ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.265 by root, Wed Aug 26 17:11:42 2009 UTC vs.
Revision 1.354 by root, Tue Jan 11 01:40:25 2011 UTC

…		…
26	puts ("stdin ready");	26	puts ("stdin ready");
27	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
28	// with its corresponding stop function.	28	// with its corresponding stop function.
29	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
30		30
31	// this causes all nested ev_loop's to stop iterating	31	// this causes all nested ev_run's to stop iterating
32	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_break (EV_A_ EVBREAK_ALL);
33	}	33	}
34		34
35	// another callback, this time for a time-out	35	// another callback, this time for a time-out
36	static void	36	static void
37	timeout_cb (EV_P_ ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
38	{	38	{
39	puts ("timeout");	39	puts ("timeout");
40	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_run to stop iterating
41	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_break (EV_A_ EVBREAK_ONE);
42	}	42	}
43		43
44	int	44	int
45	main (void)	45	main (void)
46	{	46	{
47	// use the default event loop unless you have special needs	47	// use the default event loop unless you have special needs
48	struct ev_loop *loop = ev_default_loop (0);	48	struct ev_loop *loop = EV_DEFAULT;
49		49
50	// 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	51	// this one will watch for stdin to become readable
52	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);
53	ev_io_start (loop, &stdin_watcher);	53	ev_io_start (loop, &stdin_watcher);
…		…
56	// simple non-repeating 5.5 second timeout	56	// simple non-repeating 5.5 second timeout
57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);	57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
58	ev_timer_start (loop, &timeout_watcher);	58	ev_timer_start (loop, &timeout_watcher);
59		59
60	// now wait for events to arrive	60	// now wait for events to arrive
61	ev_loop (loop, 0);	61	ev_run (loop, 0);
62		62
63	// unloop was called, so exit	63	// unloop was called, so exit
64	return 0;	64	return 0;
65	}	65	}
66		66
…		…
75	While this document tries to be as complete as possible in documenting	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	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	77	on event-based programming, nor will it introduce event-based programming
78	with libev.	78	with libev.
79		79
80	Familarity with event based programming techniques in general is assumed	80	Familiarity with event based programming techniques in general is assumed
81	throughout this document.	81	throughout this document.
		82
		83	=head1 WHAT TO READ WHEN IN A HURRY
		84
		85	This manual tries to be very detailed, but unfortunately, this also makes
		86	it very long. If you just want to know the basics of libev, I suggest
		87	reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and
		88	look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and
		89	C<ev_timer> sections in L<WATCHER TYPES>.
82		90
83	=head1 ABOUT LIBEV	91	=head1 ABOUT LIBEV
84		92
85	Libev is an event loop: you register interest in certain events (such as a	93	Libev is an event loop: you register interest in certain events (such as a
86	file descriptor being readable or a timeout occurring), and it will manage	94	file descriptor being readable or a timeout occurring), and it will manage
…		…
118	Libev is very configurable. In this manual the default (and most common)	126	Libev is very configurable. In this manual the default (and most common)
119	configuration will be described, which supports multiple event loops. For	127	configuration will be described, which supports multiple event loops. For
120	more info about various configuration options please have a look at	128	more info about various configuration options please have a look at
121	B<EMBED> section in this manual. If libev was configured without support	129	B<EMBED> section in this manual. If libev was configured without support
122	for multiple event loops, then all functions taking an initial argument of	130	for multiple event loops, then all functions taking an initial argument of
123	name C<loop> (which is always of type C<ev_loop *>) will not have	131	name C<loop> (which is always of type C<struct ev_loop *>) will not have
124	this argument.	132	this argument.
125		133
126	=head2 TIME REPRESENTATION	134	=head2 TIME REPRESENTATION
127		135
128	Libev represents time as a single floating point number, representing	136	Libev represents time as a single floating point number, representing
129	the (fractional) number of seconds since the (POSIX) epoch (somewhere	137	the (fractional) number of seconds since the (POSIX) epoch (in practice
130	near the beginning of 1970, details are complicated, don't ask). This	138	somewhere near the beginning of 1970, details are complicated, don't
131	type is called C<ev_tstamp>, which is what you should use too. It usually	139	ask). This type is called C<ev_tstamp>, which is what you should use
132	aliases to the C<double> type in C. When you need to do any calculations	140	too. It usually aliases to the C<double> type in C. When you need to do
133	on it, you should treat it as some floating point value. Unlike the name	141	any calculations on it, you should treat it as some floating point value.
		142
134	component C<stamp> might indicate, it is also used for time differences	143	Unlike the name component C<stamp> might indicate, it is also used for
135	throughout libev.	144	time differences (e.g. delays) throughout libev.
136		145
137	=head1 ERROR HANDLING	146	=head1 ERROR HANDLING
138		147
139	Libev knows three classes of errors: operating system errors, usage errors	148	Libev knows three classes of errors: operating system errors, usage errors
140	and internal errors (bugs).	149	and internal errors (bugs).
…		…
164		173
165	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
166		175
167	Returns the current time as libev would use it. Please note that the	176	Returns the current time as libev would use it. Please note that the
168	C<ev_now> function is usually faster and also often returns the timestamp	177	C<ev_now> function is usually faster and also often returns the timestamp
169	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_update_now> and C<ev_now>.
170		180
171	=item ev_sleep (ev_tstamp interval)	181	=item ev_sleep (ev_tstamp interval)
172		182
173	Sleep for the given interval: The current thread will be blocked until	183	Sleep for the given interval: The current thread will be blocked until
174	either it is interrupted or the given time interval has passed. Basically	184	either it is interrupted or the given time interval has passed. Basically
…		…
191	as this indicates an incompatible change. Minor versions are usually	201	as this indicates an incompatible change. Minor versions are usually
192	compatible to older versions, so a larger minor version alone is usually	202	compatible to older versions, so a larger minor version alone is usually
193	not a problem.	203	not a problem.
194		204
195	Example: Make sure we haven't accidentally been linked against the wrong	205	Example: Make sure we haven't accidentally been linked against the wrong
196	version.	206	version (note, however, that this will not detect other ABI mismatches,
		207	such as LFS or reentrancy).
197		208
198	assert (("libev version mismatch",	209	assert (("libev version mismatch",
199	ev_version_major () == EV_VERSION_MAJOR	210	ev_version_major () == EV_VERSION_MAJOR
200	&& ev_version_minor () >= EV_VERSION_MINOR));	211	&& ev_version_minor () >= EV_VERSION_MINOR));
201		212
…		…
212	assert (("sorry, no epoll, no sex",	223	assert (("sorry, no epoll, no sex",
213	ev_supported_backends () & EVBACKEND_EPOLL));	224	ev_supported_backends () & EVBACKEND_EPOLL));
214		225
215	=item unsigned int ev_recommended_backends ()	226	=item unsigned int ev_recommended_backends ()
216		227
217	Return the set of all backends compiled into this binary of libev and also	228	Return the set of all backends compiled into this binary of libev and
218	recommended for this platform. This set is often smaller than the one	229	also recommended for this platform, meaning it will work for most file
		230	descriptor types. This set is often smaller than the one returned by
219	returned by C<ev_supported_backends>, as for example kqueue is broken on	231	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
220	most BSDs and will not be auto-detected unless you explicitly request it	232	and will not be auto-detected unless you explicitly request it (assuming
221	(assuming you know what you are doing). This is the set of backends that	233	you know what you are doing). This is the set of backends that libev will
222	libev will probe for if you specify no backends explicitly.	234	probe for if you specify no backends explicitly.
223		235
224	=item unsigned int ev_embeddable_backends ()	236	=item unsigned int ev_embeddable_backends ()
225		237
226	Returns the set of backends that are embeddable in other event loops. This	238	Returns the set of backends that are embeddable in other event loops. This
227	is the theoretical, all-platform, value. To find which backends	239	value is platform-specific but can include backends not available on the
228	might be supported on the current system, you would need to look at	240	current system. To find which embeddable backends might be supported on
229	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	241	the current system, you would need to look at C<ev_embeddable_backends ()
230	recommended ones.	242	& ev_supported_backends ()>, likewise for recommended ones.
231		243
232	See the description of C<ev_embed> watchers for more info.	244	See the description of C<ev_embed> watchers for more info.
233		245
234	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]	246	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
235		247
236	Sets the allocation function to use (the prototype is similar - the	248	Sets the allocation function to use (the prototype is similar - the
237	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	249	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
238	used to allocate and free memory (no surprises here). If it returns zero	250	used to allocate and free memory (no surprises here). If it returns zero
239	when memory needs to be allocated (C<size != 0>), the library might abort	251	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
265	}	277	}
266		278
267	...	279	...
268	ev_set_allocator (persistent_realloc);	280	ev_set_allocator (persistent_realloc);
269		281
270	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]	282	=item ev_set_syserr_cb (void (*cb)(const char *msg))
271		283
272	Set the callback function to call on a retryable system call error (such	284	Set the callback function to call on a retryable system call error (such
273	as failed select, poll, epoll_wait). The message is a printable string	285	as failed select, poll, epoll_wait). The message is a printable string
274	indicating the system call or subsystem causing the problem. If this	286	indicating the system call or subsystem causing the problem. If this
275	callback is set, then libev will expect it to remedy the situation, no	287	callback is set, then libev will expect it to remedy the situation, no
…		…
287	}	299	}
288		300
289	...	301	...
290	ev_set_syserr_cb (fatal_error);	302	ev_set_syserr_cb (fatal_error);
291		303
		304	=item ev_feed_signal (int signum)
		305
		306	This function can be used to "simulate" a signal receive. It is completely
		307	safe to call this function at any time, from any context, including signal
		308	handlers or random threads.
		309
		310	Its main use is to customise signal handling in your process, especially
		311	in the presence of threads. For example, you could block signals
		312	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		313	creating any loops), and in one thread, use C<sigwait> or any other
		314	mechanism to wait for signals, then "deliver" them to libev by calling
		315	C<ev_feed_signal>.
		316
292	=back	317	=back
293		318
294	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	319	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
295		320
296	An event loop is described by a C<struct ev_loop *> (the C<struct>	321	An event loop is described by a C<struct ev_loop *> (the C<struct> is
297	is I<not> optional in this case, as there is also an C<ev_loop>	322	I<not> optional in this case unless libev 3 compatibility is disabled, as
298	I<function>).	323	libev 3 had an C<ev_loop> function colliding with the struct name).
299		324
300	The library knows two types of such loops, the I<default> loop, which	325	The library knows two types of such loops, the I<default> loop, which
301	supports signals and child events, and dynamically created loops which do	326	supports child process events, and dynamically created event loops which
302	not.	327	do not.
303		328
304	=over 4	329	=over 4
305		330
306	=item struct ev_loop *ev_default_loop (unsigned int flags)	331	=item struct ev_loop *ev_default_loop (unsigned int flags)
307		332
308	This will initialise the default event loop if it hasn't been initialised	333	This returns the "default" event loop object, which is what you should
309	yet and return it. If the default loop could not be initialised, returns	334	normally use when you just need "the event loop". Event loop objects and
310	false. If it already was initialised it simply returns it (and ignores the	335	the C<flags> parameter are described in more detail in the entry for
311	flags. If that is troubling you, check C<ev_backend ()> afterwards).	336	C<ev_loop_new>.
		337
		338	If the default loop is already initialised then this function simply
		339	returns it (and ignores the flags. If that is troubling you, check
		340	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		341	flags, which should almost always be C<0>, unless the caller is also the
		342	one calling C<ev_run> or otherwise qualifies as "the main program".
312		343
313	If you don't know what event loop to use, use the one returned from this	344	If you don't know what event loop to use, use the one returned from this
314	function.	345	function (or via the C<EV_DEFAULT> macro).
315		346
316	Note that this function is I<not> thread-safe, so if you want to use it	347	Note that this function is I<not> thread-safe, so if you want to use it
317	from multiple threads, you have to lock (note also that this is unlikely,	348	from multiple threads, you have to employ some kind of mutex (note also
318	as loops cannot be shared easily between threads anyway).	349	that this case is unlikely, as loops cannot be shared easily between
		350	threads anyway).
319		351
320	The default loop is the only loop that can handle C<ev_signal> and	352	The default loop is the only loop that can handle C<ev_child> watchers,
321	C<ev_child> watchers, and to do this, it always registers a handler	353	and to do this, it always registers a handler for C<SIGCHLD>. If this is
322	for C<SIGCHLD>. If this is a problem for your application you can either	354	a problem for your application you can either create a dynamic loop with
323	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	355	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
324	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	356	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
325	C<ev_default_init>.	357
		358	Example: This is the most typical usage.
		359
		360	if (!ev_default_loop (0))
		361	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		362
		363	Example: Restrict libev to the select and poll backends, and do not allow
		364	environment settings to be taken into account:
		365
		366	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		367
		368	=item struct ev_loop *ev_loop_new (unsigned int flags)
		369
		370	This will create and initialise a new event loop object. If the loop
		371	could not be initialised, returns false.
		372
		373	This function is thread-safe, and one common way to use libev with
		374	threads is indeed to create one loop per thread, and using the default
		375	loop in the "main" or "initial" thread.
326		376
327	The flags argument can be used to specify special behaviour or specific	377	The flags argument can be used to specify special behaviour or specific
328	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	378	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
329		379
330	The following flags are supported:	380	The following flags are supported:
…		…
345	useful to try out specific backends to test their performance, or to work	395	useful to try out specific backends to test their performance, or to work
346	around bugs.	396	around bugs.
347		397
348	=item C<EVFLAG_FORKCHECK>	398	=item C<EVFLAG_FORKCHECK>
349		399
350	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	400	Instead of calling C<ev_loop_fork> manually after a fork, you can also
351	a fork, you can also make libev check for a fork in each iteration by	401	make libev check for a fork in each iteration by enabling this flag.
352	enabling this flag.
353		402
354	This works by calling C<getpid ()> on every iteration of the loop,	403	This works by calling C<getpid ()> on every iteration of the loop,
355	and thus this might slow down your event loop if you do a lot of loop	404	and thus this might slow down your event loop if you do a lot of loop
356	iterations and little real work, but is usually not noticeable (on my	405	iterations and little real work, but is usually not noticeable (on my
357	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	406	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
…		…
366	environment variable.	415	environment variable.
367		416
368	=item C<EVFLAG_NOINOTIFY>	417	=item C<EVFLAG_NOINOTIFY>
369		418
370	When this flag is specified, then libev will not attempt to use the	419	When this flag is specified, then libev will not attempt to use the
371	I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and	420	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
372	testing, this flag can be useful to conserve inotify file descriptors, as	421	testing, this flag can be useful to conserve inotify file descriptors, as
373	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.	422	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
374		423
375	=item C<EVFLAG_NOSIGNALFD>	424	=item C<EVFLAG_SIGNALFD>
376		425
377	When this flag is specified, then libev will not attempt to use the	426	When this flag is specified, then libev will attempt to use the
378	I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This is	427	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
379	probably only useful to work around any bugs in libev. Consequently, this	428	delivers signals synchronously, which makes it both faster and might make
380	flag might go away once the signalfd functionality is considered stable,	429	it possible to get the queued signal data. It can also simplify signal
381	so it's useful mostly in environment variables and not in program code.	430	handling with threads, as long as you properly block signals in your
		431	threads that are not interested in handling them.
		432
		433	Signalfd will not be used by default as this changes your signal mask, and
		434	there are a lot of shoddy libraries and programs (glib's threadpool for
		435	example) that can't properly initialise their signal masks.
		436
		437	=item C<EVFLAG_NOSIGMASK>
		438
		439	When this flag is specified, then libev will avoid to modify the signal
		440	mask. Specifically, this means you ahve to make sure signals are unblocked
		441	when you want to receive them.
		442
		443	This behaviour is useful when you want to do your own signal handling, or
		444	want to handle signals only in specific threads and want to avoid libev
		445	unblocking the signals.
		446
		447	This flag's behaviour will become the default in future versions of libev.
382		448
383	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	449	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
384		450
385	This is your standard select(2) backend. Not I<completely> standard, as	451	This is your standard select(2) backend. Not I<completely> standard, as
386	libev tries to roll its own fd_set with no limits on the number of fds,	452	libev tries to roll its own fd_set with no limits on the number of fds,
…		…
411	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and	477	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
412	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.	478	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
413		479
414	=item C<EVBACKEND_EPOLL> (value 4, Linux)	480	=item C<EVBACKEND_EPOLL> (value 4, Linux)
415		481
		482	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		483	kernels).
		484
416	For few fds, this backend is a bit little slower than poll and select,	485	For few fds, this backend is a bit little slower than poll and select,
417	but it scales phenomenally better. While poll and select usually scale	486	but it scales phenomenally better. While poll and select usually scale
418	like O(total_fds) where n is the total number of fds (or the highest fd),	487	like O(total_fds) where n is the total number of fds (or the highest fd),
419	epoll scales either O(1) or O(active_fds).	488	epoll scales either O(1) or O(active_fds).
420		489
421	The epoll mechanism deserves honorable mention as the most misdesigned	490	The epoll mechanism deserves honorable mention as the most misdesigned
422	of the more advanced event mechanisms: mere annoyances include silently	491	of the more advanced event mechanisms: mere annoyances include silently
423	dropping file descriptors, requiring a system call per change per file	492	dropping file descriptors, requiring a system call per change per file
424	descriptor (and unnecessary guessing of parameters), problems with dup and	493	descriptor (and unnecessary guessing of parameters), problems with dup,
		494	returning before the timeout value, resulting in additional iterations
		495	(and only giving 5ms accuracy while select on the same platform gives
425	so on. The biggest issue is fork races, however - if a program forks then	496	0.1ms) and so on. The biggest issue is fork races, however - if a program
426	I<both> parent and child process have to recreate the epoll set, which can	497	forks then I<both> parent and child process have to recreate the epoll
427	take considerable time (one syscall per file descriptor) and is of course	498	set, which can take considerable time (one syscall per file descriptor)
428	hard to detect.	499	and is of course hard to detect.
