ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.298 by sf-exg, Sat Jul 31 23:00:11 2010 UTC vs.
Revision 1.406 by root, Thu May 3 16:00:47 2012 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	// break was called, so exit
64	return 0;	64	return 0;
65	}	65	}
66		66
67	=head1 ABOUT THIS DOCUMENT	67	=head1 ABOUT THIS DOCUMENT
68		68
…		…
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	Familiarity 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
…		…
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 (in practise	137	the (fractional) number of seconds since the (POSIX) epoch (in practice
130	somewhere near the beginning of 1970, details are complicated, don't	138	somewhere near the beginning of 1970, details are complicated, don't
131	ask). This type is called C<ev_tstamp>, which is what you should use	139	ask). This type is called C<ev_tstamp>, which is what you should use
132	too. It usually aliases to the C<double> type in C. When you need to do	140	too. It usually aliases to the C<double> type in C. When you need to do
133	any calculations on it, you should treat it as some floating point value.	141	any calculations on it, you should treat it as some floating point value.
134		142
…		…
165		173
166	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
167		175
168	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
169	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
170	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_now_update> and C<ev_now>.
171		180
172	=item ev_sleep (ev_tstamp interval)	181	=item ev_sleep (ev_tstamp interval)
173		182
174	Sleep for the given interval: The current thread will be blocked until	183	Sleep for the given interval: The current thread will be blocked
175	either it is interrupted or the given time interval has passed. Basically	184	until either it is interrupted or the given time interval has
		185	passed (approximately - it might return a bit earlier even if not
		186	interrupted). Returns immediately if C<< interval <= 0 >>.
		187
176	this is a sub-second-resolution C<sleep ()>.	188	Basically this is a sub-second-resolution C<sleep ()>.
		189
		190	The range of the C<interval> is limited - libev only guarantees to work
		191	with sleep times of up to one day (C<< interval <= 86400 >>).
177		192
178	=item int ev_version_major ()	193	=item int ev_version_major ()
179		194
180	=item int ev_version_minor ()	195	=item int ev_version_minor ()
181		196
…		…
192	as this indicates an incompatible change. Minor versions are usually	207	as this indicates an incompatible change. Minor versions are usually
193	compatible to older versions, so a larger minor version alone is usually	208	compatible to older versions, so a larger minor version alone is usually
194	not a problem.	209	not a problem.
195		210
196	Example: Make sure we haven't accidentally been linked against the wrong	211	Example: Make sure we haven't accidentally been linked against the wrong
197	version (note, however, that this will not detect ABI mismatches :).	212	version (note, however, that this will not detect other ABI mismatches,
		213	such as LFS or reentrancy).
198		214
199	assert (("libev version mismatch",	215	assert (("libev version mismatch",
200	ev_version_major () == EV_VERSION_MAJOR	216	ev_version_major () == EV_VERSION_MAJOR
201	&& ev_version_minor () >= EV_VERSION_MINOR));	217	&& ev_version_minor () >= EV_VERSION_MINOR));
202		218
…		…
213	assert (("sorry, no epoll, no sex",	229	assert (("sorry, no epoll, no sex",
214	ev_supported_backends () & EVBACKEND_EPOLL));	230	ev_supported_backends () & EVBACKEND_EPOLL));
215		231
216	=item unsigned int ev_recommended_backends ()	232	=item unsigned int ev_recommended_backends ()
217		233
218	Return the set of all backends compiled into this binary of libev and also	234	Return the set of all backends compiled into this binary of libev and
219	recommended for this platform. This set is often smaller than the one	235	also recommended for this platform, meaning it will work for most file
		236	descriptor types. This set is often smaller than the one returned by
220	returned by C<ev_supported_backends>, as for example kqueue is broken on	237	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
221	most BSDs and will not be auto-detected unless you explicitly request it	238	and will not be auto-detected unless you explicitly request it (assuming
222	(assuming you know what you are doing). This is the set of backends that	239	you know what you are doing). This is the set of backends that libev will
223	libev will probe for if you specify no backends explicitly.	240	probe for if you specify no backends explicitly.
