless you explicitly specify it in the flags (i.e. using
371	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	456	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
372	system like NetBSD.	457	system like NetBSD.
373		458
374	You still can embed kqueue into a normal poll or select backend and use it	459	You still can embed kqueue into a normal poll or select backend and use it
375	only for sockets (after having made sure that sockets work with kqueue on	460	only for sockets (after having made sure that sockets work with kqueue on
376	the target platform). See C<ev_embed> watchers for more info.	461	the target platform). See C<ev_embed> watchers for more info.
377		462
378	It scales in the same way as the epoll backend, but the interface to the	463	It scales in the same way as the epoll backend, but the interface to the
379	kernel is more efficient (which says nothing about its actual speed, of	464	kernel is more efficient (which says nothing about its actual speed, of
380	course). While stopping, setting and starting an I/O watcher does never	465	course). While stopping, setting and starting an I/O watcher does never
381	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to	466	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
382	two event changes per incident, support for C<fork ()> is very bad and it	467	two event changes per incident. Support for C<fork ()> is very bad (but
383	drops fds silently in similarly hard-to-detect cases.	468	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		469	cases
384		470
385	This backend usually performs well under most conditions.	471	This backend usually performs well under most conditions.
386		472
387	While nominally embeddable in other event loops, this doesn't work	473	While nominally embeddable in other event loops, this doesn't work
388	everywhere, so you might need to test for this. And since it is broken	474	everywhere, so you might need to test for this. And since it is broken
389	almost everywhere, you should only use it when you have a lot of sockets	475	almost everywhere, you should only use it when you have a lot of sockets
390	(for which it usually works), by embedding it into another event loop	476	(for which it usually works), by embedding it into another event loop
391	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	477	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
392	sockets.	478	also broken on OS X)) and, did I mention it, using it only for sockets.
		479
		480	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		481	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		482	C<NOTE_EOF>.
393		483
394	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	484	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
395		485
396	This is not implemented yet (and might never be, unless you send me an	486	This is not implemented yet (and might never be, unless you send me an
397	implementation). According to reports, C</dev/poll> only supports sockets	487	implementation). According to reports, C</dev/poll> only supports sockets
…		…
401	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	491	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
402		492
403	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	493	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
404	it's really slow, but it still scales very well (O(active_fds)).	494	it's really slow, but it still scales very well (O(active_fds)).
405		495
406	Please note that solaris event ports can deliver a lot of spurious	496	Please note that Solaris event ports can deliver a lot of spurious
407	notifications, so you need to use non-blocking I/O or other means to avoid	497	notifications, so you need to use non-blocking I/O or other means to avoid
408	blocking when no data (or space) is available.	498	blocking when no data (or space) is available.
409		499
410	While this backend scales well, it requires one system call per active	500	While this backend scales well, it requires one system call per active
411	file descriptor per loop iteration. For small and medium numbers of file	501	file descriptor per loop iteration. For small and medium numbers of file
412	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	502	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
413	might perform better.	503	might perform better.
414		504
415	On the positive side, ignoring the spurious readyness notifications, this	505	On the positive side, with the exception of the spurious readiness
416	backend actually performed to specification in all tests and is fully	506	notifications, this backend actually performed fully to specification
417	embeddable, which is a rare feat among the OS-specific backends.	507	in all tests and is fully embeddable, which is a rare feat among the
		508	OS-specific backends (I vastly prefer correctness over speed hacks).
		509
		510	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		511	C<EVBACKEND_POLL>.
418		512
419	=item C<EVBACKEND_ALL>	513	=item C<EVBACKEND_ALL>
420		514
421	Try all backends (even potentially broken ones that wouldn't be tried	515	Try all backends (even potentially broken ones that wouldn't be tried
422	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	516	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
…		…
424		518
425	It is definitely not recommended to use this flag.	519	It is definitely not recommended to use this flag.
426		520
427	=back	521	=back
428		522
429	If one or more of these are ored into the flags value, then only these	523	If one or more of these are or'ed into the flags value, then only these
430	backends will be tried (in the reverse order as listed here). If none are	524	backends will be tried (in the reverse order as listed here). If none are
431	specified, all backends in C<ev_recommended_backends ()> will be tried.	525	specified, all backends in C<ev_recommended_backends ()> will be tried.
432		526
433	The most typical usage is like this:	527	Example: This is the most typical usage.
434		528
435	if (!ev_default_loop (0))	529	if (!ev_default_loop (0))
436	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	530	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
437		531
438	Restrict libev to the select and poll backends, and do not allow	532	Example: Restrict libev to the select and poll backends, and do not allow
439	environment settings to be taken into account:	533	environment settings to be taken into account:
440		534
441	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	535	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
442		536
443	Use whatever libev has to offer, but make sure that kqueue is used if	537	Example: Use whatever libev has to offer, but make sure that kqueue is
444	available (warning, breaks stuff, best use only with your own private	538	used if available (warning, breaks stuff, best use only with your own
445	event loop and only if you know the OS supports your types of fds):	539	private event loop and only if you know the OS supports your types of
		540	fds):
446		541
447	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	542	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
448		543
449	=item struct ev_loop *ev_loop_new (unsigned int flags)	544	=item struct ev_loop *ev_loop_new (unsigned int flags)
450		545
451	Similar to C<ev_default_loop>, but always creates a new event loop that is	546	Similar to C<ev_default_loop>, but always creates a new event loop that is
452	always distinct from the default loop. Unlike the default loop, it cannot	547	always distinct from the default loop. Unlike the default loop, it cannot
453	handle signal and child watchers, and attempts to do so will be greeted by	548	handle signal and child watchers, and attempts to do so will be greeted by
454	undefined behaviour (or a failed assertion if assertions are enabled).	549	undefined behaviour (or a failed assertion if assertions are enabled).
455		550
		551	Note that this function I<is> thread-safe, and the recommended way to use
		552	libev with threads is indeed to create one loop per thread, and using the
		553	default loop in the "main" or "initial" thread.
		554
456	Example: Try to create a event loop that uses epoll and nothing else.	555	Example: Try to create a event loop that uses epoll and nothing else.
457		556
458	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	557	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
459	if (!epoller)	558	if (!epoller)
460	fatal ("no epoll found here, maybe it hides under your chair");	559	fatal ("no epoll found here, maybe it hides under your chair");
461		560
462	=item ev_default_destroy ()	561	=item ev_default_destroy ()
463		562
464	Destroys the default loop again (frees all memory and kernel state	563	Destroys the default loop again (frees all memory and kernel state
465	etc.). None of the active event watchers will be stopped in the normal	564	etc.). None of the active event watchers will be stopped in the normal
466	sense, so e.g. C<ev_is_active> might still return true. It is your	565	sense, so e.g. C<ev_is_active> might still return true. It is your
467	responsibility to either stop all watchers cleanly yoursef I<before>	566	responsibility to either stop all watchers cleanly yourself I<before>
468	calling this function, or cope with the fact afterwards (which is usually	567	calling this function, or cope with the fact afterwards (which is usually
469	the easiest thing, you can just ignore the watchers and/or C<free ()> them	568	the easiest thing, you can just ignore the watchers and/or C<free ()> them
470	for example).	569	for example).
471		570
472	Note that certain global state, such as signal state, will not be freed by	571	Note that certain global state, such as signal state (and installed signal
473	this function, and related watchers (such as signal and child watchers)	572	handlers), will not be freed by this function, and related watchers (such
474	would need to be stopped manually.	573	as signal and child watchers) would need to be stopped manually.
475		574
476	In general it is not advisable to call this function except in the	575	In general it is not advisable to call this function except in the
477	rare occasion where you really need to free e.g. the signal handling	576	rare occasion where you really need to free e.g. the signal handling
478	pipe fds. If you need dynamically allocated loops it is better to use	577	pipe fds. If you need dynamically allocated loops it is better to use
479	C<ev_loop_new> and C<ev_loop_destroy>).	578	C<ev_loop_new> and C<ev_loop_destroy>).
…		…
504		603
505	=item ev_loop_fork (loop)	604	=item ev_loop_fork (loop)
506		605
507	Like C<ev_default_fork>, but acts on an event loop created by	606	Like C<ev_default_fork>, but acts on an event loop created by
508	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	607	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
509	after fork, and how you do this is entirely your own problem.	608	after fork that you want to re-use in the child, and how you do this is
		609	entirely your own problem.
510		610
511	=item int ev_is_default_loop (loop)	611	=item int ev_is_default_loop (loop)
512		612
513	Returns true when the given loop actually is the default loop, false otherwise.	613	Returns true when the given loop is, in fact, the default loop, and false
		614	otherwise.
514		615
515	=item unsigned int ev_loop_count (loop)	616	=item unsigned int ev_loop_count (loop)
516		617
517	Returns the count of loop iterations for the loop, which is identical to	618	Returns the count of loop iterations for the loop, which is identical to
518	the number of times libev did poll for new events. It starts at C<0> and	619	the number of times libev did poll for new events. It starts at C<0> and
519	happily wraps around with enough iterations.	620	happily wraps around with enough iterations.
520		621
521	This value can sometimes be useful as a generation counter of sorts (it	622	This value can sometimes be useful as a generation counter of sorts (it
522	"ticks" the number of loop iterations), as it roughly corresponds with	623	"ticks" the number of loop iterations), as it roughly corresponds with
523	C<ev_prepare> and C<ev_check> calls.	624	C<ev_prepare> and C<ev_check> calls.
		625
		626	=item unsigned int ev_loop_depth (loop)
		627
		628	Returns the number of times C<ev_loop> was entered minus the number of
		629	times C<ev_loop> was exited, in other words, the recursion depth.
		630
		631	Outside C<ev_loop>, this number is zero. In a callback, this number is
		632	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		633	in which case it is higher.
		634
		635	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		636	etc.), doesn't count as exit.
524		637
525	=item unsigned int ev_backend (loop)	638	=item unsigned int ev_backend (loop)
526		639
527	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	640	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
528	use.	641	use.
…		…
533	received events and started processing them. This timestamp does not	646	received events and started processing them. This timestamp does not
534	change as long as callbacks are being processed, and this is also the base	647	change as long as callbacks are being processed, and this is also the base
535	time used for relative timers. You can treat it as the timestamp of the	648	time used for relative timers. You can treat it as the timestamp of the
536	event occurring (or more correctly, libev finding out about it).	649	event occurring (or more correctly, libev finding out about it).
537		650
		651	=item ev_now_update (loop)
		652
		653	Establishes the current time by querying the kernel, updating the time
		654	returned by C<ev_now ()> in the progress. This is a costly operation and
		655	is usually done automatically within C<ev_loop ()>.
		656
		657	This function is rarely useful, but when some event callback runs for a
		658	very long time without entering the event loop, updating libev's idea of
		659	the current time is a good idea.
		660
		661	See also L<The special problem of time updates> in the C<ev_timer> section.
		662
		663	=item ev_suspend (loop)
		664
		665	=item ev_resume (loop)
		666
		667	These two functions suspend and resume a loop, for use when the loop is
		668	not used for a while and timeouts should not be processed.
		669
		670	A typical use case would be an interactive program such as a game: When
		671	the user presses C<^Z> to suspend the game and resumes it an hour later it
		672	would be best to handle timeouts as if no time had actually passed while
		673	the program was suspended. This can be achieved by calling C<ev_suspend>
		674	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		675	C<ev_resume> directly afterwards to resume timer processing.
		676
		677	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		678	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		679	will be rescheduled (that is, they will lose any events that would have
		680	occured while suspended).
		681
		682	After calling C<ev_suspend> you B<must not> call I<any> function on the
		683	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		684	without a previous call to C<ev_suspend>.
		685
		686	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		687	event loop time (see C<ev_now_update>).
		688
538	=item ev_loop (loop, int flags)	689	=item ev_loop (loop, int flags)
539		690
540	Finally, this is it, the event handler. This function usually is called	691	Finally, this is it, the event handler. This function usually is called
541	after you initialised all your watchers and you want to start handling	692	after you initialised all your watchers and you want to start handling
542	events.	693	events.
…		…
544	If the flags argument is specified as C<0>, it will not return until	695	If the flags argument is specified as C<0>, it will not return until
545	either no event watchers are active anymore or C<ev_unloop> was called.	696	either no event watchers are active anymore or C<ev_unloop> was called.
546		697
547	Please note that an explicit C<ev_unloop> is usually better than	698	Please note that an explicit C<ev_unloop> is usually better than
548	relying on all watchers to be stopped when deciding when a program has	699	relying on all watchers to be stopped when deciding when a program has
549	finished (especially in interactive programs), but having a program that	700	finished (especially in interactive programs), but having a program
550	automatically loops as long as it has to and no longer by virtue of	701	that automatically loops as long as it has to and no longer by virtue
551	relying on its watchers stopping correctly is a thing of beauty.	702	of relying on its watchers stopping correctly, that is truly a thing of
		703	beauty.
552		704
553	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	705	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
554	those events and any outstanding ones, but will not block your process in	706	those events and any already outstanding ones, but will not block your
555	case there are no events and will return after one iteration of the loop.	707	process in case there are no events and will return after one iteration of
		708	the loop.
556		709
557	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	710	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
558	neccessary) and will handle those and any outstanding ones. It will block	711	necessary) and will handle those and any already outstanding ones. It
559	your process until at least one new event arrives, and will return after	712	will block your process until at least one new event arrives (which could
560	one iteration of the loop. This is useful if you are waiting for some	713	be an event internal to libev itself, so there is no guarantee that a
561	external event in conjunction with something not expressible using other	714	user-registered callback will be called), and will return after one
		715	iteration of the loop.
		716
		717	This is useful if you are waiting for some external event in conjunction
		718	with something not expressible using other libev watchers (i.e. "roll your
562	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	719	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
563	usually a better approach for this kind of thing.	720	usually a better approach for this kind of thing.
564		721
565	Here are the gory details of what C<ev_loop> does:	722	Here are the gory details of what C<ev_loop> does:
566		723
567	- Before the first iteration, call any pending watchers.	724	- Before the first iteration, call any pending watchers.
568	* If EVFLAG_FORKCHECK was used, check for a fork.	725	* If EVFLAG_FORKCHECK was used, check for a fork.
569	- If a fork was detected, queue and call all fork watchers.	726	- If a fork was detected (by any means), queue and call all fork watchers.
570	- Queue and call all prepare watchers.	727	- Queue and call all prepare watchers.
571	- If we have been forked, recreate the kernel state.	728	- If we have been forked, detach and recreate the kernel state
		729	as to not disturb the other process.
572	- Update the kernel state with all outstanding changes.	730	- Update the kernel state with all outstanding changes.
573	- Update the "event loop time".	731	- Update the "event loop time" (ev_now ()).
574	- Calculate for how long to sleep or block, if at all	732	- Calculate for how long to sleep or block, if at all
575	(active idle watchers, EVLOOP_NONBLOCK or not having	733	(active idle watchers, EVLOOP_NONBLOCK or not having
576	any active watchers at all will result in not sleeping).	734	any active watchers at all will result in not sleeping).
577	- Sleep if the I/O and timer collect interval say so.	735	- Sleep if the I/O and timer collect interval say so.
578	- Block the process, waiting for any events.	736	- Block the process, waiting for any events.
579	- Queue all outstanding I/O (fd) events.	737	- Queue all outstanding I/O (fd) events.
580	- Update the "event loop time" and do time jump handling.	738	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
581	- Queue all outstanding timers.	739	- Queue all expired timers.
582	- Queue all outstanding periodics.	740	- Queue all expired periodics.
583	- If no events are pending now, queue all idle watchers.	741	- Unless any events are pending now, queue all idle watchers.
584	- Queue all check watchers.	742	- Queue all check watchers.
585	- Call all queued watchers in reverse order (i.e. check watchers first).	743	- Call all queued watchers in reverse order (i.e. check watchers first).
586	Signals and child watchers are implemented as I/O watchers, and will	744	Signals and child watchers are implemented as I/O watchers, and will
587	be handled here by queueing them when their watcher gets executed.	745	be handled here by queueing them when their watcher gets executed.
588	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	746	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
…		…
593	anymore.	751	anymore.
594		752
595	... queue jobs here, make sure they register event watchers as long	753	... queue jobs here, make sure they register event watchers as long
596	... as they still have work to do (even an idle watcher will do..)	754	... as they still have work to do (even an idle watcher will do..)
597	ev_loop (my_loop, 0);	755	ev_loop (my_loop, 0);
598	... jobs done. yeah!	756	... jobs done or somebody called unloop. yeah!
599		757
600	=item ev_unloop (loop, how)	758	=item ev_unloop (loop, how)
601		759
602	Can be used to make a call to C<ev_loop> return early (but only after it	760	Can be used to make a call to C<ev_loop> return early (but only after it
603	has processed all outstanding events). The C<how> argument must be either	761	has processed all outstanding events). The C<how> argument must be either
604	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	762	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
605	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	763	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
606		764
607	This "unloop state" will be cleared when entering C<ev_loop> again.	765	This "unloop state" will be cleared when entering C<ev_loop> again.
608		766
		767	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		768
609	=item ev_ref (loop)	769	=item ev_ref (loop)
610		770
611	=item ev_unref (loop)	771	=item ev_unref (loop)
612		772
613	Ref/unref can be used to add or remove a reference count on the event	773	Ref/unref can be used to add or remove a reference count on the event
614	loop: Every watcher keeps one reference, and as long as the reference	774	loop: Every watcher keeps one reference, and as long as the reference
615	count is nonzero, C<ev_loop> will not return on its own. If you have	775	count is nonzero, C<ev_loop> will not return on its own.
		776
616	a watcher you never unregister that should not keep C<ev_loop> from	777	If you have a watcher you never unregister that should not keep C<ev_loop>
617	returning, ev_unref() after starting, and ev_ref() before stopping it. For	778	from returning, call ev_unref() after starting, and ev_ref() before
		779	stopping it.
		780
618	example, libev itself uses this for its internal signal pipe: It is not	781	As an example, libev itself uses this for its internal signal pipe: It
619	visible to the libev user and should not keep C<ev_loop> from exiting if	782	is not visible to the libev user and should not keep C<ev_loop> from
620	no event watchers registered by it are active. It is also an excellent	783	exiting if no event watchers registered by it are active. It is also an
621	way to do this for generic recurring timers or from within third-party	784	excellent way to do this for generic recurring timers or from within
622	libraries. Just remember to I<unref after start> and I<ref before stop>	785	third-party libraries. Just remember to I<unref after start> and I<ref
623	(but only if the watcher wasn't active before, or was active before,	786	before stop> (but only if the watcher wasn't active before, or was active
624	respectively).	787	before, respectively. Note also that libev might stop watchers itself
		788	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		789	in the callback).
625		790
626	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	791	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
627	running when nothing else is active.	792	running when nothing else is active.
628		793
629	struct ev_signal exitsig;	794	ev_signal exitsig;
630	ev_signal_init (&exitsig, sig_cb, SIGINT);	795	ev_signal_init (&exitsig, sig_cb, SIGINT);
631	ev_signal_start (loop, &exitsig);	796	ev_signal_start (loop, &exitsig);
632	evf_unref (loop);	797	evf_unref (loop);
633		798
634	Example: For some weird reason, unregister the above signal handler again.	799	Example: For some weird reason, unregister the above signal handler again.
635		800
636	ev_ref (loop);	801	ev_ref (loop);
637	ev_signal_stop (loop, &exitsig);	802	ev_signal_stop (loop, &exitsig);
638		803
639	=item ev_set_io_collect_interval (loop, ev_tstamp interval)	804	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
640		805
641	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)	806	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
642		807
643	These advanced functions influence the time that libev will spend waiting	808	These advanced functions influence the time that libev will spend waiting
644	for events. Both are by default C<0>, meaning that libev will try to	809	for events. Both time intervals are by default C<0>, meaning that libev
645	invoke timer/periodic callbacks and I/O callbacks with minimum latency.	810	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		811	latency.
646		812
647	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	813	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
648	allows libev to delay invocation of I/O and timer/periodic callbacks to	814	allows libev to delay invocation of I/O and timer/periodic callbacks
649	increase efficiency of loop iterations.	815	to increase efficiency of loop iterations (or to increase power-saving
		816	opportunities).
650		817
651	The background is that sometimes your program runs just fast enough to	818	The idea is that sometimes your program runs just fast enough to handle
652	handle one (or very few) event(s) per loop iteration. While this makes	819	one (or very few) event(s) per loop iteration. While this makes the
653	the program responsive, it also wastes a lot of CPU time to poll for new	820	program responsive, it also wastes a lot of CPU time to poll for new
654	events, especially with backends like C<select ()> which have a high	821	events, especially with backends like C<select ()> which have a high
655	overhead for the actual polling but can deliver many events at once.	822	overhead for the actual polling but can deliver many events at once.
656		823
657	By setting a higher I<io collect interval> you allow libev to spend more	824	By setting a higher I<io collect interval> you allow libev to spend more
658	time collecting I/O events, so you can handle more events per iteration,	825	time collecting I/O events, so you can handle more events per iteration,
659	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	826	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
660	C<ev_timer>) will be not affected. Setting this to a non-null value will	827	C<ev_timer>) will be not affected. Setting this to a non-null value will
661	introduce an additional C<ev_sleep ()> call into most loop iterations.	828	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		829	sleep time ensures that libev will not poll for I/O events more often then
		830	once per this interval, on average.
662		831
663	Likewise, by setting a higher I<timeout collect interval> you allow libev	832	Likewise, by setting a higher I<timeout collect interval> you allow libev
664	to spend more time collecting timeouts, at the expense of increased	833	to spend more time collecting timeouts, at the expense of increased
665	latency (the watcher callback will be called later). C<ev_io> watchers	834	latency/jitter/inexactness (the watcher callback will be called
666	will not be affected. Setting this to a non-null value will not introduce	835	later). C<ev_io> watchers will not be affected. Setting this to a non-null
667	any overhead in libev.	836	value will not introduce any overhead in libev.
668		837
669	Many (busy) programs can usually benefit by setting the io collect	838	Many (busy) programs can usually benefit by setting the I/O collect
670	interval to a value near C<0.1> or so, which is often enough for	839	interval to a value near C<0.1> or so, which is often enough for
671	interactive servers (of course not for games), likewise for timeouts. It	840	interactive servers (of course not for games), likewise for timeouts. It
672	usually doesn't make much sense to set it to a lower value than C<0.01>,	841	usually doesn't make much sense to set it to a lower value than C<0.01>,
673	as this approsaches the timing granularity of most systems.	842	as this approaches the timing granularity of most systems. Note that if
		843	you do transactions with the outside world and you can't increase the
		844	parallelity, then this setting will limit your transaction rate (if you
		845	need to poll once per transaction and the I/O collect interval is 0.01,
		846	then you can't do more than 100 transations per second).
		847
		848	Setting the I<timeout collect interval> can improve the opportunity for
		849	saving power, as the program will "bundle" timer callback invocations that
		850	are "near" in time together, by delaying some, thus reducing the number of
		851	times the process sleeps and wakes up again. Another useful technique to
		852	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		853	they fire on, say, one-second boundaries only.
		854
		855	Example: we only need 0.1s timeout granularity, and we wish not to poll
		856	more often than 100 times per second:
		857
		858	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		859	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		860
		861	=item ev_invoke_pending (loop)
		862
		863	This call will simply invoke all pending watchers while resetting their
		864	pending state. Normally, C<ev_loop> does this automatically when required,
		865	but when overriding the invoke callback this call comes handy.
		866
		867	=item int ev_pending_count (loop)
		868
		869	Returns the number of pending watchers - zero indicates that no watchers
		870	are pending.
		871
		872	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		873
		874	This overrides the invoke pending functionality of the loop: Instead of
		875	invoking all pending watchers when there are any, C<ev_loop> will call
		876	this callback instead. This is useful, for example, when you want to
		877	invoke the actual watchers inside another context (another thread etc.).
		878
		879	If you want to reset the callback, use C<ev_invoke_pending> as new
		880	callback.
		881
		882	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		883
		884	Sometimes you want to share the same loop between multiple threads. This
		885	can be done relatively simply by putting mutex_lock/unlock calls around
		886	each call to a libev function.