429		500
430	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but	501	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
431	of course I<doesn't>, and epoll just loves to report events for totally	502	of course I<doesn't>, and epoll just loves to report events for totally
432	I<different> file descriptors (even already closed ones, so one cannot	503	I<different> file descriptors (even already closed ones, so one cannot
433	even remove them from the set) than registered in the set (especially	504	even remove them from the set) than registered in the set (especially
434	on SMP systems). Libev tries to counter these spurious notifications by	505	on SMP systems). Libev tries to counter these spurious notifications by
435	employing an additional generation counter and comparing that against the	506	employing an additional generation counter and comparing that against the
436	events to filter out spurious ones, recreating the set when required.	507	events to filter out spurious ones, recreating the set when required. Last
		508	not least, it also refuses to work with some file descriptors which work
		509	perfectly fine with C<select> (files, many character devices...).
		510
		511	Epoll is truly the train wreck analog among event poll mechanisms,
		512	a frankenpoll, cobbled together in a hurry, no thought to design or
		513	interaction with others.
437		514
438	While stopping, setting and starting an I/O watcher in the same iteration	515	While stopping, setting and starting an I/O watcher in the same iteration
439	will result in some caching, there is still a system call per such	516	will result in some caching, there is still a system call per such
440	incident (because the same I<file descriptor> could point to a different	517	incident (because the same I<file descriptor> could point to a different
441	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	518	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
507	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	584	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
508		585
509	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	586	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
510	it's really slow, but it still scales very well (O(active_fds)).	587	it's really slow, but it still scales very well (O(active_fds)).
511		588
512	Please note that Solaris event ports can deliver a lot of spurious
513	notifications, so you need to use non-blocking I/O or other means to avoid
514	blocking when no data (or space) is available.
515
516	While this backend scales well, it requires one system call per active	589	While this backend scales well, it requires one system call per active
517	file descriptor per loop iteration. For small and medium numbers of file	590	file descriptor per loop iteration. For small and medium numbers of file
518	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	591	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
519	might perform better.	592	might perform better.
520		593
521	On the positive side, with the exception of the spurious readiness	594	On the positive side, this backend actually performed fully to
522	notifications, this backend actually performed fully to specification
523	in all tests and is fully embeddable, which is a rare feat among the	595	specification in all tests and is fully embeddable, which is a rare feat
524	OS-specific backends (I vastly prefer correctness over speed hacks).	596	among the OS-specific backends (I vastly prefer correctness over speed
		597	hacks).
		598
		599	On the negative side, the interface is I<bizarre> - so bizarre that
		600	even sun itself gets it wrong in their code examples: The event polling
		601	function sometimes returning events to the caller even though an error
		602	occurred, but with no indication whether it has done so or not (yes, it's
		603	even documented that way) - deadly for edge-triggered interfaces where
		604	you absolutely have to know whether an event occurred or not because you
		605	have to re-arm the watcher.
		606
		607	Fortunately libev seems to be able to work around these idiocies.
525		608
526	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	609	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
527	C<EVBACKEND_POLL>.	610	C<EVBACKEND_POLL>.
528		611
529	=item C<EVBACKEND_ALL>	612	=item C<EVBACKEND_ALL>
530		613
531	Try all backends (even potentially broken ones that wouldn't be tried	614	Try all backends (even potentially broken ones that wouldn't be tried
532	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	615	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
533	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	616	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
534		617
535	It is definitely not recommended to use this flag.	618	It is definitely not recommended to use this flag, use whatever
		619	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		620	at all.
		621
		622	=item C<EVBACKEND_MASK>
		623
		624	Not a backend at all, but a mask to select all backend bits from a
		625	C<flags> value, in case you want to mask out any backends from a flags
		626	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
536		627
537	=back	628	=back
538		629
539	If one or more of the backend flags are or'ed into the flags value,	630	If one or more of the backend flags are or'ed into the flags value,
540	then only these backends will be tried (in the reverse order as listed	631	then only these backends will be tried (in the reverse order as listed
541	here). If none are specified, all backends in C<ev_recommended_backends	632	here). If none are specified, all backends in C<ev_recommended_backends
542	()> will be tried.	633	()> will be tried.
543		634
544	Example: This is the most typical usage.
545
546	if (!ev_default_loop (0))
547	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
548
549	Example: Restrict libev to the select and poll backends, and do not allow
550	environment settings to be taken into account:
551
552	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
553
554	Example: Use whatever libev has to offer, but make sure that kqueue is
555	used if available (warning, breaks stuff, best use only with your own
556	private event loop and only if you know the OS supports your types of
557	fds):
558
559	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
560
561	=item struct ev_loop *ev_loop_new (unsigned int flags)
562
563	Similar to C<ev_default_loop>, but always creates a new event loop that is
564	always distinct from the default loop. Unlike the default loop, it cannot
565	handle signal and child watchers, and attempts to do so will be greeted by
566	undefined behaviour (or a failed assertion if assertions are enabled).
567
568	Note that this function I<is> thread-safe, and the recommended way to use
569	libev with threads is indeed to create one loop per thread, and using the
570	default loop in the "main" or "initial" thread.
571
572	Example: Try to create a event loop that uses epoll and nothing else.	635	Example: Try to create a event loop that uses epoll and nothing else.
573		636
574	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	637	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
575	if (!epoller)	638	if (!epoller)
576	fatal ("no epoll found here, maybe it hides under your chair");	639	fatal ("no epoll found here, maybe it hides under your chair");
577		640
		641	Example: Use whatever libev has to offer, but make sure that kqueue is
		642	used if available.
		643
		644	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		645
578	=item ev_default_destroy ()	646	=item ev_loop_destroy (loop)
579		647
580	Destroys the default loop again (frees all memory and kernel state	648	Destroys an event loop object (frees all memory and kernel state
581	etc.). None of the active event watchers will be stopped in the normal	649	etc.). None of the active event watchers will be stopped in the normal
582	sense, so e.g. C<ev_is_active> might still return true. It is your	650	sense, so e.g. C<ev_is_active> might still return true. It is your
583	responsibility to either stop all watchers cleanly yourself I<before>	651	responsibility to either stop all watchers cleanly yourself I<before>
584	calling this function, or cope with the fact afterwards (which is usually	652	calling this function, or cope with the fact afterwards (which is usually
585	the easiest thing, you can just ignore the watchers and/or C<free ()> them	653	the easiest thing, you can just ignore the watchers and/or C<free ()> them
…		…
587		655
588	Note that certain global state, such as signal state (and installed signal	656	Note that certain global state, such as signal state (and installed signal
589	handlers), will not be freed by this function, and related watchers (such	657	handlers), will not be freed by this function, and related watchers (such
590	as signal and child watchers) would need to be stopped manually.	658	as signal and child watchers) would need to be stopped manually.
591		659
592	In general it is not advisable to call this function except in the	660	This function is normally used on loop objects allocated by
593	rare occasion where you really need to free e.g. the signal handling	661	C<ev_loop_new>, but it can also be used on the default loop returned by
		662	C<ev_default_loop>, in which case it is not thread-safe.
		663
		664	Note that it is not advisable to call this function on the default loop
		665	except in the rare occasion where you really need to free its resources.
594	pipe fds. If you need dynamically allocated loops it is better to use	666	If you need dynamically allocated loops it is better to use C<ev_loop_new>
595	C<ev_loop_new> and C<ev_loop_destroy>).	667	and C<ev_loop_destroy>.
596		668
597	=item ev_loop_destroy (loop)	669	=item ev_loop_fork (loop)
598		670
599	Like C<ev_default_destroy>, but destroys an event loop created by an
600	earlier call to C<ev_loop_new>.
601
602	=item ev_default_fork ()
603
604	This function sets a flag that causes subsequent C<ev_loop> iterations	671	This function sets a flag that causes subsequent C<ev_run> iterations to
605	to reinitialise the kernel state for backends that have one. Despite the	672	reinitialise the kernel state for backends that have one. Despite the
606	name, you can call it anytime, but it makes most sense after forking, in	673	name, you can call it anytime, but it makes most sense after forking, in
607	the child process (or both child and parent, but that again makes little	674	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
608	sense). You I<must> call it in the child before using any of the libev	675	child before resuming or calling C<ev_run>.
609	functions, and it will only take effect at the next C<ev_loop> iteration.	676
		677	Again, you I<have> to call it on I<any> loop that you want to re-use after
		678	a fork, I<even if you do not plan to use the loop in the parent>. This is
		679	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		680	during fork.
610		681
611	On the other hand, you only need to call this function in the child	682	On the other hand, you only need to call this function in the child
612	process if and only if you want to use the event library in the child. If	683	process if and only if you want to use the event loop in the child. If
613	you just fork+exec, you don't have to call it at all.	684	you just fork+exec or create a new loop in the child, you don't have to
		685	call it at all (in fact, C<epoll> is so badly broken that it makes a
		686	difference, but libev will usually detect this case on its own and do a
		687	costly reset of the backend).
614		688
615	The function itself is quite fast and it's usually not a problem to call	689	The function itself is quite fast and it's usually not a problem to call
616	it just in case after a fork. To make this easy, the function will fit in	690	it just in case after a fork.
617	quite nicely into a call to C<pthread_atfork>:
618		691
		692	Example: Automate calling C<ev_loop_fork> on the default loop when
		693	using pthreads.
		694
		695	static void
		696	post_fork_child (void)
		697	{
		698	ev_loop_fork (EV_DEFAULT);
		699	}
		700
		701	...
619	pthread_atfork (0, 0, ev_default_fork);	702	pthread_atfork (0, 0, post_fork_child);
620
621	=item ev_loop_fork (loop)
622
623	Like C<ev_default_fork>, but acts on an event loop created by
624	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
625	after fork that you want to re-use in the child, and how you do this is
626	entirely your own problem.
627		703
628	=item int ev_is_default_loop (loop)	704	=item int ev_is_default_loop (loop)
629		705
630	Returns true when the given loop is, in fact, the default loop, and false	706	Returns true when the given loop is, in fact, the default loop, and false
631	otherwise.	707	otherwise.
632		708
633	=item unsigned int ev_loop_count (loop)	709	=item unsigned int ev_iteration (loop)
634		710
635	Returns the count of loop iterations for the loop, which is identical to	711	Returns the current iteration count for the event loop, which is identical
636	the number of times libev did poll for new events. It starts at C<0> and	712	to the number of times libev did poll for new events. It starts at C<0>
637	happily wraps around with enough iterations.	713	and happily wraps around with enough iterations.
638		714
639	This value can sometimes be useful as a generation counter of sorts (it	715	This value can sometimes be useful as a generation counter of sorts (it
640	"ticks" the number of loop iterations), as it roughly corresponds with	716	"ticks" the number of loop iterations), as it roughly corresponds with
641	C<ev_prepare> and C<ev_check> calls.	717	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		718	prepare and check phases.
642		719
643	=item unsigned int ev_loop_depth (loop)	720	=item unsigned int ev_depth (loop)
644		721
645	Returns the number of times C<ev_loop> was entered minus the number of	722	Returns the number of times C<ev_run> was entered minus the number of
646	times C<ev_loop> was exited, in other words, the recursion depth.	723	times C<ev_run> was exited normally, in other words, the recursion depth.
647		724
648	Outside C<ev_loop>, this number is zero. In a callback, this number is	725	Outside C<ev_run>, this number is zero. In a callback, this number is
649	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),	726	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
650	in which case it is higher.	727	in which case it is higher.
651		728
652	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread	729	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
653	etc.), doesn't count as exit.	730	throwing an exception etc.), doesn't count as "exit" - consider this
		731	as a hint to avoid such ungentleman-like behaviour unless it's really
		732	convenient, in which case it is fully supported.
654		733
655	=item unsigned int ev_backend (loop)	734	=item unsigned int ev_backend (loop)
656		735
657	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	736	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
658	use.	737	use.
…		…
667		746
668	=item ev_now_update (loop)	747	=item ev_now_update (loop)
669		748
670	Establishes the current time by querying the kernel, updating the time	749	Establishes the current time by querying the kernel, updating the time
671	returned by C<ev_now ()> in the progress. This is a costly operation and	750	returned by C<ev_now ()> in the progress. This is a costly operation and
672	is usually done automatically within C<ev_loop ()>.	751	is usually done automatically within C<ev_run ()>.
673		752
674	This function is rarely useful, but when some event callback runs for a	753	This function is rarely useful, but when some event callback runs for a
675	very long time without entering the event loop, updating libev's idea of	754	very long time without entering the event loop, updating libev's idea of
676	the current time is a good idea.	755	the current time is a good idea.
677		756
…		…
679		758
680	=item ev_suspend (loop)	759	=item ev_suspend (loop)
681		760
682	=item ev_resume (loop)	761	=item ev_resume (loop)
683		762
684	These two functions suspend and resume a loop, for use when the loop is	763	These two functions suspend and resume an event loop, for use when the
685	not used for a while and timeouts should not be processed.	764	loop is not used for a while and timeouts should not be processed.
686		765
687	A typical use case would be an interactive program such as a game: When	766	A typical use case would be an interactive program such as a game: When
688	the user presses C<^Z> to suspend the game and resumes it an hour later it	767	the user presses C<^Z> to suspend the game and resumes it an hour later it
689	would be best to handle timeouts as if no time had actually passed while	768	would be best to handle timeouts as if no time had actually passed while
690	the program was suspended. This can be achieved by calling C<ev_suspend>	769	the program was suspended. This can be achieved by calling C<ev_suspend>
…		…
692	C<ev_resume> directly afterwards to resume timer processing.	771	C<ev_resume> directly afterwards to resume timer processing.
693		772
694	Effectively, all C<ev_timer> watchers will be delayed by the time spend	773	Effectively, all C<ev_timer> watchers will be delayed by the time spend
695	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers	774	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
696	will be rescheduled (that is, they will lose any events that would have	775	will be rescheduled (that is, they will lose any events that would have
697	occured while suspended).	776	occurred while suspended).
698		777
699	After calling C<ev_suspend> you B<must not> call I<any> function on the	778	After calling C<ev_suspend> you B<must not> call I<any> function on the
700	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>	779	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
701	without a previous call to C<ev_suspend>.	780	without a previous call to C<ev_suspend>.
702		781
703	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the	782	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
704	event loop time (see C<ev_now_update>).	783	event loop time (see C<ev_now_update>).
705		784
706	=item ev_loop (loop, int flags)	785	=item ev_run (loop, int flags)
707		786
708	Finally, this is it, the event handler. This function usually is called	787	Finally, this is it, the event handler. This function usually is called
709	after you initialised all your watchers and you want to start handling	788	after you have initialised all your watchers and you want to start
710	events.	789	handling events. It will ask the operating system for any new events, call
		790	the watcher callbacks, an then repeat the whole process indefinitely: This
		791	is why event loops are called I<loops>.
711		792
712	If the flags argument is specified as C<0>, it will not return until	793	If the flags argument is specified as C<0>, it will keep handling events
713	either no event watchers are active anymore or C<ev_unloop> was called.	794	until either no event watchers are active anymore or C<ev_break> was
		795	called.
714		796
715	Please note that an explicit C<ev_unloop> is usually better than	797	Please note that an explicit C<ev_break> is usually better than
716	relying on all watchers to be stopped when deciding when a program has	798	relying on all watchers to be stopped when deciding when a program has
717	finished (especially in interactive programs), but having a program	799	finished (especially in interactive programs), but having a program
718	that automatically loops as long as it has to and no longer by virtue	800	that automatically loops as long as it has to and no longer by virtue
719	of relying on its watchers stopping correctly, that is truly a thing of	801	of relying on its watchers stopping correctly, that is truly a thing of
720	beauty.	802	beauty.
721		803
		804	This function is also I<mostly> exception-safe - you can break out of
		805	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		806	exception and so on. This does not decrement the C<ev_depth> value, nor
		807	will it clear any outstanding C<EVBREAK_ONE> breaks.
		808
722	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	809	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
723	those events and any already outstanding ones, but will not block your	810	those events and any already outstanding ones, but will not wait and
724	process in case there are no events and will return after one iteration of	811	block your process in case there are no events and will return after one
725	the loop.	812	iteration of the loop. This is sometimes useful to poll and handle new
		813	events while doing lengthy calculations, to keep the program responsive.
726		814
727	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	815	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
728	necessary) and will handle those and any already outstanding ones. It	816	necessary) and will handle those and any already outstanding ones. It
729	will block your process until at least one new event arrives (which could	817	will block your process until at least one new event arrives (which could
730	be an event internal to libev itself, so there is no guarantee that a	818	be an event internal to libev itself, so there is no guarantee that a
731	user-registered callback will be called), and will return after one	819	user-registered callback will be called), and will return after one
732	iteration of the loop.	820	iteration of the loop.
733		821
734	This is useful if you are waiting for some external event in conjunction	822	This is useful if you are waiting for some external event in conjunction
735	with something not expressible using other libev watchers (i.e. "roll your	823	with something not expressible using other libev watchers (i.e. "roll your
736	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	824	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
737	usually a better approach for this kind of thing.	825	usually a better approach for this kind of thing.
738		826
739	Here are the gory details of what C<ev_loop> does:	827	Here are the gory details of what C<ev_run> does:
740		828
		829	- Increment loop depth.
		830	- Reset the ev_break status.
741	- Before the first iteration, call any pending watchers.	831	- Before the first iteration, call any pending watchers.
		832	LOOP:
742	* If EVFLAG_FORKCHECK was used, check for a fork.	833	- If EVFLAG_FORKCHECK was used, check for a fork.
743	- If a fork was detected (by any means), queue and call all fork watchers.	834	- If a fork was detected (by any means), queue and call all fork watchers.
744	- Queue and call all prepare watchers.	835	- Queue and call all prepare watchers.
		836	- If ev_break was called, goto FINISH.
745	- If we have been forked, detach and recreate the kernel state	837	- If we have been forked, detach and recreate the kernel state
746	as to not disturb the other process.	838	as to not disturb the other process.
747	- Update the kernel state with all outstanding changes.	839	- Update the kernel state with all outstanding changes.
748	- Update the "event loop time" (ev_now ()).	840	- Update the "event loop time" (ev_now ()).
749	- Calculate for how long to sleep or block, if at all	841	- Calculate for how long to sleep or block, if at all
750	(active idle watchers, EVLOOP_NONBLOCK or not having	842	(active idle watchers, EVRUN_NOWAIT or not having
751	any active watchers at all will result in not sleeping).	843	any active watchers at all will result in not sleeping).
752	- Sleep if the I/O and timer collect interval say so.	844	- Sleep if the I/O and timer collect interval say so.