224		241
225	=item unsigned int ev_embeddable_backends ()	242	=item unsigned int ev_embeddable_backends ()
226		243
227	Returns the set of backends that are embeddable in other event loops. This	244	Returns the set of backends that are embeddable in other event loops. This
228	is the theoretical, all-platform, value. To find which backends	245	value is platform-specific but can include backends not available on the
229	might be supported on the current system, you would need to look at	246	current system. To find which embeddable backends might be supported on
230	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	247	the current system, you would need to look at C<ev_embeddable_backends ()
231	recommended ones.	248	& ev_supported_backends ()>, likewise for recommended ones.
232		249
233	See the description of C<ev_embed> watchers for more info.	250	See the description of C<ev_embed> watchers for more info.
234		251
235	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]	252	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
236		253
237	Sets the allocation function to use (the prototype is similar - the	254	Sets the allocation function to use (the prototype is similar - the
238	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	255	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
239	used to allocate and free memory (no surprises here). If it returns zero	256	used to allocate and free memory (no surprises here). If it returns zero
240	when memory needs to be allocated (C<size != 0>), the library might abort	257	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
266	}	283	}
267		284
268	...	285	...
269	ev_set_allocator (persistent_realloc);	286	ev_set_allocator (persistent_realloc);
270		287
271	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]	288	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
272		289
273	Set the callback function to call on a retryable system call error (such	290	Set the callback function to call on a retryable system call error (such
274	as failed select, poll, epoll_wait). The message is a printable string	291	as failed select, poll, epoll_wait). The message is a printable string
275	indicating the system call or subsystem causing the problem. If this	292	indicating the system call or subsystem causing the problem. If this
276	callback is set, then libev will expect it to remedy the situation, no	293	callback is set, then libev will expect it to remedy the situation, no
…		…
288	}	305	}
289		306
290	...	307	...
291	ev_set_syserr_cb (fatal_error);	308	ev_set_syserr_cb (fatal_error);
292		309
		310	=item ev_feed_signal (int signum)
		311
		312	This function can be used to "simulate" a signal receive. It is completely
		313	safe to call this function at any time, from any context, including signal
		314	handlers or random threads.
		315
		316	Its main use is to customise signal handling in your process, especially
		317	in the presence of threads. For example, you could block signals
		318	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		319	creating any loops), and in one thread, use C<sigwait> or any other
		320	mechanism to wait for signals, then "deliver" them to libev by calling
		321	C<ev_feed_signal>.
		322
293	=back	323	=back
294		324
295	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	325	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
296		326
297	An event loop is described by a C<struct ev_loop *> (the C<struct>	327	An event loop is described by a C<struct ev_loop *> (the C<struct> is
298	is I<not> optional in this case, as there is also an C<ev_loop>	328	I<not> optional in this case unless libev 3 compatibility is disabled, as
299	I<function>).	329	libev 3 had an C<ev_loop> function colliding with the struct name).
300		330
301	The library knows two types of such loops, the I<default> loop, which	331	The library knows two types of such loops, the I<default> loop, which
302	supports signals and child events, and dynamically created loops which do	332	supports child process events, and dynamically created event loops which
303	not.	333	do not.
304		334
305	=over 4	335	=over 4
306		336
307	=item struct ev_loop *ev_default_loop (unsigned int flags)	337	=item struct ev_loop *ev_default_loop (unsigned int flags)
308		338
309	This will initialise the default event loop if it hasn't been initialised	339	This returns the "default" event loop object, which is what you should
310	yet and return it. If the default loop could not be initialised, returns	340	normally use when you just need "the event loop". Event loop objects and
311	false. If it already was initialised it simply returns it (and ignores the	341	the C<flags> parameter are described in more detail in the entry for
312	flags. If that is troubling you, check C<ev_backend ()> afterwards).	342	C<ev_loop_new>.