		887
		888	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		889	wait for it to return. One way around this is to wake up the loop via
		890	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		891	and I<acquire> callbacks on the loop.
		892
		893	When set, then C<release> will be called just before the thread is
		894	suspended waiting for new events, and C<acquire> is called just
		895	afterwards.
		896
		897	Ideally, C<release> will just call your mutex_unlock function, and
		898	C<acquire> will just call the mutex_lock function again.
		899
		900	While event loop modifications are allowed between invocations of
		901	C<release> and C<acquire> (that's their only purpose after all), no
		902	modifications done will affect the event loop, i.e. adding watchers will
		903	have no effect on the set of file descriptors being watched, or the time
		904	waited. USe an C<ev_async> watcher to wake up C<ev_loop> when you want it
		905	to take note of any changes you made.
		906
		907	In theory, threads executing C<ev_loop> will be async-cancel safe between
		908	invocations of C<release> and C<acquire>.
		909
		910	See also the locking example in the C<THREADS> section later in this
		911	document.
		912
		913	=item ev_set_userdata (loop, void *data)
		914
		915	=item ev_userdata (loop)
		916
		917	Set and retrieve a single C<void *> associated with a loop. When
		918	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		919	C<0.>
		920
		921	These two functions can be used to associate arbitrary data with a loop,
		922	and are intended solely for the C<invoke_pending_cb>, C<release> and
		923	C<acquire> callbacks described above, but of course can be (ab-)used for
		924	any other purpose as well.
		925
		926	=item ev_loop_verify (loop)
		927
		928	This function only does something when C<EV_VERIFY> support has been
		929	compiled in, which is the default for non-minimal builds. It tries to go
		930	through all internal structures and checks them for validity. If anything
		931	is found to be inconsistent, it will print an error message to standard
		932	error and call C<abort ()>.
		933
		934	This can be used to catch bugs inside libev itself: under normal
		935	circumstances, this function will never abort as of course libev keeps its
		936	data structures consistent.
674		937
675	=back	938	=back
676		939
677		940
678	=head1 ANATOMY OF A WATCHER	941	=head1 ANATOMY OF A WATCHER
		942
		943	In the following description, uppercase C<TYPE> in names stands for the
		944	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		945	watchers and C<ev_io_start> for I/O watchers.
679		946
680	A watcher is a structure that you create and register to record your	947	A watcher is a structure that you create and register to record your
681	interest in some event. For instance, if you want to wait for STDIN to	948	interest in some event. For instance, if you want to wait for STDIN to
682	become readable, you would create an C<ev_io> watcher for that:	949	become readable, you would create an C<ev_io> watcher for that:
683		950
684	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	951	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
685	{	952	{
686	ev_io_stop (w);	953	ev_io_stop (w);
687	ev_unloop (loop, EVUNLOOP_ALL);	954	ev_unloop (loop, EVUNLOOP_ALL);
688	}	955	}
689		956
690	struct ev_loop *loop = ev_default_loop (0);	957	struct ev_loop *loop = ev_default_loop (0);
		958
691	struct ev_io stdin_watcher;	959	ev_io stdin_watcher;
		960
692	ev_init (&stdin_watcher, my_cb);	961	ev_init (&stdin_watcher, my_cb);
693	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	962	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
694	ev_io_start (loop, &stdin_watcher);	963	ev_io_start (loop, &stdin_watcher);
		964
695	ev_loop (loop, 0);	965	ev_loop (loop, 0);
696		966
697	As you can see, you are responsible for allocating the memory for your	967	As you can see, you are responsible for allocating the memory for your
698	watcher structures (and it is usually a bad idea to do this on the stack,	968	watcher structures (and it is I<usually> a bad idea to do this on the
699	although this can sometimes be quite valid).	969	stack).
		970
		971	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		972	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
700		973
701	Each watcher structure must be initialised by a call to C<ev_init	974	Each watcher structure must be initialised by a call to C<ev_init
702	(watcher *, callback)>, which expects a callback to be provided. This	975	(watcher *, callback)>, which expects a callback to be provided. This
703	callback gets invoked each time the event occurs (or, in the case of io	976	callback gets invoked each time the event occurs (or, in the case of I/O
704	watchers, each time the event loop detects that the file descriptor given	977	watchers, each time the event loop detects that the file descriptor given
705	is readable and/or writable).	978	is readable and/or writable).
706		979
707	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	980	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
708	with arguments specific to this watcher type. There is also a macro	981	macro to configure it, with arguments specific to the watcher type. There
709	to combine initialisation and setting in one call: C<< ev_<type>_init	982	is also a macro to combine initialisation and setting in one call: C<<
710	(watcher *, callback, ...) >>.	983	ev_TYPE_init (watcher *, callback, ...) >>.
711		984
712	To make the watcher actually watch out for events, you have to start it	985	To make the watcher actually watch out for events, you have to start it
713	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	986	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
714	*) >>), and you can stop watching for events at any time by calling the	987	*) >>), and you can stop watching for events at any time by calling the
715	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	988	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
716		989
717	As long as your watcher is active (has been started but not stopped) you	990	As long as your watcher is active (has been started but not stopped) you
718	must not touch the values stored in it. Most specifically you must never	991	must not touch the values stored in it. Most specifically you must never
719	reinitialise it or call its C<set> macro.	992	reinitialise it or call its C<ev_TYPE_set> macro.
720		993
721	Each and every callback receives the event loop pointer as first, the	994	Each and every callback receives the event loop pointer as first, the
722	registered watcher structure as second, and a bitset of received events as	995	registered watcher structure as second, and a bitset of received events as
723	third argument.	996	third argument.
724		997
…		…
782		1055
783	=item C<EV_ASYNC>	1056	=item C<EV_ASYNC>
784		1057
785	The given async watcher has been asynchronously notified (see C<ev_async>).	1058	The given async watcher has been asynchronously notified (see C<ev_async>).
786		1059
		1060	=item C<EV_CUSTOM>
		1061
		1062	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1063	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1064
787	=item C<EV_ERROR>	1065	=item C<EV_ERROR>
788		1066
789	An unspecified error has occured, the watcher has been stopped. This might	1067	An unspecified error has occurred, the watcher has been stopped. This might
790	happen because the watcher could not be properly started because libev	1068	happen because the watcher could not be properly started because libev
791	ran out of memory, a file descriptor was found to be closed or any other	1069	ran out of memory, a file descriptor was found to be closed or any other
		1070	problem. Libev considers these application bugs.
		1071
792	problem. You best act on it by reporting the problem and somehow coping	1072	You best act on it by reporting the problem and somehow coping with the
793	with the watcher being stopped.	1073	watcher being stopped. Note that well-written programs should not receive
		1074	an error ever, so when your watcher receives it, this usually indicates a
		1075	bug in your program.
794		1076
795	Libev will usually signal a few "dummy" events together with an error,	1077	Libev will usually signal a few "dummy" events together with an error, for
796	for example it might indicate that a fd is readable or writable, and if	1078	example it might indicate that a fd is readable or writable, and if your
797	your callbacks is well-written it can just attempt the operation and cope	1079	callbacks is well-written it can just attempt the operation and cope with
798	with the error from read() or write(). This will not work in multithreaded	1080	the error from read() or write(). This will not work in multi-threaded
799	programs, though, so beware.	1081	programs, though, as the fd could already be closed and reused for another
		1082	thing, so beware.
800		1083
801	=back	1084	=back
802		1085
803	=head2 GENERIC WATCHER FUNCTIONS	1086	=head2 GENERIC WATCHER FUNCTIONS
804
805	In the following description, C<TYPE> stands for the watcher type,
806	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
807		1087
808	=over 4	1088	=over 4
809		1089
810	=item C<ev_init> (ev_TYPE *watcher, callback)	1090	=item C<ev_init> (ev_TYPE *watcher, callback)
811		1091
…		…
817	which rolls both calls into one.	1097	which rolls both calls into one.
818		1098
819	You can reinitialise a watcher at any time as long as it has been stopped	1099	You can reinitialise a watcher at any time as long as it has been stopped
820	(or never started) and there are no pending events outstanding.	1100	(or never started) and there are no pending events outstanding.
821		1101
822	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1102	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
823	int revents)>.	1103	int revents)>.
		1104
		1105	Example: Initialise an C<ev_io> watcher in two steps.
		1106
		1107	ev_io w;
		1108	ev_init (&w, my_cb);
		1109	ev_io_set (&w, STDIN_FILENO, EV_READ);
824		1110
825	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1111	=item C<ev_TYPE_set> (ev_TYPE *, [args])
826		1112
827	This macro initialises the type-specific parts of a watcher. You need to	1113	This macro initialises the type-specific parts of a watcher. You need to
828	call C<ev_init> at least once before you call this macro, but you can	1114	call C<ev_init> at least once before you call this macro, but you can
…		…
831	difference to the C<ev_init> macro).	1117	difference to the C<ev_init> macro).
832		1118
833	Although some watcher types do not have type-specific arguments	1119	Although some watcher types do not have type-specific arguments
834	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1120	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
835		1121
		1122	See C<ev_init>, above, for an example.
		1123
836	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1124	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
837		1125
838	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1126	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
839	calls into a single call. This is the most convinient method to initialise	1127	calls into a single call. This is the most convenient method to initialise
840	a watcher. The same limitations apply, of course.	1128	a watcher. The same limitations apply, of course.
		1129
		1130	Example: Initialise and set an C<ev_io> watcher in one step.
		1131
		1132	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
841		1133
842	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1134	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
843		1135
844	Starts (activates) the given watcher. Only active watchers will receive	1136	Starts (activates) the given watcher. Only active watchers will receive
845	events. If the watcher is already active nothing will happen.	1137	events. If the watcher is already active nothing will happen.
846		1138
		1139	Example: Start the C<ev_io> watcher that is being abused as example in this
		1140	whole section.
		1141
		1142	ev_io_start (EV_DEFAULT_UC, &w);
		1143
847	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1144	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
848		1145
849	Stops the given watcher again (if active) and clears the pending	1146	Stops the given watcher if active, and clears the pending status (whether
		1147	the watcher was active or not).
		1148
850	status. It is possible that stopped watchers are pending (for example,	1149	It is possible that stopped watchers are pending - for example,
851	non-repeating timers are being stopped when they become pending), but	1150	non-repeating timers are being stopped when they become pending - but
852	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1151	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
853	you want to free or reuse the memory used by the watcher it is therefore a	1152	pending. If you want to free or reuse the memory used by the watcher it is
854	good idea to always call its C<ev_TYPE_stop> function.	1153	therefore a good idea to always call its C<ev_TYPE_stop> function.
855		1154
856	=item bool ev_is_active (ev_TYPE *watcher)	1155	=item bool ev_is_active (ev_TYPE *watcher)
857		1156
858	Returns a true value iff the watcher is active (i.e. it has been started	1157	Returns a true value iff the watcher is active (i.e. it has been started
859	and not yet been stopped). As long as a watcher is active you must not modify	1158	and not yet been stopped). As long as a watcher is active you must not modify
…		…
885	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1184	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
886	(default: C<-2>). Pending watchers with higher priority will be invoked	1185	(default: C<-2>). Pending watchers with higher priority will be invoked
887	before watchers with lower priority, but priority will not keep watchers	1186	before watchers with lower priority, but priority will not keep watchers
888	from being executed (except for C<ev_idle> watchers).	1187	from being executed (except for C<ev_idle> watchers).
889		1188
890	This means that priorities are I<only> used for ordering callback
891	invocation after new events have been received. This is useful, for
892	example, to reduce latency after idling, or more often, to bind two
893	watchers on the same event and make sure one is called first.
894
895	If you need to suppress invocation when higher priority events are pending	1189	If you need to suppress invocation when higher priority events are pending
896	you need to look at C<ev_idle> watchers, which provide this functionality.	1190	you need to look at C<ev_idle> watchers, which provide this functionality.
897		1191
898	You I<must not> change the priority of a watcher as long as it is active or	1192	You I<must not> change the priority of a watcher as long as it is active or
899	pending.	1193	pending.
900		1194
		1195	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1196	fine, as long as you do not mind that the priority value you query might
		1197	or might not have been clamped to the valid range.
		1198
901	The default priority used by watchers when no priority has been set is	1199	The default priority used by watchers when no priority has been set is
902	always C<0>, which is supposed to not be too high and not be too low :).	1200	always C<0>, which is supposed to not be too high and not be too low :).
903		1201
904	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1202	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
905	fine, as long as you do not mind that the priority value you query might	1203	priorities.
906	or might not have been adjusted to be within valid range.
907		1204
908	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1205	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
909		1206
910	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1207	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
911	C<loop> nor C<revents> need to be valid as long as the watcher callback	1208	C<loop> nor C<revents> need to be valid as long as the watcher callback
912	can deal with that fact.	1209	can deal with that fact, as both are simply passed through to the
		1210	callback.
913		1211
914	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1212	=item int ev_clear_pending (loop, ev_TYPE *watcher)
915		1213
916	If the watcher is pending, this function returns clears its pending status	1214	If the watcher is pending, this function clears its pending status and
917	and returns its C<revents> bitset (as if its callback was invoked). If the	1215	returns its C<revents> bitset (as if its callback was invoked). If the
918	watcher isn't pending it does nothing and returns C<0>.	1216	watcher isn't pending it does nothing and returns C<0>.
919		1217
		1218	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1219	callback to be invoked, which can be accomplished with this function.
		1220
920	=back	1221	=back
921		1222
922		1223
923	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1224	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
924		1225
925	Each watcher has, by default, a member C<void *data> that you can change	1226	Each watcher has, by default, a member C<void *data> that you can change
926	and read at any time, libev will completely ignore it. This can be used	1227	and read at any time: libev will completely ignore it. This can be used
927	to associate arbitrary data with your watcher. If you need more data and	1228	to associate arbitrary data with your watcher. If you need more data and
928	don't want to allocate memory and store a pointer to it in that data	1229	don't want to allocate memory and store a pointer to it in that data
929	member, you can also "subclass" the watcher type and provide your own	1230	member, you can also "subclass" the watcher type and provide your own
930	data:	1231	data:
931		1232
932	struct my_io	1233	struct my_io
933	{	1234	{
934	struct ev_io io;	1235	ev_io io;
935	int otherfd;	1236	int otherfd;
936	void *somedata;	1237	void *somedata;
937	struct whatever *mostinteresting;	1238	struct whatever *mostinteresting;
938	}	1239	};
		1240
		1241	...
		1242	struct my_io w;
		1243	ev_io_init (&w.io, my_cb, fd, EV_READ);
939		1244
940	And since your callback will be called with a pointer to the watcher, you	1245	And since your callback will be called with a pointer to the watcher, you
941	can cast it back to your own type:	1246	can cast it back to your own type:
942		1247
943	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1248	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
944	{	1249	{
945	struct my_io *w = (struct my_io *)w_;	1250	struct my_io *w = (struct my_io *)w_;
946	...	1251	...
947	}	1252	}
948		1253
949	More interesting and less C-conformant ways of casting your callback type	1254	More interesting and less C-conformant ways of casting your callback type
950	instead have been omitted.	1255	instead have been omitted.
951		1256
952	Another common scenario is having some data structure with multiple	1257	Another common scenario is to use some data structure with multiple
953	watchers:	1258	embedded watchers:
954		1259
955	struct my_biggy	1260	struct my_biggy
956	{	1261	{
957	int some_data;	1262	int some_data;
958	ev_timer t1;	1263	ev_timer t1;
959	ev_timer t2;	1264	ev_timer t2;
960	}	1265	}
961		1266
962	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1267	In this case getting the pointer to C<my_biggy> is a bit more
963	you need to use C<offsetof>:	1268	complicated: Either you store the address of your C<my_biggy> struct
		1269	in the C<data> member of the watcher (for woozies), or you need to use
		1270	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1271	programmers):
964		1272
965	#include <stddef.h>	1273	#include <stddef.h>
966		1274
967	static void	1275	static void
968	t1_cb (EV_P_ struct ev_timer *w, int revents)	1276	t1_cb (EV_P_ ev_timer *w, int revents)
969	{	1277	{
970	struct my_biggy big = (struct my_biggy *	1278	struct my_biggy big = (struct my_biggy *)
971	(((char *)w) - offsetof (struct my_biggy, t1));	1279	(((char *)w) - offsetof (struct my_biggy, t1));
972	}	1280	}
973		1281
974	static void	1282	static void
975	t2_cb (EV_P_ struct ev_timer *w, int revents)	1283	t2_cb (EV_P_ ev_timer *w, int revents)
976	{	1284	{
977	struct my_biggy big = (struct my_biggy *	1285	struct my_biggy big = (struct my_biggy *)
978	(((char *)w) - offsetof (struct my_biggy, t2));	1286	(((char *)w) - offsetof (struct my_biggy, t2));
979	}	1287	}
		1288
		1289	=head2 WATCHER PRIORITY MODELS
		1290
		1291	Many event loops support I<watcher priorities>, which are usually small
		1292	integers that influence the ordering of event callback invocation
		1293	between watchers in some way, all else being equal.
		1294
		1295	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1296	description for the more technical details such as the actual priority
		1297	range.
		1298
		1299	There are two common ways how these these priorities are being interpreted
		1300	by event loops:
		1301
		1302	In the more common lock-out model, higher priorities "lock out" invocation
		1303	of lower priority watchers, which means as long as higher priority
		1304	watchers receive events, lower priority watchers are not being invoked.
		1305
		1306	The less common only-for-ordering model uses priorities solely to order
		1307	callback invocation within a single event loop iteration: Higher priority
		1308	watchers are invoked before lower priority ones, but they all get invoked
		1309	before polling for new events.
		1310
		1311	Libev uses the second (only-for-ordering) model for all its watchers
		1312	except for idle watchers (which use the lock-out model).
		1313
		1314	The rationale behind this is that implementing the lock-out model for
		1315	watchers is not well supported by most kernel interfaces, and most event
		1316	libraries will just poll for the same events again and again as long as
		1317	their callbacks have not been executed, which is very inefficient in the
		1318	common case of one high-priority watcher locking out a mass of lower
		1319	priority ones.
		1320
		1321	Static (ordering) priorities are most useful when you have two or more
		1322	watchers handling the same resource: a typical usage example is having an
		1323	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1324	timeouts. Under load, data might be received while the program handles
		1325	other jobs, but since timers normally get invoked first, the timeout
		1326	handler will be executed before checking for data. In that case, giving
		1327	the timer a lower priority than the I/O watcher ensures that I/O will be
		1328	handled first even under adverse conditions (which is usually, but not
		1329	always, what you want).
		1330
		1331	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1332	will only be executed when no same or higher priority watchers have
		1333	received events, they can be used to implement the "lock-out" model when
		1334	required.
		1335
		1336	For example, to emulate how many other event libraries handle priorities,
		1337	you can associate an C<ev_idle> watcher to each such watcher, and in
		1338	the normal watcher callback, you just start the idle watcher. The real
		1339	processing is done in the idle watcher callback. This causes libev to
		1340	continously poll and process kernel event data for the watcher, but when
		1341	the lock-out case is known to be rare (which in turn is rare :), this is
		1342	workable.
		1343
		1344	Usually, however, the lock-out model implemented that way will perform
		1345	miserably under the type of load it was designed to handle. In that case,
		1346	it might be preferable to stop the real watcher before starting the
		1347	idle watcher, so the kernel will not have to process the event in case
		1348	the actual processing will be delayed for considerable time.
		1349
		1350	Here is an example of an I/O watcher that should run at a strictly lower
		1351	priority than the default, and which should only process data when no
		1352	other events are pending:
		1353
		1354	ev_idle idle; // actual processing watcher
		1355	ev_io io; // actual event watcher
		1356
		1357	static void
		1358	io_cb (EV_P_ ev_io *w, int revents)
		1359	{
		1360	// stop the I/O watcher, we received the event, but
		1361	// are not yet ready to handle it.
		1362	ev_io_stop (EV_A_ w);
		1363
		1364	// start the idle watcher to ahndle the actual event.
		1365	// it will not be executed as long as other watchers
		1366	// with the default priority are receiving events.
		1367	ev_idle_start (EV_A_ &idle);
		1368	}
		1369
		1370	static void
		1371	idle_cb (EV_P_ ev_idle *w, int revents)
		1372	{
		1373	// actual processing
		1374	read (STDIN_FILENO, ...);
		1375
		1376	// have to start the I/O watcher again, as
		1377	// we have handled the event
		1378	ev_io_start (EV_P_ &io);
		1379	}
		1380
		1381	// initialisation
		1382	ev_idle_init (&idle, idle_cb);
		1383	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1384	ev_io_start (EV_DEFAULT_ &io);
		1385
		1386	In the "real" world, it might also be beneficial to start a timer, so that
		1387	low-priority connections can not be locked out forever under load. This
		1388	enables your program to keep a lower latency for important connections
		1389	during short periods of high load, while not completely locking out less
		1390	important ones.
980		1391
981		1392
982	=head1 WATCHER TYPES	1393	=head1 WATCHER TYPES
983		1394
984	This section describes each watcher in detail, but will not repeat	1395	This section describes each watcher in detail, but will not repeat
…		…
1008	In general you can register as many read and/or write event watchers per	1419	In general you can register as many read and/or write event watchers per
1009	fd as you want (as long as you don't confuse yourself). Setting all file	1420	fd as you want (as long as you don't confuse yourself). Setting all file
1010	descriptors to non-blocking mode is also usually a good idea (but not	1421	descriptors to non-blocking mode is also usually a good idea (but not
1011	required if you know what you are doing).	1422	required if you know what you are doing).
1012		1423
1013	If you must do this, then force the use of a known-to-be-good backend	1424	If you cannot use non-blocking mode, then force the use of a
1014	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1425	known-to-be-good backend (at the time of this writing, this includes only
1015	C<EVBACKEND_POLL>).	1426	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1427	descriptors for which non-blocking operation makes no sense (such as
		1428	files) - libev doesn't guarentee any specific behaviour in that case.
1016		1429
1017	Another thing you have to watch out for is that it is quite easy to	1430	Another thing you have to watch out for is that it is quite easy to
1018	receive "spurious" readyness notifications, that is your callback might	1431	receive "spurious" readiness notifications, that is your callback might
1019	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1432	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1020	because there is no data. Not only are some backends known to create a	1433	because there is no data. Not only are some backends known to create a
1021	lot of those (for example solaris ports), it is very easy to get into	1434	lot of those (for example Solaris ports), it is very easy to get into
1022	this situation even with a relatively standard program structure. Thus	1435	this situation even with a relatively standard program structure. Thus
1023	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1436	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1024	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1437	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1025		1438
1026	If you cannot run the fd in non-blocking mode (for example you should not	1439	If you cannot run the fd in non-blocking mode (for example you should
1027	play around with an Xlib connection), then you have to seperately re-test	1440	not play around with an Xlib connection), then you have to separately
1028	whether a file descriptor is really ready with a known-to-be good interface	1441	re-test whether a file descriptor is really ready with a known-to-be good
1029	such as poll (fortunately in our Xlib example, Xlib already does this on	1442	interface such as poll (fortunately in our Xlib example, Xlib already
1030	its own, so its quite safe to use).	1443	does this on its own, so its quite safe to use). Some people additionally
		1444	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1445	indefinitely.
		1446
		1447	But really, best use non-blocking mode.
1031		1448
1032	=head3 The special problem of disappearing file descriptors	1449	=head3 The special problem of disappearing file descriptors
1033		1450
1034	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1451	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1035	descriptor (either by calling C<close> explicitly or by any other means,	1452	descriptor (either due to calling C<close> explicitly or any other means,
1036	such as C<dup>). The reason is that you register interest in some file	1453	such as C<dup2>). The reason is that you register interest in some file
1037	descriptor, but when it goes away, the operating system will silently drop	1454	descriptor, but when it goes away, the operating system will silently drop
1038	this interest. If another file descriptor with the same number then is	1455	this interest. If another file descriptor with the same number then is
1039	registered with libev, there is no efficient way to see that this is, in	1456	registered with libev, there is no efficient way to see that this is, in
1040	fact, a different file descriptor.	1457	fact, a different file descriptor.