		845	- Increment loop iteration counter.
753	- Block the process, waiting for any events.	846	- Block the process, waiting for any events.
754	- Queue all outstanding I/O (fd) events.	847	- Queue all outstanding I/O (fd) events.
755	- Update the "event loop time" (ev_now ()), and do time jump adjustments.	848	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
756	- Queue all expired timers.	849	- Queue all expired timers.
757	- Queue all expired periodics.	850	- Queue all expired periodics.
758	- Unless any events are pending now, queue all idle watchers.	851	- Queue all idle watchers with priority higher than that of pending events.
759	- Queue all check watchers.	852	- Queue all check watchers.
760	- Call all queued watchers in reverse order (i.e. check watchers first).	853	- Call all queued watchers in reverse order (i.e. check watchers first).
761	Signals and child watchers are implemented as I/O watchers, and will	854	Signals and child watchers are implemented as I/O watchers, and will
762	be handled here by queueing them when their watcher gets executed.	855	be handled here by queueing them when their watcher gets executed.
763	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	856	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
764	were used, or there are no active watchers, return, otherwise	857	were used, or there are no active watchers, goto FINISH, otherwise
765	continue with step *.	858	continue with step LOOP.
		859	FINISH:
		860	- Reset the ev_break status iff it was EVBREAK_ONE.
		861	- Decrement the loop depth.
		862	- Return.
766		863
767	Example: Queue some jobs and then loop until no events are outstanding	864	Example: Queue some jobs and then loop until no events are outstanding
768	anymore.	865	anymore.
769		866
770	... queue jobs here, make sure they register event watchers as long	867	... queue jobs here, make sure they register event watchers as long
771	... as they still have work to do (even an idle watcher will do..)	868	... as they still have work to do (even an idle watcher will do..)
772	ev_loop (my_loop, 0);	869	ev_run (my_loop, 0);
773	... jobs done or somebody called unloop. yeah!	870	... jobs done or somebody called unloop. yeah!
774		871
775	=item ev_unloop (loop, how)	872	=item ev_break (loop, how)
776		873
777	Can be used to make a call to C<ev_loop> return early (but only after it	874	Can be used to make a call to C<ev_run> return early (but only after it
778	has processed all outstanding events). The C<how> argument must be either	875	has processed all outstanding events). The C<how> argument must be either
779	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	876	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
780	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	877	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
781		878
782	This "unloop state" will be cleared when entering C<ev_loop> again.	879	This "break state" will be cleared on the next call to C<ev_run>.
783		880
784	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.	881	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		882	which case it will have no effect.
785		883
786	=item ev_ref (loop)	884	=item ev_ref (loop)
787		885
788	=item ev_unref (loop)	886	=item ev_unref (loop)
789		887
790	Ref/unref can be used to add or remove a reference count on the event	888	Ref/unref can be used to add or remove a reference count on the event
791	loop: Every watcher keeps one reference, and as long as the reference	889	loop: Every watcher keeps one reference, and as long as the reference
792	count is nonzero, C<ev_loop> will not return on its own.	890	count is nonzero, C<ev_run> will not return on its own.
793		891
794	If you have a watcher you never unregister that should not keep C<ev_loop>	892	This is useful when you have a watcher that you never intend to
795	from returning, call ev_unref() after starting, and ev_ref() before	893	unregister, but that nevertheless should not keep C<ev_run> from
		894	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
796	stopping it.	895	before stopping it.
797		896
798	As an example, libev itself uses this for its internal signal pipe: It	897	As an example, libev itself uses this for its internal signal pipe: It
799	is not visible to the libev user and should not keep C<ev_loop> from	898	is not visible to the libev user and should not keep C<ev_run> from
800	exiting if no event watchers registered by it are active. It is also an	899	exiting if no event watchers registered by it are active. It is also an
801	excellent way to do this for generic recurring timers or from within	900	excellent way to do this for generic recurring timers or from within
802	third-party libraries. Just remember to I<unref after start> and I<ref	901	third-party libraries. Just remember to I<unref after start> and I<ref
803	before stop> (but only if the watcher wasn't active before, or was active	902	before stop> (but only if the watcher wasn't active before, or was active
804	before, respectively. Note also that libev might stop watchers itself	903	before, respectively. Note also that libev might stop watchers itself
805	(e.g. non-repeating timers) in which case you have to C<ev_ref>	904	(e.g. non-repeating timers) in which case you have to C<ev_ref>
806	in the callback).	905	in the callback).
807		906
808	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	907	Example: Create a signal watcher, but keep it from keeping C<ev_run>
809	running when nothing else is active.	908	running when nothing else is active.
810		909
811	ev_signal exitsig;	910	ev_signal exitsig;
812	ev_signal_init (&exitsig, sig_cb, SIGINT);	911	ev_signal_init (&exitsig, sig_cb, SIGINT);
813	ev_signal_start (loop, &exitsig);	912	ev_signal_start (loop, &exitsig);
814	evf_unref (loop);	913	ev_unref (loop);
815		914
816	Example: For some weird reason, unregister the above signal handler again.	915	Example: For some weird reason, unregister the above signal handler again.
817		916
818	ev_ref (loop);	917	ev_ref (loop);
819	ev_signal_stop (loop, &exitsig);	918	ev_signal_stop (loop, &exitsig);
…		…
858	usually doesn't make much sense to set it to a lower value than C<0.01>,	957	usually doesn't make much sense to set it to a lower value than C<0.01>,
859	as this approaches the timing granularity of most systems. Note that if	958	as this approaches the timing granularity of most systems. Note that if
860	you do transactions with the outside world and you can't increase the	959	you do transactions with the outside world and you can't increase the
861	parallelity, then this setting will limit your transaction rate (if you	960	parallelity, then this setting will limit your transaction rate (if you
862	need to poll once per transaction and the I/O collect interval is 0.01,	961	need to poll once per transaction and the I/O collect interval is 0.01,
863	then you can't do more than 100 transations per second).	962	then you can't do more than 100 transactions per second).
864		963
865	Setting the I<timeout collect interval> can improve the opportunity for	964	Setting the I<timeout collect interval> can improve the opportunity for
866	saving power, as the program will "bundle" timer callback invocations that	965	saving power, as the program will "bundle" timer callback invocations that
867	are "near" in time together, by delaying some, thus reducing the number of	966	are "near" in time together, by delaying some, thus reducing the number of
868	times the process sleeps and wakes up again. Another useful technique to	967	times the process sleeps and wakes up again. Another useful technique to
…		…
876	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);	975	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
877		976
878	=item ev_invoke_pending (loop)	977	=item ev_invoke_pending (loop)
879		978
880	This call will simply invoke all pending watchers while resetting their	979	This call will simply invoke all pending watchers while resetting their
881	pending state. Normally, C<ev_loop> does this automatically when required,	980	pending state. Normally, C<ev_run> does this automatically when required,
882	but when overriding the invoke callback this call comes handy.	981	but when overriding the invoke callback this call comes handy. This
		982	function can be invoked from a watcher - this can be useful for example
		983	when you want to do some lengthy calculation and want to pass further
		984	event handling to another thread (you still have to make sure only one
		985	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
883		986
884	=item int ev_pending_count (loop)	987	=item int ev_pending_count (loop)
885		988
886	Returns the number of pending watchers - zero indicates that no watchers	989	Returns the number of pending watchers - zero indicates that no watchers
887	are pending.	990	are pending.
888		991
889	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))	992	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
890		993
891	This overrides the invoke pending functionality of the loop: Instead of	994	This overrides the invoke pending functionality of the loop: Instead of
892	invoking all pending watchers when there are any, C<ev_loop> will call	995	invoking all pending watchers when there are any, C<ev_run> will call
893	this callback instead. This is useful, for example, when you want to	996	this callback instead. This is useful, for example, when you want to
894	invoke the actual watchers inside another context (another thread etc.).	997	invoke the actual watchers inside another context (another thread etc.).
895		998
896	If you want to reset the callback, use C<ev_invoke_pending> as new	999	If you want to reset the callback, use C<ev_invoke_pending> as new
897	callback.	1000	callback.
…		…
900		1003
901	Sometimes you want to share the same loop between multiple threads. This	1004	Sometimes you want to share the same loop between multiple threads. This
902	can be done relatively simply by putting mutex_lock/unlock calls around	1005	can be done relatively simply by putting mutex_lock/unlock calls around
903	each call to a libev function.	1006	each call to a libev function.
904		1007
905	However, C<ev_loop> can run an indefinite time, so it is not feasible to	1008	However, C<ev_run> can run an indefinite time, so it is not feasible
906	wait for it to return. One way around this is to wake up the loop via	1009	to wait for it to return. One way around this is to wake up the event
907	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>	1010	loop via C<ev_break> and C<av_async_send>, another way is to set these
908	and I<acquire> callbacks on the loop.	1011	I<release> and I<acquire> callbacks on the loop.
909		1012
910	When set, then C<release> will be called just before the thread is	1013	When set, then C<release> will be called just before the thread is
911	suspended waiting for new events, and C<acquire> is called just	1014	suspended waiting for new events, and C<acquire> is called just
912	afterwards.	1015	afterwards.
913		1016
…		…
916		1019
917	While event loop modifications are allowed between invocations of	1020	While event loop modifications are allowed between invocations of
918	C<release> and C<acquire> (that's their only purpose after all), no	1021	C<release> and C<acquire> (that's their only purpose after all), no
919	modifications done will affect the event loop, i.e. adding watchers will	1022	modifications done will affect the event loop, i.e. adding watchers will
920	have no effect on the set of file descriptors being watched, or the time	1023	have no effect on the set of file descriptors being watched, or the time
921	waited. USe an C<ev_async> watcher to wake up C<ev_loop> when you want it	1024	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
922	to take note of any changes you made.	1025	to take note of any changes you made.
923		1026
924	In theory, threads executing C<ev_loop> will be async-cancel safe between	1027	In theory, threads executing C<ev_run> will be async-cancel safe between
925	invocations of C<release> and C<acquire>.	1028	invocations of C<release> and C<acquire>.
926		1029
927	See also the locking example in the C<THREADS> section later in this	1030	See also the locking example in the C<THREADS> section later in this
928	document.	1031	document.
929		1032
930	=item ev_set_userdata (loop, void *data)	1033	=item ev_set_userdata (loop, void *data)
931		1034
932	=item ev_userdata (loop)	1035	=item void *ev_userdata (loop)
933		1036
934	Set and retrieve a single C<void *> associated with a loop. When	1037	Set and retrieve a single C<void *> associated with a loop. When
935	C<ev_set_userdata> has never been called, then C<ev_userdata> returns	1038	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
936	C<0.>	1039	C<0>.
937		1040
938	These two functions can be used to associate arbitrary data with a loop,	1041	These two functions can be used to associate arbitrary data with a loop,
939	and are intended solely for the C<invoke_pending_cb>, C<release> and	1042	and are intended solely for the C<invoke_pending_cb>, C<release> and
940	C<acquire> callbacks described above, but of course can be (ab-)used for	1043	C<acquire> callbacks described above, but of course can be (ab-)used for
941	any other purpose as well.	1044	any other purpose as well.
942		1045
943	=item ev_loop_verify (loop)	1046	=item ev_verify (loop)
944		1047
945	This function only does something when C<EV_VERIFY> support has been	1048	This function only does something when C<EV_VERIFY> support has been
946	compiled in, which is the default for non-minimal builds. It tries to go	1049	compiled in, which is the default for non-minimal builds. It tries to go
947	through all internal structures and checks them for validity. If anything	1050	through all internal structures and checks them for validity. If anything
948	is found to be inconsistent, it will print an error message to standard	1051	is found to be inconsistent, it will print an error message to standard
…		…
959		1062
960	In the following description, uppercase C<TYPE> in names stands for the	1063	In the following description, uppercase C<TYPE> in names stands for the
961	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer	1064	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
962	watchers and C<ev_io_start> for I/O watchers.	1065	watchers and C<ev_io_start> for I/O watchers.
963		1066
964	A watcher is a structure that you create and register to record your	1067	A watcher is an opaque structure that you allocate and register to record
965	interest in some event. For instance, if you want to wait for STDIN to	1068	your interest in some event. To make a concrete example, imagine you want
966	become readable, you would create an C<ev_io> watcher for that:	1069	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1070	for that:
967		1071
968	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)	1072	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
969	{	1073	{
970	ev_io_stop (w);	1074	ev_io_stop (w);
971	ev_unloop (loop, EVUNLOOP_ALL);	1075	ev_break (loop, EVBREAK_ALL);
972	}	1076	}
973		1077
974	struct ev_loop *loop = ev_default_loop (0);	1078	struct ev_loop *loop = ev_default_loop (0);
975		1079
976	ev_io stdin_watcher;	1080	ev_io stdin_watcher;
977		1081
978	ev_init (&stdin_watcher, my_cb);	1082	ev_init (&stdin_watcher, my_cb);
979	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1083	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
980	ev_io_start (loop, &stdin_watcher);	1084	ev_io_start (loop, &stdin_watcher);
981		1085
982	ev_loop (loop, 0);	1086	ev_run (loop, 0);
983		1087
984	As you can see, you are responsible for allocating the memory for your	1088	As you can see, you are responsible for allocating the memory for your
985	watcher structures (and it is I<usually> a bad idea to do this on the	1089	watcher structures (and it is I<usually> a bad idea to do this on the
986	stack).	1090	stack).
987		1091
988	Each watcher has an associated watcher structure (called C<struct ev_TYPE>	1092	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
989	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).	1093	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
990		1094
991	Each watcher structure must be initialised by a call to C<ev_init	1095	Each watcher structure must be initialised by a call to C<ev_init (watcher
992	(watcher *, callback)>, which expects a callback to be provided. This	1096	*, callback)>, which expects a callback to be provided. This callback is
993	callback gets invoked each time the event occurs (or, in the case of I/O	1097	invoked each time the event occurs (or, in the case of I/O watchers, each
994	watchers, each time the event loop detects that the file descriptor given	1098	time the event loop detects that the file descriptor given is readable
995	is readable and/or writable).	1099	and/or writable).
996		1100
997	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>	1101	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
998	macro to configure it, with arguments specific to the watcher type. There	1102	macro to configure it, with arguments specific to the watcher type. There
999	is also a macro to combine initialisation and setting in one call: C<<	1103	is also a macro to combine initialisation and setting in one call: C<<
1000	ev_TYPE_init (watcher *, callback, ...) >>.	1104	ev_TYPE_init (watcher *, callback, ...) >>.
…		…
1023	=item C<EV_WRITE>	1127	=item C<EV_WRITE>
1024		1128
1025	The file descriptor in the C<ev_io> watcher has become readable and/or	1129	The file descriptor in the C<ev_io> watcher has become readable and/or
1026	writable.	1130	writable.
1027		1131
1028	=item C<EV_TIMEOUT>	1132	=item C<EV_TIMER>
1029		1133
1030	The C<ev_timer> watcher has timed out.	1134	The C<ev_timer> watcher has timed out.
1031		1135
1032	=item C<EV_PERIODIC>	1136	=item C<EV_PERIODIC>
1033		1137
…		…
1051		1155
1052	=item C<EV_PREPARE>	1156	=item C<EV_PREPARE>
1053		1157
1054	=item C<EV_CHECK>	1158	=item C<EV_CHECK>
1055		1159
1056	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1160	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts
1057	to gather new events, and all C<ev_check> watchers are invoked just after	1161	to gather new events, and all C<ev_check> watchers are invoked just after
1058	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1162	C<ev_run> has gathered them, but before it invokes any callbacks for any
1059	received events. Callbacks of both watcher types can start and stop as	1163	received events. Callbacks of both watcher types can start and stop as
1060	many watchers as they want, and all of them will be taken into account	1164	many watchers as they want, and all of them will be taken into account
1061	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1165	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
1062	C<ev_loop> from blocking).	1166	C<ev_run> from blocking).
1063		1167
1064	=item C<EV_EMBED>	1168	=item C<EV_EMBED>
1065		1169
1066	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1170	The embedded event loop specified in the C<ev_embed> watcher needs attention.
1067		1171
1068	=item C<EV_FORK>	1172	=item C<EV_FORK>
1069		1173
1070	The event loop has been resumed in the child process after fork (see	1174	The event loop has been resumed in the child process after fork (see
1071	C<ev_fork>).	1175	C<ev_fork>).
		1176
		1177	=item C<EV_CLEANUP>
		1178
		1179	The event loop is about to be destroyed (see C<ev_cleanup>).
1072		1180
1073	=item C<EV_ASYNC>	1181	=item C<EV_ASYNC>
1074		1182
1075	The given async watcher has been asynchronously notified (see C<ev_async>).	1183	The given async watcher has been asynchronously notified (see C<ev_async>).
1076		1184
…		…
1123		1231
1124	ev_io w;	1232	ev_io w;
1125	ev_init (&w, my_cb);	1233	ev_init (&w, my_cb);
1126	ev_io_set (&w, STDIN_FILENO, EV_READ);	1234	ev_io_set (&w, STDIN_FILENO, EV_READ);
1127		1235
1128	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1236	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
1129		1237
1130	This macro initialises the type-specific parts of a watcher. You need to	1238	This macro initialises the type-specific parts of a watcher. You need to
1131	call C<ev_init> at least once before you call this macro, but you can	1239	call C<ev_init> at least once before you call this macro, but you can
1132	call C<ev_TYPE_set> any number of times. You must not, however, call this	1240	call C<ev_TYPE_set> any number of times. You must not, however, call this
1133	macro on a watcher that is active (it can be pending, however, which is a	1241	macro on a watcher that is active (it can be pending, however, which is a
…		…
1146		1254
1147	Example: Initialise and set an C<ev_io> watcher in one step.	1255	Example: Initialise and set an C<ev_io> watcher in one step.
1148		1256
1149	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);	1257	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
1150		1258
1151	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1259	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
1152		1260
1153	Starts (activates) the given watcher. Only active watchers will receive	1261	Starts (activates) the given watcher. Only active watchers will receive
1154	events. If the watcher is already active nothing will happen.	1262	events. If the watcher is already active nothing will happen.
1155		1263
1156	Example: Start the C<ev_io> watcher that is being abused as example in this	1264	Example: Start the C<ev_io> watcher that is being abused as example in this
1157	whole section.	1265	whole section.