		343
		344	If the default loop is already initialised then this function simply
		345	returns it (and ignores the flags. If that is troubling you, check
		346	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		347	flags, which should almost always be C<0>, unless the caller is also the
		348	one calling C<ev_run> or otherwise qualifies as "the main program".
313		349
314	If you don't know what event loop to use, use the one returned from this	350	If you don't know what event loop to use, use the one returned from this
315	function.	351	function (or via the C<EV_DEFAULT> macro).
316		352
317	Note that this function is I<not> thread-safe, so if you want to use it	353	Note that this function is I<not> thread-safe, so if you want to use it
318	from multiple threads, you have to lock (note also that this is unlikely,	354	from multiple threads, you have to employ some kind of mutex (note also
319	as loops cannot be shared easily between threads anyway).	355	that this case is unlikely, as loops cannot be shared easily between
		356	threads anyway).
320		357
321	The default loop is the only loop that can handle C<ev_signal> and	358	The default loop is the only loop that can handle C<ev_child> watchers,
322	C<ev_child> watchers, and to do this, it always registers a handler	359	and to do this, it always registers a handler for C<SIGCHLD>. If this is
323	for C<SIGCHLD>. If this is a problem for your application you can either	360	a problem for your application you can either create a dynamic loop with
324	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	361	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
325	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	362	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
326	C<ev_default_init>.	363
		364	Example: This is the most typical usage.
		365
		366	if (!ev_default_loop (0))
		367	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		368
		369	Example: Restrict libev to the select and poll backends, and do not allow
		370	environment settings to be taken into account:
		371
		372	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		373
		374	=item struct ev_loop *ev_loop_new (unsigned int flags)
		375
		376	This will create and initialise a new event loop object. If the loop
		377	could not be initialised, returns false.
		378
		379	This function is thread-safe, and one common way to use libev with
		380	threads is indeed to create one loop per thread, and using the default
		381	loop in the "main" or "initial" thread.
327		382
328	The flags argument can be used to specify special behaviour or specific	383	The flags argument can be used to specify special behaviour or specific
329	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	384	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
330		385
331	The following flags are supported:	386	The following flags are supported:
…		…
366	environment variable.	421	environment variable.
367		422
368	=item C<EVFLAG_NOINOTIFY>	423	=item C<EVFLAG_NOINOTIFY>
369		424
370	When this flag is specified, then libev will not attempt to use the	425	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	426	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	427	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.	428	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
374		429
375	=item C<EVFLAG_SIGNALFD>	430	=item C<EVFLAG_SIGNALFD>
376		431
377	When this flag is specified, then libev will attempt to use the	432	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 API	433	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
379	delivers signals synchronously, which makes it both faster and might make	434	delivers signals synchronously, which makes it both faster and might make
380	it possible to get the queued signal data. It can also simplify signal	435	it possible to get the queued signal data. It can also simplify signal
381	handling with threads, as long as you properly block signals in your	436	handling with threads, as long as you properly block signals in your
382	threads that are not interested in handling them.	437	threads that are not interested in handling them.
383		438
384	Signalfd will not be used by default as this changes your signal mask, and	439	Signalfd will not be used by default as this changes your signal mask, and
385	there are a lot of shoddy libraries and programs (glib's threadpool for	440	there are a lot of shoddy libraries and programs (glib's threadpool for
386	example) that can't properly initialise their signal masks.	441	example) that can't properly initialise their signal masks.
		442
		443	=item C<EVFLAG_NOSIGMASK>
		444
		445	When this flag is specified, then libev will avoid to modify the signal
		446	mask. Specifically, this means you have to make sure signals are unblocked
		447	when you want to receive them.
		448
		449	This behaviour is useful when you want to do your own signal handling, or
		450	want to handle signals only in specific threads and want to avoid libev
		451	unblocking the signals.