1041		1458
…		…
1070	To support fork in your programs, you either have to call	1487	To support fork in your programs, you either have to call
1071	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1488	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
1072	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1489	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1073	C<EVBACKEND_POLL>.	1490	C<EVBACKEND_POLL>.
1074		1491
		1492	=head3 The special problem of SIGPIPE
		1493
		1494	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1495	when writing to a pipe whose other end has been closed, your program gets
		1496	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1497	this is sensible behaviour, for daemons, this is usually undesirable.
		1498
		1499	So when you encounter spurious, unexplained daemon exits, make sure you
		1500	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1501	somewhere, as that would have given you a big clue).
		1502
1075		1503
1076	=head3 Watcher-Specific Functions	1504	=head3 Watcher-Specific Functions
1077		1505
1078	=over 4	1506	=over 4
1079		1507
1080	=item ev_io_init (ev_io *, callback, int fd, int events)	1508	=item ev_io_init (ev_io *, callback, int fd, int events)
1081		1509
1082	=item ev_io_set (ev_io *, int fd, int events)	1510	=item ev_io_set (ev_io *, int fd, int events)
1083		1511
1084	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1512	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1085	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or	1513	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1086	C<EV_READ | EV_WRITE> to receive the given events.	1514	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1087		1515
1088	=item int fd [read-only]	1516	=item int fd [read-only]
1089		1517
1090	The file descriptor being watched.	1518	The file descriptor being watched.
1091		1519
…		…
1099		1527
1100	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1528	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1101	readable, but only once. Since it is likely line-buffered, you could	1529	readable, but only once. Since it is likely line-buffered, you could
1102	attempt to read a whole line in the callback.	1530	attempt to read a whole line in the callback.
1103		1531
1104	static void	1532	static void
1105	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1533	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1106	{	1534	{
1107	ev_io_stop (loop, w);	1535	ev_io_stop (loop, w);
1108	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1536	.. read from stdin here (or from w->fd) and handle any I/O errors
1109	}	1537	}
1110		1538
1111	...	1539	...
1112	struct ev_loop *loop = ev_default_init (0);	1540	struct ev_loop *loop = ev_default_init (0);
1113	struct ev_io stdin_readable;	1541	ev_io stdin_readable;
1114	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1542	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1115	ev_io_start (loop, &stdin_readable);	1543	ev_io_start (loop, &stdin_readable);
1116	ev_loop (loop, 0);	1544	ev_loop (loop, 0);
1117		1545
1118		1546
1119	=head2 C<ev_timer> - relative and optionally repeating timeouts	1547	=head2 C<ev_timer> - relative and optionally repeating timeouts
1120		1548
1121	Timer watchers are simple relative timers that generate an event after a	1549	Timer watchers are simple relative timers that generate an event after a
1122	given time, and optionally repeating in regular intervals after that.	1550	given time, and optionally repeating in regular intervals after that.
1123		1551
1124	The timers are based on real time, that is, if you register an event that	1552	The timers are based on real time, that is, if you register an event that
1125	times out after an hour and you reset your system clock to last years	1553	times out after an hour and you reset your system clock to January last
1126	time, it will still time out after (roughly) and hour. "Roughly" because	1554	year, it will still time out after (roughly) one hour. "Roughly" because
1127	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1555	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1128	monotonic clock option helps a lot here).	1556	monotonic clock option helps a lot here).
		1557
		1558	The callback is guaranteed to be invoked only I<after> its timeout has
		1559	passed (not I<at>, so on systems with very low-resolution clocks this
		1560	might introduce a small delay). If multiple timers become ready during the
		1561	same loop iteration then the ones with earlier time-out values are invoked
		1562	before ones of the same priority with later time-out values (but this is
		1563	no longer true when a callback calls C<ev_loop> recursively).
		1564
		1565	=head3 Be smart about timeouts
		1566
		1567	Many real-world problems involve some kind of timeout, usually for error
		1568	recovery. A typical example is an HTTP request - if the other side hangs,
		1569	you want to raise some error after a while.
		1570
		1571	What follows are some ways to handle this problem, from obvious and
		1572	inefficient to smart and efficient.
		1573
		1574	In the following, a 60 second activity timeout is assumed - a timeout that
		1575	gets reset to 60 seconds each time there is activity (e.g. each time some
		1576	data or other life sign was received).
		1577
		1578	=over 4
		1579
		1580	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1581
		1582	This is the most obvious, but not the most simple way: In the beginning,
		1583	start the watcher:
		1584
		1585	ev_timer_init (timer, callback, 60., 0.);
		1586	ev_timer_start (loop, timer);
		1587
		1588	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1589	and start it again:
		1590
		1591	ev_timer_stop (loop, timer);
		1592	ev_timer_set (timer, 60., 0.);
		1593	ev_timer_start (loop, timer);
		1594
		1595	This is relatively simple to implement, but means that each time there is
		1596	some activity, libev will first have to remove the timer from its internal
		1597	data structure and then add it again. Libev tries to be fast, but it's
		1598	still not a constant-time operation.
		1599
		1600	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1601
		1602	This is the easiest way, and involves using C<ev_timer_again> instead of
		1603	C<ev_timer_start>.
		1604
		1605	To implement this, configure an C<ev_timer> with a C<repeat> value
		1606	of C<60> and then call C<ev_timer_again> at start and each time you
		1607	successfully read or write some data. If you go into an idle state where
		1608	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1609	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1610
		1611	That means you can ignore both the C<ev_timer_start> function and the
		1612	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1613	member and C<ev_timer_again>.
		1614
		1615	At start:
		1616
		1617	ev_init (timer, callback);
		1618	timer->repeat = 60.;
		1619	ev_timer_again (loop, timer);
		1620
		1621	Each time there is some activity:
		1622
		1623	ev_timer_again (loop, timer);
		1624
		1625	It is even possible to change the time-out on the fly, regardless of
		1626	whether the watcher is active or not:
		1627
		1628	timer->repeat = 30.;
		1629	ev_timer_again (loop, timer);
		1630
		1631	This is slightly more efficient then stopping/starting the timer each time
		1632	you want to modify its timeout value, as libev does not have to completely
		1633	remove and re-insert the timer from/into its internal data structure.
		1634
		1635	It is, however, even simpler than the "obvious" way to do it.
		1636
		1637	=item 3. Let the timer time out, but then re-arm it as required.
		1638
		1639	This method is more tricky, but usually most efficient: Most timeouts are
		1640	relatively long compared to the intervals between other activity - in
		1641	our example, within 60 seconds, there are usually many I/O events with
		1642	associated activity resets.
		1643
		1644	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1645	but remember the time of last activity, and check for a real timeout only
		1646	within the callback:
		1647
		1648	ev_tstamp last_activity; // time of last activity
		1649
		1650	static void
		1651	callback (EV_P_ ev_timer *w, int revents)
		1652	{
		1653	ev_tstamp now = ev_now (EV_A);
		1654	ev_tstamp timeout = last_activity + 60.;
		1655
		1656	// if last_activity + 60. is older than now, we did time out
		1657	if (timeout < now)
		1658	{
		1659	// timeout occured, take action
		1660	}
		1661	else
		1662	{
		1663	// callback was invoked, but there was some activity, re-arm
		1664	// the watcher to fire in last_activity + 60, which is
		1665	// guaranteed to be in the future, so "again" is positive:
		1666	w->repeat = timeout - now;
		1667	ev_timer_again (EV_A_ w);
		1668	}
		1669	}
		1670
		1671	To summarise the callback: first calculate the real timeout (defined
		1672	as "60 seconds after the last activity"), then check if that time has
		1673	been reached, which means something I<did>, in fact, time out. Otherwise
		1674	the callback was invoked too early (C<timeout> is in the future), so
		1675	re-schedule the timer to fire at that future time, to see if maybe we have
		1676	a timeout then.
		1677
		1678	Note how C<ev_timer_again> is used, taking advantage of the
		1679	C<ev_timer_again> optimisation when the timer is already running.
		1680
		1681	This scheme causes more callback invocations (about one every 60 seconds
		1682	minus half the average time between activity), but virtually no calls to
		1683	libev to change the timeout.
		1684
		1685	To start the timer, simply initialise the watcher and set C<last_activity>
		1686	to the current time (meaning we just have some activity :), then call the
		1687	callback, which will "do the right thing" and start the timer:
		1688
		1689	ev_init (timer, callback);
		1690	last_activity = ev_now (loop);
		1691	callback (loop, timer, EV_TIMEOUT);
		1692
		1693	And when there is some activity, simply store the current time in
		1694	C<last_activity>, no libev calls at all:
		1695
		1696	last_actiivty = ev_now (loop);
		1697
		1698	This technique is slightly more complex, but in most cases where the
		1699	time-out is unlikely to be triggered, much more efficient.
		1700
		1701	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1702	callback :) - just change the timeout and invoke the callback, which will
		1703	fix things for you.
		1704
		1705	=item 4. Wee, just use a double-linked list for your timeouts.
		1706
		1707	If there is not one request, but many thousands (millions...), all
		1708	employing some kind of timeout with the same timeout value, then one can
		1709	do even better:
		1710
		1711	When starting the timeout, calculate the timeout value and put the timeout
		1712	at the I<end> of the list.
		1713
		1714	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1715	the list is expected to fire (for example, using the technique #3).
		1716
		1717	When there is some activity, remove the timer from the list, recalculate
		1718	the timeout, append it to the end of the list again, and make sure to
		1719	update the C<ev_timer> if it was taken from the beginning of the list.
		1720
		1721	This way, one can manage an unlimited number of timeouts in O(1) time for
		1722	starting, stopping and updating the timers, at the expense of a major
		1723	complication, and having to use a constant timeout. The constant timeout
		1724	ensures that the list stays sorted.
		1725
		1726	=back
		1727
		1728	So which method the best?
		1729
		1730	Method #2 is a simple no-brain-required solution that is adequate in most
		1731	situations. Method #3 requires a bit more thinking, but handles many cases
		1732	better, and isn't very complicated either. In most case, choosing either
		1733	one is fine, with #3 being better in typical situations.
		1734
		1735	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1736	rather complicated, but extremely efficient, something that really pays
		1737	off after the first million or so of active timers, i.e. it's usually
		1738	overkill :)
		1739
		1740	=head3 The special problem of time updates
		1741
		1742	Establishing the current time is a costly operation (it usually takes at
		1743	least two system calls): EV therefore updates its idea of the current
		1744	time only before and after C<ev_loop> collects new events, which causes a
		1745	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1746	lots of events in one iteration.
1129		1747
1130	The relative timeouts are calculated relative to the C<ev_now ()>	1748	The relative timeouts are calculated relative to the C<ev_now ()>
1131	time. This is usually the right thing as this timestamp refers to the time	1749	time. This is usually the right thing as this timestamp refers to the time
1132	of the event triggering whatever timeout you are modifying/starting. If	1750	of the event triggering whatever timeout you are modifying/starting. If
1133	you suspect event processing to be delayed and you I<need> to base the timeout	1751	you suspect event processing to be delayed and you I<need> to base the
1134	on the current time, use something like this to adjust for this:	1752	timeout on the current time, use something like this to adjust for this:
1135		1753
1136	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1754	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1137		1755
1138	The callback is guarenteed to be invoked only when its timeout has passed,	1756	If the event loop is suspended for a long time, you can also force an
1139	but if multiple timers become ready during the same loop iteration then	1757	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1140	order of execution is undefined.	1758	()>.
		1759
		1760	=head3 The special problems of suspended animation
		1761
		1762	When you leave the server world it is quite customary to hit machines that
		1763	can suspend/hibernate - what happens to the clocks during such a suspend?
		1764
		1765	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1766	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1767	to run until the system is suspended, but they will not advance while the
		1768	system is suspended. That means, on resume, it will be as if the program
		1769	was frozen for a few seconds, but the suspend time will not be counted
		1770	towards C<ev_timer> when a monotonic clock source is used. The real time
		1771	clock advanced as expected, but if it is used as sole clocksource, then a
		1772	long suspend would be detected as a time jump by libev, and timers would
		1773	be adjusted accordingly.
		1774
		1775	I would not be surprised to see different behaviour in different between
		1776	operating systems, OS versions or even different hardware.
		1777
		1778	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1779	time jump in the monotonic clocks and the realtime clock. If the program
		1780	is suspended for a very long time, and monotonic clock sources are in use,
		1781	then you can expect C<ev_timer>s to expire as the full suspension time
		1782	will be counted towards the timers. When no monotonic clock source is in
		1783	use, then libev will again assume a timejump and adjust accordingly.
		1784
		1785	It might be beneficial for this latter case to call C<ev_suspend>
		1786	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1787	deterministic behaviour in this case (you can do nothing against
		1788	C<SIGSTOP>).
1141		1789
1142	=head3 Watcher-Specific Functions and Data Members	1790	=head3 Watcher-Specific Functions and Data Members
1143		1791
1144	=over 4	1792	=over 4
1145		1793
1146	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1794	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1147		1795
1148	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1796	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1149		1797
1150	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1798	Configure the timer to trigger after C<after> seconds. If C<repeat>
1151	C<0.>, then it will automatically be stopped. If it is positive, then the	1799	is C<0.>, then it will automatically be stopped once the timeout is
1152	timer will automatically be configured to trigger again C<repeat> seconds	1800	reached. If it is positive, then the timer will automatically be
1153	later, again, and again, until stopped manually.	1801	configured to trigger again C<repeat> seconds later, again, and again,
		1802	until stopped manually.
1154		1803
1155	The timer itself will do a best-effort at avoiding drift, that is, if you	1804	The timer itself will do a best-effort at avoiding drift, that is, if
1156	configure a timer to trigger every 10 seconds, then it will trigger at	1805	you configure a timer to trigger every 10 seconds, then it will normally
1157	exactly 10 second intervals. If, however, your program cannot keep up with	1806	trigger at exactly 10 second intervals. If, however, your program cannot
1158	the timer (because it takes longer than those 10 seconds to do stuff) the	1807	keep up with the timer (because it takes longer than those 10 seconds to
1159	timer will not fire more than once per event loop iteration.	1808	do stuff) the timer will not fire more than once per event loop iteration.
1160		1809
1161	=item ev_timer_again (loop)	1810	=item ev_timer_again (loop, ev_timer *)
1162		1811
1163	This will act as if the timer timed out and restart it again if it is	1812	This will act as if the timer timed out and restart it again if it is
1164	repeating. The exact semantics are:	1813	repeating. The exact semantics are:
1165		1814
1166	If the timer is pending, its pending status is cleared.	1815	If the timer is pending, its pending status is cleared.
1167		1816
1168	If the timer is started but nonrepeating, stop it (as if it timed out).	1817	If the timer is started but non-repeating, stop it (as if it timed out).
1169		1818
1170	If the timer is repeating, either start it if necessary (with the	1819	If the timer is repeating, either start it if necessary (with the
1171	C<repeat> value), or reset the running timer to the C<repeat> value.	1820	C<repeat> value), or reset the running timer to the C<repeat> value.
1172		1821
1173	This sounds a bit complicated, but here is a useful and typical	1822	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1174	example: Imagine you have a tcp connection and you want a so-called idle	1823	usage example.
1175	timeout, that is, you want to be called when there have been, say, 60
1176	seconds of inactivity on the socket. The easiest way to do this is to
1177	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1178	C<ev_timer_again> each time you successfully read or write some data. If
1179	you go into an idle state where you do not expect data to travel on the
1180	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1181	automatically restart it if need be.
1182		1824
1183	That means you can ignore the C<after> value and C<ev_timer_start>	1825	=item ev_timer_remaining (loop, ev_timer *)
1184	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1185		1826
1186	ev_timer_init (timer, callback, 0., 5.);	1827	Returns the remaining time until a timer fires. If the timer is active,
1187	ev_timer_again (loop, timer);	1828	then this time is relative to the current event loop time, otherwise it's
1188	...	1829	the timeout value currently configured.
1189	timer->again = 17.;
1190	ev_timer_again (loop, timer);
1191	...
1192	timer->again = 10.;
1193	ev_timer_again (loop, timer);
1194		1830
1195	This is more slightly efficient then stopping/starting the timer each time	1831	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1196	you want to modify its timeout value.	1832	C<5>. When the timer is started and one second passes, C<ev_timer_remain>
		1833	will return C<4>. When the timer expires and is restarted, it will return
		1834	roughly C<7> (likely slightly less as callback invocation takes some time,
		1835	too), and so on.
1197		1836
1198	=item ev_tstamp repeat [read-write]	1837	=item ev_tstamp repeat [read-write]
1199		1838
1200	The current C<repeat> value. Will be used each time the watcher times out	1839	The current C<repeat> value. Will be used each time the watcher times out
1201	or C<ev_timer_again> is called and determines the next timeout (if any),	1840	or C<ev_timer_again> is called, and determines the next timeout (if any),
1202	which is also when any modifications are taken into account.	1841	which is also when any modifications are taken into account.
1203		1842
1204	=back	1843	=back
1205		1844
1206	=head3 Examples	1845	=head3 Examples
1207		1846
1208	Example: Create a timer that fires after 60 seconds.	1847	Example: Create a timer that fires after 60 seconds.
1209		1848
1210	static void	1849	static void
1211	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1850	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1212	{	1851	{
1213	.. one minute over, w is actually stopped right here	1852	.. one minute over, w is actually stopped right here
1214	}	1853	}
1215		1854
1216	struct ev_timer mytimer;	1855	ev_timer mytimer;
1217	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1856	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1218	ev_timer_start (loop, &mytimer);	1857	ev_timer_start (loop, &mytimer);
1219		1858
1220	Example: Create a timeout timer that times out after 10 seconds of	1859	Example: Create a timeout timer that times out after 10 seconds of
1221	inactivity.	1860	inactivity.
1222		1861
1223	static void	1862	static void
1224	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1863	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1225	{	1864	{
1226	.. ten seconds without any activity	1865	.. ten seconds without any activity
1227	}	1866	}
1228		1867
1229	struct ev_timer mytimer;	1868	ev_timer mytimer;
1230	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1869	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1231	ev_timer_again (&mytimer); /* start timer */	1870	ev_timer_again (&mytimer); /* start timer */
1232	ev_loop (loop, 0);	1871	ev_loop (loop, 0);
1233		1872
1234	// and in some piece of code that gets executed on any "activity":	1873	// and in some piece of code that gets executed on any "activity":
1235	// reset the timeout to start ticking again at 10 seconds	1874	// reset the timeout to start ticking again at 10 seconds
1236	ev_timer_again (&mytimer);	1875	ev_timer_again (&mytimer);
1237		1876
1238		1877
1239	=head2 C<ev_periodic> - to cron or not to cron?	1878	=head2 C<ev_periodic> - to cron or not to cron?
1240		1879
1241	Periodic watchers are also timers of a kind, but they are very versatile	1880	Periodic watchers are also timers of a kind, but they are very versatile
1242	(and unfortunately a bit complex).	1881	(and unfortunately a bit complex).
1243		1882
1244	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1883	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1245	but on wallclock time (absolute time). You can tell a periodic watcher	1884	relative time, the physical time that passes) but on wall clock time
1246	to trigger "at" some specific point in time. For example, if you tell a	1885	(absolute time, the thing you can read on your calender or clock). The
1247	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1886	difference is that wall clock time can run faster or slower than real
1248	+ 10.>) and then reset your system clock to the last year, then it will	1887	time, and time jumps are not uncommon (e.g. when you adjust your
1249	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1888	wrist-watch).
1250	roughly 10 seconds later).
1251		1889
1252	They can also be used to implement vastly more complex timers, such as	1890	You can tell a periodic watcher to trigger after some specific point
1253	triggering an event on each midnight, local time or other, complicated,	1891	in time: for example, if you tell a periodic watcher to trigger "in 10
1254	rules.	1892	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1893	not a delay) and then reset your system clock to January of the previous
		1894	year, then it will take a year or more to trigger the event (unlike an
		1895	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1896	it, as it uses a relative timeout).
1255		1897
		1898	C<ev_periodic> watchers can also be used to implement vastly more complex
		1899	timers, such as triggering an event on each "midnight, local time", or
		1900	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1901	those cannot react to time jumps.
		1902
1256	As with timers, the callback is guarenteed to be invoked only when the	1903	As with timers, the callback is guaranteed to be invoked only when the
1257	time (C<at>) has been passed, but if multiple periodic timers become ready	1904	point in time where it is supposed to trigger has passed. If multiple
1258	during the same loop iteration then order of execution is undefined.	1905	timers become ready during the same loop iteration then the ones with
		1906	earlier time-out values are invoked before ones with later time-out values
		1907	(but this is no longer true when a callback calls C<ev_loop> recursively).
1259		1908
1260	=head3 Watcher-Specific Functions and Data Members	1909	=head3 Watcher-Specific Functions and Data Members
1261		1910
1262	=over 4	1911	=over 4
1263		1912
1264	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1913	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1265		1914
1266	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1915	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1267		1916
1268	Lots of arguments, lets sort it out... There are basically three modes of	1917	Lots of arguments, let's sort it out... There are basically three modes of
1269	operation, and we will explain them from simplest to complex:	1918	operation, and we will explain them from simplest to most complex:
1270		1919
1271	=over 4	1920	=over 4
1272		1921
1273	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1922	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1274		1923
1275	In this configuration the watcher triggers an event at the wallclock time	1924	In this configuration the watcher triggers an event after the wall clock
1276	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1925	time C<offset> has passed. It will not repeat and will not adjust when a
1277	that is, if it is to be run at January 1st 2011 then it will run when the	1926	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1278	system time reaches or surpasses this time.	1927	will be stopped and invoked when the system clock reaches or surpasses
		1928	this point in time.
1279		1929
1280	=item * non-repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1930	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1281		1931
1282	In this mode the watcher will always be scheduled to time out at the next	1932	In this mode the watcher will always be scheduled to time out at the next
1283	C<at + N * interval> time (for some integer N, which can also be negative)	1933	C<offset + N * interval> time (for some integer N, which can also be
1284	and then repeat, regardless of any time jumps.	1934	negative) and then repeat, regardless of any time jumps. The C<offset>
		1935	argument is merely an offset into the C<interval> periods.
1285		1936
1286	This can be used to create timers that do not drift with respect to system	1937	This can be used to create timers that do not drift with respect to the
1287	time:	1938	system clock, for example, here is an C<ev_periodic> that triggers each
		1939	hour, on the hour (with respect to UTC):
1288		1940
1289	ev_periodic_set (&periodic, 0., 3600., 0);	1941	ev_periodic_set (&periodic, 0., 3600., 0);
1290		1942
1291	This doesn't mean there will always be 3600 seconds in between triggers,	1943	This doesn't mean there will always be 3600 seconds in between triggers,
1292	but only that the the callback will be called when the system time shows a	1944	but only that the callback will be called when the system time shows a
1293	full hour (UTC), or more correctly, when the system time is evenly divisible	1945	full hour (UTC), or more correctly, when the system time is evenly divisible
1294	by 3600.	1946	by 3600.
1295		1947
1296	Another way to think about it (for the mathematically inclined) is that	1948	Another way to think about it (for the mathematically inclined) is that
1297	C<ev_periodic> will try to run the callback in this mode at the next possible	1949	C<ev_periodic> will try to run the callback in this mode at the next possible
1298	time where C<time = at (mod interval)>, regardless of any time jumps.	1950	time where C<time = offset (mod interval)>, regardless of any time jumps.
1299		1951
1300	For numerical stability it is preferable that the C<at> value is near	1952	For numerical stability it is preferable that the C<offset> value is near
1301	C<ev_now ()> (the current time), but there is no range requirement for	1953	C<ev_now ()> (the current time), but there is no range requirement for
1302	this value.	1954	this value, and in fact is often specified as zero.
1303		1955
		1956	Note also that there is an upper limit to how often a timer can fire (CPU
		1957	speed for example), so if C<interval> is very small then timing stability
		1958	will of course deteriorate. Libev itself tries to be exact to be about one
		1959	millisecond (if the OS supports it and the machine is fast enough).