1158		1266
1159	ev_io_start (EV_DEFAULT_UC, &w);	1267	ev_io_start (EV_DEFAULT_UC, &w);
1160		1268
1161	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1269	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
1162		1270
1163	Stops the given watcher if active, and clears the pending status (whether	1271	Stops the given watcher if active, and clears the pending status (whether
1164	the watcher was active or not).	1272	the watcher was active or not).
1165		1273
1166	It is possible that stopped watchers are pending - for example,	1274	It is possible that stopped watchers are pending - for example,
…		…
1191	=item ev_cb_set (ev_TYPE *watcher, callback)	1299	=item ev_cb_set (ev_TYPE *watcher, callback)
1192		1300
1193	Change the callback. You can change the callback at virtually any time	1301	Change the callback. You can change the callback at virtually any time
1194	(modulo threads).	1302	(modulo threads).
1195		1303
1196	=item ev_set_priority (ev_TYPE *watcher, priority)	1304	=item ev_set_priority (ev_TYPE *watcher, int priority)
1197		1305
1198	=item int ev_priority (ev_TYPE *watcher)	1306	=item int ev_priority (ev_TYPE *watcher)
1199		1307
1200	Set and query the priority of the watcher. The priority is a small	1308	Set and query the priority of the watcher. The priority is a small
1201	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1309	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
…		…
1233	watcher isn't pending it does nothing and returns C<0>.	1341	watcher isn't pending it does nothing and returns C<0>.
1234		1342
1235	Sometimes it can be useful to "poll" a watcher instead of waiting for its	1343	Sometimes it can be useful to "poll" a watcher instead of waiting for its
1236	callback to be invoked, which can be accomplished with this function.	1344	callback to be invoked, which can be accomplished with this function.
1237		1345
		1346	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1347
		1348	Feeds the given event set into the event loop, as if the specified event
		1349	had happened for the specified watcher (which must be a pointer to an
		1350	initialised but not necessarily started event watcher). Obviously you must
		1351	not free the watcher as long as it has pending events.
		1352
		1353	Stopping the watcher, letting libev invoke it, or calling
		1354	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1355	not started in the first place.
		1356
		1357	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1358	functions that do not need a watcher.
		1359
1238	=back	1360	=back
1239
1240		1361
1241	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1362	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
1242		1363
1243	Each watcher has, by default, a member C<void *data> that you can change	1364	Each watcher has, by default, a member C<void *data> that you can change
1244	and read at any time: libev will completely ignore it. This can be used	1365	and read at any time: libev will completely ignore it. This can be used
…		…
1300	t2_cb (EV_P_ ev_timer *w, int revents)	1421	t2_cb (EV_P_ ev_timer *w, int revents)
1301	{	1422	{
1302	struct my_biggy big = (struct my_biggy *)	1423	struct my_biggy big = (struct my_biggy *)
1303	(((char *)w) - offsetof (struct my_biggy, t2));	1424	(((char *)w) - offsetof (struct my_biggy, t2));
1304	}	1425	}
		1426
		1427	=head2 WATCHER STATES
		1428
		1429	There are various watcher states mentioned throughout this manual -
		1430	active, pending and so on. In this section these states and the rules to
		1431	transition between them will be described in more detail - and while these
		1432	rules might look complicated, they usually do "the right thing".
		1433
		1434	=over 4
		1435
		1436	=item initialiased
		1437
		1438	Before a watcher can be registered with the event looop it has to be
		1439	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1440	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
		1441
		1442	In this state it is simply some block of memory that is suitable for use
		1443	in an event loop. It can be moved around, freed, reused etc. at will.
		1444
		1445	=item started/running/active
		1446
		1447	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
		1448	property of the event loop, and is actively waiting for events. While in
		1449	this state it cannot be accessed (except in a few documented ways), moved,
		1450	freed or anything else - the only legal thing is to keep a pointer to it,
		1451	and call libev functions on it that are documented to work on active watchers.
		1452
		1453	=item pending
		1454
		1455	If a watcher is active and libev determines that an event it is interested
		1456	in has occurred (such as a timer expiring), it will become pending. It will
		1457	stay in this pending state until either it is stopped or its callback is
		1458	about to be invoked, so it is not normally pending inside the watcher
		1459	callback.
		1460
		1461	The watcher might or might not be active while it is pending (for example,
		1462	an expired non-repeating timer can be pending but no longer active). If it
		1463	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1464	but it is still property of the event loop at this time, so cannot be
		1465	moved, freed or reused. And if it is active the rules described in the
		1466	previous item still apply.
		1467
		1468	It is also possible to feed an event on a watcher that is not active (e.g.
		1469	via C<ev_feed_event>), in which case it becomes pending without being
		1470	active.
		1471
		1472	=item stopped
		1473
		1474	A watcher can be stopped implicitly by libev (in which case it might still
		1475	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1476	latter will clear any pending state the watcher might be in, regardless
		1477	of whether it was active or not, so stopping a watcher explicitly before
		1478	freeing it is often a good idea.
		1479
		1480	While stopped (and not pending) the watcher is essentially in the
		1481	initialised state, that is it can be reused, moved, modified in any way
		1482	you wish.
		1483
		1484	=back
1305		1485
1306	=head2 WATCHER PRIORITY MODELS	1486	=head2 WATCHER PRIORITY MODELS
1307		1487
1308	Many event loops support I<watcher priorities>, which are usually small	1488	Many event loops support I<watcher priorities>, which are usually small
1309	integers that influence the ordering of event callback invocation	1489	integers that influence the ordering of event callback invocation
…		…
1352		1532
1353	For example, to emulate how many other event libraries handle priorities,	1533	For example, to emulate how many other event libraries handle priorities,
1354	you can associate an C<ev_idle> watcher to each such watcher, and in	1534	you can associate an C<ev_idle> watcher to each such watcher, and in
1355	the normal watcher callback, you just start the idle watcher. The real	1535	the normal watcher callback, you just start the idle watcher. The real
1356	processing is done in the idle watcher callback. This causes libev to	1536	processing is done in the idle watcher callback. This causes libev to
1357	continously poll and process kernel event data for the watcher, but when	1537	continuously poll and process kernel event data for the watcher, but when
1358	the lock-out case is known to be rare (which in turn is rare :), this is	1538	the lock-out case is known to be rare (which in turn is rare :), this is
1359	workable.	1539	workable.
1360		1540
1361	Usually, however, the lock-out model implemented that way will perform	1541	Usually, however, the lock-out model implemented that way will perform
1362	miserably under the type of load it was designed to handle. In that case,	1542	miserably under the type of load it was designed to handle. In that case,
…		…
1376	{	1556	{
1377	// stop the I/O watcher, we received the event, but	1557	// stop the I/O watcher, we received the event, but
1378	// are not yet ready to handle it.	1558	// are not yet ready to handle it.
1379	ev_io_stop (EV_A_ w);	1559	ev_io_stop (EV_A_ w);
1380		1560
1381	// start the idle watcher to ahndle the actual event.	1561	// start the idle watcher to handle the actual event.
1382	// it will not be executed as long as other watchers	1562	// it will not be executed as long as other watchers
1383	// with the default priority are receiving events.	1563	// with the default priority are receiving events.
1384	ev_idle_start (EV_A_ &idle);	1564	ev_idle_start (EV_A_ &idle);
1385	}	1565	}
1386		1566
…		…
1436	In general you can register as many read and/or write event watchers per	1616	In general you can register as many read and/or write event watchers per
1437	fd as you want (as long as you don't confuse yourself). Setting all file	1617	fd as you want (as long as you don't confuse yourself). Setting all file
1438	descriptors to non-blocking mode is also usually a good idea (but not	1618	descriptors to non-blocking mode is also usually a good idea (but not
1439	required if you know what you are doing).	1619	required if you know what you are doing).
1440		1620
1441	If you cannot use non-blocking mode, then force the use of a
1442	known-to-be-good backend (at the time of this writing, this includes only
1443	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1444	descriptors for which non-blocking operation makes no sense (such as
1445	files) - libev doesn't guarentee any specific behaviour in that case.
1446
1447	Another thing you have to watch out for is that it is quite easy to	1621	Another thing you have to watch out for is that it is quite easy to
1448	receive "spurious" readiness notifications, that is your callback might	1622	receive "spurious" readiness notifications, that is, your callback might
1449	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1623	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1450	because there is no data. Not only are some backends known to create a	1624	because there is no data. It is very easy to get into this situation even
1451	lot of those (for example Solaris ports), it is very easy to get into	1625	with a relatively standard program structure. Thus it is best to always
1452	this situation even with a relatively standard program structure. Thus	1626	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1453	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1454	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1627	preferable to a program hanging until some data arrives.
1455		1628
1456	If you cannot run the fd in non-blocking mode (for example you should	1629	If you cannot run the fd in non-blocking mode (for example you should
1457	not play around with an Xlib connection), then you have to separately	1630	not play around with an Xlib connection), then you have to separately
1458	re-test whether a file descriptor is really ready with a known-to-be good	1631	re-test whether a file descriptor is really ready with a known-to-be good
1459	interface such as poll (fortunately in our Xlib example, Xlib already	1632	interface such as poll (fortunately in the case of Xlib, it already does
1460	does this on its own, so its quite safe to use). Some people additionally	1633	this on its own, so its quite safe to use). Some people additionally
1461	use C<SIGALRM> and an interval timer, just to be sure you won't block	1634	use C<SIGALRM> and an interval timer, just to be sure you won't block
1462	indefinitely.	1635	indefinitely.
1463		1636
1464	But really, best use non-blocking mode.	1637	But really, best use non-blocking mode.
1465		1638
…		…
1493		1666
1494	There is no workaround possible except not registering events	1667	There is no workaround possible except not registering events
1495	for potentially C<dup ()>'ed file descriptors, or to resort to	1668	for potentially C<dup ()>'ed file descriptors, or to resort to
1496	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1669	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1497		1670
		1671	=head3 The special problem of files
		1672
		1673	Many people try to use C<select> (or libev) on file descriptors
		1674	representing files, and expect it to become ready when their program
		1675	doesn't block on disk accesses (which can take a long time on their own).
		1676
		1677	However, this cannot ever work in the "expected" way - you get a readiness
		1678	notification as soon as the kernel knows whether and how much data is
		1679	there, and in the case of open files, that's always the case, so you
		1680	always get a readiness notification instantly, and your read (or possibly
		1681	write) will still block on the disk I/O.
		1682
		1683	Another way to view it is that in the case of sockets, pipes, character
		1684	devices and so on, there is another party (the sender) that delivers data
		1685	on it's own, but in the case of files, there is no such thing: the disk
		1686	will not send data on it's own, simply because it doesn't know what you
		1687	wish to read - you would first have to request some data.
		1688
		1689	Since files are typically not-so-well supported by advanced notification
		1690	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1691	to files, even though you should not use it. The reason for this is
		1692	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1693	usually a tty, often a pipe, but also sometimes files or special devices
		1694	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1695	F</dev/urandom>), and even though the file might better be served with
		1696	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1697	it "just works" instead of freezing.
		1698
		1699	So avoid file descriptors pointing to files when you know it (e.g. use
		1700	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1701	when you rarely read from a file instead of from a socket, and want to
		1702	reuse the same code path.
		1703
1498	=head3 The special problem of fork	1704	=head3 The special problem of fork
1499		1705
1500	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1706	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1501	useless behaviour. Libev fully supports fork, but needs to be told about	1707	useless behaviour. Libev fully supports fork, but needs to be told about
1502	it in the child.	1708	it in the child if you want to continue to use it in the child.
1503		1709
1504	To support fork in your programs, you either have to call	1710	To support fork in your child processes, you have to call C<ev_loop_fork
1505	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1711	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1506	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1712	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1507	C<EVBACKEND_POLL>.
1508		1713
1509	=head3 The special problem of SIGPIPE	1714	=head3 The special problem of SIGPIPE
1510		1715
1511	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1716	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1512	when writing to a pipe whose other end has been closed, your program gets	1717	when writing to a pipe whose other end has been closed, your program gets
…		…
1515		1720
1516	So when you encounter spurious, unexplained daemon exits, make sure you	1721	So when you encounter spurious, unexplained daemon exits, make sure you
1517	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1722	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1518	somewhere, as that would have given you a big clue).	1723	somewhere, as that would have given you a big clue).
1519		1724
		1725	=head3 The special problem of accept()ing when you can't
		1726
		1727	Many implementations of the POSIX C<accept> function (for example,
		1728	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1729	connection from the pending queue in all error cases.
		1730
		1731	For example, larger servers often run out of file descriptors (because
		1732	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1733	rejecting the connection, leading to libev signalling readiness on
		1734	the next iteration again (the connection still exists after all), and
		1735	typically causing the program to loop at 100% CPU usage.
		1736
		1737	Unfortunately, the set of errors that cause this issue differs between
		1738	operating systems, there is usually little the app can do to remedy the
		1739	situation, and no known thread-safe method of removing the connection to
		1740	cope with overload is known (to me).
		1741
		1742	One of the easiest ways to handle this situation is to just ignore it
		1743	- when the program encounters an overload, it will just loop until the
		1744	situation is over. While this is a form of busy waiting, no OS offers an
		1745	event-based way to handle this situation, so it's the best one can do.
		1746
		1747	A better way to handle the situation is to log any errors other than
		1748	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1749	messages, and continue as usual, which at least gives the user an idea of
		1750	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1751	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1752	usage.
		1753
		1754	If your program is single-threaded, then you could also keep a dummy file
		1755	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1756	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1757	close that fd, and create a new dummy fd. This will gracefully refuse
		1758	clients under typical overload conditions.
		1759
		1760	The last way to handle it is to simply log the error and C<exit>, as
		1761	is often done with C<malloc> failures, but this results in an easy
		1762	opportunity for a DoS attack.
1520		1763
1521	=head3 Watcher-Specific Functions	1764	=head3 Watcher-Specific Functions
1522		1765
1523	=over 4	1766	=over 4
1524		1767
…		…
1556	...	1799	...
1557	struct ev_loop *loop = ev_default_init (0);	1800	struct ev_loop *loop = ev_default_init (0);
1558	ev_io stdin_readable;	1801	ev_io stdin_readable;
1559	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1802	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1560	ev_io_start (loop, &stdin_readable);	1803	ev_io_start (loop, &stdin_readable);
1561	ev_loop (loop, 0);	1804	ev_run (loop, 0);
1562		1805
1563		1806
1564	=head2 C<ev_timer> - relative and optionally repeating timeouts	1807	=head2 C<ev_timer> - relative and optionally repeating timeouts
1565		1808
1566	Timer watchers are simple relative timers that generate an event after a	1809	Timer watchers are simple relative timers that generate an event after a
…		…
1575	The callback is guaranteed to be invoked only I<after> its timeout has	1818	The callback is guaranteed to be invoked only I<after> its timeout has
1576	passed (not I<at>, so on systems with very low-resolution clocks this	1819	passed (not I<at>, so on systems with very low-resolution clocks this
1577	might introduce a small delay). If multiple timers become ready during the	1820	might introduce a small delay). If multiple timers become ready during the
1578	same loop iteration then the ones with earlier time-out values are invoked	1821	same loop iteration then the ones with earlier time-out values are invoked
1579	before ones of the same priority with later time-out values (but this is	1822	before ones of the same priority with later time-out values (but this is
1580	no longer true when a callback calls C<ev_loop> recursively).	1823	no longer true when a callback calls C<ev_run> recursively).
1581		1824
1582	=head3 Be smart about timeouts	1825	=head3 Be smart about timeouts
1583		1826
1584	Many real-world problems involve some kind of timeout, usually for error	1827	Many real-world problems involve some kind of timeout, usually for error
1585	recovery. A typical example is an HTTP request - if the other side hangs,	1828	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1671	ev_tstamp timeout = last_activity + 60.;	1914	ev_tstamp timeout = last_activity + 60.;
1672		1915
1673	// if last_activity + 60. is older than now, we did time out	1916	// if last_activity + 60. is older than now, we did time out
1674	if (timeout < now)	1917	if (timeout < now)
1675	{	1918	{
1676	// timeout occured, take action	1919	// timeout occurred, take action
1677	}	1920	}
1678	else	1921	else
1679	{	1922	{
1680	// callback was invoked, but there was some activity, re-arm	1923	// callback was invoked, but there was some activity, re-arm
1681	// the watcher to fire in last_activity + 60, which is	1924	// the watcher to fire in last_activity + 60, which is
…		…
1703	to the current time (meaning we just have some activity :), then call the	1946	to the current time (meaning we just have some activity :), then call the
1704	callback, which will "do the right thing" and start the timer:	1947	callback, which will "do the right thing" and start the timer:
1705		1948
1706	ev_init (timer, callback);	1949	ev_init (timer, callback);
1707	last_activity = ev_now (loop);	1950	last_activity = ev_now (loop);
1708	callback (loop, timer, EV_TIMEOUT);	1951	callback (loop, timer, EV_TIMER);
1709		1952
1710	And when there is some activity, simply store the current time in	1953	And when there is some activity, simply store the current time in
1711	C<last_activity>, no libev calls at all:	1954	C<last_activity>, no libev calls at all:
1712		1955
1713	last_actiivty = ev_now (loop);	1956	last_activity = ev_now (loop);
1714		1957
1715	This technique is slightly more complex, but in most cases where the	1958	This technique is slightly more complex, but in most cases where the
1716	time-out is unlikely to be triggered, much more efficient.	1959	time-out is unlikely to be triggered, much more efficient.
1717		1960
1718	Changing the timeout is trivial as well (if it isn't hard-coded in the	1961	Changing the timeout is trivial as well (if it isn't hard-coded in the
…		…
1756		1999
1757	=head3 The special problem of time updates	2000	=head3 The special problem of time updates
1758		2001
1759	Establishing the current time is a costly operation (it usually takes at	2002	Establishing the current time is a costly operation (it usually takes at
1760	least two system calls): EV therefore updates its idea of the current	2003	least two system calls): EV therefore updates its idea of the current
1761	time only before and after C<ev_loop> collects new events, which causes a	2004	time only before and after C<ev_run> collects new events, which causes a
1762	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2005	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1763	lots of events in one iteration.	2006	lots of events in one iteration.