		452
		453	It's also required by POSIX in a threaded program, as libev calls
		454	C<sigprocmask>, whose behaviour is officially unspecified.
		455
		456	This flag's behaviour will become the default in future versions of libev.
387		457
388	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	458	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
389		459
390	This is your standard select(2) backend. Not I<completely> standard, as	460	This is your standard select(2) backend. Not I<completely> standard, as
391	libev tries to roll its own fd_set with no limits on the number of fds,	461	libev tries to roll its own fd_set with no limits on the number of fds,
…		…
419	=item C<EVBACKEND_EPOLL> (value 4, Linux)	489	=item C<EVBACKEND_EPOLL> (value 4, Linux)
420		490
421	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9	491	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
422	kernels).	492	kernels).
423		493
424	For few fds, this backend is a bit little slower than poll and select,	494	For few fds, this backend is a bit little slower than poll and select, but
425	but it scales phenomenally better. While poll and select usually scale	495	it scales phenomenally better. While poll and select usually scale like
426	like O(total_fds) where n is the total number of fds (or the highest fd),	496	O(total_fds) where total_fds is the total number of fds (or the highest
427	epoll scales either O(1) or O(active_fds).	497	fd), epoll scales either O(1) or O(active_fds).
428		498
429	The epoll mechanism deserves honorable mention as the most misdesigned	499	The epoll mechanism deserves honorable mention as the most misdesigned
430	of the more advanced event mechanisms: mere annoyances include silently	500	of the more advanced event mechanisms: mere annoyances include silently
431	dropping file descriptors, requiring a system call per change per file	501	dropping file descriptors, requiring a system call per change per file
432	descriptor (and unnecessary guessing of parameters), problems with dup and	502	descriptor (and unnecessary guessing of parameters), problems with dup,
		503	returning before the timeout value, resulting in additional iterations
		504	(and only giving 5ms accuracy while select on the same platform gives
433	so on. The biggest issue is fork races, however - if a program forks then	505	0.1ms) and so on. The biggest issue is fork races, however - if a program
434	I<both> parent and child process have to recreate the epoll set, which can	506	forks then I<both> parent and child process have to recreate the epoll
435	take considerable time (one syscall per file descriptor) and is of course	507	set, which can take considerable time (one syscall per file descriptor)
436	hard to detect.	508	and is of course hard to detect.
437		509
438	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but	510	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
439	of course I<doesn't>, and epoll just loves to report events for totally	511	but of course I<doesn't>, and epoll just loves to report events for
440	I<different> file descriptors (even already closed ones, so one cannot	512	totally I<different> file descriptors (even already closed ones, so
441	even remove them from the set) than registered in the set (especially	513	one cannot even remove them from the set) than registered in the set
442	on SMP systems). Libev tries to counter these spurious notifications by	514	(especially on SMP systems). Libev tries to counter these spurious
443	employing an additional generation counter and comparing that against the	515	notifications by employing an additional generation counter and comparing
444	events to filter out spurious ones, recreating the set when required.	516	that against the events to filter out spurious ones, recreating the set
		517	when required. Epoll also erroneously rounds down timeouts, but gives you
		518	no way to know when and by how much, so sometimes you have to busy-wait
		519	because epoll returns immediately despite a nonzero timeout. And last
		520	not least, it also refuses to work with some file descriptors which work
		521	perfectly fine with C<select> (files, many character devices...).
		522
		523	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		524	cobbled together in a hurry, no thought to design or interaction with
		525	others. Oh, the pain, will it ever stop...