		1960
1304	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1961	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1305		1962
1306	In this mode the values for C<interval> and C<at> are both being	1963	In this mode the values for C<interval> and C<offset> are both being
1307	ignored. Instead, each time the periodic watcher gets scheduled, the	1964	ignored. Instead, each time the periodic watcher gets scheduled, the
1308	reschedule callback will be called with the watcher as first, and the	1965	reschedule callback will be called with the watcher as first, and the
1309	current time as second argument.	1966	current time as second argument.
1310		1967
1311	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1968	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1312	ever, or make any event loop modifications>. If you need to stop it,	1969	or make ANY other event loop modifications whatsoever, unless explicitly
1313	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	1970	allowed by documentation here>.
1314	starting an C<ev_prepare> watcher, which is legal).
1315		1971
		1972	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1973	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1974	only event loop modification you are allowed to do).
		1975
1316	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1976	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1317	ev_tstamp now)>, e.g.:	1977	*w, ev_tstamp now)>, e.g.:
1318		1978
		1979	static ev_tstamp
1319	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1980	my_rescheduler (ev_periodic *w, ev_tstamp now)
1320	{	1981	{
1321	return now + 60.;	1982	return now + 60.;
1322	}	1983	}
1323		1984
1324	It must return the next time to trigger, based on the passed time value	1985	It must return the next time to trigger, based on the passed time value
1325	(that is, the lowest time value larger than to the second argument). It	1986	(that is, the lowest time value larger than to the second argument). It
1326	will usually be called just before the callback will be triggered, but	1987	will usually be called just before the callback will be triggered, but
1327	might be called at other times, too.	1988	might be called at other times, too.
1328		1989
1329	NOTE: I<< This callback must always return a time that is later than the	1990	NOTE: I<< This callback must always return a time that is higher than or
1330	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1991	equal to the passed C<now> value >>.
1331		1992
1332	This can be used to create very complex timers, such as a timer that	1993	This can be used to create very complex timers, such as a timer that
1333	triggers on each midnight, local time. To do this, you would calculate the	1994	triggers on "next midnight, local time". To do this, you would calculate the
1334	next midnight after C<now> and return the timestamp value for this. How	1995	next midnight after C<now> and return the timestamp value for this. How
1335	you do this is, again, up to you (but it is not trivial, which is the main	1996	you do this is, again, up to you (but it is not trivial, which is the main
1336	reason I omitted it as an example).	1997	reason I omitted it as an example).
1337		1998
1338	=back	1999	=back
…		…
1342	Simply stops and restarts the periodic watcher again. This is only useful	2003	Simply stops and restarts the periodic watcher again. This is only useful
1343	when you changed some parameters or the reschedule callback would return	2004	when you changed some parameters or the reschedule callback would return
1344	a different time than the last time it was called (e.g. in a crond like	2005	a different time than the last time it was called (e.g. in a crond like
1345	program when the crontabs have changed).	2006	program when the crontabs have changed).
1346		2007
		2008	=item ev_tstamp ev_periodic_at (ev_periodic *)
		2009
		2010	When active, returns the absolute time that the watcher is supposed
		2011	to trigger next. This is not the same as the C<offset> argument to
		2012	C<ev_periodic_set>, but indeed works even in interval and manual
		2013	rescheduling modes.
		2014
1347	=item ev_tstamp offset [read-write]	2015	=item ev_tstamp offset [read-write]
1348		2016
1349	When repeating, this contains the offset value, otherwise this is the	2017	When repeating, this contains the offset value, otherwise this is the
1350	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2018	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2019	although libev might modify this value for better numerical stability).
1351		2020
1352	Can be modified any time, but changes only take effect when the periodic	2021	Can be modified any time, but changes only take effect when the periodic
1353	timer fires or C<ev_periodic_again> is being called.	2022	timer fires or C<ev_periodic_again> is being called.
1354		2023
1355	=item ev_tstamp interval [read-write]	2024	=item ev_tstamp interval [read-write]
1356		2025
1357	The current interval value. Can be modified any time, but changes only	2026	The current interval value. Can be modified any time, but changes only
1358	take effect when the periodic timer fires or C<ev_periodic_again> is being	2027	take effect when the periodic timer fires or C<ev_periodic_again> is being
1359	called.	2028	called.
1360		2029
1361	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2030	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1362		2031
1363	The current reschedule callback, or C<0>, if this functionality is	2032	The current reschedule callback, or C<0>, if this functionality is
1364	switched off. Can be changed any time, but changes only take effect when	2033	switched off. Can be changed any time, but changes only take effect when
1365	the periodic timer fires or C<ev_periodic_again> is being called.	2034	the periodic timer fires or C<ev_periodic_again> is being called.
1366		2035
1367	=item ev_tstamp at [read-only]
1368
1369	When active, contains the absolute time that the watcher is supposed to
1370	trigger next.
1371
1372	=back	2036	=back
1373		2037
1374	=head3 Examples	2038	=head3 Examples
1375		2039
1376	Example: Call a callback every hour, or, more precisely, whenever the	2040	Example: Call a callback every hour, or, more precisely, whenever the
1377	system clock is divisible by 3600. The callback invocation times have	2041	system time is divisible by 3600. The callback invocation times have
1378	potentially a lot of jittering, but good long-term stability.	2042	potentially a lot of jitter, but good long-term stability.
1379		2043
1380	static void	2044	static void
1381	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2045	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1382	{	2046	{
1383	... its now a full hour (UTC, or TAI or whatever your clock follows)	2047	... its now a full hour (UTC, or TAI or whatever your clock follows)
1384	}	2048	}
1385		2049
1386	struct ev_periodic hourly_tick;	2050	ev_periodic hourly_tick;
1387	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2051	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1388	ev_periodic_start (loop, &hourly_tick);	2052	ev_periodic_start (loop, &hourly_tick);
1389		2053
1390	Example: The same as above, but use a reschedule callback to do it:	2054	Example: The same as above, but use a reschedule callback to do it:
1391		2055
1392	#include <math.h>	2056	#include <math.h>
1393		2057
1394	static ev_tstamp	2058	static ev_tstamp
1395	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2059	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1396	{	2060	{
1397	return fmod (now, 3600.) + 3600.;	2061	return now + (3600. - fmod (now, 3600.));
1398	}	2062	}
1399		2063
1400	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2064	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1401		2065
1402	Example: Call a callback every hour, starting now:	2066	Example: Call a callback every hour, starting now:
1403		2067
1404	struct ev_periodic hourly_tick;	2068	ev_periodic hourly_tick;
1405	ev_periodic_init (&hourly_tick, clock_cb,	2069	ev_periodic_init (&hourly_tick, clock_cb,
1406	fmod (ev_now (loop), 3600.), 3600., 0);	2070	fmod (ev_now (loop), 3600.), 3600., 0);
1407	ev_periodic_start (loop, &hourly_tick);	2071	ev_periodic_start (loop, &hourly_tick);
1408		2072
1409		2073
1410	=head2 C<ev_signal> - signal me when a signal gets signalled!	2074	=head2 C<ev_signal> - signal me when a signal gets signalled!
1411		2075
1412	Signal watchers will trigger an event when the process receives a specific	2076	Signal watchers will trigger an event when the process receives a specific
1413	signal one or more times. Even though signals are very asynchronous, libev	2077	signal one or more times. Even though signals are very asynchronous, libev
1414	will try it's best to deliver signals synchronously, i.e. as part of the	2078	will try it's best to deliver signals synchronously, i.e. as part of the
1415	normal event processing, like any other event.	2079	normal event processing, like any other event.
1416		2080
		2081	Note that only the default loop supports registering signal watchers
		2082	currently.
		2083
		2084	If you want signals asynchronously, just use C<sigaction> as you would
		2085	do without libev and forget about sharing the signal. You can even use
		2086	C<ev_async> from a signal handler to synchronously wake up an event loop.
		2087
1417	You can configure as many watchers as you like per signal. Only when the	2088	You can configure as many watchers as you like per signal. Only when the
1418	first watcher gets started will libev actually register a signal watcher	2089	first watcher gets started will libev actually register something with
1419	with the kernel (thus it coexists with your own signal handlers as long	2090	the kernel (thus it coexists with your own signal handlers as long as you
1420	as you don't register any with libev). Similarly, when the last signal	2091	don't register any with libev for the same signal).
1421	watcher for a signal is stopped libev will reset the signal handler to	2092
1422	SIG_DFL (regardless of what it was set to before).	2093	Both the signal mask state (C<sigprocmask>) and the signal handler state
		2094	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2095	sotpping it again), that is, libev might or might not block the signal,
		2096	and might or might not set or restore the installed signal handler.
		2097
		2098	If possible and supported, libev will install its handlers with
		2099	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
		2100	not be unduly interrupted. If you have a problem with system calls getting
		2101	interrupted by signals you can block all signals in an C<ev_check> watcher
		2102	and unblock them in an C<ev_prepare> watcher.
1423		2103
1424	=head3 Watcher-Specific Functions and Data Members	2104	=head3 Watcher-Specific Functions and Data Members
1425		2105
1426	=over 4	2106	=over 4
1427		2107
…		…
1436		2116
1437	The signal the watcher watches out for.	2117	The signal the watcher watches out for.
1438		2118
1439	=back	2119	=back
1440		2120
		2121	=head3 Examples
		2122
		2123	Example: Try to exit cleanly on SIGINT.
		2124
		2125	static void
		2126	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		2127	{
		2128	ev_unloop (loop, EVUNLOOP_ALL);
		2129	}
		2130
		2131	ev_signal signal_watcher;
		2132	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		2133	ev_signal_start (loop, &signal_watcher);
		2134
1441		2135
1442	=head2 C<ev_child> - watch out for process status changes	2136	=head2 C<ev_child> - watch out for process status changes
1443		2137
1444	Child watchers trigger when your process receives a SIGCHLD in response to	2138	Child watchers trigger when your process receives a SIGCHLD in response to
1445	some child status changes (most typically when a child of yours dies).	2139	some child status changes (most typically when a child of yours dies or
		2140	exits). It is permissible to install a child watcher I<after> the child
		2141	has been forked (which implies it might have already exited), as long
		2142	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2143	forking and then immediately registering a watcher for the child is fine,
		2144	but forking and registering a watcher a few event loop iterations later or
		2145	in the next callback invocation is not.
		2146
		2147	Only the default event loop is capable of handling signals, and therefore
		2148	you can only register child watchers in the default event loop.
		2149
		2150	Due to some design glitches inside libev, child watchers will always be
		2151	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2152	libev)
		2153
		2154	=head3 Process Interaction
		2155
		2156	Libev grabs C<SIGCHLD> as soon as the default event loop is
		2157	initialised. This is necessary to guarantee proper behaviour even if the
		2158	first child watcher is started after the child exits. The occurrence
		2159	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		2160	synchronously as part of the event loop processing. Libev always reaps all
		2161	children, even ones not watched.
		2162
		2163	=head3 Overriding the Built-In Processing
		2164
		2165	Libev offers no special support for overriding the built-in child
		2166	processing, but if your application collides with libev's default child
		2167	handler, you can override it easily by installing your own handler for
		2168	C<SIGCHLD> after initialising the default loop, and making sure the
		2169	default loop never gets destroyed. You are encouraged, however, to use an
		2170	event-based approach to child reaping and thus use libev's support for
		2171	that, so other libev users can use C<ev_child> watchers freely.
		2172
		2173	=head3 Stopping the Child Watcher
		2174
		2175	Currently, the child watcher never gets stopped, even when the
		2176	child terminates, so normally one needs to stop the watcher in the
		2177	callback. Future versions of libev might stop the watcher automatically
		2178	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2179	problem).
1446		2180
1447	=head3 Watcher-Specific Functions and Data Members	2181	=head3 Watcher-Specific Functions and Data Members
1448		2182
1449	=over 4	2183	=over 4
1450		2184
…		…
1476		2210
1477	=back	2211	=back
1478		2212
1479	=head3 Examples	2213	=head3 Examples
1480		2214
1481	Example: Try to exit cleanly on SIGINT and SIGTERM.	2215	Example: C<fork()> a new process and install a child handler to wait for
		2216	its completion.
1482		2217
		2218	ev_child cw;
		2219
1483	static void	2220	static void
1484	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2221	child_cb (EV_P_ ev_child *w, int revents)
1485	{	2222	{
1486	ev_unloop (loop, EVUNLOOP_ALL);	2223	ev_child_stop (EV_A_ w);
		2224	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1487	}	2225	}
1488		2226
1489	struct ev_signal signal_watcher;	2227	pid_t pid = fork ();
1490	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2228
1491	ev_signal_start (loop, &sigint_cb);	2229	if (pid < 0)
		2230	// error
		2231	else if (pid == 0)
		2232	{
		2233	// the forked child executes here
		2234	exit (1);
		2235	}
		2236	else
		2237	{
		2238	ev_child_init (&cw, child_cb, pid, 0);
		2239	ev_child_start (EV_DEFAULT_ &cw);
		2240	}
1492		2241
1493		2242
1494	=head2 C<ev_stat> - did the file attributes just change?	2243	=head2 C<ev_stat> - did the file attributes just change?
1495		2244
1496	This watches a filesystem path for attribute changes. That is, it calls	2245	This watches a file system path for attribute changes. That is, it calls
1497	C<stat> regularly (or when the OS says it changed) and sees if it changed	2246	C<stat> on that path in regular intervals (or when the OS says it changed)
1498	compared to the last time, invoking the callback if it did.	2247	and sees if it changed compared to the last time, invoking the callback if
		2248	it did.
1499		2249
1500	The path does not need to exist: changing from "path exists" to "path does	2250	The path does not need to exist: changing from "path exists" to "path does
1501	not exist" is a status change like any other. The condition "path does	2251	not exist" is a status change like any other. The condition "path does not
1502	not exist" is signified by the C<st_nlink> field being zero (which is	2252	exist" (or more correctly "path cannot be stat'ed") is signified by the
1503	otherwise always forced to be at least one) and all the other fields of	2253	C<st_nlink> field being zero (which is otherwise always forced to be at
1504	the stat buffer having unspecified contents.	2254	least one) and all the other fields of the stat buffer having unspecified
		2255	contents.
1505		2256
1506	The path I<should> be absolute and I<must not> end in a slash. If it is	2257	The path I<must not> end in a slash or contain special components such as
		2258	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1507	relative and your working directory changes, the behaviour is undefined.	2259	your working directory changes, then the behaviour is undefined.
1508		2260
1509	Since there is no standard to do this, the portable implementation simply	2261	Since there is no portable change notification interface available, the
1510	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2262	portable implementation simply calls C<stat(2)> regularly on the path
1511	can specify a recommended polling interval for this case. If you specify	2263	to see if it changed somehow. You can specify a recommended polling
1512	a polling interval of C<0> (highly recommended!) then a I<suitable,	2264	interval for this case. If you specify a polling interval of C<0> (highly
1513	unspecified default> value will be used (which you can expect to be around	2265	recommended!) then a I<suitable, unspecified default> value will be used
1514	five seconds, although this might change dynamically). Libev will also	2266	(which you can expect to be around five seconds, although this might
1515	impose a minimum interval which is currently around C<0.1>, but thats	2267	change dynamically). Libev will also impose a minimum interval which is
1516	usually overkill.	2268	currently around C<0.1>, but that's usually overkill.
1517		2269
1518	This watcher type is not meant for massive numbers of stat watchers,	2270	This watcher type is not meant for massive numbers of stat watchers,
1519	as even with OS-supported change notifications, this can be	2271	as even with OS-supported change notifications, this can be
1520	resource-intensive.	2272	resource-intensive.
1521		2273
1522	At the time of this writing, only the Linux inotify interface is	2274	At the time of this writing, the only OS-specific interface implemented
1523	implemented (implementing kqueue support is left as an exercise for the	2275	is the Linux inotify interface (implementing kqueue support is left as an
1524	reader). Inotify will be used to give hints only and should not change the	2276	exercise for the reader. Note, however, that the author sees no way of
1525	semantics of C<ev_stat> watchers, which means that libev sometimes needs	2277	implementing C<ev_stat> semantics with kqueue, except as a hint).
1526	to fall back to regular polling again even with inotify, but changes are
1527	usually detected immediately, and if the file exists there will be no
1528	polling.
1529		2278
1530	=head3 Inotify	2279	=head3 ABI Issues (Largefile Support)
1531		2280
		2281	Libev by default (unless the user overrides this) uses the default
		2282	compilation environment, which means that on systems with large file
		2283	support disabled by default, you get the 32 bit version of the stat
		2284	structure. When using the library from programs that change the ABI to
		2285	use 64 bit file offsets the programs will fail. In that case you have to
		2286	compile libev with the same flags to get binary compatibility. This is
		2287	obviously the case with any flags that change the ABI, but the problem is
		2288	most noticeably displayed with ev_stat and large file support.
		2289
		2290	The solution for this is to lobby your distribution maker to make large
		2291	file interfaces available by default (as e.g. FreeBSD does) and not
		2292	optional. Libev cannot simply switch on large file support because it has
		2293	to exchange stat structures with application programs compiled using the
		2294	default compilation environment.
		2295
		2296	=head3 Inotify and Kqueue
		2297
1532	When C<inotify (7)> support has been compiled into libev (generally only	2298	When C<inotify (7)> support has been compiled into libev and present at
1533	available on Linux) and present at runtime, it will be used to speed up	2299	runtime, it will be used to speed up change detection where possible. The
1534	change detection where possible. The inotify descriptor will be created lazily	2300	inotify descriptor will be created lazily when the first C<ev_stat>
1535	when the first C<ev_stat> watcher is being started.	2301	watcher is being started.
1536		2302
1537	Inotify presense does not change the semantics of C<ev_stat> watchers	2303	Inotify presence does not change the semantics of C<ev_stat> watchers
1538	except that changes might be detected earlier, and in some cases, to avoid	2304	except that changes might be detected earlier, and in some cases, to avoid
1539	making regular C<stat> calls. Even in the presense of inotify support	2305	making regular C<stat> calls. Even in the presence of inotify support
1540	there are many cases where libev has to resort to regular C<stat> polling.	2306	there are many cases where libev has to resort to regular C<stat> polling,
		2307	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2308	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2309	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2310	xfs are fully working) libev usually gets away without polling.
1541		2311
1542	(There is no support for kqueue, as apparently it cannot be used to	2312	There is no support for kqueue, as apparently it cannot be used to
1543	implement this functionality, due to the requirement of having a file	2313	implement this functionality, due to the requirement of having a file
1544	descriptor open on the object at all times).	2314	descriptor open on the object at all times, and detecting renames, unlinks
		2315	etc. is difficult.
		2316
		2317	=head3 C<stat ()> is a synchronous operation
		2318
		2319	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2320	the process. The exception are C<ev_stat> watchers - those call C<stat
		2321	()>, which is a synchronous operation.
		2322
		2323	For local paths, this usually doesn't matter: unless the system is very
		2324	busy or the intervals between stat's are large, a stat call will be fast,
		2325	as the path data is usually in memory already (except when starting the
		2326	watcher).
		2327
		2328	For networked file systems, calling C<stat ()> can block an indefinite
		2329	time due to network issues, and even under good conditions, a stat call
		2330	often takes multiple milliseconds.
		2331
		2332	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2333	paths, although this is fully supported by libev.
1545		2334
1546	=head3 The special problem of stat time resolution	2335	=head3 The special problem of stat time resolution
1547		2336
1548	The C<stat ()> syscall only supports full-second resolution portably, and	2337	The C<stat ()> system call only supports full-second resolution portably,
1549	even on systems where the resolution is higher, many filesystems still	2338	and even on systems where the resolution is higher, most file systems
1550	only support whole seconds.	2339	still only support whole seconds.
1551		2340
1552	That means that, if the time is the only thing that changes, you might	2341	That means that, if the time is the only thing that changes, you can
1553	miss updates: on the first update, C<ev_stat> detects a change and calls	2342	easily miss updates: on the first update, C<ev_stat> detects a change and
1554	your callback, which does something. When there is another update within	2343	calls your callback, which does something. When there is another update
1555	the same second, C<ev_stat> will be unable to detect it.	2344	within the same second, C<ev_stat> will be unable to detect unless the
		2345	stat data does change in other ways (e.g. file size).
1556		2346
1557	The solution to this is to delay acting on a change for a second (or till	2347	The solution to this is to delay acting on a change for slightly more
1558	the next second boundary), using a roughly one-second delay C<ev_timer>	2348	than a second (or till slightly after the next full second boundary), using
1559	(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>	2349	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1560	is added to work around small timing inconsistencies of some operating	2350	ev_timer_again (loop, w)>).
1561	systems.	2351
		2352	The C<.02> offset is added to work around small timing inconsistencies
		2353	of some operating systems (where the second counter of the current time
		2354	might be be delayed. One such system is the Linux kernel, where a call to
		2355	C<gettimeofday> might return a timestamp with a full second later than
		2356	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2357	update file times then there will be a small window where the kernel uses
		2358	the previous second to update file times but libev might already execute
		2359	the timer callback).
1562		2360
1563	=head3 Watcher-Specific Functions and Data Members	2361	=head3 Watcher-Specific Functions and Data Members
1564		2362
1565	=over 4	2363	=over 4
1566		2364
…		…
1572	C<path>. The C<interval> is a hint on how quickly a change is expected to	2370	C<path>. The C<interval> is a hint on how quickly a change is expected to
1573	be detected and should normally be specified as C<0> to let libev choose	2371	be detected and should normally be specified as C<0> to let libev choose
1574	a suitable value. The memory pointed to by C<path> must point to the same	2372	a suitable value. The memory pointed to by C<path> must point to the same
1575	path for as long as the watcher is active.	2373	path for as long as the watcher is active.
1576		2374
1577	The callback will be receive C<EV_STAT> when a change was detected,	2375	The callback will receive an C<EV_STAT> event when a change was detected,
1578	relative to the attributes at the time the watcher was started (or the	2376	relative to the attributes at the time the watcher was started (or the
1579	last change was detected).	2377	last change was detected).
1580		2378
1581	=item ev_stat_stat (ev_stat *)	2379	=item ev_stat_stat (loop, ev_stat *)
1582		2380
1583	Updates the stat buffer immediately with new values. If you change the	2381	Updates the stat buffer immediately with new values. If you change the
1584	watched path in your callback, you could call this fucntion to avoid	2382	watched path in your callback, you could call this function to avoid
1585	detecting this change (while introducing a race condition). Can also be	2383	detecting this change (while introducing a race condition if you are not
1586	useful simply to find out the new values.	2384	the only one changing the path). Can also be useful simply to find out the
		2385	new values.
1587		2386
1588	=item ev_statdata attr [read-only]	2387	=item ev_statdata attr [read-only]
1589		2388
1590	The most-recently detected attributes of the file. Although the type is of	2389	The most-recently detected attributes of the file. Although the type is
1591	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types	2390	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
1592	suitable for your system. If the C<st_nlink> member is C<0>, then there	2391	suitable for your system, but you can only rely on the POSIX-standardised
		2392	members to be present. If the C<st_nlink> member is C<0>, then there was
1593	was some error while C<stat>ing the file.	2393	some error while C<stat>ing the file.
1594		2394
1595	=item ev_statdata prev [read-only]	2395	=item ev_statdata prev [read-only]
1596		2396
1597	The previous attributes of the file. The callback gets invoked whenever	2397	The previous attributes of the file. The callback gets invoked whenever
1598	C<prev> != C<attr>.	2398	C<prev> != C<attr>, or, more precisely, one or more of these members
		2399	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2400	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1599		2401
1600	=item ev_tstamp interval [read-only]	2402	=item ev_tstamp interval [read-only]
1601		2403
1602	The specified interval.	2404	The specified interval.
1603		2405
1604	=item const char *path [read-only]	2406	=item const char *path [read-only]
1605		2407
1606	The filesystem path that is being watched.	2408	The file system path that is being watched.
1607		2409
1608	=back	2410	=back
1609		2411
1610	=head3 Examples	2412	=head3 Examples
1611		2413
1612	Example: Watch C</etc/passwd> for attribute changes.	2414	Example: Watch C</etc/passwd> for attribute changes.
1613		2415
1614	static void	2416	static void
1615	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	2417	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1616	{	2418	{
1617	/* /etc/passwd changed in some way */	2419	/* /etc/passwd changed in some way */
1618	if (w->attr.st_nlink)	2420	if (w->attr.st_nlink)
1619	{	2421	{
1620	printf ("passwd current size %ld\n", (long)w->attr.st_size);	2422	printf ("passwd current size %ld\n", (long)w->attr.st_size);
1621	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);	2423	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
1622	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);	2424	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
1623	}	2425	}
1624	else	2426	else
1625	/* you shalt not abuse printf for puts */	2427	/* you shalt not abuse printf for puts */
1626	puts ("wow, /etc/passwd is not there, expect problems. "	2428	puts ("wow, /etc/passwd is not there, expect problems. "
1627	"if this is windows, they already arrived\n");	2429	"if this is windows, they already arrived\n");
1628	}	2430	}
1629		2431
1630	...	2432	...