1764		2007
1765	The relative timeouts are calculated relative to the C<ev_now ()>	2008	The relative timeouts are calculated relative to the C<ev_now ()>
1766	time. This is usually the right thing as this timestamp refers to the time	2009	time. This is usually the right thing as this timestamp refers to the time
…		…
1837	C<repeat> value), or reset the running timer to the C<repeat> value.	2080	C<repeat> value), or reset the running timer to the C<repeat> value.
1838		2081
1839	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a	2082	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1840	usage example.	2083	usage example.
1841		2084
1842	=item ev_timer_remaining (loop, ev_timer *)	2085	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1843		2086
1844	Returns the remaining time until a timer fires. If the timer is active,	2087	Returns the remaining time until a timer fires. If the timer is active,
1845	then this time is relative to the current event loop time, otherwise it's	2088	then this time is relative to the current event loop time, otherwise it's
1846	the timeout value currently configured.	2089	the timeout value currently configured.
1847		2090
1848	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns	2091	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1849	C<5>. When the timer is started and one second passes, C<ev_timer_remain>	2092	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
1850	will return C<4>. When the timer expires and is restarted, it will return	2093	will return C<4>. When the timer expires and is restarted, it will return
1851	roughly C<7> (likely slightly less as callback invocation takes some time,	2094	roughly C<7> (likely slightly less as callback invocation takes some time,
1852	too), and so on.	2095	too), and so on.
1853		2096
1854	=item ev_tstamp repeat [read-write]	2097	=item ev_tstamp repeat [read-write]
…		…
1883	}	2126	}
1884		2127
1885	ev_timer mytimer;	2128	ev_timer mytimer;
1886	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2129	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1887	ev_timer_again (&mytimer); /* start timer */	2130	ev_timer_again (&mytimer); /* start timer */
1888	ev_loop (loop, 0);	2131	ev_run (loop, 0);
1889		2132
1890	// and in some piece of code that gets executed on any "activity":	2133	// and in some piece of code that gets executed on any "activity":
1891	// reset the timeout to start ticking again at 10 seconds	2134	// reset the timeout to start ticking again at 10 seconds
1892	ev_timer_again (&mytimer);	2135	ev_timer_again (&mytimer);
1893		2136
…		…
1919		2162
1920	As with timers, the callback is guaranteed to be invoked only when the	2163	As with timers, the callback is guaranteed to be invoked only when the
1921	point in time where it is supposed to trigger has passed. If multiple	2164	point in time where it is supposed to trigger has passed. If multiple
1922	timers become ready during the same loop iteration then the ones with	2165	timers become ready during the same loop iteration then the ones with
1923	earlier time-out values are invoked before ones with later time-out values	2166	earlier time-out values are invoked before ones with later time-out values
1924	(but this is no longer true when a callback calls C<ev_loop> recursively).	2167	(but this is no longer true when a callback calls C<ev_run> recursively).
1925		2168
1926	=head3 Watcher-Specific Functions and Data Members	2169	=head3 Watcher-Specific Functions and Data Members
1927		2170
1928	=over 4	2171	=over 4
1929		2172
…		…
2057	Example: Call a callback every hour, or, more precisely, whenever the	2300	Example: Call a callback every hour, or, more precisely, whenever the
2058	system time is divisible by 3600. The callback invocation times have	2301	system time is divisible by 3600. The callback invocation times have
2059	potentially a lot of jitter, but good long-term stability.	2302	potentially a lot of jitter, but good long-term stability.
2060		2303
2061	static void	2304	static void
2062	clock_cb (struct ev_loop *loop, ev_io *w, int revents)	2305	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
2063	{	2306	{
2064	... its now a full hour (UTC, or TAI or whatever your clock follows)	2307	... its now a full hour (UTC, or TAI or whatever your clock follows)
2065	}	2308	}
2066		2309
2067	ev_periodic hourly_tick;	2310	ev_periodic hourly_tick;
…		…
2090		2333
2091	=head2 C<ev_signal> - signal me when a signal gets signalled!	2334	=head2 C<ev_signal> - signal me when a signal gets signalled!
2092		2335
2093	Signal watchers will trigger an event when the process receives a specific	2336	Signal watchers will trigger an event when the process receives a specific
2094	signal one or more times. Even though signals are very asynchronous, libev	2337	signal one or more times. Even though signals are very asynchronous, libev
2095	will try it's best to deliver signals synchronously, i.e. as part of the	2338	will try its best to deliver signals synchronously, i.e. as part of the
2096	normal event processing, like any other event.	2339	normal event processing, like any other event.
2097		2340
2098	If you want signals to be delivered truly asynchronously, just use	2341	If you want signals to be delivered truly asynchronously, just use
2099	C<sigaction> as you would do without libev and forget about sharing	2342	C<sigaction> as you would do without libev and forget about sharing
2100	the signal. You can even use C<ev_async> from a signal handler to	2343	the signal. You can even use C<ev_async> from a signal handler to
…		…
2114	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should	2357	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2115	not be unduly interrupted. If you have a problem with system calls getting	2358	not be unduly interrupted. If you have a problem with system calls getting
2116	interrupted by signals you can block all signals in an C<ev_check> watcher	2359	interrupted by signals you can block all signals in an C<ev_check> watcher
2117	and unblock them in an C<ev_prepare> watcher.	2360	and unblock them in an C<ev_prepare> watcher.
2118		2361
2119	=head3 The special problem of inheritance over execve	2362	=head3 The special problem of inheritance over fork/execve/pthread_create
2120		2363
2121	Both the signal mask (C<sigprocmask>) and the signal disposition	2364	Both the signal mask (C<sigprocmask>) and the signal disposition
2122	(C<sigaction>) are unspecified after starting a signal watcher (and after	2365	(C<sigaction>) are unspecified after starting a signal watcher (and after
2123	stopping it again), that is, libev might or might not block the signal,	2366	stopping it again), that is, libev might or might not block the signal,
2124	and might or might not set or restore the installed signal handler.	2367	and might or might not set or restore the installed signal handler.
2125		2368
2126	While this does not matter for the signal disposition (libev never	2369	While this does not matter for the signal disposition (libev never
2127	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on	2370	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2128	C<execve>), this matters for the signal mask: many programs do not expect	2371	C<execve>), this matters for the signal mask: many programs do not expect
2129	many signals to be blocked.	2372	certain signals to be blocked.
2130		2373
2131	This means that before calling C<exec> (from the child) you should reset	2374	This means that before calling C<exec> (from the child) you should reset
2132	the signal mask to whatever "default" you expect (all clear is a good	2375	the signal mask to whatever "default" you expect (all clear is a good
2133	choice usually).	2376	choice usually).
2134		2377
		2378	The simplest way to ensure that the signal mask is reset in the child is
		2379	to install a fork handler with C<pthread_atfork> that resets it. That will
		2380	catch fork calls done by libraries (such as the libc) as well.
		2381
		2382	In current versions of libev, the signal will not be blocked indefinitely
		2383	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2384	the window of opportunity for problems, it will not go away, as libev
		2385	I<has> to modify the signal mask, at least temporarily.
		2386
		2387	So I can't stress this enough: I<If you do not reset your signal mask when
		2388	you expect it to be empty, you have a race condition in your code>. This
		2389	is not a libev-specific thing, this is true for most event libraries.
		2390
		2391	=head3 The special problem of threads signal handling
		2392
		2393	POSIX threads has problematic signal handling semantics, specifically,
		2394	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2395	threads in a process block signals, which is hard to achieve.
		2396
		2397	When you want to use sigwait (or mix libev signal handling with your own
		2398	for the same signals), you can tackle this problem by globally blocking
		2399	all signals before creating any threads (or creating them with a fully set
		2400	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2401	loops. Then designate one thread as "signal receiver thread" which handles
		2402	these signals. You can pass on any signals that libev might be interested
		2403	in by calling C<ev_feed_signal>.
		2404
2135	=head3 Watcher-Specific Functions and Data Members	2405	=head3 Watcher-Specific Functions and Data Members
2136		2406
2137	=over 4	2407	=over 4
2138		2408
2139	=item ev_signal_init (ev_signal *, callback, int signum)	2409	=item ev_signal_init (ev_signal *, callback, int signum)
…		…
2154	Example: Try to exit cleanly on SIGINT.	2424	Example: Try to exit cleanly on SIGINT.
2155		2425
2156	static void	2426	static void
2157	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)	2427	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
2158	{	2428	{
2159	ev_unloop (loop, EVUNLOOP_ALL);	2429	ev_break (loop, EVBREAK_ALL);
2160	}	2430	}
2161		2431
2162	ev_signal signal_watcher;	2432	ev_signal signal_watcher;
2163	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2433	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
2164	ev_signal_start (loop, &signal_watcher);	2434	ev_signal_start (loop, &signal_watcher);
…		…
2550		2820
2551	Prepare and check watchers are usually (but not always) used in pairs:	2821	Prepare and check watchers are usually (but not always) used in pairs:
2552	prepare watchers get invoked before the process blocks and check watchers	2822	prepare watchers get invoked before the process blocks and check watchers
2553	afterwards.	2823	afterwards.
2554		2824
2555	You I<must not> call C<ev_loop> or similar functions that enter	2825	You I<must not> call C<ev_run> or similar functions that enter
2556	the current event loop from either C<ev_prepare> or C<ev_check>	2826	the current event loop from either C<ev_prepare> or C<ev_check>
2557	watchers. Other loops than the current one are fine, however. The	2827	watchers. Other loops than the current one are fine, however. The
2558	rationale behind this is that you do not need to check for recursion in	2828	rationale behind this is that you do not need to check for recursion in
2559	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2829	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
2560	C<ev_check> so if you have one watcher of each kind they will always be	2830	C<ev_check> so if you have one watcher of each kind they will always be
…		…
2728		2998
2729	if (timeout >= 0)	2999	if (timeout >= 0)
2730	// create/start timer	3000	// create/start timer
2731		3001
2732	// poll	3002	// poll
2733	ev_loop (EV_A_ 0);	3003	ev_run (EV_A_ 0);
2734		3004
2735	// stop timer again	3005	// stop timer again
2736	if (timeout >= 0)	3006	if (timeout >= 0)
2737	ev_timer_stop (EV_A_ &to);	3007	ev_timer_stop (EV_A_ &to);
2738		3008
…		…
2816	if you do not want that, you need to temporarily stop the embed watcher).	3086	if you do not want that, you need to temporarily stop the embed watcher).
2817		3087
2818	=item ev_embed_sweep (loop, ev_embed *)	3088	=item ev_embed_sweep (loop, ev_embed *)
2819		3089
2820	Make a single, non-blocking sweep over the embedded loop. This works	3090	Make a single, non-blocking sweep over the embedded loop. This works
2821	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3091	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
2822	appropriate way for embedded loops.	3092	appropriate way for embedded loops.
2823		3093
2824	=item struct ev_loop *other [read-only]	3094	=item struct ev_loop *other [read-only]
2825		3095
2826	The embedded event loop.	3096	The embedded event loop.
…		…
2886	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3156	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2887	handlers will be invoked, too, of course.	3157	handlers will be invoked, too, of course.
2888		3158
2889	=head3 The special problem of life after fork - how is it possible?	3159	=head3 The special problem of life after fork - how is it possible?
2890		3160
2891	Most uses of C<fork()> consist of forking, then some simple calls to ste	3161	Most uses of C<fork()> consist of forking, then some simple calls to set
2892	up/change the process environment, followed by a call to C<exec()>. This	3162	up/change the process environment, followed by a call to C<exec()>. This
2893	sequence should be handled by libev without any problems.	3163	sequence should be handled by libev without any problems.
2894		3164
2895	This changes when the application actually wants to do event handling	3165	This changes when the application actually wants to do event handling
2896	in the child, or both parent in child, in effect "continuing" after the	3166	in the child, or both parent in child, in effect "continuing" after the
…		…
2912	disadvantage of having to use multiple event loops (which do not support	3182	disadvantage of having to use multiple event loops (which do not support
2913	signal watchers).	3183	signal watchers).
2914		3184
2915	When this is not possible, or you want to use the default loop for	3185	When this is not possible, or you want to use the default loop for
2916	other reasons, then in the process that wants to start "fresh", call	3186	other reasons, then in the process that wants to start "fresh", call
2917	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying	3187	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
2918	the default loop will "orphan" (not stop) all registered watchers, so you	3188	Destroying the default loop will "orphan" (not stop) all registered
2919	have to be careful not to execute code that modifies those watchers. Note	3189	watchers, so you have to be careful not to execute code that modifies
2920	also that in that case, you have to re-register any signal watchers.	3190	those watchers. Note also that in that case, you have to re-register any
		3191	signal watchers.
2921		3192
2922	=head3 Watcher-Specific Functions and Data Members	3193	=head3 Watcher-Specific Functions and Data Members
2923		3194
2924	=over 4	3195	=over 4
2925		3196
2926	=item ev_fork_init (ev_signal *, callback)	3197	=item ev_fork_init (ev_fork *, callback)
2927		3198
2928	Initialises and configures the fork watcher - it has no parameters of any	3199	Initialises and configures the fork watcher - it has no parameters of any
2929	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3200	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
2930	believe me.	3201	really.
2931		3202
2932	=back	3203	=back
2933		3204
2934		3205
		3206	=head2 C<ev_cleanup> - even the best things end
		3207
		3208	Cleanup watchers are called just before the event loop is being destroyed
		3209	by a call to C<ev_loop_destroy>.
		3210
		3211	While there is no guarantee that the event loop gets destroyed, cleanup
		3212	watchers provide a convenient method to install cleanup hooks for your
		3213	program, worker threads and so on - you just to make sure to destroy the
		3214	loop when you want them to be invoked.
		3215
		3216	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3217	all other watchers, they do not keep a reference to the event loop (which
		3218	makes a lot of sense if you think about it). Like all other watchers, you
		3219	can call libev functions in the callback, except C<ev_cleanup_start>.
		3220
		3221	=head3 Watcher-Specific Functions and Data Members
		3222
		3223	=over 4
		3224
		3225	=item ev_cleanup_init (ev_cleanup *, callback)
		3226
		3227	Initialises and configures the cleanup watcher - it has no parameters of
		3228	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3229	pointless, I assure you.
		3230
		3231	=back
		3232
		3233	Example: Register an atexit handler to destroy the default loop, so any
		3234	cleanup functions are called.
		3235
		3236	static void
		3237	program_exits (void)
		3238	{
		3239	ev_loop_destroy (EV_DEFAULT_UC);
		3240	}
		3241
		3242	...
		3243	atexit (program_exits);
		3244
		3245
2935	=head2 C<ev_async> - how to wake up another event loop	3246	=head2 C<ev_async> - how to wake up an event loop
2936		3247
2937	In general, you cannot use an C<ev_loop> from multiple threads or other	3248	In general, you cannot use an C<ev_run> from multiple threads or other
2938	asynchronous sources such as signal handlers (as opposed to multiple event	3249	asynchronous sources such as signal handlers (as opposed to multiple event
2939	loops - those are of course safe to use in different threads).	3250	loops - those are of course safe to use in different threads).
2940		3251
2941	Sometimes, however, you need to wake up another event loop you do not	3252	Sometimes, however, you need to wake up an event loop you do not control,
2942	control, for example because it belongs to another thread. This is what	3253	for example because it belongs to another thread. This is what C<ev_async>
2943	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3254	watchers do: as long as the C<ev_async> watcher is active, you can signal
2944	can signal it by calling C<ev_async_send>, which is thread- and signal	3255	it by calling C<ev_async_send>, which is thread- and signal safe.
2945	safe.
2946		3256
2947	This functionality is very similar to C<ev_signal> watchers, as signals,	3257	This functionality is very similar to C<ev_signal> watchers, as signals,
2948	too, are asynchronous in nature, and signals, too, will be compressed	3258	too, are asynchronous in nature, and signals, too, will be compressed
2949	(i.e. the number of callback invocations may be less than the number of	3259	(i.e. the number of callback invocations may be less than the number of
2950	C<ev_async_sent> calls).	3260	C<ev_async_sent> calls). In fact, you could use signal watchers as a kind
		3261	of "global async watchers" by using a watcher on an otherwise unused
		3262	signal, and C<ev_feed_signal> to signal this watcher from another thread,
		3263	even without knowing which loop owns the signal.
2951		3264
2952	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not	3265	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
2953	just the default loop.	3266	just the default loop.
2954		3267
2955	=head3 Queueing	3268	=head3 Queueing
2956		3269
2957	C<ev_async> does not support queueing of data in any way. The reason	3270	C<ev_async> does not support queueing of data in any way. The reason
2958	is that the author does not know of a simple (or any) algorithm for a	3271	is that the author does not know of a simple (or any) algorithm for a
2959	multiple-writer-single-reader queue that works in all cases and doesn't	3272	multiple-writer-single-reader queue that works in all cases and doesn't
2960	need elaborate support such as pthreads.	3273	need elaborate support such as pthreads or unportable memory access
		3274	semantics.
2961		3275
2962	That means that if you want to queue data, you have to provide your own	3276	That means that if you want to queue data, you have to provide your own
2963	queue. But at least I can tell you how to implement locking around your	3277	queue. But at least I can tell you how to implement locking around your
2964	queue:	3278	queue:
2965		3279
…		…
3104		3418
3105	If C<timeout> is less than 0, then no timeout watcher will be	3419	If C<timeout> is less than 0, then no timeout watcher will be
3106	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3420	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
3107	repeat = 0) will be started. C<0> is a valid timeout.	3421	repeat = 0) will be started. C<0> is a valid timeout.
3108		3422
3109	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3423	The callback has the type C<void (*cb)(int revents, void *arg)> and is
3110	passed an C<revents> set like normal event callbacks (a combination of	3424	passed an C<revents> set like normal event callbacks (a combination of
3111	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3425	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
3112	value passed to C<ev_once>. Note that it is possible to receive I<both>	3426	value passed to C<ev_once>. Note that it is possible to receive I<both>
3113	a timeout and an io event at the same time - you probably should give io	3427	a timeout and an io event at the same time - you probably should give io
3114	events precedence.	3428	events precedence.