445		526
446	While stopping, setting and starting an I/O watcher in the same iteration	527	While stopping, setting and starting an I/O watcher in the same iteration
447	will result in some caching, there is still a system call per such	528	will result in some caching, there is still a system call per such
448	incident (because the same I<file descriptor> could point to a different	529	incident (because the same I<file descriptor> could point to a different
449	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	530	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
486		567
487	It scales in the same way as the epoll backend, but the interface to the	568	It scales in the same way as the epoll backend, but the interface to the
488	kernel is more efficient (which says nothing about its actual speed, of	569	kernel is more efficient (which says nothing about its actual speed, of
489	course). While stopping, setting and starting an I/O watcher does never	570	course). While stopping, setting and starting an I/O watcher does never
490	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	571	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
491	two event changes per incident. Support for C<fork ()> is very bad (but	572	two event changes per incident. Support for C<fork ()> is very bad (you
492	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect	573	might have to leak fd's on fork, but it's more sane than epoll) and it
493	cases	574	drops fds silently in similarly hard-to-detect cases
494		575
495	This backend usually performs well under most conditions.	576	This backend usually performs well under most conditions.
496		577
497	While nominally embeddable in other event loops, this doesn't work	578	While nominally embeddable in other event loops, this doesn't work
498	everywhere, so you might need to test for this. And since it is broken	579	everywhere, so you might need to test for this. And since it is broken
…		…
515	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	596	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
516		597
517	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	598	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
518	it's really slow, but it still scales very well (O(active_fds)).	599	it's really slow, but it still scales very well (O(active_fds)).
519		600
520	Please note that Solaris event ports can deliver a lot of spurious
521	notifications, so you need to use non-blocking I/O or other means to avoid
522	blocking when no data (or space) is available.
523
524	While this backend scales well, it requires one system call per active	601	While this backend scales well, it requires one system call per active
525	file descriptor per loop iteration. For small and medium numbers of file	602	file descriptor per loop iteration. For small and medium numbers of file
526	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	603	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
527	might perform better.	604	might perform better.
528		605
529	On the positive side, with the exception of the spurious readiness	606	On the positive side, this backend actually performed fully to
530	notifications, this backend actually performed fully to specification
531	in all tests and is fully embeddable, which is a rare feat among the	607	specification in all tests and is fully embeddable, which is a rare feat
532	OS-specific backends (I vastly prefer correctness over speed hacks).	608	among the OS-specific backends (I vastly prefer correctness over speed
		609	hacks).
		610
		611	On the negative side, the interface is I<bizarre> - so bizarre that
		612	even sun itself gets it wrong in their code examples: The event polling
		613	function sometimes returns events to the caller even though an error
		614	occurred, but with no indication whether it has done so or not (yes, it's
		615	even documented that way) - deadly for edge-triggered interfaces where you
		616	absolutely have to know whether an event occurred or not because you have
		617	to re-arm the watcher.
		618
		619	Fortunately libev seems to be able to work around these idiocies.
533		620
534	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	621	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
535	C<EVBACKEND_POLL>.	622	C<EVBACKEND_POLL>.
536		623
537	=item C<EVBACKEND_ALL>	624	=item C<EVBACKEND_ALL>
538		625
539	Try all backends (even potentially broken ones that wouldn't be tried	626	Try all backends (even potentially broken ones that wouldn't be tried
540	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	627	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
541	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	628	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
542		629
543	It is definitely not recommended to use this flag.	630	It is definitely not recommended to use this flag, use whatever
		631	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		632	at all.
		633
		634	=item C<EVBACKEND_MASK>
		635
		636	Not a backend at all, but a mask to select all backend bits from a
		637	C<flags> value, in case you want to mask out any backends from a flags
		638	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
544		639
545	=back	640	=back
546		641
547	If one or more of the backend flags are or'ed into the flags value,	642	If one or more of the backend flags are or'ed into the flags value,
548	then only these backends will be tried (in the reverse order as listed	643	then only these backends will be tried (in the reverse order as listed
549	here). If none are specified, all backends in C<ev_recommended_backends	644	here). If none are specified, all backends in C<ev_recommended_backends
550	()> will be tried.	645	()> will be tried.