1631	ev_stat passwd;	2433	ev_stat passwd;
1632		2434
1633	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);	2435	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1634	ev_stat_start (loop, &passwd);	2436	ev_stat_start (loop, &passwd);
1635		2437
1636	Example: Like above, but additionally use a one-second delay so we do not	2438	Example: Like above, but additionally use a one-second delay so we do not
1637	miss updates (however, frequent updates will delay processing, too, so	2439	miss updates (however, frequent updates will delay processing, too, so
1638	one might do the work both on C<ev_stat> callback invocation I<and> on	2440	one might do the work both on C<ev_stat> callback invocation I<and> on
1639	C<ev_timer> callback invocation).	2441	C<ev_timer> callback invocation).
1640		2442
1641	static ev_stat passwd;	2443	static ev_stat passwd;
1642	static ev_timer timer;	2444	static ev_timer timer;
1643		2445
1644	static void	2446	static void
1645	timer_cb (EV_P_ ev_timer *w, int revents)	2447	timer_cb (EV_P_ ev_timer *w, int revents)
1646	{	2448	{
1647	ev_timer_stop (EV_A_ w);	2449	ev_timer_stop (EV_A_ w);
1648		2450
1649	/* now it's one second after the most recent passwd change */	2451	/* now it's one second after the most recent passwd change */
1650	}	2452	}
1651		2453
1652	static void	2454	static void
1653	stat_cb (EV_P_ ev_stat *w, int revents)	2455	stat_cb (EV_P_ ev_stat *w, int revents)
1654	{	2456	{
1655	/* reset the one-second timer */	2457	/* reset the one-second timer */
1656	ev_timer_again (EV_A_ &timer);	2458	ev_timer_again (EV_A_ &timer);
1657	}	2459	}
1658		2460
1659	...	2461	...
1660	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	2462	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1661	ev_stat_start (loop, &passwd);	2463	ev_stat_start (loop, &passwd);
1662	ev_timer_init (&timer, timer_cb, 0., 1.01);	2464	ev_timer_init (&timer, timer_cb, 0., 1.02);
1663		2465
1664		2466
1665	=head2 C<ev_idle> - when you've got nothing better to do...	2467	=head2 C<ev_idle> - when you've got nothing better to do...
1666		2468
1667	Idle watchers trigger events when no other events of the same or higher	2469	Idle watchers trigger events when no other events of the same or higher
1668	priority are pending (prepare, check and other idle watchers do not	2470	priority are pending (prepare, check and other idle watchers do not count
1669	count).	2471	as receiving "events").
1670		2472
1671	That is, as long as your process is busy handling sockets or timeouts	2473	That is, as long as your process is busy handling sockets or timeouts
1672	(or even signals, imagine) of the same or higher priority it will not be	2474	(or even signals, imagine) of the same or higher priority it will not be
1673	triggered. But when your process is idle (or only lower-priority watchers	2475	triggered. But when your process is idle (or only lower-priority watchers
1674	are pending), the idle watchers are being called once per event loop	2476	are pending), the idle watchers are being called once per event loop
…		…
1685		2487
1686	=head3 Watcher-Specific Functions and Data Members	2488	=head3 Watcher-Specific Functions and Data Members
1687		2489
1688	=over 4	2490	=over 4
1689		2491
1690	=item ev_idle_init (ev_signal *, callback)	2492	=item ev_idle_init (ev_idle *, callback)
1691		2493
1692	Initialises and configures the idle watcher - it has no parameters of any	2494	Initialises and configures the idle watcher - it has no parameters of any
1693	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2495	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1694	believe me.	2496	believe me.
1695		2497
…		…
1698	=head3 Examples	2500	=head3 Examples
1699		2501
1700	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2502	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1701	callback, free it. Also, use no error checking, as usual.	2503	callback, free it. Also, use no error checking, as usual.
1702		2504
1703	static void	2505	static void
1704	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2506	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1705	{	2507	{
1706	free (w);	2508	free (w);
1707	// now do something you wanted to do when the program has	2509	// now do something you wanted to do when the program has
1708	// no longer anything immediate to do.	2510	// no longer anything immediate to do.
1709	}	2511	}
1710		2512
1711	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2513	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1712	ev_idle_init (idle_watcher, idle_cb);	2514	ev_idle_init (idle_watcher, idle_cb);
1713	ev_idle_start (loop, idle_cb);	2515	ev_idle_start (loop, idle_watcher);
1714		2516
1715		2517
1716	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2518	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1717		2519
1718	Prepare and check watchers are usually (but not always) used in tandem:	2520	Prepare and check watchers are usually (but not always) used in pairs:
1719	prepare watchers get invoked before the process blocks and check watchers	2521	prepare watchers get invoked before the process blocks and check watchers
1720	afterwards.	2522	afterwards.
1721		2523
1722	You I<must not> call C<ev_loop> or similar functions that enter	2524	You I<must not> call C<ev_loop> or similar functions that enter
1723	the current event loop from either C<ev_prepare> or C<ev_check>	2525	the current event loop from either C<ev_prepare> or C<ev_check>
…		…
1726	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2528	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1727	C<ev_check> so if you have one watcher of each kind they will always be	2529	C<ev_check> so if you have one watcher of each kind they will always be
1728	called in pairs bracketing the blocking call.	2530	called in pairs bracketing the blocking call.
1729		2531
1730	Their main purpose is to integrate other event mechanisms into libev and	2532	Their main purpose is to integrate other event mechanisms into libev and
1731	their use is somewhat advanced. This could be used, for example, to track	2533	their use is somewhat advanced. They could be used, for example, to track
1732	variable changes, implement your own watchers, integrate net-snmp or a	2534	variable changes, implement your own watchers, integrate net-snmp or a
1733	coroutine library and lots more. They are also occasionally useful if	2535	coroutine library and lots more. They are also occasionally useful if
1734	you cache some data and want to flush it before blocking (for example,	2536	you cache some data and want to flush it before blocking (for example,
1735	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2537	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1736	watcher).	2538	watcher).
1737		2539
1738	This is done by examining in each prepare call which file descriptors need	2540	This is done by examining in each prepare call which file descriptors
1739	to be watched by the other library, registering C<ev_io> watchers for	2541	need to be watched by the other library, registering C<ev_io> watchers
1740	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2542	for them and starting an C<ev_timer> watcher for any timeouts (many
1741	provide just this functionality). Then, in the check watcher you check for	2543	libraries provide exactly this functionality). Then, in the check watcher,
1742	any events that occured (by checking the pending status of all watchers	2544	you check for any events that occurred (by checking the pending status
1743	and stopping them) and call back into the library. The I/O and timer	2545	of all watchers and stopping them) and call back into the library. The
1744	callbacks will never actually be called (but must be valid nevertheless,	2546	I/O and timer callbacks will never actually be called (but must be valid
1745	because you never know, you know?).	2547	nevertheless, because you never know, you know?).
1746		2548
1747	As another example, the Perl Coro module uses these hooks to integrate	2549	As another example, the Perl Coro module uses these hooks to integrate
1748	coroutines into libev programs, by yielding to other active coroutines	2550	coroutines into libev programs, by yielding to other active coroutines
1749	during each prepare and only letting the process block if no coroutines	2551	during each prepare and only letting the process block if no coroutines
1750	are ready to run (it's actually more complicated: it only runs coroutines	2552	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1753	loop from blocking if lower-priority coroutines are active, thus mapping	2555	loop from blocking if lower-priority coroutines are active, thus mapping
1754	low-priority coroutines to idle/background tasks).	2556	low-priority coroutines to idle/background tasks).
1755		2557
1756	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2558	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1757	priority, to ensure that they are being run before any other watchers	2559	priority, to ensure that they are being run before any other watchers
		2560	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2561
1758	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2562	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1759	too) should not activate ("feed") events into libev. While libev fully	2563	activate ("feed") events into libev. While libev fully supports this, they
1760	supports this, they will be called before other C<ev_check> watchers	2564	might get executed before other C<ev_check> watchers did their job. As
1761	did their job. As C<ev_check> watchers are often used to embed other	2565	C<ev_check> watchers are often used to embed other (non-libev) event
1762	(non-libev) event loops those other event loops might be in an unusable	2566	loops those other event loops might be in an unusable state until their
1763	state until their C<ev_check> watcher ran (always remind yourself to	2567	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1764	coexist peacefully with others).	2568	others).
1765		2569
1766	=head3 Watcher-Specific Functions and Data Members	2570	=head3 Watcher-Specific Functions and Data Members
1767		2571
1768	=over 4	2572	=over 4
1769		2573
…		…
1771		2575
1772	=item ev_check_init (ev_check *, callback)	2576	=item ev_check_init (ev_check *, callback)
1773		2577
1774	Initialises and configures the prepare or check watcher - they have no	2578	Initialises and configures the prepare or check watcher - they have no
1775	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2579	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1776	macros, but using them is utterly, utterly and completely pointless.	2580	macros, but using them is utterly, utterly, utterly and completely
		2581	pointless.
1777		2582
1778	=back	2583	=back
1779		2584
1780	=head3 Examples	2585	=head3 Examples
1781		2586
1782	There are a number of principal ways to embed other event loops or modules	2587	There are a number of principal ways to embed other event loops or modules
1783	into libev. Here are some ideas on how to include libadns into libev	2588	into libev. Here are some ideas on how to include libadns into libev
1784	(there is a Perl module named C<EV::ADNS> that does this, which you could	2589	(there is a Perl module named C<EV::ADNS> that does this, which you could
1785	use for an actually working example. Another Perl module named C<EV::Glib>	2590	use as a working example. Another Perl module named C<EV::Glib> embeds a
1786	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV	2591	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
1787	into the Glib event loop).	2592	Glib event loop).
1788		2593
1789	Method 1: Add IO watchers and a timeout watcher in a prepare handler,	2594	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1790	and in a check watcher, destroy them and call into libadns. What follows	2595	and in a check watcher, destroy them and call into libadns. What follows
1791	is pseudo-code only of course. This requires you to either use a low	2596	is pseudo-code only of course. This requires you to either use a low
1792	priority for the check watcher or use C<ev_clear_pending> explicitly, as	2597	priority for the check watcher or use C<ev_clear_pending> explicitly, as
1793	the callbacks for the IO/timeout watchers might not have been called yet.	2598	the callbacks for the IO/timeout watchers might not have been called yet.
1794		2599
1795	static ev_io iow [nfd];	2600	static ev_io iow [nfd];
1796	static ev_timer tw;	2601	static ev_timer tw;
1797		2602
1798	static void	2603	static void
1799	io_cb (ev_loop *loop, ev_io *w, int revents)	2604	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1800	{	2605	{
1801	}	2606	}
1802		2607
1803	// create io watchers for each fd and a timer before blocking	2608	// create io watchers for each fd and a timer before blocking
1804	static void	2609	static void
1805	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2610	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1806	{	2611	{
1807	int timeout = 3600000;	2612	int timeout = 3600000;
1808	struct pollfd fds [nfd];	2613	struct pollfd fds [nfd];
1809	// actual code will need to loop here and realloc etc.	2614	// actual code will need to loop here and realloc etc.
1810	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2615	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1811		2616
1812	/* the callback is illegal, but won't be called as we stop during check */	2617	/* the callback is illegal, but won't be called as we stop during check */
1813	ev_timer_init (&tw, 0, timeout * 1e-3);	2618	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
1814	ev_timer_start (loop, &tw);	2619	ev_timer_start (loop, &tw);
1815		2620
1816	// create one ev_io per pollfd	2621	// create one ev_io per pollfd
1817	for (int i = 0; i < nfd; ++i)	2622	for (int i = 0; i < nfd; ++i)
1818	{	2623	{
1819	ev_io_init (iow + i, io_cb, fds [i].fd,	2624	ev_io_init (iow + i, io_cb, fds [i].fd,
1820	((fds [i].events & POLLIN ? EV_READ : 0)	2625	((fds [i].events & POLLIN ? EV_READ : 0)
1821	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	2626	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1822		2627
1823	fds [i].revents = 0;	2628	fds [i].revents = 0;
1824	ev_io_start (loop, iow + i);	2629	ev_io_start (loop, iow + i);
1825	}	2630	}
1826	}	2631	}
1827		2632
1828	// stop all watchers after blocking	2633	// stop all watchers after blocking
1829	static void	2634	static void
1830	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2635	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
1831	{	2636	{
1832	ev_timer_stop (loop, &tw);	2637	ev_timer_stop (loop, &tw);
1833		2638
1834	for (int i = 0; i < nfd; ++i)	2639	for (int i = 0; i < nfd; ++i)
1835	{	2640	{
1836	// set the relevant poll flags	2641	// set the relevant poll flags
1837	// could also call adns_processreadable etc. here	2642	// could also call adns_processreadable etc. here
1838	struct pollfd *fd = fds + i;	2643	struct pollfd *fd = fds + i;
1839	int revents = ev_clear_pending (iow + i);	2644	int revents = ev_clear_pending (iow + i);
1840	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;	2645	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1841	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;	2646	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1842		2647
1843	// now stop the watcher	2648	// now stop the watcher
1844	ev_io_stop (loop, iow + i);	2649	ev_io_stop (loop, iow + i);
1845	}	2650	}
1846		2651
1847	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	2652	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1848	}	2653	}
1849		2654
1850	Method 2: This would be just like method 1, but you run C<adns_afterpoll>	2655	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
1851	in the prepare watcher and would dispose of the check watcher.	2656	in the prepare watcher and would dispose of the check watcher.
1852		2657
1853	Method 3: If the module to be embedded supports explicit event	2658	Method 3: If the module to be embedded supports explicit event
1854	notification (adns does), you can also make use of the actual watcher	2659	notification (libadns does), you can also make use of the actual watcher
1855	callbacks, and only destroy/create the watchers in the prepare watcher.	2660	callbacks, and only destroy/create the watchers in the prepare watcher.
1856		2661
1857	static void	2662	static void
1858	timer_cb (EV_P_ ev_timer *w, int revents)	2663	timer_cb (EV_P_ ev_timer *w, int revents)
1859	{	2664	{
1860	adns_state ads = (adns_state)w->data;	2665	adns_state ads = (adns_state)w->data;
1861	update_now (EV_A);	2666	update_now (EV_A);
1862		2667
1863	adns_processtimeouts (ads, &tv_now);	2668	adns_processtimeouts (ads, &tv_now);
1864	}	2669	}
1865		2670
1866	static void	2671	static void
1867	io_cb (EV_P_ ev_io *w, int revents)	2672	io_cb (EV_P_ ev_io *w, int revents)
1868	{	2673	{
1869	adns_state ads = (adns_state)w->data;	2674	adns_state ads = (adns_state)w->data;
1870	update_now (EV_A);	2675	update_now (EV_A);
1871		2676
1872	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);	2677	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
1873	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);	2678	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
1874	}	2679	}
1875		2680
1876	// do not ever call adns_afterpoll	2681	// do not ever call adns_afterpoll
1877		2682
1878	Method 4: Do not use a prepare or check watcher because the module you	2683	Method 4: Do not use a prepare or check watcher because the module you
1879	want to embed is too inflexible to support it. Instead, youc na override	2684	want to embed is not flexible enough to support it. Instead, you can
1880	their poll function. The drawback with this solution is that the main	2685	override their poll function. The drawback with this solution is that the
1881	loop is now no longer controllable by EV. The C<Glib::EV> module does	2686	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
1882	this.	2687	this approach, effectively embedding EV as a client into the horrible
		2688	libglib event loop.
1883		2689
1884	static gint	2690	static gint
1885	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2691	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
1886	{	2692	{
1887	int got_events = 0;	2693	int got_events = 0;
1888		2694
1889	for (n = 0; n < nfds; ++n)	2695	for (n = 0; n < nfds; ++n)
1890	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events	2696	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
1891		2697
1892	if (timeout >= 0)	2698	if (timeout >= 0)
1893	// create/start timer	2699	// create/start timer
1894		2700
1895	// poll	2701	// poll
1896	ev_loop (EV_A_ 0);	2702	ev_loop (EV_A_ 0);
1897		2703
1898	// stop timer again	2704	// stop timer again
1899	if (timeout >= 0)	2705	if (timeout >= 0)
1900	ev_timer_stop (EV_A_ &to);	2706	ev_timer_stop (EV_A_ &to);
1901		2707
1902	// stop io watchers again - their callbacks should have set	2708	// stop io watchers again - their callbacks should have set
1903	for (n = 0; n < nfds; ++n)	2709	for (n = 0; n < nfds; ++n)
1904	ev_io_stop (EV_A_ iow [n]);	2710	ev_io_stop (EV_A_ iow [n]);
1905		2711
1906	return got_events;	2712	return got_events;
1907	}	2713	}
1908		2714
1909		2715
1910	=head2 C<ev_embed> - when one backend isn't enough...	2716	=head2 C<ev_embed> - when one backend isn't enough...
1911		2717
1912	This is a rather advanced watcher type that lets you embed one event loop	2718	This is a rather advanced watcher type that lets you embed one event loop
…		…
1918	prioritise I/O.	2724	prioritise I/O.
1919		2725
1920	As an example for a bug workaround, the kqueue backend might only support	2726	As an example for a bug workaround, the kqueue backend might only support
1921	sockets on some platform, so it is unusable as generic backend, but you	2727	sockets on some platform, so it is unusable as generic backend, but you
1922	still want to make use of it because you have many sockets and it scales	2728	still want to make use of it because you have many sockets and it scales
1923	so nicely. In this case, you would create a kqueue-based loop and embed it	2729	so nicely. In this case, you would create a kqueue-based loop and embed
1924	into your default loop (which might use e.g. poll). Overall operation will	2730	it into your default loop (which might use e.g. poll). Overall operation
1925	be a bit slower because first libev has to poll and then call kevent, but	2731	will be a bit slower because first libev has to call C<poll> and then
1926	at least you can use both at what they are best.	2732	C<kevent>, but at least you can use both mechanisms for what they are
		2733	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1927		2734
1928	As for prioritising I/O: rarely you have the case where some fds have	2735	As for prioritising I/O: under rare circumstances you have the case where
1929	to be watched and handled very quickly (with low latency), and even	2736	some fds have to be watched and handled very quickly (with low latency),
1930	priorities and idle watchers might have too much overhead. In this case	2737	and even priorities and idle watchers might have too much overhead. In
1931	you would put all the high priority stuff in one loop and all the rest in	2738	this case you would put all the high priority stuff in one loop and all
1932	a second one, and embed the second one in the first.	2739	the rest in a second one, and embed the second one in the first.
1933		2740
1934	As long as the watcher is active, the callback will be invoked every time	2741	As long as the watcher is active, the callback will be invoked every
1935	there might be events pending in the embedded loop. The callback must then	2742	time there might be events pending in the embedded loop. The callback
1936	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2743	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
1937	their callbacks (you could also start an idle watcher to give the embedded	2744	sweep and invoke their callbacks (the callback doesn't need to invoke the
1938	loop strictly lower priority for example). You can also set the callback	2745	C<ev_embed_sweep> function directly, it could also start an idle watcher
1939	to C<0>, in which case the embed watcher will automatically execute the	2746	to give the embedded loop strictly lower priority for example).
1940	embedded loop sweep.
1941		2747
1942	As long as the watcher is started it will automatically handle events. The	2748	You can also set the callback to C<0>, in which case the embed watcher
1943	callback will be invoked whenever some events have been handled. You can	2749	will automatically execute the embedded loop sweep whenever necessary.
1944	set the callback to C<0> to avoid having to specify one if you are not
1945	interested in that.
1946		2750
1947	Also, there have not currently been made special provisions for forking:	2751	Fork detection will be handled transparently while the C<ev_embed> watcher
1948	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2752	is active, i.e., the embedded loop will automatically be forked when the
1949	but you will also have to stop and restart any C<ev_embed> watchers	2753	embedding loop forks. In other cases, the user is responsible for calling
1950	yourself.	2754	C<ev_loop_fork> on the embedded loop.
1951		2755
1952	Unfortunately, not all backends are embeddable, only the ones returned by	2756	Unfortunately, not all backends are embeddable: only the ones returned by
1953	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2757	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1954	portable one.	2758	portable one.
1955		2759
1956	So when you want to use this feature you will always have to be prepared	2760	So when you want to use this feature you will always have to be prepared
1957	that you cannot get an embeddable loop. The recommended way to get around	2761	that you cannot get an embeddable loop. The recommended way to get around
1958	this is to have a separate variables for your embeddable loop, try to	2762	this is to have a separate variables for your embeddable loop, try to
1959	create it, and if that fails, use the normal loop for everything.	2763	create it, and if that fails, use the normal loop for everything.
1960		2764
		2765	=head3 C<ev_embed> and fork
		2766
		2767	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2768	automatically be applied to the embedded loop as well, so no special
		2769	fork handling is required in that case. When the watcher is not running,
		2770	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2771	as applicable.
		2772
1961	=head3 Watcher-Specific Functions and Data Members	2773	=head3 Watcher-Specific Functions and Data Members
1962		2774
1963	=over 4	2775	=over 4
1964		2776
1965	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2777	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
…		…
1968		2780
1969	Configures the watcher to embed the given loop, which must be	2781	Configures the watcher to embed the given loop, which must be
1970	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	2782	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1971	invoked automatically, otherwise it is the responsibility of the callback	2783	invoked automatically, otherwise it is the responsibility of the callback
1972	to invoke it (it will continue to be called until the sweep has been done,	2784	to invoke it (it will continue to be called until the sweep has been done,
1973	if you do not want thta, you need to temporarily stop the embed watcher).	2785	if you do not want that, you need to temporarily stop the embed watcher).
1974		2786
1975	=item ev_embed_sweep (loop, ev_embed *)	2787	=item ev_embed_sweep (loop, ev_embed *)
1976		2788
1977	Make a single, non-blocking sweep over the embedded loop. This works	2789	Make a single, non-blocking sweep over the embedded loop. This works
1978	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	2790	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1979	apropriate way for embedded loops.	2791	appropriate way for embedded loops.
1980		2792
1981	=item struct ev_loop *other [read-only]	2793	=item struct ev_loop *other [read-only]
1982		2794
1983	The embedded event loop.	2795	The embedded event loop.
1984		2796
…		…
1986		2798
1987	=head3 Examples	2799	=head3 Examples
1988		2800
1989	Example: Try to get an embeddable event loop and embed it into the default	2801	Example: Try to get an embeddable event loop and embed it into the default
1990	event loop. If that is not possible, use the default loop. The default	2802	event loop. If that is not possible, use the default loop. The default
1991	loop is stored in C<loop_hi>, while the mebeddable loop is stored in	2803	loop is stored in C<loop_hi>, while the embeddable loop is stored in
1992	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be	2804	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
1993	used).	2805	used).