3115		3429
3116	Example: wait up to ten seconds for data to appear on STDIN_FILENO.	3430	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
3117		3431
3118	static void stdin_ready (int revents, void *arg)	3432	static void stdin_ready (int revents, void *arg)
3119	{	3433	{
3120	if (revents & EV_READ)	3434	if (revents & EV_READ)
3121	/* stdin might have data for us, joy! */;	3435	/* stdin might have data for us, joy! */;
3122	else if (revents & EV_TIMEOUT)	3436	else if (revents & EV_TIMER)
3123	/* doh, nothing entered */;	3437	/* doh, nothing entered */;
3124	}	3438	}
3125		3439
3126	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3440	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3127		3441
3128	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
3129
3130	Feeds the given event set into the event loop, as if the specified event
3131	had happened for the specified watcher (which must be a pointer to an
3132	initialised but not necessarily started event watcher).
3133
3134	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)	3442	=item ev_feed_fd_event (loop, int fd, int revents)
3135		3443
3136	Feed an event on the given fd, as if a file descriptor backend detected	3444	Feed an event on the given fd, as if a file descriptor backend detected
3137	the given events it.	3445	the given events it.
3138		3446
3139	=item ev_feed_signal_event (struct ev_loop *loop, int signum)	3447	=item ev_feed_signal_event (loop, int signum)
3140		3448
3141	Feed an event as if the given signal occurred (C<loop> must be the default	3449	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3142	loop!).	3450	which is async-safe.
		3451
		3452	=back
		3453
		3454
		3455	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3456
		3457	This section explains some common idioms that are not immediately
		3458	obvious. Note that examples are sprinkled over the whole manual, and this
		3459	section only contains stuff that wouldn't fit anywhere else.
		3460
		3461	=over 4
		3462
		3463	=item Model/nested event loop invocations and exit conditions.
		3464
		3465	Often (especially in GUI toolkits) there are places where you have
		3466	I<modal> interaction, which is most easily implemented by recursively
		3467	invoking C<ev_run>.
		3468
		3469	This brings the problem of exiting - a callback might want to finish the
		3470	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3471	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3472	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3473	other combination: In these cases, C<ev_break> will not work alone.
		3474
		3475	The solution is to maintain "break this loop" variable for each C<ev_run>
		3476	invocation, and use a loop around C<ev_run> until the condition is
		3477	triggered, using C<EVRUN_ONCE>:
		3478
		3479	// main loop
		3480	int exit_main_loop = 0;
		3481
		3482	while (!exit_main_loop)
		3483	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3484
		3485	// in a model watcher
		3486	int exit_nested_loop = 0;
		3487
		3488	while (!exit_nested_loop)
		3489	ev_run (EV_A_ EVRUN_ONCE);
		3490
		3491	To exit from any of these loops, just set the corresponding exit variable:
		3492
		3493	// exit modal loop
		3494	exit_nested_loop = 1;
		3495
		3496	// exit main program, after modal loop is finished
		3497	exit_main_loop = 1;
		3498
		3499	// exit both
		3500	exit_main_loop = exit_nested_loop = 1;
		3501
		3502	=item Thread locking example
		3503
		3504	Here is a fictitious example of how to run an event loop in a different
		3505	thread than where callbacks are being invoked and watchers are
		3506	created/added/removed.
		3507
		3508	For a real-world example, see the C<EV::Loop::Async> perl module,
		3509	which uses exactly this technique (which is suited for many high-level
		3510	languages).
		3511
		3512	The example uses a pthread mutex to protect the loop data, a condition
		3513	variable to wait for callback invocations, an async watcher to notify the
		3514	event loop thread and an unspecified mechanism to wake up the main thread.
		3515
		3516	First, you need to associate some data with the event loop:
		3517
		3518	typedef struct {
		3519	mutex_t lock; /* global loop lock */
		3520	ev_async async_w;
		3521	thread_t tid;
		3522	cond_t invoke_cv;
		3523	} userdata;
		3524
		3525	void prepare_loop (EV_P)
		3526	{
		3527	// for simplicity, we use a static userdata struct.
		3528	static userdata u;
		3529
		3530	ev_async_init (&u->async_w, async_cb);
		3531	ev_async_start (EV_A_ &u->async_w);
		3532
		3533	pthread_mutex_init (&u->lock, 0);
		3534	pthread_cond_init (&u->invoke_cv, 0);
		3535
		3536	// now associate this with the loop
		3537	ev_set_userdata (EV_A_ u);
		3538	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3539	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3540
		3541	// then create the thread running ev_loop
		3542	pthread_create (&u->tid, 0, l_run, EV_A);
		3543	}
		3544
		3545	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3546	solely to wake up the event loop so it takes notice of any new watchers
		3547	that might have been added:
		3548
		3549	static void
		3550	async_cb (EV_P_ ev_async *w, int revents)
		3551	{
		3552	// just used for the side effects
		3553	}
		3554
		3555	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3556	protecting the loop data, respectively.
		3557
		3558	static void
		3559	l_release (EV_P)
		3560	{
		3561	userdata *u = ev_userdata (EV_A);
		3562	pthread_mutex_unlock (&u->lock);
		3563	}
		3564
		3565	static void
		3566	l_acquire (EV_P)
		3567	{
		3568	userdata *u = ev_userdata (EV_A);
		3569	pthread_mutex_lock (&u->lock);
		3570	}
		3571
		3572	The event loop thread first acquires the mutex, and then jumps straight
		3573	into C<ev_run>:
		3574
		3575	void *
		3576	l_run (void *thr_arg)
		3577	{
		3578	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3579
		3580	l_acquire (EV_A);
		3581	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3582	ev_run (EV_A_ 0);
		3583	l_release (EV_A);
		3584
		3585	return 0;
		3586	}
		3587
		3588	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3589	signal the main thread via some unspecified mechanism (signals? pipe
		3590	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3591	have been called (in a while loop because a) spurious wakeups are possible
		3592	and b) skipping inter-thread-communication when there are no pending
		3593	watchers is very beneficial):
		3594
		3595	static void
		3596	l_invoke (EV_P)
		3597	{
		3598	userdata *u = ev_userdata (EV_A);
		3599
		3600	while (ev_pending_count (EV_A))
		3601	{
		3602	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3603	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3604	}
		3605	}
		3606
		3607	Now, whenever the main thread gets told to invoke pending watchers, it
		3608	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3609	thread to continue:
		3610
		3611	static void
		3612	real_invoke_pending (EV_P)
		3613	{
		3614	userdata *u = ev_userdata (EV_A);
		3615
		3616	pthread_mutex_lock (&u->lock);
		3617	ev_invoke_pending (EV_A);
		3618	pthread_cond_signal (&u->invoke_cv);
		3619	pthread_mutex_unlock (&u->lock);
		3620	}
		3621
		3622	Whenever you want to start/stop a watcher or do other modifications to an
		3623	event loop, you will now have to lock:
		3624
		3625	ev_timer timeout_watcher;
		3626	userdata *u = ev_userdata (EV_A);
		3627
		3628	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3629
		3630	pthread_mutex_lock (&u->lock);
		3631	ev_timer_start (EV_A_ &timeout_watcher);
		3632	ev_async_send (EV_A_ &u->async_w);
		3633	pthread_mutex_unlock (&u->lock);
		3634
		3635	Note that sending the C<ev_async> watcher is required because otherwise
		3636	an event loop currently blocking in the kernel will have no knowledge
		3637	about the newly added timer. By waking up the loop it will pick up any new
		3638	watchers in the next event loop iteration.
3143		3639
3144	=back	3640	=back
3145		3641
3146		3642
3147	=head1 LIBEVENT EMULATION	3643	=head1 LIBEVENT EMULATION
3148		3644
3149	Libev offers a compatibility emulation layer for libevent. It cannot	3645	Libev offers a compatibility emulation layer for libevent. It cannot
3150	emulate the internals of libevent, so here are some usage hints:	3646	emulate the internals of libevent, so here are some usage hints:
3151		3647
3152	=over 4	3648	=over 4
		3649
		3650	=item * Only the libevent-1.4.1-beta API is being emulated.
		3651
		3652	This was the newest libevent version available when libev was implemented,
		3653	and is still mostly unchanged in 2010.
3153		3654
3154	=item * Use it by including <event.h>, as usual.	3655	=item * Use it by including <event.h>, as usual.
3155		3656
3156	=item * The following members are fully supported: ev_base, ev_callback,	3657	=item * The following members are fully supported: ev_base, ev_callback,
3157	ev_arg, ev_fd, ev_res, ev_events.	3658	ev_arg, ev_fd, ev_res, ev_events.
…		…
3163	=item * Priorities are not currently supported. Initialising priorities	3664	=item * Priorities are not currently supported. Initialising priorities
3164	will fail and all watchers will have the same priority, even though there	3665	will fail and all watchers will have the same priority, even though there
3165	is an ev_pri field.	3666	is an ev_pri field.
3166		3667
3167	=item * In libevent, the last base created gets the signals, in libev, the	3668	=item * In libevent, the last base created gets the signals, in libev, the
3168	first base created (== the default loop) gets the signals.	3669	base that registered the signal gets the signals.
3169		3670
3170	=item * Other members are not supported.	3671	=item * Other members are not supported.
3171		3672
3172	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3673	=item * The libev emulation is I<not> ABI compatible to libevent, you need
3173	to use the libev header file and library.	3674	to use the libev header file and library.
…		…
3192	Care has been taken to keep the overhead low. The only data member the C++	3693	Care has been taken to keep the overhead low. The only data member the C++
3193	classes add (compared to plain C-style watchers) is the event loop pointer	3694	classes add (compared to plain C-style watchers) is the event loop pointer
3194	that the watcher is associated with (or no additional members at all if	3695	that the watcher is associated with (or no additional members at all if
3195	you disable C<EV_MULTIPLICITY> when embedding libev).	3696	you disable C<EV_MULTIPLICITY> when embedding libev).
3196		3697
3197	Currently, functions, and static and non-static member functions can be	3698	Currently, functions, static and non-static member functions and classes
3198	used as callbacks. Other types should be easy to add as long as they only	3699	with C<operator ()> can be used as callbacks. Other types should be easy
3199	need one additional pointer for context. If you need support for other	3700	to add as long as they only need one additional pointer for context. If
3200	types of functors please contact the author (preferably after implementing	3701	you need support for other types of functors please contact the author
3201	it).	3702	(preferably after implementing it).
3202		3703
3203	Here is a list of things available in the C<ev> namespace:	3704	Here is a list of things available in the C<ev> namespace:
3204		3705
3205	=over 4	3706	=over 4
3206		3707
…		…
3224		3725
3225	=over 4	3726	=over 4
3226		3727
3227	=item ev::TYPE::TYPE ()	3728	=item ev::TYPE::TYPE ()
3228		3729
3229	=item ev::TYPE::TYPE (struct ev_loop *)	3730	=item ev::TYPE::TYPE (loop)
3230		3731
3231	=item ev::TYPE::~TYPE	3732	=item ev::TYPE::~TYPE
3232		3733
3233	The constructor (optionally) takes an event loop to associate the watcher	3734	The constructor (optionally) takes an event loop to associate the watcher
3234	with. If it is omitted, it will use C<EV_DEFAULT>.	3735	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
3267	myclass obj;	3768	myclass obj;
3268	ev::io iow;	3769	ev::io iow;
3269	iow.set <myclass, &myclass::io_cb> (&obj);	3770	iow.set <myclass, &myclass::io_cb> (&obj);
3270		3771
3271	=item w->set (object *)	3772	=item w->set (object *)
3272
3273	This is an B<experimental> feature that might go away in a future version.
3274		3773
3275	This is a variation of a method callback - leaving out the method to call	3774	This is a variation of a method callback - leaving out the method to call
3276	will default the method to C<operator ()>, which makes it possible to use	3775	will default the method to C<operator ()>, which makes it possible to use
3277	functor objects without having to manually specify the C<operator ()> all	3776	functor objects without having to manually specify the C<operator ()> all
3278	the time. Incidentally, you can then also leave out the template argument	3777	the time. Incidentally, you can then also leave out the template argument
…		…
3311	Example: Use a plain function as callback.	3810	Example: Use a plain function as callback.
3312		3811
3313	static void io_cb (ev::io &w, int revents) { }	3812	static void io_cb (ev::io &w, int revents) { }
3314	iow.set <io_cb> ();	3813	iow.set <io_cb> ();
3315		3814
3316	=item w->set (struct ev_loop *)	3815	=item w->set (loop)
3317		3816
3318	Associates a different C<struct ev_loop> with this watcher. You can only	3817	Associates a different C<struct ev_loop> with this watcher. You can only
3319	do this when the watcher is inactive (and not pending either).	3818	do this when the watcher is inactive (and not pending either).
3320		3819
3321	=item w->set ([arguments])	3820	=item w->set ([arguments])
3322		3821
3323	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be	3822	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
3324	called at least once. Unlike the C counterpart, an active watcher gets	3823	method or a suitable start method must be called at least once. Unlike the
3325	automatically stopped and restarted when reconfiguring it with this	3824	C counterpart, an active watcher gets automatically stopped and restarted
3326	method.	3825	when reconfiguring it with this method.
3327		3826
3328	=item w->start ()	3827	=item w->start ()
3329		3828
3330	Starts the watcher. Note that there is no C<loop> argument, as the	3829	Starts the watcher. Note that there is no C<loop> argument, as the
3331	constructor already stores the event loop.	3830	constructor already stores the event loop.
3332		3831
		3832	=item w->start ([arguments])
		3833
		3834	Instead of calling C<set> and C<start> methods separately, it is often
		3835	convenient to wrap them in one call. Uses the same type of arguments as
		3836	the configure C<set> method of the watcher.
		3837
3333	=item w->stop ()	3838	=item w->stop ()
3334		3839
3335	Stops the watcher if it is active. Again, no C<loop> argument.	3840	Stops the watcher if it is active. Again, no C<loop> argument.
3336		3841
3337	=item w->again () (C<ev::timer>, C<ev::periodic> only)	3842	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
3349		3854
3350	=back	3855	=back
3351		3856
3352	=back	3857	=back
3353		3858
3354	Example: Define a class with an IO and idle watcher, start one of them in	3859	Example: Define a class with two I/O and idle watchers, start the I/O
3355	the constructor.	3860	watchers in the constructor.
3356		3861
3357	class myclass	3862	class myclass
3358	{	3863	{
3359	ev::io io ; void io_cb (ev::io &w, int revents);	3864	ev::io io ; void io_cb (ev::io &w, int revents);
		3865	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);
3360	ev::idle idle; void idle_cb (ev::idle &w, int revents);	3866	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3361		3867
3362	myclass (int fd)	3868	myclass (int fd)
3363	{	3869	{
3364	io .set <myclass, &myclass::io_cb > (this);	3870	io .set <myclass, &myclass::io_cb > (this);
		3871	io2 .set <myclass, &myclass::io2_cb > (this);
3365	idle.set <myclass, &myclass::idle_cb> (this);	3872	idle.set <myclass, &myclass::idle_cb> (this);
3366		3873
3367	io.start (fd, ev::READ);	3874	io.set (fd, ev::WRITE); // configure the watcher
		3875	io.start (); // start it whenever convenient
		3876
		3877	io2.start (fd, ev::READ); // set + start in one call
3368	}	3878	}
3369	};	3879	};
3370		3880
3371		3881
3372	=head1 OTHER LANGUAGE BINDINGS	3882	=head1 OTHER LANGUAGE BINDINGS
…		…
3420	Erkki Seppala has written Ocaml bindings for libev, to be found at	3930	Erkki Seppala has written Ocaml bindings for libev, to be found at
3421	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	3931	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
3422		3932
3423	=item Lua	3933	=item Lua
3424		3934
3425	Brian Maher has written a partial interface to libev	3935	Brian Maher has written a partial interface to libev for lua (at the
3426	for lua (only C<ev_io> and C<ev_timer>), to be found at	3936	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
3427	L<http://github.com/brimworks/lua-ev>.	3937	L<http://github.com/brimworks/lua-ev>.
3428		3938
3429	=back	3939	=back
3430		3940
3431		3941
…		…
3446	loop argument"). The C<EV_A> form is used when this is the sole argument,	3956	loop argument"). The C<EV_A> form is used when this is the sole argument,
3447	C<EV_A_> is used when other arguments are following. Example:	3957	C<EV_A_> is used when other arguments are following. Example:
3448		3958
3449	ev_unref (EV_A);	3959	ev_unref (EV_A);
3450	ev_timer_add (EV_A_ watcher);	3960	ev_timer_add (EV_A_ watcher);
3451	ev_loop (EV_A_ 0);	3961	ev_run (EV_A_ 0);
3452		3962
3453	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	3963	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
3454	which is often provided by the following macro.	3964	which is often provided by the following macro.
3455		3965
3456	=item C<EV_P>, C<EV_P_>	3966	=item C<EV_P>, C<EV_P_>
…		…
3496	}	4006	}
3497		4007
3498	ev_check check;	4008	ev_check check;
3499	ev_check_init (&check, check_cb);	4009	ev_check_init (&check, check_cb);
3500	ev_check_start (EV_DEFAULT_ &check);	4010	ev_check_start (EV_DEFAULT_ &check);
3501	ev_loop (EV_DEFAULT_ 0);	4011	ev_run (EV_DEFAULT_ 0);
3502		4012
3503	=head1 EMBEDDING	4013	=head1 EMBEDDING
3504		4014
3505	Libev can (and often is) directly embedded into host	4015	Libev can (and often is) directly embedded into host
3506	applications. Examples of applications that embed it include the Deliantra	4016	applications. Examples of applications that embed it include the Deliantra
…		…
3586	libev.m4	4096	libev.m4
3587		4097
3588	=head2 PREPROCESSOR SYMBOLS/MACROS	4098	=head2 PREPROCESSOR SYMBOLS/MACROS
3589		4099
3590	Libev can be configured via a variety of preprocessor symbols you have to	4100	Libev can be configured via a variety of preprocessor symbols you have to
3591	define before including any of its files. The default in the absence of	4101	define before including (or compiling) any of its files. The default in
3592	autoconf is documented for every option.	4102	the absence of autoconf is documented for every option.
		4103
		4104	Symbols marked with "(h)" do not change the ABI, and can have different
		4105	values when compiling libev vs. including F<ev.h>, so it is permissible
		4106	to redefine them before including F<ev.h> without breaking compatibility
		4107	to a compiled library. All other symbols change the ABI, which means all
		4108	users of libev and the libev code itself must be compiled with compatible
		4109	settings.