551		646
552	Example: This is the most typical usage.
553
554	if (!ev_default_loop (0))
555	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
556
557	Example: Restrict libev to the select and poll backends, and do not allow
558	environment settings to be taken into account:
559
560	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
561
562	Example: Use whatever libev has to offer, but make sure that kqueue is
563	used if available (warning, breaks stuff, best use only with your own
564	private event loop and only if you know the OS supports your types of
565	fds):
566
567	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
568
569	=item struct ev_loop *ev_loop_new (unsigned int flags)
570
571	Similar to C<ev_default_loop>, but always creates a new event loop that is
572	always distinct from the default loop.
573
574	Note that this function I<is> thread-safe, and one common way to use
575	libev with threads is indeed to create one loop per thread, and using the
576	default loop in the "main" or "initial" thread.
577
578	Example: Try to create a event loop that uses epoll and nothing else.	647	Example: Try to create a event loop that uses epoll and nothing else.
579		648
580	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	649	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
581	if (!epoller)	650	if (!epoller)
582	fatal ("no epoll found here, maybe it hides under your chair");	651	fatal ("no epoll found here, maybe it hides under your chair");
583		652
		653	Example: Use whatever libev has to offer, but make sure that kqueue is
		654	used if available.
		655
		656	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		657
584	=item ev_default_destroy ()	658	=item ev_loop_destroy (loop)
585		659
586	Destroys the default loop (frees all memory and kernel state etc.). None	660	Destroys an event loop object (frees all memory and kernel state
587	of the active event watchers will be stopped in the normal sense, so	661	etc.). None of the active event watchers will be stopped in the normal
588	e.g. C<ev_is_active> might still return true. It is your responsibility to	662	sense, so e.g. C<ev_is_active> might still return true. It is your
589	either stop all watchers cleanly yourself I<before> calling this function,	663	responsibility to either stop all watchers cleanly yourself I<before>
590	or cope with the fact afterwards (which is usually the easiest thing, you	664	calling this function, or cope with the fact afterwards (which is usually
591	can just ignore the watchers and/or C<free ()> them for example).	665	the easiest thing, you can just ignore the watchers and/or C<free ()> them
		666	for example).
592		667
593	Note that certain global state, such as signal state (and installed signal	668	Note that certain global state, such as signal state (and installed signal
594	handlers), will not be freed by this function, and related watchers (such	669	handlers), will not be freed by this function, and related watchers (such
595	as signal and child watchers) would need to be stopped manually.	670	as signal and child watchers) would need to be stopped manually.
596		671
597	In general it is not advisable to call this function except in the	672	This function is normally used on loop objects allocated by
598	rare occasion where you really need to free e.g. the signal handling	673	C<ev_loop_new>, but it can also be used on the default loop returned by
		674	C<ev_default_loop>, in which case it is not thread-safe.
		675
		676	Note that it is not advisable to call this function on the default loop
		677	except in the rare occasion where you really need to free its resources.
599	pipe fds. If you need dynamically allocated loops it is better to use	678	If you need dynamically allocated loops it is better to use C<ev_loop_new>
600	C<ev_loop_new> and C<ev_loop_destroy>.	679	and C<ev_loop_destroy>.
601		680
602	=item ev_loop_destroy (loop)	681	=item ev_loop_fork (loop)
603		682
604	Like C<ev_default_destroy>, but destroys an event loop created by an
605	earlier call to C<ev_loop_new>.
606
607	=item ev_default_fork ()
608
609	This function sets a flag that causes subsequent C<ev_loop> iterations	683	This function sets a flag that causes subsequent C<ev_run> iterations to
610	to reinitialise the kernel state for backends that have one. Despite the	684	reinitialise the kernel state for backends that have one. Despite the
611	name, you can call it anytime, but it makes most sense after forking, in	685	name, you can call it anytime, but it makes most sense after forking, in
612	the child process (or both child and parent, but that again makes little	686	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
613	sense). You I<must> call it in the child before using any of the libev	687	child before resuming or calling C<ev_run>.