1994		2806
1995	struct ev_loop *loop_hi = ev_default_init (0);	2807	struct ev_loop *loop_hi = ev_default_init (0);
1996	struct ev_loop *loop_lo = 0;	2808	struct ev_loop *loop_lo = 0;
1997	struct ev_embed embed;	2809	ev_embed embed;
1998		2810
1999	// see if there is a chance of getting one that works	2811	// see if there is a chance of getting one that works
2000	// (remember that a flags value of 0 means autodetection)	2812	// (remember that a flags value of 0 means autodetection)
2001	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2813	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2002	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2814	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
2003	: 0;	2815	: 0;
2004		2816
2005	// if we got one, then embed it, otherwise default to loop_hi	2817	// if we got one, then embed it, otherwise default to loop_hi
2006	if (loop_lo)	2818	if (loop_lo)
2007	{	2819	{
2008	ev_embed_init (&embed, 0, loop_lo);	2820	ev_embed_init (&embed, 0, loop_lo);
2009	ev_embed_start (loop_hi, &embed);	2821	ev_embed_start (loop_hi, &embed);
2010	}	2822	}
2011	else	2823	else
2012	loop_lo = loop_hi;	2824	loop_lo = loop_hi;
2013		2825
2014	Example: Check if kqueue is available but not recommended and create	2826	Example: Check if kqueue is available but not recommended and create
2015	a kqueue backend for use with sockets (which usually work with any	2827	a kqueue backend for use with sockets (which usually work with any
2016	kqueue implementation). Store the kqueue/socket-only event loop in	2828	kqueue implementation). Store the kqueue/socket-only event loop in
2017	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2829	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2018		2830
2019	struct ev_loop *loop = ev_default_init (0);	2831	struct ev_loop *loop = ev_default_init (0);
2020	struct ev_loop *loop_socket = 0;	2832	struct ev_loop *loop_socket = 0;
2021	struct ev_embed embed;	2833	ev_embed embed;
2022		2834
2023	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2835	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2024	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2836	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2025	{	2837	{
2026	ev_embed_init (&embed, 0, loop_socket);	2838	ev_embed_init (&embed, 0, loop_socket);
2027	ev_embed_start (loop, &embed);	2839	ev_embed_start (loop, &embed);
2028	}	2840	}
2029		2841
2030	if (!loop_socket)	2842	if (!loop_socket)
2031	loop_socket = loop;	2843	loop_socket = loop;
2032		2844
2033	// now use loop_socket for all sockets, and loop for everything else	2845	// now use loop_socket for all sockets, and loop for everything else
2034		2846
2035		2847
2036	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	2848	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
2037		2849
2038	Fork watchers are called when a C<fork ()> was detected (usually because	2850	Fork watchers are called when a C<fork ()> was detected (usually because
…		…
2041	event loop blocks next and before C<ev_check> watchers are being called,	2853	event loop blocks next and before C<ev_check> watchers are being called,
2042	and only in the child after the fork. If whoever good citizen calling	2854	and only in the child after the fork. If whoever good citizen calling
2043	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2855	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2044	handlers will be invoked, too, of course.	2856	handlers will be invoked, too, of course.
2045		2857
		2858	=head3 The special problem of life after fork - how is it possible?
		2859
		2860	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2861	up/change the process environment, followed by a call to C<exec()>. This
		2862	sequence should be handled by libev without any problems.
		2863
		2864	This changes when the application actually wants to do event handling
		2865	in the child, or both parent in child, in effect "continuing" after the
		2866	fork.
		2867
		2868	The default mode of operation (for libev, with application help to detect
		2869	forks) is to duplicate all the state in the child, as would be expected
		2870	when I<either> the parent I<or> the child process continues.
		2871
		2872	When both processes want to continue using libev, then this is usually the
		2873	wrong result. In that case, usually one process (typically the parent) is
		2874	supposed to continue with all watchers in place as before, while the other
		2875	process typically wants to start fresh, i.e. without any active watchers.
		2876
		2877	The cleanest and most efficient way to achieve that with libev is to
		2878	simply create a new event loop, which of course will be "empty", and
		2879	use that for new watchers. This has the advantage of not touching more
		2880	memory than necessary, and thus avoiding the copy-on-write, and the
		2881	disadvantage of having to use multiple event loops (which do not support
		2882	signal watchers).
		2883
		2884	When this is not possible, or you want to use the default loop for
		2885	other reasons, then in the process that wants to start "fresh", call
		2886	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2887	the default loop will "orphan" (not stop) all registered watchers, so you
		2888	have to be careful not to execute code that modifies those watchers. Note
		2889	also that in that case, you have to re-register any signal watchers.
		2890
2046	=head3 Watcher-Specific Functions and Data Members	2891	=head3 Watcher-Specific Functions and Data Members
2047		2892
2048	=over 4	2893	=over 4
2049		2894
2050	=item ev_fork_init (ev_signal *, callback)	2895	=item ev_fork_init (ev_signal *, callback)
…		…
2082	is that the author does not know of a simple (or any) algorithm for a	2927	is that the author does not know of a simple (or any) algorithm for a
2083	multiple-writer-single-reader queue that works in all cases and doesn't	2928	multiple-writer-single-reader queue that works in all cases and doesn't
2084	need elaborate support such as pthreads.	2929	need elaborate support such as pthreads.
2085		2930
2086	That means that if you want to queue data, you have to provide your own	2931	That means that if you want to queue data, you have to provide your own
2087	queue. But at least I can tell you would implement locking around your	2932	queue. But at least I can tell you how to implement locking around your
2088	queue:	2933	queue:
2089		2934
2090	=over 4	2935	=over 4
2091		2936
2092	=item queueing from a signal handler context	2937	=item queueing from a signal handler context
2093		2938
2094	To implement race-free queueing, you simply add to the queue in the signal	2939	To implement race-free queueing, you simply add to the queue in the signal
2095	handler but you block the signal handler in the watcher callback. Here is an example that does that for	2940	handler but you block the signal handler in the watcher callback. Here is
2096	some fictitiuous SIGUSR1 handler:	2941	an example that does that for some fictitious SIGUSR1 handler:
2097		2942
2098	static ev_async mysig;	2943	static ev_async mysig;
2099		2944
2100	static void	2945	static void
2101	sigusr1_handler (void)	2946	sigusr1_handler (void)
2102	{	2947	{
2103	sometype data;	2948	sometype data;
2104		2949
2105	// no locking etc.	2950	// no locking etc.
2106	queue_put (data);	2951	queue_put (data);
2107	ev_async_send (DEFAULT_ &mysig);	2952	ev_async_send (EV_DEFAULT_ &mysig);
2108	}	2953	}
2109		2954
2110	static void	2955	static void
2111	mysig_cb (EV_P_ ev_async *w, int revents)	2956	mysig_cb (EV_P_ ev_async *w, int revents)
2112	{	2957	{
…		…
2143	// only need to lock the actual queueing operation	2988	// only need to lock the actual queueing operation
2144	pthread_mutex_lock (&mymutex);	2989	pthread_mutex_lock (&mymutex);
2145	queue_put (data);	2990	queue_put (data);
2146	pthread_mutex_unlock (&mymutex);	2991	pthread_mutex_unlock (&mymutex);
2147		2992
2148	ev_async_send (DEFAULT_ &mysig);	2993	ev_async_send (EV_DEFAULT_ &mysig);
2149	}	2994	}
2150		2995
2151	static void	2996	static void
2152	mysig_cb (EV_P_ ev_async *w, int revents)	2997	mysig_cb (EV_P_ ev_async *w, int revents)
2153	{	2998	{
…		…
2167	=over 4	3012	=over 4
2168		3013
2169	=item ev_async_init (ev_async *, callback)	3014	=item ev_async_init (ev_async *, callback)
2170		3015
2171	Initialises and configures the async watcher - it has no parameters of any	3016	Initialises and configures the async watcher - it has no parameters of any
2172	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3017	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2173	believe me.	3018	trust me.
2174		3019
2175	=item ev_async_send (loop, ev_async *)	3020	=item ev_async_send (loop, ev_async *)
2176		3021
2177	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3022	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2178	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3023	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2179	C<ev_feed_event>, this call is safe to do in other threads, signal or	3024	C<ev_feed_event>, this call is safe to do from other threads, signal or
2180	similar contexts (see the dicusssion of C<EV_ATOMIC_T> in the embedding	3025	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2181	section below on what exactly this means).	3026	section below on what exactly this means).
2182		3027
		3028	Note that, as with other watchers in libev, multiple events might get
		3029	compressed into a single callback invocation (another way to look at this
		3030	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3031	reset when the event loop detects that).
		3032
2183	This call incurs the overhead of a syscall only once per loop iteration,	3033	This call incurs the overhead of a system call only once per event loop
2184	so while the overhead might be noticable, it doesn't apply to repeated	3034	iteration, so while the overhead might be noticeable, it doesn't apply to
2185	calls to C<ev_async_send>.	3035	repeated calls to C<ev_async_send> for the same event loop.
		3036
		3037	=item bool = ev_async_pending (ev_async *)
		3038
		3039	Returns a non-zero value when C<ev_async_send> has been called on the
		3040	watcher but the event has not yet been processed (or even noted) by the
		3041	event loop.
		3042
		3043	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		3044	the loop iterates next and checks for the watcher to have become active,
		3045	it will reset the flag again. C<ev_async_pending> can be used to very
		3046	quickly check whether invoking the loop might be a good idea.
		3047
		3048	Not that this does I<not> check whether the watcher itself is pending,
		3049	only whether it has been requested to make this watcher pending: there
		3050	is a time window between the event loop checking and resetting the async
		3051	notification, and the callback being invoked.
2186		3052
2187	=back	3053	=back
2188		3054
2189		3055
2190	=head1 OTHER FUNCTIONS	3056	=head1 OTHER FUNCTIONS
…		…
2194	=over 4	3060	=over 4
2195		3061
2196	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3062	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2197		3063
2198	This function combines a simple timer and an I/O watcher, calls your	3064	This function combines a simple timer and an I/O watcher, calls your
2199	callback on whichever event happens first and automatically stop both	3065	callback on whichever event happens first and automatically stops both
2200	watchers. This is useful if you want to wait for a single event on an fd	3066	watchers. This is useful if you want to wait for a single event on an fd
2201	or timeout without having to allocate/configure/start/stop/free one or	3067	or timeout without having to allocate/configure/start/stop/free one or
2202	more watchers yourself.	3068	more watchers yourself.
2203		3069
2204	If C<fd> is less than 0, then no I/O watcher will be started and events	3070	If C<fd> is less than 0, then no I/O watcher will be started and the
2205	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3071	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2206	C<events> set will be craeted and started.	3072	the given C<fd> and C<events> set will be created and started.
2207		3073
2208	If C<timeout> is less than 0, then no timeout watcher will be	3074	If C<timeout> is less than 0, then no timeout watcher will be
2209	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3075	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2210	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3076	repeat = 0) will be started. C<0> is a valid timeout.
2211	dubious value.
2212		3077
2213	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3078	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2214	passed an C<revents> set like normal event callbacks (a combination of	3079	passed an C<revents> set like normal event callbacks (a combination of
2215	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3080	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2216	value passed to C<ev_once>:	3081	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3082	a timeout and an io event at the same time - you probably should give io
		3083	events precedence.
2217		3084
		3085	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3086
2218	static void stdin_ready (int revents, void *arg)	3087	static void stdin_ready (int revents, void *arg)
2219	{	3088	{
2220	if (revents & EV_TIMEOUT)
2221	/* doh, nothing entered */;
2222	else if (revents & EV_READ)	3089	if (revents & EV_READ)
2223	/* stdin might have data for us, joy! */;	3090	/* stdin might have data for us, joy! */;
		3091	else if (revents & EV_TIMEOUT)
		3092	/* doh, nothing entered */;
2224	}	3093	}
2225		3094
2226	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3095	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2227		3096
2228	=item ev_feed_event (ev_loop *, watcher *, int revents)	3097	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
2229		3098
2230	Feeds the given event set into the event loop, as if the specified event	3099	Feeds the given event set into the event loop, as if the specified event
2231	had happened for the specified watcher (which must be a pointer to an	3100	had happened for the specified watcher (which must be a pointer to an
2232	initialised but not necessarily started event watcher).	3101	initialised but not necessarily started event watcher).
2233		3102
2234	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3103	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)
2235		3104
2236	Feed an event on the given fd, as if a file descriptor backend detected	3105	Feed an event on the given fd, as if a file descriptor backend detected
2237	the given events it.	3106	the given events it.
2238		3107
2239	=item ev_feed_signal_event (ev_loop *loop, int signum)	3108	=item ev_feed_signal_event (struct ev_loop *loop, int signum)
2240		3109
2241	Feed an event as if the given signal occured (C<loop> must be the default	3110	Feed an event as if the given signal occurred (C<loop> must be the default
2242	loop!).	3111	loop!).
2243		3112
2244	=back	3113	=back
2245		3114
2246		3115
…		…
2262		3131
2263	=item * Priorities are not currently supported. Initialising priorities	3132	=item * Priorities are not currently supported. Initialising priorities
2264	will fail and all watchers will have the same priority, even though there	3133	will fail and all watchers will have the same priority, even though there
2265	is an ev_pri field.	3134	is an ev_pri field.
2266		3135
		3136	=item * In libevent, the last base created gets the signals, in libev, the
		3137	first base created (== the default loop) gets the signals.
		3138
2267	=item * Other members are not supported.	3139	=item * Other members are not supported.
2268		3140
2269	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3141	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2270	to use the libev header file and library.	3142	to use the libev header file and library.
2271		3143
2272	=back	3144	=back
2273		3145
2274	=head1 C++ SUPPORT	3146	=head1 C++ SUPPORT
2275		3147
2276	Libev comes with some simplistic wrapper classes for C++ that mainly allow	3148	Libev comes with some simplistic wrapper classes for C++ that mainly allow
2277	you to use some convinience methods to start/stop watchers and also change	3149	you to use some convenience methods to start/stop watchers and also change
2278	the callback model to a model using method callbacks on objects.	3150	the callback model to a model using method callbacks on objects.
2279		3151
2280	To use it,	3152	To use it,
2281		3153
2282	#include <ev++.h>	3154	#include <ev++.h>
2283		3155
2284	This automatically includes F<ev.h> and puts all of its definitions (many	3156	This automatically includes F<ev.h> and puts all of its definitions (many
2285	of them macros) into the global namespace. All C++ specific things are	3157	of them macros) into the global namespace. All C++ specific things are
2286	put into the C<ev> namespace. It should support all the same embedding	3158	put into the C<ev> namespace. It should support all the same embedding
2287	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.	3159	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
…		…
2354	your compiler is good :), then the method will be fully inlined into the	3226	your compiler is good :), then the method will be fully inlined into the
2355	thunking function, making it as fast as a direct C callback.	3227	thunking function, making it as fast as a direct C callback.
2356		3228
2357	Example: simple class declaration and watcher initialisation	3229	Example: simple class declaration and watcher initialisation
2358		3230
2359	struct myclass	3231	struct myclass
2360	{	3232	{
2361	void io_cb (ev::io &w, int revents) { }	3233	void io_cb (ev::io &w, int revents) { }
2362	}	3234	}
2363		3235
2364	myclass obj;	3236	myclass obj;
2365	ev::io iow;	3237	ev::io iow;
2366	iow.set <myclass, &myclass::io_cb> (&obj);	3238	iow.set <myclass, &myclass::io_cb> (&obj);
		3239
		3240	=item w->set (object *)
		3241
		3242	This is an B<experimental> feature that might go away in a future version.
		3243
		3244	This is a variation of a method callback - leaving out the method to call
		3245	will default the method to C<operator ()>, which makes it possible to use
		3246	functor objects without having to manually specify the C<operator ()> all
		3247	the time. Incidentally, you can then also leave out the template argument
		3248	list.
		3249
		3250	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3251	int revents)>.
		3252
		3253	See the method-C<set> above for more details.
		3254
		3255	Example: use a functor object as callback.
		3256
		3257	struct myfunctor
		3258	{
		3259	void operator() (ev::io &w, int revents)
		3260	{
		3261	...
		3262	}
		3263	}
		3264
		3265	myfunctor f;
		3266
		3267	ev::io w;
		3268	w.set (&f);
2367		3269
2368	=item w->set<function> (void *data = 0)	3270	=item w->set<function> (void *data = 0)
2369		3271
2370	Also sets a callback, but uses a static method or plain function as	3272	Also sets a callback, but uses a static method or plain function as
2371	callback. The optional C<data> argument will be stored in the watcher's	3273	callback. The optional C<data> argument will be stored in the watcher's
…		…
2373		3275
2374	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	3276	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2375		3277
2376	See the method-C<set> above for more details.	3278	See the method-C<set> above for more details.
2377		3279
2378	Example:	3280	Example: Use a plain function as callback.
2379		3281
2380	static void io_cb (ev::io &w, int revents) { }	3282	static void io_cb (ev::io &w, int revents) { }
2381	iow.set <io_cb> ();	3283	iow.set <io_cb> ();
2382		3284
2383	=item w->set (struct ev_loop *)	3285	=item w->set (struct ev_loop *)
2384		3286
2385	Associates a different C<struct ev_loop> with this watcher. You can only	3287	Associates a different C<struct ev_loop> with this watcher. You can only
2386	do this when the watcher is inactive (and not pending either).	3288	do this when the watcher is inactive (and not pending either).
2387		3289
2388	=item w->set ([args])	3290	=item w->set ([arguments])
2389		3291
2390	Basically the same as C<ev_TYPE_set>, with the same args. Must be	3292	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be
2391	called at least once. Unlike the C counterpart, an active watcher gets	3293	called at least once. Unlike the C counterpart, an active watcher gets
2392	automatically stopped and restarted when reconfiguring it with this	3294	automatically stopped and restarted when reconfiguring it with this
2393	method.	3295	method.
2394		3296
2395	=item w->start ()	3297	=item w->start ()
…		…
2419	=back	3321	=back
2420		3322
2421	Example: Define a class with an IO and idle watcher, start one of them in	3323	Example: Define a class with an IO and idle watcher, start one of them in
2422	the constructor.	3324	the constructor.
2423		3325
2424	class myclass	3326	class myclass
2425	{	3327	{
2426	ev::io io; void io_cb (ev::io &w, int revents);	3328	ev::io io ; void io_cb (ev::io &w, int revents);
2427	ev:idle idle void idle_cb (ev::idle &w, int revents);	3329	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2428		3330
2429	myclass (int fd)	3331	myclass (int fd)
2430	{	3332	{
2431	io .set <myclass, &myclass::io_cb > (this);	3333	io .set <myclass, &myclass::io_cb > (this);
2432	idle.set <myclass, &myclass::idle_cb> (this);	3334	idle.set <myclass, &myclass::idle_cb> (this);
2433		3335
2434	io.start (fd, ev::READ);	3336	io.start (fd, ev::READ);
2435	}	3337	}
2436	};	3338	};
		3339
		3340
		3341	=head1 OTHER LANGUAGE BINDINGS
		3342
		3343	Libev does not offer other language bindings itself, but bindings for a
		3344	number of languages exist in the form of third-party packages. If you know
		3345	any interesting language binding in addition to the ones listed here, drop
		3346	me a note.
		3347
		3348	=over 4
		3349
		3350	=item Perl
		3351
		3352	The EV module implements the full libev API and is actually used to test
		3353	libev. EV is developed together with libev. Apart from the EV core module,
		3354	there are additional modules that implement libev-compatible interfaces
		3355	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		3356	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3357	and C<EV::Glib>).
		3358
		3359	It can be found and installed via CPAN, its homepage is at
		3360	L<http://software.schmorp.de/pkg/EV>.
		3361
		3362	=item Python
		3363
		3364	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		3365	seems to be quite complete and well-documented.
		3366
		3367	=item Ruby
		3368
		3369	Tony Arcieri has written a ruby extension that offers access to a subset
		3370	of the libev API and adds file handle abstractions, asynchronous DNS and
		3371	more on top of it. It can be found via gem servers. Its homepage is at
		3372	L<http://rev.rubyforge.org/>.
		3373
		3374	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3375	makes rev work even on mingw.
		3376
		3377	=item Haskell
		3378
		3379	A haskell binding to libev is available at
		3380	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3381
		3382	=item D
		3383
		3384	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		3385	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3386
		3387	=item Ocaml
		3388
		3389	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3390	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3391
		3392	=back
2437		3393
2438		3394
2439	=head1 MACRO MAGIC	3395	=head1 MACRO MAGIC
2440		3396
2441	Libev can be compiled with a variety of options, the most fundamantal	3397	Libev can be compiled with a variety of options, the most fundamental
2442	of which is C<EV_MULTIPLICITY>. This option determines whether (most)	3398	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
2443	functions and callbacks have an initial C<struct ev_loop *> argument.	3399	functions and callbacks have an initial C<struct ev_loop *> argument.
2444		3400
2445	To make it easier to write programs that cope with either variant, the	3401	To make it easier to write programs that cope with either variant, the
2446	following macros are defined:	3402	following macros are defined:
…		…
2451		3407
2452	This provides the loop I<argument> for functions, if one is required ("ev	3408	This provides the loop I<argument> for functions, if one is required ("ev
2453	loop argument"). The C<EV_A> form is used when this is the sole argument,	3409	loop argument"). The C<EV_A> form is used when this is the sole argument,
2454	C<EV_A_> is used when other arguments are following. Example:	3410	C<EV_A_> is used when other arguments are following. Example:
2455		3411
2456	ev_unref (EV_A);	3412	ev_unref (EV_A);
2457	ev_timer_add (EV_A_ watcher);	3413	ev_timer_add (EV_A_ watcher);
2458	ev_loop (EV_A_ 0);	3414	ev_loop (EV_A_ 0);
2459		3415
2460	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	3416	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
2461	which is often provided by the following macro.	3417	which is often provided by the following macro.
2462		3418
2463	=item C<EV_P>, C<EV_P_>	3419	=item C<EV_P>, C<EV_P_>
2464		3420
2465	This provides the loop I<parameter> for functions, if one is required ("ev	3421	This provides the loop I<parameter> for functions, if one is required ("ev
2466	loop parameter"). The C<EV_P> form is used when this is the sole parameter,	3422	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
2467	C<EV_P_> is used when other parameters are following. Example:	3423	C<EV_P_> is used when other parameters are following. Example:
2468		3424
2469	// this is how ev_unref is being declared	3425	// this is how ev_unref is being declared
2470	static void ev_unref (EV_P);	3426	static void ev_unref (EV_P);
2471		3427
2472	// this is how you can declare your typical callback	3428	// this is how you can declare your typical callback
2473	static void cb (EV_P_ ev_timer *w, int revents)	3429	static void cb (EV_P_ ev_timer *w, int revents)
2474		3430
2475	It declares a parameter C<loop> of type C<struct ev_loop *>, quite	3431	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
2476	suitable for use with C<EV_A>.	3432	suitable for use with C<EV_A>.
2477		3433
2478	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	3434	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
2479		3435
2480	Similar to the other two macros, this gives you the value of the default	3436	Similar to the other two macros, this gives you the value of the default
2481	loop, if multiple loops are supported ("ev loop default").	3437	loop, if multiple loops are supported ("ev loop default").
		3438
		3439	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		3440
		3441	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		3442	default loop has been initialised (C<UC> == unchecked). Their behaviour
		3443	is undefined when the default loop has not been initialised by a previous
		3444	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		3445
		3446	It is often prudent to use C<EV_DEFAULT> when initialising the first
		3447	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
2482		3448
2483	=back	3449	=back
2484		3450
2485	Example: Declare and initialise a check watcher, utilising the above	3451	Example: Declare and initialise a check watcher, utilising the above
2486	macros so it will work regardless of whether multiple loops are supported	3452	macros so it will work regardless of whether multiple loops are supported
2487	or not.	3453	or not.
2488		3454
2489	static void	3455	static void
2490	check_cb (EV_P_ ev_timer *w, int revents)	3456	check_cb (EV_P_ ev_timer *w, int revents)
2491	{	3457	{
2492	ev_check_stop (EV_A_ w);	3458	ev_check_stop (EV_A_ w);
2493	}	3459	}
2494		3460
2495	ev_check check;	3461	ev_check check;
2496	ev_check_init (&check, check_cb);	3462	ev_check_init (&check, check_cb);
2497	ev_check_start (EV_DEFAULT_ &check);	3463	ev_check_start (EV_DEFAULT_ &check);
2498	ev_loop (EV_DEFAULT_ 0);	3464	ev_loop (EV_DEFAULT_ 0);
2499		3465
2500	=head1 EMBEDDING	3466	=head1 EMBEDDING
2501		3467
2502	Libev can (and often is) directly embedded into host	3468	Libev can (and often is) directly embedded into host
2503	applications. Examples of applications that embed it include the Deliantra	3469	applications. Examples of applications that embed it include the Deliantra
…		…
2510	libev somewhere in your source tree).	3476	libev somewhere in your source tree).