3593		4110
3594	=over 4	4111	=over 4
3595		4112
		4113	=item EV_COMPAT3 (h)
		4114
		4115	Backwards compatibility is a major concern for libev. This is why this
		4116	release of libev comes with wrappers for the functions and symbols that
		4117	have been renamed between libev version 3 and 4.
		4118
		4119	You can disable these wrappers (to test compatibility with future
		4120	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4121	sources. This has the additional advantage that you can drop the C<struct>
		4122	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4123	typedef in that case.
		4124
		4125	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4126	and in some even more future version the compatibility code will be
		4127	removed completely.
		4128
3596	=item EV_STANDALONE	4129	=item EV_STANDALONE (h)
3597		4130
3598	Must always be C<1> if you do not use autoconf configuration, which	4131	Must always be C<1> if you do not use autoconf configuration, which
3599	keeps libev from including F<config.h>, and it also defines dummy	4132	keeps libev from including F<config.h>, and it also defines dummy
3600	implementations for some libevent functions (such as logging, which is not	4133	implementations for some libevent functions (such as logging, which is not
3601	supported). It will also not define any of the structs usually found in	4134	supported). It will also not define any of the structs usually found in
…		…
3751	as well as for signal and thread safety in C<ev_async> watchers.	4284	as well as for signal and thread safety in C<ev_async> watchers.
3752		4285
3753	In the absence of this define, libev will use C<sig_atomic_t volatile>	4286	In the absence of this define, libev will use C<sig_atomic_t volatile>
3754	(from F<signal.h>), which is usually good enough on most platforms.	4287	(from F<signal.h>), which is usually good enough on most platforms.
3755		4288
3756	=item EV_H	4289	=item EV_H (h)
3757		4290
3758	The name of the F<ev.h> header file used to include it. The default if	4291	The name of the F<ev.h> header file used to include it. The default if
3759	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	4292	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
3760	used to virtually rename the F<ev.h> header file in case of conflicts.	4293	used to virtually rename the F<ev.h> header file in case of conflicts.
3761		4294
3762	=item EV_CONFIG_H	4295	=item EV_CONFIG_H (h)
3763		4296
3764	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	4297	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
3765	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	4298	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
3766	C<EV_H>, above.	4299	C<EV_H>, above.
3767		4300
3768	=item EV_EVENT_H	4301	=item EV_EVENT_H (h)
3769		4302
3770	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	4303	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
3771	of how the F<event.h> header can be found, the default is C<"event.h">.	4304	of how the F<event.h> header can be found, the default is C<"event.h">.
3772		4305
3773	=item EV_PROTOTYPES	4306	=item EV_PROTOTYPES (h)
3774		4307
3775	If defined to be C<0>, then F<ev.h> will not define any function	4308	If defined to be C<0>, then F<ev.h> will not define any function
3776	prototypes, but still define all the structs and other symbols. This is	4309	prototypes, but still define all the structs and other symbols. This is
3777	occasionally useful if you want to provide your own wrapper functions	4310	occasionally useful if you want to provide your own wrapper functions
3778	around libev functions.	4311	around libev functions.
…		…
3800	fine.	4333	fine.
3801		4334
3802	If your embedding application does not need any priorities, defining these	4335	If your embedding application does not need any priorities, defining these
3803	both to C<0> will save some memory and CPU.	4336	both to C<0> will save some memory and CPU.
3804		4337
3805	=item EV_PERIODIC_ENABLE	4338	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4339	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4340	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3806		4341
3807	If undefined or defined to be C<1>, then periodic timers are supported. If	4342	If undefined or defined to be C<1> (and the platform supports it), then
3808	defined to be C<0>, then they are not. Disabling them saves a few kB of	4343	the respective watcher type is supported. If defined to be C<0>, then it
3809	code.	4344	is not. Disabling watcher types mainly saves code size.
3810		4345
3811	=item EV_IDLE_ENABLE	4346	=item EV_FEATURES
3812
3813	If undefined or defined to be C<1>, then idle watchers are supported. If
3814	defined to be C<0>, then they are not. Disabling them saves a few kB of
3815	code.
3816
3817	=item EV_EMBED_ENABLE
3818
3819	If undefined or defined to be C<1>, then embed watchers are supported. If
3820	defined to be C<0>, then they are not. Embed watchers rely on most other
3821	watcher types, which therefore must not be disabled.
3822
3823	=item EV_STAT_ENABLE
3824
3825	If undefined or defined to be C<1>, then stat watchers are supported. If
3826	defined to be C<0>, then they are not.
3827
3828	=item EV_FORK_ENABLE
3829
3830	If undefined or defined to be C<1>, then fork watchers are supported. If
3831	defined to be C<0>, then they are not.
3832
3833	=item EV_ASYNC_ENABLE
3834
3835	If undefined or defined to be C<1>, then async watchers are supported. If
3836	defined to be C<0>, then they are not.
3837
3838	=item EV_MINIMAL
3839		4347
3840	If you need to shave off some kilobytes of code at the expense of some	4348	If you need to shave off some kilobytes of code at the expense of some
3841	speed (but with the full API), define this symbol to C<1>. Currently this	4349	speed (but with the full API), you can define this symbol to request
3842	is used to override some inlining decisions, saves roughly 30% code size	4350	certain subsets of functionality. The default is to enable all features
3843	on amd64. It also selects a much smaller 2-heap for timer management over	4351	that can be enabled on the platform.
3844	the default 4-heap.
3845		4352
3846	You can save even more by disabling watcher types you do not need	4353	A typical way to use this symbol is to define it to C<0> (or to a bitset
3847	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>	4354	with some broad features you want) and then selectively re-enable
3848	(C<-DNDEBUG>) will usually reduce code size a lot.	4355	additional parts you want, for example if you want everything minimal,
		4356	but multiple event loop support, async and child watchers and the poll
		4357	backend, use this:
3849		4358
3850	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to	4359	#define EV_FEATURES 0
3851	provide a bare-bones event library. See C<ev.h> for details on what parts	4360	#define EV_MULTIPLICITY 1
3852	of the API are still available, and do not complain if this subset changes	4361	#define EV_USE_POLL 1
3853	over time.	4362	#define EV_CHILD_ENABLE 1
		4363	#define EV_ASYNC_ENABLE 1
		4364
		4365	The actual value is a bitset, it can be a combination of the following
		4366	values:
		4367
		4368	=over 4
		4369
		4370	=item C<1> - faster/larger code
		4371
		4372	Use larger code to speed up some operations.
		4373
		4374	Currently this is used to override some inlining decisions (enlarging the
		4375	code size by roughly 30% on amd64).
		4376
		4377	When optimising for size, use of compiler flags such as C<-Os> with
		4378	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4379	assertions.
		4380
		4381	=item C<2> - faster/larger data structures
		4382
		4383	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4384	hash table sizes and so on. This will usually further increase code size
		4385	and can additionally have an effect on the size of data structures at
		4386	runtime.
		4387
		4388	=item C<4> - full API configuration
		4389
		4390	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4391	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4392
		4393	=item C<8> - full API
		4394
		4395	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4396	details on which parts of the API are still available without this
		4397	feature, and do not complain if this subset changes over time.
		4398
		4399	=item C<16> - enable all optional watcher types
		4400
		4401	Enables all optional watcher types. If you want to selectively enable
		4402	only some watcher types other than I/O and timers (e.g. prepare,
		4403	embed, async, child...) you can enable them manually by defining
		4404	C<EV_watchertype_ENABLE> to C<1> instead.
		4405
		4406	=item C<32> - enable all backends
		4407
		4408	This enables all backends - without this feature, you need to enable at
		4409	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4410
		4411	=item C<64> - enable OS-specific "helper" APIs
		4412
		4413	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4414	default.
		4415
		4416	=back
		4417
		4418	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4419	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4420	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4421	watchers, timers and monotonic clock support.
		4422
		4423	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4424	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4425	your program might be left out as well - a binary starting a timer and an
		4426	I/O watcher then might come out at only 5Kb.
		4427
		4428	=item EV_AVOID_STDIO
		4429
		4430	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4431	functions (printf, scanf, perror etc.). This will increase the code size
		4432	somewhat, but if your program doesn't otherwise depend on stdio and your
		4433	libc allows it, this avoids linking in the stdio library which is quite
		4434	big.
		4435
		4436	Note that error messages might become less precise when this option is
		4437	enabled.
3854		4438
3855	=item EV_NSIG	4439	=item EV_NSIG
3856		4440
3857	The highest supported signal number, +1 (or, the number of	4441	The highest supported signal number, +1 (or, the number of
3858	signals): Normally, libev tries to deduce the maximum number of signals	4442	signals): Normally, libev tries to deduce the maximum number of signals
3859	automatically, but sometimes this fails, in which case it can be	4443	automatically, but sometimes this fails, in which case it can be
3860	specified. Also, using a lower number than detected (C<32> should be	4444	specified. Also, using a lower number than detected (C<32> should be
3861	good for about any system in existance) can save some memory, as libev	4445	good for about any system in existence) can save some memory, as libev
3862	statically allocates some 12-24 bytes per signal number.	4446	statically allocates some 12-24 bytes per signal number.
3863		4447
3864	=item EV_PID_HASHSIZE	4448	=item EV_PID_HASHSIZE
3865		4449
3866	C<ev_child> watchers use a small hash table to distribute workload by	4450	C<ev_child> watchers use a small hash table to distribute workload by
3867	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	4451	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3868	than enough. If you need to manage thousands of children you might want to	4452	usually more than enough. If you need to manage thousands of children you
3869	increase this value (I<must> be a power of two).	4453	might want to increase this value (I<must> be a power of two).
3870		4454
3871	=item EV_INOTIFY_HASHSIZE	4455	=item EV_INOTIFY_HASHSIZE
3872		4456
3873	C<ev_stat> watchers use a small hash table to distribute workload by	4457	C<ev_stat> watchers use a small hash table to distribute workload by
3874	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4458	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3875	usually more than enough. If you need to manage thousands of C<ev_stat>	4459	disabled), usually more than enough. If you need to manage thousands of
3876	watchers you might want to increase this value (I<must> be a power of	4460	C<ev_stat> watchers you might want to increase this value (I<must> be a
3877	two).	4461	power of two).
3878		4462
3879	=item EV_USE_4HEAP	4463	=item EV_USE_4HEAP
3880		4464
3881	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4465	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3882	timer and periodics heaps, libev uses a 4-heap when this symbol is defined	4466	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3883	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably	4467	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3884	faster performance with many (thousands) of watchers.	4468	faster performance with many (thousands) of watchers.
3885		4469
3886	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4470	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3887	(disabled).	4471	will be C<0>.
3888		4472
3889	=item EV_HEAP_CACHE_AT	4473	=item EV_HEAP_CACHE_AT
3890		4474
3891	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4475	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3892	timer and periodics heaps, libev can cache the timestamp (I<at>) within	4476	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3893	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4477	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3894	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4478	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3895	but avoids random read accesses on heap changes. This improves performance	4479	but avoids random read accesses on heap changes. This improves performance
3896	noticeably with many (hundreds) of watchers.	4480	noticeably with many (hundreds) of watchers.
3897		4481
3898	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4482	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3899	(disabled).	4483	will be C<0>.
3900		4484
3901	=item EV_VERIFY	4485	=item EV_VERIFY
3902		4486
3903	Controls how much internal verification (see C<ev_loop_verify ()>) will	4487	Controls how much internal verification (see C<ev_verify ()>) will
3904	be done: If set to C<0>, no internal verification code will be compiled	4488	be done: If set to C<0>, no internal verification code will be compiled
3905	in. If set to C<1>, then verification code will be compiled in, but not	4489	in. If set to C<1>, then verification code will be compiled in, but not
3906	called. If set to C<2>, then the internal verification code will be	4490	called. If set to C<2>, then the internal verification code will be
3907	called once per loop, which can slow down libev. If set to C<3>, then the	4491	called once per loop, which can slow down libev. If set to C<3>, then the
3908	verification code will be called very frequently, which will slow down	4492	verification code will be called very frequently, which will slow down
3909	libev considerably.	4493	libev considerably.
3910		4494
3911	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4495	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3912	C<0>.	4496	will be C<0>.
3913		4497
3914	=item EV_COMMON	4498	=item EV_COMMON
3915		4499
3916	By default, all watchers have a C<void *data> member. By redefining	4500	By default, all watchers have a C<void *data> member. By redefining
3917	this macro to a something else you can include more and other types of	4501	this macro to something else you can include more and other types of
3918	members. You have to define it each time you include one of the files,	4502	members. You have to define it each time you include one of the files,
3919	though, and it must be identical each time.	4503	though, and it must be identical each time.
3920		4504
3921	For example, the perl EV module uses something like this:	4505	For example, the perl EV module uses something like this:
3922		4506
…		…
3975	file.	4559	file.
3976		4560
3977	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4561	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3978	that everybody includes and which overrides some configure choices:	4562	that everybody includes and which overrides some configure choices:
3979		4563
3980	#define EV_MINIMAL 1	4564	#define EV_FEATURES 8
3981	#define EV_USE_POLL 0	4565	#define EV_USE_SELECT 1
3982	#define EV_MULTIPLICITY 0
3983	#define EV_PERIODIC_ENABLE 0	4566	#define EV_PREPARE_ENABLE 1
		4567	#define EV_IDLE_ENABLE 1
3984	#define EV_STAT_ENABLE 0	4568	#define EV_SIGNAL_ENABLE 1
3985	#define EV_FORK_ENABLE 0	4569	#define EV_CHILD_ENABLE 1
		4570	#define EV_USE_STDEXCEPT 0
3986	#define EV_CONFIG_H <config.h>	4571	#define EV_CONFIG_H <config.h>
3987	#define EV_MINPRI 0
3988	#define EV_MAXPRI 0
3989		4572
3990	#include "ev++.h"	4573	#include "ev++.h"
3991		4574
3992	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4575	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3993		4576
…		…
4053	default loop and triggering an C<ev_async> watcher from the default loop	4636	default loop and triggering an C<ev_async> watcher from the default loop
4054	watcher callback into the event loop interested in the signal.	4637	watcher callback into the event loop interested in the signal.
4055		4638
4056	=back	4639	=back
4057		4640
4058	=head4 THREAD LOCKING EXAMPLE	4641	See also L<Thread locking example>.
4059
4060	Here is a fictitious example of how to run an event loop in a different
4061	thread than where callbacks are being invoked and watchers are
4062	created/added/removed.
4063
4064	For a real-world example, see the C<EV::Loop::Async> perl module,
4065	which uses exactly this technique (which is suited for many high-level
4066	languages).
4067
4068	The example uses a pthread mutex to protect the loop data, a condition
4069	variable to wait for callback invocations, an async watcher to notify the
4070	event loop thread and an unspecified mechanism to wake up the main thread.
4071
4072	First, you need to associate some data with the event loop:
4073
4074	typedef struct {
4075	mutex_t lock; /* global loop lock */
4076	ev_async async_w;
4077	thread_t tid;
4078	cond_t invoke_cv;
4079	} userdata;
4080
4081	void prepare_loop (EV_P)
4082	{
4083	// for simplicity, we use a static userdata struct.
4084	static userdata u;
4085
4086	ev_async_init (&u->async_w, async_cb);
4087	ev_async_start (EV_A_ &u->async_w);
4088
4089	pthread_mutex_init (&u->lock, 0);
4090	pthread_cond_init (&u->invoke_cv, 0);
4091
4092	// now associate this with the loop
4093	ev_set_userdata (EV_A_ u);
4094	ev_set_invoke_pending_cb (EV_A_ l_invoke);
4095	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4096
4097	// then create the thread running ev_loop
4098	pthread_create (&u->tid, 0, l_run, EV_A);
4099	}
4100
4101	The callback for the C<ev_async> watcher does nothing: the watcher is used
4102	solely to wake up the event loop so it takes notice of any new watchers
4103	that might have been added:
4104
4105	static void
4106	async_cb (EV_P_ ev_async *w, int revents)
4107	{
4108	// just used for the side effects
4109	}
4110
4111	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4112	protecting the loop data, respectively.
4113
4114	static void
4115	l_release (EV_P)
4116	{
4117	userdata *u = ev_userdata (EV_A);
4118	pthread_mutex_unlock (&u->lock);
4119	}
4120
4121	static void
4122	l_acquire (EV_P)
4123	{
4124	userdata *u = ev_userdata (EV_A);
4125	pthread_mutex_lock (&u->lock);
4126	}
4127
4128	The event loop thread first acquires the mutex, and then jumps straight
4129	into C<ev_loop>:
4130
4131	void *
4132	l_run (void *thr_arg)
4133	{
4134	struct ev_loop *loop = (struct ev_loop *)thr_arg;
4135
4136	l_acquire (EV_A);
4137	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4138	ev_loop (EV_A_ 0);
4139	l_release (EV_A);
4140
4141	return 0;
4142	}
4143
4144	Instead of invoking all pending watchers, the C<l_invoke> callback will
4145	signal the main thread via some unspecified mechanism (signals? pipe
4146	writes? C<Async::Interrupt>?) and then waits until all pending watchers
4147	have been called (in a while loop because a) spurious wakeups are possible
4148	and b) skipping inter-thread-communication when there are no pending
4149	watchers is very beneficial):
4150
4151	static void
4152	l_invoke (EV_P)
4153	{
4154	userdata *u = ev_userdata (EV_A);
4155
4156	while (ev_pending_count (EV_A))
4157	{
4158	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4159	pthread_cond_wait (&u->invoke_cv, &u->lock);
4160	}
4161	}
4162
4163	Now, whenever the main thread gets told to invoke pending watchers, it
4164	will grab the lock, call C<ev_invoke_pending> and then signal the loop
4165	thread to continue:
4166
4167	static void
4168	real_invoke_pending (EV_P)
4169	{
4170	userdata *u = ev_userdata (EV_A);
4171
4172	pthread_mutex_lock (&u->lock);
4173	ev_invoke_pending (EV_A);
4174	pthread_cond_signal (&u->invoke_cv);
4175	pthread_mutex_unlock (&u->lock);
4176	}
4177
4178	Whenever you want to start/stop a watcher or do other modifications to an
4179	event loop, you will now have to lock:
4180
4181	ev_timer timeout_watcher;
4182	userdata *u = ev_userdata (EV_A);
4183
4184	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4185
4186	pthread_mutex_lock (&u->lock);
4187	ev_timer_start (EV_A_ &timeout_watcher);
4188	ev_async_send (EV_A_ &u->async_w);
4189	pthread_mutex_unlock (&u->lock);
4190
4191	Note that sending the C<ev_async> watcher is required because otherwise
4192	an event loop currently blocking in the kernel will have no knowledge
4193	about the newly added timer. By waking up the loop it will pick up any new
4194	watchers in the next event loop iteration.