614	functions, and it will only take effect at the next C<ev_loop> iteration.
615		688
616	Again, you I<have> to call it on I<any> loop that you want to re-use after	689	Again, you I<have> to call it on I<any> loop that you want to re-use after
617	a fork, I<even if you do not plan to use the loop in the parent>. This is	690	a fork, I<even if you do not plan to use the loop in the parent>. This is
618	because some kernel interfaces *cough* I<kqueue> *cough* do funny things	691	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
619	during fork.	692	during fork.
620		693
621	On the other hand, you only need to call this function in the child	694	On the other hand, you only need to call this function in the child
622	process if and only if you want to use the event loop in the child. If you	695	process if and only if you want to use the event loop in the child. If
623	just fork+exec or create a new loop in the child, you don't have to call	696	you just fork+exec or create a new loop in the child, you don't have to
624	it at all.	697	call it at all (in fact, C<epoll> is so badly broken that it makes a
		698	difference, but libev will usually detect this case on its own and do a
		699	costly reset of the backend).
625		700
626	The function itself is quite fast and it's usually not a problem to call	701	The function itself is quite fast and it's usually not a problem to call
627	it just in case after a fork. To make this easy, the function will fit in	702	it just in case after a fork.
628	quite nicely into a call to C<pthread_atfork>:
629		703
		704	Example: Automate calling C<ev_loop_fork> on the default loop when
		705	using pthreads.
		706
		707	static void
		708	post_fork_child (void)
		709	{
		710	ev_loop_fork (EV_DEFAULT);
		711	}
		712
		713	...
630	pthread_atfork (0, 0, ev_default_fork);	714	pthread_atfork (0, 0, post_fork_child);
631
632	=item ev_loop_fork (loop)
633
634	Like C<ev_default_fork>, but acts on an event loop created by
635	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
636	after fork that you want to re-use in the child, and how you keep track of
637	them is entirely your own problem.
638		715
639	=item int ev_is_default_loop (loop)	716	=item int ev_is_default_loop (loop)
640		717
641	Returns true when the given loop is, in fact, the default loop, and false	718	Returns true when the given loop is, in fact, the default loop, and false
642	otherwise.	719	otherwise.
643		720
644	=item unsigned int ev_iteration (loop)	721	=item unsigned int ev_iteration (loop)
645		722
646	Returns the current iteration count for the loop, which is identical to	723	Returns the current iteration count for the event loop, which is identical
647	the number of times libev did poll for new events. It starts at C<0> and	724	to the number of times libev did poll for new events. It starts at C<0>
648	happily wraps around with enough iterations.	725	and happily wraps around with enough iterations.
649		726
650	This value can sometimes be useful as a generation counter of sorts (it	727	This value can sometimes be useful as a generation counter of sorts (it
651	"ticks" the number of loop iterations), as it roughly corresponds with	728	"ticks" the number of loop iterations), as it roughly corresponds with
652	C<ev_prepare> and C<ev_check> calls - and is incremented between the	729	C<ev_prepare> and C<ev_check> calls - and is incremented between the
653	prepare and check phases.	730	prepare and check phases.
654		731
655	=item unsigned int ev_depth (loop)	732	=item unsigned int ev_depth (loop)
656		733
657	Returns the number of times C<ev_loop> was entered minus the number of	734	Returns the number of times C<ev_run> was entered minus the number of
658	times C<ev_loop> was exited, in other words, the recursion depth.	735	times C<ev_run> was exited normally, in other words, the recursion depth.
659		736
660	Outside C<ev_loop>, this number is zero. In a callback, this number is	737	Outside C<ev_run>, this number is zero. In a callback, this number is
661	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),	738	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
662	in which case it is higher.	739	in which case it is higher.