2511		3477
2512	=head2 FILESETS	3478	=head2 FILESETS
2513		3479
2514	Depending on what features you need you need to include one or more sets of files	3480	Depending on what features you need you need to include one or more sets of files
2515	in your app.	3481	in your application.
2516		3482
2517	=head3 CORE EVENT LOOP	3483	=head3 CORE EVENT LOOP
2518		3484
2519	To include only the libev core (all the C<ev_*> functions), with manual	3485	To include only the libev core (all the C<ev_*> functions), with manual
2520	configuration (no autoconf):	3486	configuration (no autoconf):
2521		3487
2522	#define EV_STANDALONE 1	3488	#define EV_STANDALONE 1
2523	#include "ev.c"	3489	#include "ev.c"
2524		3490
2525	This will automatically include F<ev.h>, too, and should be done in a	3491	This will automatically include F<ev.h>, too, and should be done in a
2526	single C source file only to provide the function implementations. To use	3492	single C source file only to provide the function implementations. To use
2527	it, do the same for F<ev.h> in all files wishing to use this API (best	3493	it, do the same for F<ev.h> in all files wishing to use this API (best
2528	done by writing a wrapper around F<ev.h> that you can include instead and	3494	done by writing a wrapper around F<ev.h> that you can include instead and
2529	where you can put other configuration options):	3495	where you can put other configuration options):
2530		3496
2531	#define EV_STANDALONE 1	3497	#define EV_STANDALONE 1
2532	#include "ev.h"	3498	#include "ev.h"
2533		3499
2534	Both header files and implementation files can be compiled with a C++	3500	Both header files and implementation files can be compiled with a C++
2535	compiler (at least, thats a stated goal, and breakage will be treated	3501	compiler (at least, that's a stated goal, and breakage will be treated
2536	as a bug).	3502	as a bug).
2537		3503
2538	You need the following files in your source tree, or in a directory	3504	You need the following files in your source tree, or in a directory
2539	in your include path (e.g. in libev/ when using -Ilibev):	3505	in your include path (e.g. in libev/ when using -Ilibev):
2540		3506
2541	ev.h	3507	ev.h
2542	ev.c	3508	ev.c
2543	ev_vars.h	3509	ev_vars.h
2544	ev_wrap.h	3510	ev_wrap.h
2545		3511
2546	ev_win32.c required on win32 platforms only	3512	ev_win32.c required on win32 platforms only
2547		3513
2548	ev_select.c only when select backend is enabled (which is enabled by default)	3514	ev_select.c only when select backend is enabled (which is enabled by default)
2549	ev_poll.c only when poll backend is enabled (disabled by default)	3515	ev_poll.c only when poll backend is enabled (disabled by default)
2550	ev_epoll.c only when the epoll backend is enabled (disabled by default)	3516	ev_epoll.c only when the epoll backend is enabled (disabled by default)
2551	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	3517	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
2552	ev_port.c only when the solaris port backend is enabled (disabled by default)	3518	ev_port.c only when the solaris port backend is enabled (disabled by default)
2553		3519
2554	F<ev.c> includes the backend files directly when enabled, so you only need	3520	F<ev.c> includes the backend files directly when enabled, so you only need
2555	to compile this single file.	3521	to compile this single file.
2556		3522
2557	=head3 LIBEVENT COMPATIBILITY API	3523	=head3 LIBEVENT COMPATIBILITY API
2558		3524
2559	To include the libevent compatibility API, also include:	3525	To include the libevent compatibility API, also include:
2560		3526
2561	#include "event.c"	3527	#include "event.c"
2562		3528
2563	in the file including F<ev.c>, and:	3529	in the file including F<ev.c>, and:
2564		3530
2565	#include "event.h"	3531	#include "event.h"
2566		3532
2567	in the files that want to use the libevent API. This also includes F<ev.h>.	3533	in the files that want to use the libevent API. This also includes F<ev.h>.
2568		3534
2569	You need the following additional files for this:	3535	You need the following additional files for this:
2570		3536
2571	event.h	3537	event.h
2572	event.c	3538	event.c
2573		3539
2574	=head3 AUTOCONF SUPPORT	3540	=head3 AUTOCONF SUPPORT
2575		3541
2576	Instead of using C<EV_STANDALONE=1> and providing your config in	3542	Instead of using C<EV_STANDALONE=1> and providing your configuration in
2577	whatever way you want, you can also C<m4_include([libev.m4])> in your	3543	whatever way you want, you can also C<m4_include([libev.m4])> in your
2578	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then	3544	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
2579	include F<config.h> and configure itself accordingly.	3545	include F<config.h> and configure itself accordingly.
2580		3546
2581	For this of course you need the m4 file:	3547	For this of course you need the m4 file:
2582		3548
2583	libev.m4	3549	libev.m4
2584		3550
2585	=head2 PREPROCESSOR SYMBOLS/MACROS	3551	=head2 PREPROCESSOR SYMBOLS/MACROS
2586		3552
2587	Libev can be configured via a variety of preprocessor symbols you have to define	3553	Libev can be configured via a variety of preprocessor symbols you have to
2588	before including any of its files. The default is not to build for multiplicity	3554	define before including any of its files. The default in the absence of
2589	and only include the select backend.	3555	autoconf is documented for every option.
2590		3556
2591	=over 4	3557	=over 4
2592		3558
2593	=item EV_STANDALONE	3559	=item EV_STANDALONE
2594		3560
…		…
2596	keeps libev from including F<config.h>, and it also defines dummy	3562	keeps libev from including F<config.h>, and it also defines dummy
2597	implementations for some libevent functions (such as logging, which is not	3563	implementations for some libevent functions (such as logging, which is not
2598	supported). It will also not define any of the structs usually found in	3564	supported). It will also not define any of the structs usually found in
2599	F<event.h> that are not directly supported by the libev core alone.	3565	F<event.h> that are not directly supported by the libev core alone.
2600		3566
		3567	In stanbdalone mode, libev will still try to automatically deduce the
		3568	configuration, but has to be more conservative.
		3569
2601	=item EV_USE_MONOTONIC	3570	=item EV_USE_MONOTONIC
2602		3571
2603	If defined to be C<1>, libev will try to detect the availability of the	3572	If defined to be C<1>, libev will try to detect the availability of the
2604	monotonic clock option at both compiletime and runtime. Otherwise no use	3573	monotonic clock option at both compile time and runtime. Otherwise no
2605	of the monotonic clock option will be attempted. If you enable this, you	3574	use of the monotonic clock option will be attempted. If you enable this,
2606	usually have to link against librt or something similar. Enabling it when	3575	you usually have to link against librt or something similar. Enabling it
2607	the functionality isn't available is safe, though, although you have	3576	when the functionality isn't available is safe, though, although you have
2608	to make sure you link against any libraries where the C<clock_gettime>	3577	to make sure you link against any libraries where the C<clock_gettime>
2609	function is hiding in (often F<-lrt>).	3578	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2610		3579
2611	=item EV_USE_REALTIME	3580	=item EV_USE_REALTIME
2612		3581
2613	If defined to be C<1>, libev will try to detect the availability of the	3582	If defined to be C<1>, libev will try to detect the availability of the
2614	realtime clock option at compiletime (and assume its availability at	3583	real-time clock option at compile time (and assume its availability
2615	runtime if successful). Otherwise no use of the realtime clock option will	3584	at runtime if successful). Otherwise no use of the real-time clock
2616	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3585	option will be attempted. This effectively replaces C<gettimeofday>
2617	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3586	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2618	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3587	correctness. See the note about libraries in the description of
		3588	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3589	C<EV_USE_CLOCK_SYSCALL>.
		3590
		3591	=item EV_USE_CLOCK_SYSCALL
		3592
		3593	If defined to be C<1>, libev will try to use a direct syscall instead
		3594	of calling the system-provided C<clock_gettime> function. This option
		3595	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3596	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3597	programs needlessly. Using a direct syscall is slightly slower (in
		3598	theory), because no optimised vdso implementation can be used, but avoids
		3599	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3600	higher, as it simplifies linking (no need for C<-lrt>).
2619		3601
2620	=item EV_USE_NANOSLEEP	3602	=item EV_USE_NANOSLEEP
2621		3603
2622	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3604	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2623	and will use it for delays. Otherwise it will use C<select ()>.	3605	and will use it for delays. Otherwise it will use C<select ()>.
2624		3606
		3607	=item EV_USE_EVENTFD
		3608
		3609	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		3610	available and will probe for kernel support at runtime. This will improve
		3611	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		3612	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		3613	2.7 or newer, otherwise disabled.
		3614
2625	=item EV_USE_SELECT	3615	=item EV_USE_SELECT
2626		3616
2627	If undefined or defined to be C<1>, libev will compile in support for the	3617	If undefined or defined to be C<1>, libev will compile in support for the
2628	C<select>(2) backend. No attempt at autodetection will be done: if no	3618	C<select>(2) backend. No attempt at auto-detection will be done: if no
2629	other method takes over, select will be it. Otherwise the select backend	3619	other method takes over, select will be it. Otherwise the select backend
2630	will not be compiled in.	3620	will not be compiled in.
2631		3621
2632	=item EV_SELECT_USE_FD_SET	3622	=item EV_SELECT_USE_FD_SET
2633		3623
2634	If defined to C<1>, then the select backend will use the system C<fd_set>	3624	If defined to C<1>, then the select backend will use the system C<fd_set>
2635	structure. This is useful if libev doesn't compile due to a missing	3625	structure. This is useful if libev doesn't compile due to a missing
2636	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	3626	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2637	exotic systems. This usually limits the range of file descriptors to some	3627	on exotic systems. This usually limits the range of file descriptors to
2638	low limit such as 1024 or might have other limitations (winsocket only	3628	some low limit such as 1024 or might have other limitations (winsocket
2639	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3629	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2640	influence the size of the C<fd_set> used.	3630	configures the maximum size of the C<fd_set>.
2641		3631
2642	=item EV_SELECT_IS_WINSOCKET	3632	=item EV_SELECT_IS_WINSOCKET
2643		3633
2644	When defined to C<1>, the select backend will assume that	3634	When defined to C<1>, the select backend will assume that
2645	select/socket/connect etc. don't understand file descriptors but	3635	select/socket/connect etc. don't understand file descriptors but
…		…
2665		3655
2666	=item EV_USE_EPOLL	3656	=item EV_USE_EPOLL
2667		3657
2668	If defined to be C<1>, libev will compile in support for the Linux	3658	If defined to be C<1>, libev will compile in support for the Linux
2669	C<epoll>(7) backend. Its availability will be detected at runtime,	3659	C<epoll>(7) backend. Its availability will be detected at runtime,
2670	otherwise another method will be used as fallback. This is the	3660	otherwise another method will be used as fallback. This is the preferred
2671	preferred backend for GNU/Linux systems.	3661	backend for GNU/Linux systems. If undefined, it will be enabled if the
		3662	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2672		3663
2673	=item EV_USE_KQUEUE	3664	=item EV_USE_KQUEUE
2674		3665
2675	If defined to be C<1>, libev will compile in support for the BSD style	3666	If defined to be C<1>, libev will compile in support for the BSD style
2676	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	3667	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
2689	otherwise another method will be used as fallback. This is the preferred	3680	otherwise another method will be used as fallback. This is the preferred
2690	backend for Solaris 10 systems.	3681	backend for Solaris 10 systems.
2691		3682
2692	=item EV_USE_DEVPOLL	3683	=item EV_USE_DEVPOLL
2693		3684
2694	reserved for future expansion, works like the USE symbols above.	3685	Reserved for future expansion, works like the USE symbols above.
2695		3686
2696	=item EV_USE_INOTIFY	3687	=item EV_USE_INOTIFY
2697		3688
2698	If defined to be C<1>, libev will compile in support for the Linux inotify	3689	If defined to be C<1>, libev will compile in support for the Linux inotify
2699	interface to speed up C<ev_stat> watchers. Its actual availability will	3690	interface to speed up C<ev_stat> watchers. Its actual availability will
2700	be detected at runtime.	3691	be detected at runtime. If undefined, it will be enabled if the headers
		3692	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2701		3693
2702	=item EV_ATOMIC_T	3694	=item EV_ATOMIC_T
2703		3695
2704	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	3696	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
2705	access is atomic with respect to other threads or signal contexts. No such	3697	access is atomic with respect to other threads or signal contexts. No such
2706	type is easily found in the C language, so you can provide your own type	3698	type is easily found in the C language, so you can provide your own type
2707	that you know is safe for your purposes. It is used both for signal handler "locking"	3699	that you know is safe for your purposes. It is used both for signal handler "locking"
2708	as well as for signal and thread safety in C<ev_async> watchers.	3700	as well as for signal and thread safety in C<ev_async> watchers.
2709		3701
2710	In the absense of this define, libev will use C<sig_atomic_t volatile>	3702	In the absence of this define, libev will use C<sig_atomic_t volatile>
2711	(from F<signal.h>), which is usually good enough on most platforms.	3703	(from F<signal.h>), which is usually good enough on most platforms.
2712		3704
2713	=item EV_H	3705	=item EV_H
2714		3706
2715	The name of the F<ev.h> header file used to include it. The default if	3707	The name of the F<ev.h> header file used to include it. The default if
…		…
2754	When doing priority-based operations, libev usually has to linearly search	3746	When doing priority-based operations, libev usually has to linearly search
2755	all the priorities, so having many of them (hundreds) uses a lot of space	3747	all the priorities, so having many of them (hundreds) uses a lot of space
2756	and time, so using the defaults of five priorities (-2 .. +2) is usually	3748	and time, so using the defaults of five priorities (-2 .. +2) is usually
2757	fine.	3749	fine.
2758		3750
2759	If your embedding app does not need any priorities, defining these both to	3751	If your embedding application does not need any priorities, defining these
2760	C<0> will save some memory and cpu.	3752	both to C<0> will save some memory and CPU.
2761		3753
2762	=item EV_PERIODIC_ENABLE	3754	=item EV_PERIODIC_ENABLE
2763		3755
2764	If undefined or defined to be C<1>, then periodic timers are supported. If	3756	If undefined or defined to be C<1>, then periodic timers are supported. If
2765	defined to be C<0>, then they are not. Disabling them saves a few kB of	3757	defined to be C<0>, then they are not. Disabling them saves a few kB of
…		…
2772	code.	3764	code.
2773		3765
2774	=item EV_EMBED_ENABLE	3766	=item EV_EMBED_ENABLE
2775		3767
2776	If undefined or defined to be C<1>, then embed watchers are supported. If	3768	If undefined or defined to be C<1>, then embed watchers are supported. If
2777	defined to be C<0>, then they are not.	3769	defined to be C<0>, then they are not. Embed watchers rely on most other
		3770	watcher types, which therefore must not be disabled.
2778		3771
2779	=item EV_STAT_ENABLE	3772	=item EV_STAT_ENABLE
2780		3773
2781	If undefined or defined to be C<1>, then stat watchers are supported. If	3774	If undefined or defined to be C<1>, then stat watchers are supported. If
2782	defined to be C<0>, then they are not.	3775	defined to be C<0>, then they are not.
…		…
2792	defined to be C<0>, then they are not.	3785	defined to be C<0>, then they are not.
2793		3786
2794	=item EV_MINIMAL	3787	=item EV_MINIMAL
2795		3788
2796	If you need to shave off some kilobytes of code at the expense of some	3789	If you need to shave off some kilobytes of code at the expense of some
2797	speed, define this symbol to C<1>. Currently only used for gcc to override	3790	speed (but with the full API), define this symbol to C<1>. Currently this
2798	some inlining decisions, saves roughly 30% codesize of amd64.	3791	is used to override some inlining decisions, saves roughly 30% code size
		3792	on amd64. It also selects a much smaller 2-heap for timer management over
		3793	the default 4-heap.
		3794
		3795	You can save even more by disabling watcher types you do not need
		3796	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>
		3797	(C<-DNDEBUG>) will usually reduce code size a lot.
		3798
		3799	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to
		3800	provide a bare-bones event library. See C<ev.h> for details on what parts
		3801	of the API are still available, and do not complain if this subset changes
		3802	over time.
2799		3803
2800	=item EV_PID_HASHSIZE	3804	=item EV_PID_HASHSIZE
2801		3805
2802	C<ev_child> watchers use a small hash table to distribute workload by	3806	C<ev_child> watchers use a small hash table to distribute workload by
2803	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3807	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
2810	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	3814	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
2811	usually more than enough. If you need to manage thousands of C<ev_stat>	3815	usually more than enough. If you need to manage thousands of C<ev_stat>
2812	watchers you might want to increase this value (I<must> be a power of	3816	watchers you might want to increase this value (I<must> be a power of
2813	two).	3817	two).
2814		3818
		3819	=item EV_USE_4HEAP
		3820
		3821	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3822	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		3823	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		3824	faster performance with many (thousands) of watchers.
		3825
		3826	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3827	(disabled).
		3828
		3829	=item EV_HEAP_CACHE_AT
		3830
		3831	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3832	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		3833	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3834	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3835	but avoids random read accesses on heap changes. This improves performance
		3836	noticeably with many (hundreds) of watchers.
		3837
		3838	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3839	(disabled).
		3840
		3841	=item EV_VERIFY
		3842
		3843	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3844	be done: If set to C<0>, no internal verification code will be compiled
		3845	in. If set to C<1>, then verification code will be compiled in, but not
		3846	called. If set to C<2>, then the internal verification code will be
		3847	called once per loop, which can slow down libev. If set to C<3>, then the
		3848	verification code will be called very frequently, which will slow down
		3849	libev considerably.
		3850
		3851	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3852	C<0>.
		3853
2815	=item EV_COMMON	3854	=item EV_COMMON
2816		3855
2817	By default, all watchers have a C<void *data> member. By redefining	3856	By default, all watchers have a C<void *data> member. By redefining
2818	this macro to a something else you can include more and other types of	3857	this macro to a something else you can include more and other types of
2819	members. You have to define it each time you include one of the files,	3858	members. You have to define it each time you include one of the files,
2820	though, and it must be identical each time.	3859	though, and it must be identical each time.
2821		3860
2822	For example, the perl EV module uses something like this:	3861	For example, the perl EV module uses something like this:
2823		3862
2824	#define EV_COMMON \	3863	#define EV_COMMON \
2825	SV *self; /* contains this struct */ \	3864	SV *self; /* contains this struct */ \
2826	SV *cb_sv, *fh /* note no trailing ";" */	3865	SV *cb_sv, *fh /* note no trailing ";" */
2827		3866
2828	=item EV_CB_DECLARE (type)	3867	=item EV_CB_DECLARE (type)
2829		3868
2830	=item EV_CB_INVOKE (watcher, revents)	3869	=item EV_CB_INVOKE (watcher, revents)
2831		3870
…		…
2836	definition and a statement, respectively. See the F<ev.h> header file for	3875	definition and a statement, respectively. See the F<ev.h> header file for
2837	their default definitions. One possible use for overriding these is to	3876	their default definitions. One possible use for overriding these is to
2838	avoid the C<struct ev_loop *> as first argument in all cases, or to use	3877	avoid the C<struct ev_loop *> as first argument in all cases, or to use
2839	method calls instead of plain function calls in C++.	3878	method calls instead of plain function calls in C++.
2840		3879
		3880	=back
		3881
2841	=head2 EXPORTED API SYMBOLS	3882	=head2 EXPORTED API SYMBOLS
2842		3883
2843	If you need to re-export the API (e.g. via a dll) and you need a list of	3884	If you need to re-export the API (e.g. via a DLL) and you need a list of
2844	exported symbols, you can use the provided F<Symbol.*> files which list	3885	exported symbols, you can use the provided F<Symbol.*> files which list
2845	all public symbols, one per line:	3886	all public symbols, one per line:
2846		3887
2847	Symbols.ev for libev proper	3888	Symbols.ev for libev proper
2848	Symbols.event for the libevent emulation	3889	Symbols.event for the libevent emulation
2849		3890
2850	This can also be used to rename all public symbols to avoid clashes with	3891	This can also be used to rename all public symbols to avoid clashes with
2851	multiple versions of libev linked together (which is obviously bad in	3892	multiple versions of libev linked together (which is obviously bad in
2852	itself, but sometimes it is inconvinient to avoid this).	3893	itself, but sometimes it is inconvenient to avoid this).
2853		3894
2854	A sed command like this will create wrapper C<#define>'s that you need to	3895	A sed command like this will create wrapper C<#define>'s that you need to
2855	include before including F<ev.h>:	3896	include before including F<ev.h>:
2856		3897
2857	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h	3898	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
…		…
2874	file.	3915	file.
2875		3916
2876	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	3917	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2877	that everybody includes and which overrides some configure choices:	3918	that everybody includes and which overrides some configure choices:
2878		3919
2879	#define EV_MINIMAL 1	3920	#define EV_MINIMAL 1
2880	#define EV_USE_POLL 0	3921	#define EV_USE_POLL 0
2881	#define EV_MULTIPLICITY 0	3922	#define EV_MULTIPLICITY 0
2882	#define EV_PERIODIC_ENABLE 0	3923	#define EV_PERIODIC_ENABLE 0
2883	#define EV_STAT_ENABLE 0	3924	#define EV_STAT_ENABLE 0
2884	#define EV_FORK_ENABLE 0	3925	#define EV_FORK_ENABLE 0
2885	#define EV_CONFIG_H <config.h>	3926	#define EV_CONFIG_H <config.h>
2886	#define EV_MINPRI 0	3927	#define EV_MINPRI 0
2887	#define EV_MAXPRI 0	3928	#define EV_MAXPRI 0
2888		3929
2889	#include "ev++.h"	3930	#include "ev++.h"
2890		3931
2891	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	3932	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2892		3933
2893	#include "ev_cpp.h"	3934	#include "ev_cpp.h"
2894	#include "ev.c"	3935	#include "ev.c"
2895		3936
		3937	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
2896		3938
2897	=head1 COMPLEXITIES	3939	=head2 THREADS AND COROUTINES
2898		3940
2899	In this section the complexities of (many of) the algorithms used inside	3941	=head3 THREADS
2900	libev will be explained. For complexity discussions about backends see the
2901	documentation for C<ev_default_init>.
2902		3942
2903	All of the following are about amortised time: If an array needs to be	3943	All libev functions are reentrant and thread-safe unless explicitly
2904	extended, libev needs to realloc and move the whole array, but this	3944	documented otherwise, but libev implements no locking itself. This means
2905	happens asymptotically never with higher number of elements, so O(1) might	3945	that you can use as many loops as you want in parallel, as long as there
2906	mean it might do a lengthy realloc operation in rare cases, but on average	3946	are no concurrent calls into any libev function with the same loop
2907	it is much faster and asymptotically approaches constant time.	3947	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		3948	of course): libev guarantees that different event loops share no data
		3949	structures that need any locking.
		3950
		3951	Or to put it differently: calls with different loop parameters can be done
		3952	concurrently from multiple threads, calls with the same loop parameter
		3953	must be done serially (but can be done from different threads, as long as
		3954	only one thread ever is inside a call at any point in time, e.g. by using
		3955	a mutex per loop).
		3956
		3957	Specifically to support threads (and signal handlers), libev implements
		3958	so-called C<ev_async> watchers, which allow some limited form of
		3959	concurrency on the same event loop, namely waking it up "from the
		3960	outside".
		3961
		3962	If you want to know which design (one loop, locking, or multiple loops
		3963	without or something else still) is best for your problem, then I cannot
		3964	help you, but here is some generic advice:
2908		3965
2909	=over 4	3966	=over 4
2910		3967
2911	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	3968	=item * most applications have a main thread: use the default libev loop
		3969	in that thread, or create a separate thread running only the default loop.