4195		4642
4196	=head3 COROUTINES	4643	=head3 COROUTINES
4197		4644
4198	Libev is very accommodating to coroutines ("cooperative threads"):	4645	Libev is very accommodating to coroutines ("cooperative threads"):
4199	libev fully supports nesting calls to its functions from different	4646	libev fully supports nesting calls to its functions from different
4200	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4647	coroutines (e.g. you can call C<ev_run> on the same loop from two
4201	different coroutines, and switch freely between both coroutines running	4648	different coroutines, and switch freely between both coroutines running
4202	the loop, as long as you don't confuse yourself). The only exception is	4649	the loop, as long as you don't confuse yourself). The only exception is
4203	that you must not do this from C<ev_periodic> reschedule callbacks.	4650	that you must not do this from C<ev_periodic> reschedule callbacks.
4204		4651
4205	Care has been taken to ensure that libev does not keep local state inside	4652	Care has been taken to ensure that libev does not keep local state inside
4206	C<ev_loop>, and other calls do not usually allow for coroutine switches as	4653	C<ev_run>, and other calls do not usually allow for coroutine switches as
4207	they do not call any callbacks.	4654	they do not call any callbacks.
4208		4655
4209	=head2 COMPILER WARNINGS	4656	=head2 COMPILER WARNINGS
4210		4657
4211	Depending on your compiler and compiler settings, you might get no or a	4658	Depending on your compiler and compiler settings, you might get no or a
…		…
4222	maintainable.	4669	maintainable.
4223		4670
4224	And of course, some compiler warnings are just plain stupid, or simply	4671	And of course, some compiler warnings are just plain stupid, or simply
4225	wrong (because they don't actually warn about the condition their message	4672	wrong (because they don't actually warn about the condition their message
4226	seems to warn about). For example, certain older gcc versions had some	4673	seems to warn about). For example, certain older gcc versions had some
4227	warnings that resulted an extreme number of false positives. These have	4674	warnings that resulted in an extreme number of false positives. These have
4228	been fixed, but some people still insist on making code warn-free with	4675	been fixed, but some people still insist on making code warn-free with
4229	such buggy versions.	4676	such buggy versions.
4230		4677
4231	While libev is written to generate as few warnings as possible,	4678	While libev is written to generate as few warnings as possible,
4232	"warn-free" code is not a goal, and it is recommended not to build libev	4679	"warn-free" code is not a goal, and it is recommended not to build libev
…		…
4268	I suggest using suppression lists.	4715	I suggest using suppression lists.
4269		4716
4270		4717
4271	=head1 PORTABILITY NOTES	4718	=head1 PORTABILITY NOTES
4272		4719
		4720	=head2 GNU/LINUX 32 BIT LIMITATIONS
		4721
		4722	GNU/Linux is the only common platform that supports 64 bit file/large file
		4723	interfaces but I<disables> them by default.
		4724
		4725	That means that libev compiled in the default environment doesn't support
		4726	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		4727
		4728	Unfortunately, many programs try to work around this GNU/Linux issue
		4729	by enabling the large file API, which makes them incompatible with the
		4730	standard libev compiled for their system.
		4731
		4732	Likewise, libev cannot enable the large file API itself as this would
		4733	suddenly make it incompatible to the default compile time environment,
		4734	i.e. all programs not using special compile switches.
		4735
		4736	=head2 OS/X AND DARWIN BUGS
		4737
		4738	The whole thing is a bug if you ask me - basically any system interface
		4739	you touch is broken, whether it is locales, poll, kqueue or even the
		4740	OpenGL drivers.
		4741
		4742	=head3 C<kqueue> is buggy
		4743
		4744	The kqueue syscall is broken in all known versions - most versions support
		4745	only sockets, many support pipes.
		4746
		4747	Libev tries to work around this by not using C<kqueue> by default on this
		4748	rotten platform, but of course you can still ask for it when creating a
		4749	loop - embedding a socket-only kqueue loop into a select-based one is
		4750	probably going to work well.
		4751
		4752	=head3 C<poll> is buggy
		4753
		4754	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4755	implementation by something calling C<kqueue> internally around the 10.5.6
		4756	release, so now C<kqueue> I<and> C<poll> are broken.
		4757
		4758	Libev tries to work around this by not using C<poll> by default on
		4759	this rotten platform, but of course you can still ask for it when creating
		4760	a loop.
		4761
		4762	=head3 C<select> is buggy
		4763
		4764	All that's left is C<select>, and of course Apple found a way to fuck this
		4765	one up as well: On OS/X, C<select> actively limits the number of file
		4766	descriptors you can pass in to 1024 - your program suddenly crashes when
		4767	you use more.
		4768
		4769	There is an undocumented "workaround" for this - defining
		4770	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		4771	work on OS/X.
		4772
		4773	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		4774
		4775	=head3 C<errno> reentrancy
		4776
		4777	The default compile environment on Solaris is unfortunately so
		4778	thread-unsafe that you can't even use components/libraries compiled
		4779	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		4780	defined by default. A valid, if stupid, implementation choice.
		4781
		4782	If you want to use libev in threaded environments you have to make sure
		4783	it's compiled with C<_REENTRANT> defined.
		4784
		4785	=head3 Event port backend
		4786
		4787	The scalable event interface for Solaris is called "event
		4788	ports". Unfortunately, this mechanism is very buggy in all major
		4789	releases. If you run into high CPU usage, your program freezes or you get
		4790	a large number of spurious wakeups, make sure you have all the relevant
		4791	and latest kernel patches applied. No, I don't know which ones, but there
		4792	are multiple ones to apply, and afterwards, event ports actually work
		4793	great.
		4794
		4795	If you can't get it to work, you can try running the program by setting
		4796	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		4797	C<select> backends.
		4798
		4799	=head2 AIX POLL BUG
		4800
		4801	AIX unfortunately has a broken C<poll.h> header. Libev works around
		4802	this by trying to avoid the poll backend altogether (i.e. it's not even
		4803	compiled in), which normally isn't a big problem as C<select> works fine
		4804	with large bitsets on AIX, and AIX is dead anyway.
		4805
4273	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4806	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4807
		4808	=head3 General issues
4274		4809
4275	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4810	Win32 doesn't support any of the standards (e.g. POSIX) that libev
4276	requires, and its I/O model is fundamentally incompatible with the POSIX	4811	requires, and its I/O model is fundamentally incompatible with the POSIX
4277	model. Libev still offers limited functionality on this platform in	4812	model. Libev still offers limited functionality on this platform in
4278	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4813	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4279	descriptors. This only applies when using Win32 natively, not when using	4814	descriptors. This only applies when using Win32 natively, not when using
4280	e.g. cygwin.	4815	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		4816	as every compielr comes with a slightly differently broken/incompatible
		4817	environment.
4281		4818
4282	Lifting these limitations would basically require the full	4819	Lifting these limitations would basically require the full
4283	re-implementation of the I/O system. If you are into these kinds of	4820	re-implementation of the I/O system. If you are into this kind of thing,
4284	things, then note that glib does exactly that for you in a very portable	4821	then note that glib does exactly that for you in a very portable way (note
4285	way (note also that glib is the slowest event library known to man).	4822	also that glib is the slowest event library known to man).
4286		4823
4287	There is no supported compilation method available on windows except	4824	There is no supported compilation method available on windows except
4288	embedding it into other applications.	4825	embedding it into other applications.
4289		4826
4290	Sensible signal handling is officially unsupported by Microsoft - libev	4827	Sensible signal handling is officially unsupported by Microsoft - libev
…		…
4318	you do I<not> compile the F<ev.c> or any other embedded source files!):	4855	you do I<not> compile the F<ev.c> or any other embedded source files!):
4319		4856
4320	#include "evwrap.h"	4857	#include "evwrap.h"
4321	#include "ev.c"	4858	#include "ev.c"
4322		4859
4323	=over 4
4324
4325	=item The winsocket select function	4860	=head3 The winsocket C<select> function
4326		4861
4327	The winsocket C<select> function doesn't follow POSIX in that it	4862	The winsocket C<select> function doesn't follow POSIX in that it
4328	requires socket I<handles> and not socket I<file descriptors> (it is	4863	requires socket I<handles> and not socket I<file descriptors> (it is
4329	also extremely buggy). This makes select very inefficient, and also	4864	also extremely buggy). This makes select very inefficient, and also
4330	requires a mapping from file descriptors to socket handles (the Microsoft	4865	requires a mapping from file descriptors to socket handles (the Microsoft
…		…
4339	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	4874	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
4340		4875
4341	Note that winsockets handling of fd sets is O(n), so you can easily get a	4876	Note that winsockets handling of fd sets is O(n), so you can easily get a
4342	complexity in the O(n²) range when using win32.	4877	complexity in the O(n²) range when using win32.
4343		4878
4344	=item Limited number of file descriptors	4879	=head3 Limited number of file descriptors
4345		4880
4346	Windows has numerous arbitrary (and low) limits on things.	4881	Windows has numerous arbitrary (and low) limits on things.
4347		4882
4348	Early versions of winsocket's select only supported waiting for a maximum	4883	Early versions of winsocket's select only supported waiting for a maximum
4349	of C<64> handles (probably owning to the fact that all windows kernels	4884	of C<64> handles (probably owning to the fact that all windows kernels
…		…
4364	runtime libraries. This might get you to about C<512> or C<2048> sockets	4899	runtime libraries. This might get you to about C<512> or C<2048> sockets
4365	(depending on windows version and/or the phase of the moon). To get more,	4900	(depending on windows version and/or the phase of the moon). To get more,
4366	you need to wrap all I/O functions and provide your own fd management, but	4901	you need to wrap all I/O functions and provide your own fd management, but
4367	the cost of calling select (O(n²)) will likely make this unworkable.	4902	the cost of calling select (O(n²)) will likely make this unworkable.
4368		4903
4369	=back
4370
4371	=head2 PORTABILITY REQUIREMENTS	4904	=head2 PORTABILITY REQUIREMENTS
4372		4905
4373	In addition to a working ISO-C implementation and of course the	4906	In addition to a working ISO-C implementation and of course the
4374	backend-specific APIs, libev relies on a few additional extensions:	4907	backend-specific APIs, libev relies on a few additional extensions:
4375		4908
…		…
4381	Libev assumes not only that all watcher pointers have the same internal	4914	Libev assumes not only that all watcher pointers have the same internal
4382	structure (guaranteed by POSIX but not by ISO C for example), but it also	4915	structure (guaranteed by POSIX but not by ISO C for example), but it also
4383	assumes that the same (machine) code can be used to call any watcher	4916	assumes that the same (machine) code can be used to call any watcher
4384	callback: The watcher callbacks have different type signatures, but libev	4917	callback: The watcher callbacks have different type signatures, but libev
4385	calls them using an C<ev_watcher *> internally.	4918	calls them using an C<ev_watcher *> internally.
		4919
		4920	=item pointer accesses must be thread-atomic
		4921
		4922	Accessing a pointer value must be atomic, it must both be readable and
		4923	writable in one piece - this is the case on all current architectures.
4386		4924
4387	=item C<sig_atomic_t volatile> must be thread-atomic as well	4925	=item C<sig_atomic_t volatile> must be thread-atomic as well
4388		4926
4389	The type C<sig_atomic_t volatile> (or whatever is defined as	4927	The type C<sig_atomic_t volatile> (or whatever is defined as
4390	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different	4928	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
…		…
4413	watchers.	4951	watchers.
4414		4952
4415	=item C<double> must hold a time value in seconds with enough accuracy	4953	=item C<double> must hold a time value in seconds with enough accuracy
4416		4954
4417	The type C<double> is used to represent timestamps. It is required to	4955	The type C<double> is used to represent timestamps. It is required to
4418	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4956	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4419	enough for at least into the year 4000. This requirement is fulfilled by	4957	good enough for at least into the year 4000 with millisecond accuracy
		4958	(the design goal for libev). This requirement is overfulfilled by
4420	implementations implementing IEEE 754, which is basically all existing	4959	implementations using IEEE 754, which is basically all existing ones. With
4421	ones. With IEEE 754 doubles, you get microsecond accuracy until at least	4960	IEEE 754 doubles, you get microsecond accuracy until at least 2200.
4422	2200.
4423		4961
4424	=back	4962	=back
4425		4963
4426	If you know of other additional requirements drop me a note.	4964	If you know of other additional requirements drop me a note.
4427		4965
…		…
4495	involves iterating over all running async watchers or all signal numbers.	5033	involves iterating over all running async watchers or all signal numbers.
4496		5034
4497	=back	5035	=back
4498		5036
4499		5037
		5038	=head1 PORTING FROM LIBEV 3.X TO 4.X
		5039
		5040	The major version 4 introduced some incompatible changes to the API.
		5041
		5042	At the moment, the C<ev.h> header file provides compatibility definitions
		5043	for all changes, so most programs should still compile. The compatibility
		5044	layer might be removed in later versions of libev, so better update to the
		5045	new API early than late.
		5046
		5047	=over 4
		5048
		5049	=item C<EV_COMPAT3> backwards compatibility mechanism
		5050
		5051	The backward compatibility mechanism can be controlled by
		5052	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		5053	section.
		5054
		5055	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5056
		5057	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5058
		5059	ev_loop_destroy (EV_DEFAULT_UC);
		5060	ev_loop_fork (EV_DEFAULT);
		5061
		5062	=item function/symbol renames
		5063
		5064	A number of functions and symbols have been renamed:
		5065
		5066	ev_loop => ev_run
		5067	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5068	EVLOOP_ONESHOT => EVRUN_ONCE
		5069
		5070	ev_unloop => ev_break
		5071	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5072	EVUNLOOP_ONE => EVBREAK_ONE
		5073	EVUNLOOP_ALL => EVBREAK_ALL
		5074
		5075	EV_TIMEOUT => EV_TIMER
		5076
		5077	ev_loop_count => ev_iteration
		5078	ev_loop_depth => ev_depth
		5079	ev_loop_verify => ev_verify
		5080
		5081	Most functions working on C<struct ev_loop> objects don't have an
		5082	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5083	associated constants have been renamed to not collide with the C<struct
		5084	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5085	as all other watcher types. Note that C<ev_loop_fork> is still called
		5086	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		5087	typedef.
		5088
		5089	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		5090
		5091	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		5092	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		5093	and work, but the library code will of course be larger.
		5094
		5095	=back
		5096
		5097
4500	=head1 GLOSSARY	5098	=head1 GLOSSARY
4501		5099
4502	=over 4	5100	=over 4
4503		5101
4504	=item active	5102	=item active
4505		5103
4506	A watcher is active as long as it has been started (has been attached to	5104	A watcher is active as long as it has been started and not yet stopped.
4507	an event loop) but not yet stopped (disassociated from the event loop).	5105	See L<WATCHER STATES> for details.
4508		5106
4509	=item application	5107	=item application
4510		5108
4511	In this document, an application is whatever is using libev.	5109	In this document, an application is whatever is using libev.
		5110
		5111	=item backend
		5112
		5113	The part of the code dealing with the operating system interfaces.
4512		5114
4513	=item callback	5115	=item callback
4514		5116
4515	The address of a function that is called when some event has been	5117	The address of a function that is called when some event has been
4516	detected. Callbacks are being passed the event loop, the watcher that	5118	detected. Callbacks are being passed the event loop, the watcher that
4517	received the event, and the actual event bitset.	5119	received the event, and the actual event bitset.
4518		5120
4519	=item callback invocation	5121	=item callback/watcher invocation
4520		5122
4521	The act of calling the callback associated with a watcher.	5123	The act of calling the callback associated with a watcher.
4522		5124
4523	=item event	5125	=item event
4524		5126
4525	A change of state of some external event, such as data now being available	5127	A change of state of some external event, such as data now being available
4526	for reading on a file descriptor, time having passed or simply not having	5128	for reading on a file descriptor, time having passed or simply not having
4527	any other events happening anymore.	5129	any other events happening anymore.
4528		5130
4529	In libev, events are represented as single bits (such as C<EV_READ> or	5131	In libev, events are represented as single bits (such as C<EV_READ> or
4530	C<EV_TIMEOUT>).	5132	C<EV_TIMER>).
4531		5133
4532	=item event library	5134	=item event library
4533		5135
4534	A software package implementing an event model and loop.	5136	A software package implementing an event model and loop.
4535		5137
…		…
4543	The model used to describe how an event loop handles and processes	5145	The model used to describe how an event loop handles and processes
4544	watchers and events.	5146	watchers and events.
4545		5147
4546	=item pending	5148	=item pending
4547		5149
4548	A watcher is pending as soon as the corresponding event has been detected,	5150	A watcher is pending as soon as the corresponding event has been
4549	and stops being pending as soon as the watcher will be invoked or its	5151	detected. See L<WATCHER STATES> for details.
4550	pending status is explicitly cleared by the application.
4551
4552	A watcher can be pending, but not active. Stopping a watcher also clears
4553	its pending status.
4554		5152
4555	=item real time	5153	=item real time
4556		5154
4557	The physical time that is observed. It is apparently strictly monotonic :)	5155	The physical time that is observed. It is apparently strictly monotonic :)
4558		5156
…		…
4565	=item watcher	5163	=item watcher
4566		5164
4567	A data structure that describes interest in certain events. Watchers need	5165	A data structure that describes interest in certain events. Watchers need
4568	to be started (attached to an event loop) before they can receive events.	5166	to be started (attached to an event loop) before they can receive events.
4569		5167
4570	=item watcher invocation
4571
4572	The act of calling the callback associated with a watcher.
4573
4574	=back	5168	=back
4575		5169
4576	=head1 AUTHOR	5170	=head1 AUTHOR
4577		5171
4578	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.	5172	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5173	Magnusson and Emanuele Giaquinta.
4579		5174

Diff Legend

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