2912		3970
2913	This means that, when you have a watcher that triggers in one hour and	3971	This helps integrating other libraries or software modules that use libev
2914	there are 100 watchers that would trigger before that then inserting will	3972	themselves and don't care/know about threading.
2915	have to skip roughly seven (C<ld 100>) of these watchers.
2916		3973
2917	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	3974	=item * one loop per thread is usually a good model.
2918		3975
2919	That means that changing a timer costs less than removing/adding them	3976	Doing this is almost never wrong, sometimes a better-performance model
2920	as only the relative motion in the event queue has to be paid for.	3977	exists, but it is always a good start.
2921		3978
2922	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	3979	=item * other models exist, such as the leader/follower pattern, where one
		3980	loop is handed through multiple threads in a kind of round-robin fashion.
2923		3981
2924	These just add the watcher into an array or at the head of a list.	3982	Choosing a model is hard - look around, learn, know that usually you can do
		3983	better than you currently do :-)
2925		3984
2926	=item Stopping check/prepare/idle/fork/async watchers: O(1)	3985	=item * often you need to talk to some other thread which blocks in the
		3986	event loop.
2927		3987
2928	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	3988	C<ev_async> watchers can be used to wake them up from other threads safely
		3989	(or from signal contexts...).
2929		3990
2930	These watchers are stored in lists then need to be walked to find the	3991	An example use would be to communicate signals or other events that only
2931	correct watcher to remove. The lists are usually short (you don't usually	3992	work in the default loop by registering the signal watcher with the
2932	have many watchers waiting for the same fd or signal).	3993	default loop and triggering an C<ev_async> watcher from the default loop
2933		3994	watcher callback into the event loop interested in the signal.
2934	=item Finding the next timer in each loop iteration: O(1)
2935
2936	By virtue of using a binary heap, the next timer is always found at the
2937	beginning of the storage array.
2938
2939	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2940
2941	A change means an I/O watcher gets started or stopped, which requires
2942	libev to recalculate its status (and possibly tell the kernel, depending
2943	on backend and wether C<ev_io_set> was used).
2944
2945	=item Activating one watcher (putting it into the pending state): O(1)
2946
2947	=item Priority handling: O(number_of_priorities)
2948
2949	Priorities are implemented by allocating some space for each
2950	priority. When doing priority-based operations, libev usually has to
2951	linearly search all the priorities, but starting/stopping and activating
2952	watchers becomes O(1) w.r.t. priority handling.
2953
2954	=item Sending an ev_async: O(1)
2955
2956	=item Processing ev_async_send: O(number_of_async_watchers)
2957
2958	=item Processing signals: O(max_signal_number)
2959
2960	Sending involves a syscall I<iff> there were no other C<ev_async_send>
2961	calls in the current loop iteration. Checking for async and signal events
2962	involves iterating over all running async watchers or all signal numbers.
2963		3995
2964	=back	3996	=back
2965		3997
		3998	=head4 THREAD LOCKING EXAMPLE
2966		3999
2967	=head1 Win32 platform limitations and workarounds	4000	Here is a fictitious example of how to run an event loop in a different
		4001	thread than where callbacks are being invoked and watchers are
		4002	created/added/removed.
		4003
		4004	For a real-world example, see the C<EV::Loop::Async> perl module,
		4005	which uses exactly this technique (which is suited for many high-level
		4006	languages).
		4007
		4008	The example uses a pthread mutex to protect the loop data, a condition
		4009	variable to wait for callback invocations, an async watcher to notify the
		4010	event loop thread and an unspecified mechanism to wake up the main thread.
		4011
		4012	First, you need to associate some data with the event loop:
		4013
		4014	typedef struct {
		4015	mutex_t lock; /* global loop lock */
		4016	ev_async async_w;
		4017	thread_t tid;
		4018	cond_t invoke_cv;
		4019	} userdata;
		4020
		4021	void prepare_loop (EV_P)
		4022	{
		4023	// for simplicity, we use a static userdata struct.
		4024	static userdata u;
		4025
		4026	ev_async_init (&u->async_w, async_cb);
		4027	ev_async_start (EV_A_ &u->async_w);
		4028
		4029	pthread_mutex_init (&u->lock, 0);
		4030	pthread_cond_init (&u->invoke_cv, 0);
		4031
		4032	// now associate this with the loop
		4033	ev_set_userdata (EV_A_ u);
		4034	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4035	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4036
		4037	// then create the thread running ev_loop
		4038	pthread_create (&u->tid, 0, l_run, EV_A);
		4039	}
		4040
		4041	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4042	solely to wake up the event loop so it takes notice of any new watchers
		4043	that might have been added:
		4044
		4045	static void
		4046	async_cb (EV_P_ ev_async *w, int revents)
		4047	{
		4048	// just used for the side effects
		4049	}
		4050
		4051	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4052	protecting the loop data, respectively.
		4053
		4054	static void
		4055	l_release (EV_P)
		4056	{
		4057	userdata *u = ev_userdata (EV_A);
		4058	pthread_mutex_unlock (&u->lock);
		4059	}
		4060
		4061	static void
		4062	l_acquire (EV_P)
		4063	{
		4064	userdata *u = ev_userdata (EV_A);
		4065	pthread_mutex_lock (&u->lock);
		4066	}
		4067
		4068	The event loop thread first acquires the mutex, and then jumps straight
		4069	into C<ev_loop>:
		4070
		4071	void *
		4072	l_run (void *thr_arg)
		4073	{
		4074	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4075
		4076	l_acquire (EV_A);
		4077	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4078	ev_loop (EV_A_ 0);
		4079	l_release (EV_A);
		4080
		4081	return 0;
		4082	}
		4083
		4084	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4085	signal the main thread via some unspecified mechanism (signals? pipe
		4086	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4087	have been called (in a while loop because a) spurious wakeups are possible
		4088	and b) skipping inter-thread-communication when there are no pending
		4089	watchers is very beneficial):
		4090
		4091	static void
		4092	l_invoke (EV_P)
		4093	{
		4094	userdata *u = ev_userdata (EV_A);
		4095
		4096	while (ev_pending_count (EV_A))
		4097	{
		4098	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4099	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4100	}
		4101	}
		4102
		4103	Now, whenever the main thread gets told to invoke pending watchers, it
		4104	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4105	thread to continue:
		4106
		4107	static void
		4108	real_invoke_pending (EV_P)
		4109	{
		4110	userdata *u = ev_userdata (EV_A);
		4111
		4112	pthread_mutex_lock (&u->lock);
		4113	ev_invoke_pending (EV_A);
		4114	pthread_cond_signal (&u->invoke_cv);
		4115	pthread_mutex_unlock (&u->lock);
		4116	}
		4117
		4118	Whenever you want to start/stop a watcher or do other modifications to an
		4119	event loop, you will now have to lock:
		4120
		4121	ev_timer timeout_watcher;
		4122	userdata *u = ev_userdata (EV_A);
		4123
		4124	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4125
		4126	pthread_mutex_lock (&u->lock);
		4127	ev_timer_start (EV_A_ &timeout_watcher);
		4128	ev_async_send (EV_A_ &u->async_w);
		4129	pthread_mutex_unlock (&u->lock);
		4130
		4131	Note that sending the C<ev_async> watcher is required because otherwise
		4132	an event loop currently blocking in the kernel will have no knowledge
		4133	about the newly added timer. By waking up the loop it will pick up any new
		4134	watchers in the next event loop iteration.
		4135
		4136	=head3 COROUTINES
		4137
		4138	Libev is very accommodating to coroutines ("cooperative threads"):
		4139	libev fully supports nesting calls to its functions from different
		4140	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		4141	different coroutines, and switch freely between both coroutines running
		4142	the loop, as long as you don't confuse yourself). The only exception is
		4143	that you must not do this from C<ev_periodic> reschedule callbacks.
		4144
		4145	Care has been taken to ensure that libev does not keep local state inside
		4146	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		4147	they do not call any callbacks.
		4148
		4149	=head2 COMPILER WARNINGS
		4150
		4151	Depending on your compiler and compiler settings, you might get no or a
		4152	lot of warnings when compiling libev code. Some people are apparently
		4153	scared by this.
		4154
		4155	However, these are unavoidable for many reasons. For one, each compiler
		4156	has different warnings, and each user has different tastes regarding
		4157	warning options. "Warn-free" code therefore cannot be a goal except when
		4158	targeting a specific compiler and compiler-version.
		4159
		4160	Another reason is that some compiler warnings require elaborate
		4161	workarounds, or other changes to the code that make it less clear and less
		4162	maintainable.
		4163
		4164	And of course, some compiler warnings are just plain stupid, or simply
		4165	wrong (because they don't actually warn about the condition their message
		4166	seems to warn about). For example, certain older gcc versions had some
		4167	warnings that resulted an extreme number of false positives. These have
		4168	been fixed, but some people still insist on making code warn-free with
		4169	such buggy versions.
		4170
		4171	While libev is written to generate as few warnings as possible,
		4172	"warn-free" code is not a goal, and it is recommended not to build libev
		4173	with any compiler warnings enabled unless you are prepared to cope with
		4174	them (e.g. by ignoring them). Remember that warnings are just that:
		4175	warnings, not errors, or proof of bugs.
		4176
		4177
		4178	=head2 VALGRIND
		4179
		4180	Valgrind has a special section here because it is a popular tool that is
		4181	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		4182
		4183	If you think you found a bug (memory leak, uninitialised data access etc.)
		4184	in libev, then check twice: If valgrind reports something like:
		4185
		4186	==2274== definitely lost: 0 bytes in 0 blocks.
		4187	==2274== possibly lost: 0 bytes in 0 blocks.
		4188	==2274== still reachable: 256 bytes in 1 blocks.
		4189
		4190	Then there is no memory leak, just as memory accounted to global variables
		4191	is not a memleak - the memory is still being referenced, and didn't leak.
		4192
		4193	Similarly, under some circumstances, valgrind might report kernel bugs
		4194	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4195	although an acceptable workaround has been found here), or it might be
		4196	confused.
		4197
		4198	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		4199	make it into some kind of religion.
		4200
		4201	If you are unsure about something, feel free to contact the mailing list
		4202	with the full valgrind report and an explanation on why you think this
		4203	is a bug in libev (best check the archives, too :). However, don't be
		4204	annoyed when you get a brisk "this is no bug" answer and take the chance
		4205	of learning how to interpret valgrind properly.
		4206
		4207	If you need, for some reason, empty reports from valgrind for your project
		4208	I suggest using suppression lists.
		4209
		4210
		4211	=head1 PORTABILITY NOTES
		4212
		4213	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
2968		4214
2969	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4215	Win32 doesn't support any of the standards (e.g. POSIX) that libev
2970	requires, and its I/O model is fundamentally incompatible with the POSIX	4216	requires, and its I/O model is fundamentally incompatible with the POSIX
2971	model. Libev still offers limited functionality on this platform in	4217	model. Libev still offers limited functionality on this platform in
2972	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4218	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
2973	descriptors. This only applies when using Win32 natively, not when using	4219	descriptors. This only applies when using Win32 natively, not when using
2974	e.g. cygwin.	4220	e.g. cygwin.
2975		4221
		4222	Lifting these limitations would basically require the full
		4223	re-implementation of the I/O system. If you are into these kinds of
		4224	things, then note that glib does exactly that for you in a very portable
		4225	way (note also that glib is the slowest event library known to man).
		4226
2976	There is no supported compilation method available on windows except	4227	There is no supported compilation method available on windows except
2977	embedding it into other applications.	4228	embedding it into other applications.
2978		4229
		4230	Sensible signal handling is officially unsupported by Microsoft - libev
		4231	tries its best, but under most conditions, signals will simply not work.
		4232
		4233	Not a libev limitation but worth mentioning: windows apparently doesn't
		4234	accept large writes: instead of resulting in a partial write, windows will
		4235	either accept everything or return C<ENOBUFS> if the buffer is too large,
		4236	so make sure you only write small amounts into your sockets (less than a
		4237	megabyte seems safe, but this apparently depends on the amount of memory
		4238	available).
		4239
2979	Due to the many, low, and arbitrary limits on the win32 platform and the	4240	Due to the many, low, and arbitrary limits on the win32 platform and
2980	abysmal performance of winsockets, using a large number of sockets is not	4241	the abysmal performance of winsockets, using a large number of sockets
2981	recommended (and not reasonable). If your program needs to use more than	4242	is not recommended (and not reasonable). If your program needs to use
2982	a hundred or so sockets, then likely it needs to use a totally different	4243	more than a hundred or so sockets, then likely it needs to use a totally
2983	implementation for windows, as libev offers the POSIX model, which cannot	4244	different implementation for windows, as libev offers the POSIX readiness
2984	be implemented efficiently on windows (microsoft monopoly games).	4245	notification model, which cannot be implemented efficiently on windows
		4246	(due to Microsoft monopoly games).
		4247
		4248	A typical way to use libev under windows is to embed it (see the embedding
		4249	section for details) and use the following F<evwrap.h> header file instead
		4250	of F<ev.h>:
		4251
		4252	#define EV_STANDALONE /* keeps ev from requiring config.h */
		4253	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		4254
		4255	#include "ev.h"
		4256
		4257	And compile the following F<evwrap.c> file into your project (make sure
		4258	you do I<not> compile the F<ev.c> or any other embedded source files!):
		4259
		4260	#include "evwrap.h"
		4261	#include "ev.c"
2985		4262
2986	=over 4	4263	=over 4
2987		4264
2988	=item The winsocket select function	4265	=item The winsocket select function
2989		4266
2990	The winsocket C<select> function doesn't follow POSIX in that it requires	4267	The winsocket C<select> function doesn't follow POSIX in that it
2991	socket I<handles> and not socket I<file descriptors>. This makes select	4268	requires socket I<handles> and not socket I<file descriptors> (it is
2992	very inefficient, and also requires a mapping from file descriptors	4269	also extremely buggy). This makes select very inefficient, and also
2993	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,	4270	requires a mapping from file descriptors to socket handles (the Microsoft
2994	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor	4271	C runtime provides the function C<_open_osfhandle> for this). See the
2995	symbols for more info.	4272	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		4273	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
2996		4274
2997	The configuration for a "naked" win32 using the microsoft runtime	4275	The configuration for a "naked" win32 using the Microsoft runtime
2998	libraries and raw winsocket select is:	4276	libraries and raw winsocket select is:
2999		4277
3000	#define EV_USE_SELECT 1	4278	#define EV_USE_SELECT 1
3001	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	4279	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
3002		4280
3003	Note that winsockets handling of fd sets is O(n), so you can easily get a	4281	Note that winsockets handling of fd sets is O(n), so you can easily get a
3004	complexity in the O(n²) range when using win32.	4282	complexity in the O(n²) range when using win32.
3005		4283
3006	=item Limited number of file descriptors	4284	=item Limited number of file descriptors
3007		4285
3008	Windows has numerous arbitrary (and low) limits on things. Early versions	4286	Windows has numerous arbitrary (and low) limits on things.
3009	of winsocket's select only supported waiting for a max. of C<64> handles	4287
		4288	Early versions of winsocket's select only supported waiting for a maximum
3010	(probably owning to the fact that all windows kernels can only wait for	4289	of C<64> handles (probably owning to the fact that all windows kernels
3011	C<64> things at the same time internally; microsoft recommends spawning a	4290	can only wait for C<64> things at the same time internally; Microsoft
3012	chain of threads and wait for 63 handles and the previous thread in each).	4291	recommends spawning a chain of threads and wait for 63 handles and the
		4292	previous thread in each. Sounds great!).
3013		4293
3014	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4294	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3015	to some high number (e.g. C<2048>) before compiling the winsocket select	4295	to some high number (e.g. C<2048>) before compiling the winsocket select
3016	call (which might be in libev or elsewhere, for example, perl does its own	4296	call (which might be in libev or elsewhere, for example, perl and many
3017	select emulation on windows).	4297	other interpreters do their own select emulation on windows).
3018		4298
3019	Another limit is the number of file descriptors in the microsoft runtime	4299	Another limit is the number of file descriptors in the Microsoft runtime
3020	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4300	libraries, which by default is C<64> (there must be a hidden I<64>
3021	or something like this inside microsoft). You can increase this by calling	4301	fetish or something like this inside Microsoft). You can increase this
3022	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4302	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3023	arbitrary limit), but is broken in many versions of the microsoft runtime	4303	(another arbitrary limit), but is broken in many versions of the Microsoft
3024	libraries.
3025
3026	This might get you to about C<512> or C<2048> sockets (depending on	4304	runtime libraries. This might get you to about C<512> or C<2048> sockets
3027	windows version and/or the phase of the moon). To get more, you need to	4305	(depending on windows version and/or the phase of the moon). To get more,
3028	wrap all I/O functions and provide your own fd management, but the cost of	4306	you need to wrap all I/O functions and provide your own fd management, but
3029	calling select (O(n²)) will likely make this unworkable.	4307	the cost of calling select (O(n²)) will likely make this unworkable.
3030		4308
3031	=back	4309	=back
3032		4310
		4311	=head2 PORTABILITY REQUIREMENTS
		4312
		4313	In addition to a working ISO-C implementation and of course the
		4314	backend-specific APIs, libev relies on a few additional extensions:
		4315
		4316	=over 4
		4317
		4318	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		4319	calling conventions regardless of C<ev_watcher_type *>.
		4320
		4321	Libev assumes not only that all watcher pointers have the same internal
		4322	structure (guaranteed by POSIX but not by ISO C for example), but it also
		4323	assumes that the same (machine) code can be used to call any watcher
		4324	callback: The watcher callbacks have different type signatures, but libev
		4325	calls them using an C<ev_watcher *> internally.
		4326
		4327	=item C<sig_atomic_t volatile> must be thread-atomic as well
		4328
		4329	The type C<sig_atomic_t volatile> (or whatever is defined as
		4330	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		4331	threads. This is not part of the specification for C<sig_atomic_t>, but is
		4332	believed to be sufficiently portable.
		4333
		4334	=item C<sigprocmask> must work in a threaded environment
		4335
		4336	Libev uses C<sigprocmask> to temporarily block signals. This is not
		4337	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		4338	pthread implementations will either allow C<sigprocmask> in the "main
		4339	thread" or will block signals process-wide, both behaviours would
		4340	be compatible with libev. Interaction between C<sigprocmask> and
		4341	C<pthread_sigmask> could complicate things, however.
		4342
		4343	The most portable way to handle signals is to block signals in all threads
		4344	except the initial one, and run the default loop in the initial thread as
		4345	well.
		4346
		4347	=item C<long> must be large enough for common memory allocation sizes
		4348
		4349	To improve portability and simplify its API, libev uses C<long> internally
		4350	instead of C<size_t> when allocating its data structures. On non-POSIX
		4351	systems (Microsoft...) this might be unexpectedly low, but is still at
		4352	least 31 bits everywhere, which is enough for hundreds of millions of
		4353	watchers.
		4354
		4355	=item C<double> must hold a time value in seconds with enough accuracy
		4356
		4357	The type C<double> is used to represent timestamps. It is required to
		4358	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		4359	enough for at least into the year 4000. This requirement is fulfilled by
		4360	implementations implementing IEEE 754, which is basically all existing
		4361	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4362	2200.
		4363
		4364	=back
		4365
		4366	If you know of other additional requirements drop me a note.
		4367
		4368
		4369	=head1 ALGORITHMIC COMPLEXITIES
		4370
		4371	In this section the complexities of (many of) the algorithms used inside
		4372	libev will be documented. For complexity discussions about backends see
		4373	the documentation for C<ev_default_init>.
		4374
		4375	All of the following are about amortised time: If an array needs to be
		4376	extended, libev needs to realloc and move the whole array, but this
		4377	happens asymptotically rarer with higher number of elements, so O(1) might
		4378	mean that libev does a lengthy realloc operation in rare cases, but on
		4379	average it is much faster and asymptotically approaches constant time.
		4380
		4381	=over 4
		4382
		4383	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		4384
		4385	This means that, when you have a watcher that triggers in one hour and
		4386	there are 100 watchers that would trigger before that, then inserting will
		4387	have to skip roughly seven (C<ld 100>) of these watchers.
		4388
		4389	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		4390
		4391	That means that changing a timer costs less than removing/adding them,
		4392	as only the relative motion in the event queue has to be paid for.
		4393
		4394	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		4395
		4396	These just add the watcher into an array or at the head of a list.
		4397
		4398	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		4399
		4400	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		4401
		4402	These watchers are stored in lists, so they need to be walked to find the
		4403	correct watcher to remove. The lists are usually short (you don't usually
		4404	have many watchers waiting for the same fd or signal: one is typical, two
		4405	is rare).
		4406
		4407	=item Finding the next timer in each loop iteration: O(1)
		4408
		4409	By virtue of using a binary or 4-heap, the next timer is always found at a
		4410	fixed position in the storage array.
		4411
		4412	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4413
		4414	A change means an I/O watcher gets started or stopped, which requires
		4415	libev to recalculate its status (and possibly tell the kernel, depending
		4416	on backend and whether C<ev_io_set> was used).
		4417
		4418	=item Activating one watcher (putting it into the pending state): O(1)
		4419
		4420	=item Priority handling: O(number_of_priorities)
		4421
		4422	Priorities are implemented by allocating some space for each
		4423	priority. When doing priority-based operations, libev usually has to
		4424	linearly search all the priorities, but starting/stopping and activating
		4425	watchers becomes O(1) with respect to priority handling.
		4426
		4427	=item Sending an ev_async: O(1)
		4428
		4429	=item Processing ev_async_send: O(number_of_async_watchers)
		4430
		4431	=item Processing signals: O(max_signal_number)
		4432
		4433	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4434	calls in the current loop iteration. Checking for async and signal events
		4435	involves iterating over all running async watchers or all signal numbers.
		4436
		4437	=back
		4438
		4439
		4440	=head1 GLOSSARY
		4441
		4442	=over 4
		4443
		4444	=item active
		4445
		4446	A watcher is active as long as it has been started (has been attached to
		4447	an event loop) but not yet stopped (disassociated from the event loop).
		4448
		4449	=item application
		4450
		4451	In this document, an application is whatever is using libev.
		4452
		4453	=item callback
		4454
		4455	The address of a function that is called when some event has been
		4456	detected. Callbacks are being passed the event loop, the watcher that
		4457	received the event, and the actual event bitset.
		4458
		4459	=item callback invocation
		4460
		4461	The act of calling the callback associated with a watcher.
		4462
		4463	=item event
		4464
		4465	A change of state of some external event, such as data now being available
		4466	for reading on a file descriptor, time having passed or simply not having
		4467	any other events happening anymore.
		4468
		4469	In libev, events are represented as single bits (such as C<EV_READ> or
		4470	C<EV_TIMEOUT>).
		4471
		4472	=item event library
		4473
		4474	A software package implementing an event model and loop.
		4475
		4476	=item event loop
		4477
		4478	An entity that handles and processes external events and converts them
		4479	into callback invocations.
		4480
		4481	=item event model
		4482
		4483	The model used to describe how an event loop handles and processes
		4484	watchers and events.
		4485
		4486	=item pending
		4487
		4488	A watcher is pending as soon as the corresponding event has been detected,
		4489	and stops being pending as soon as the watcher will be invoked or its
		4490	pending status is explicitly cleared by the application.
		4491
		4492	A watcher can be pending, but not active. Stopping a watcher also clears
		4493	its pending status.
		4494
		4495	=item real time
		4496
		4497	The physical time that is observed. It is apparently strictly monotonic :)
		4498
		4499	=item wall-clock time
		4500
		4501	The time and date as shown on clocks. Unlike real time, it can actually
		4502	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4503	clock.
		4504
		4505	=item watcher
		4506
		4507	A data structure that describes interest in certain events. Watchers need
		4508	to be started (attached to an event loop) before they can receive events.
		4509
		4510	=item watcher invocation
		4511
		4512	The act of calling the callback associated with a watcher.
		4513
		4514	=back
3033		4515
3034	=head1 AUTHOR	4516	=head1 AUTHOR
3035		4517
3036	Marc Lehmann <libev@schmorp.de>.	4518	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3037		4519

Diff Legend

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