663		740
664	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread	741	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
665	etc.), doesn't count as "exit" - consider this as a hint to avoid such	742	throwing an exception etc.), doesn't count as "exit" - consider this
666	ungentleman behaviour unless it's really convenient.	743	as a hint to avoid such ungentleman-like behaviour unless it's really
		744	convenient, in which case it is fully supported.
667		745
668	=item unsigned int ev_backend (loop)	746	=item unsigned int ev_backend (loop)
669		747
670	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	748	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
671	use.	749	use.
…		…
680		758
681	=item ev_now_update (loop)	759	=item ev_now_update (loop)
682		760
683	Establishes the current time by querying the kernel, updating the time	761	Establishes the current time by querying the kernel, updating the time
684	returned by C<ev_now ()> in the progress. This is a costly operation and	762	returned by C<ev_now ()> in the progress. This is a costly operation and
685	is usually done automatically within C<ev_loop ()>.	763	is usually done automatically within C<ev_run ()>.
686		764
687	This function is rarely useful, but when some event callback runs for a	765	This function is rarely useful, but when some event callback runs for a
688	very long time without entering the event loop, updating libev's idea of	766	very long time without entering the event loop, updating libev's idea of
689	the current time is a good idea.	767	the current time is a good idea.
690		768
…		…
692		770
693	=item ev_suspend (loop)	771	=item ev_suspend (loop)
694		772
695	=item ev_resume (loop)	773	=item ev_resume (loop)
696		774
697	These two functions suspend and resume a loop, for use when the loop is	775	These two functions suspend and resume an event loop, for use when the
698	not used for a while and timeouts should not be processed.	776	loop is not used for a while and timeouts should not be processed.
699		777
700	A typical use case would be an interactive program such as a game: When	778	A typical use case would be an interactive program such as a game: When
701	the user presses C<^Z> to suspend the game and resumes it an hour later it	779	the user presses C<^Z> to suspend the game and resumes it an hour later it
702	would be best to handle timeouts as if no time had actually passed while	780	would be best to handle timeouts as if no time had actually passed while
703	the program was suspended. This can be achieved by calling C<ev_suspend>	781	the program was suspended. This can be achieved by calling C<ev_suspend>
…		…
714	without a previous call to C<ev_suspend>.	792	without a previous call to C<ev_suspend>.
715		793
716	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the	794	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
717	event loop time (see C<ev_now_update>).	795	event loop time (see C<ev_now_update>).
718		796
719	=item ev_loop (loop, int flags)	797	=item bool ev_run (loop, int flags)
720		798
721	Finally, this is it, the event handler. This function usually is called	799	Finally, this is it, the event handler. This function usually is called
722	after you have initialised all your watchers and you want to start	800	after you have initialised all your watchers and you want to start
723	handling events.	801	handling events. It will ask the operating system for any new events, call
		802	the watcher callbacks, and then repeat the whole process indefinitely: This
		803	is why event loops are called I<loops>.
724		804
725	If the flags argument is specified as C<0>, it will not return until	805	If the flags argument is specified as C<0>, it will keep handling events
726	either no event watchers are active anymore or C<ev_unloop> was called.	806	until either no event watchers are active anymore or C<ev_break> was
		807	called.
727		808
		809	The return value is false if there are no more active watchers (which
		810	usually means "all jobs done" or "deadlock"), and true in all other cases
		811	(which usually means " you should call C<ev_run> again").
		812
728	Please note that an explicit C<ev_unloop> is usually better than	813	Please note that an explicit C<ev_break> is usually better than
729	relying on all watchers to be stopped when deciding when a program has	814	relying on all watchers to be stopped when deciding when a program has
730	finished (especially in interactive programs), but having a program	815	finished (especially in interactive programs), but having a program
731	that automatically loops as long as it has to and no longer by virtue	816	that automatically loops as long as it has to and no longer by virtue
732	of relying on its watchers stopping correctly, that is truly a thing of	817	of relying on its watchers stopping correctly, that is truly a thing of
733	beauty.	818	beauty.
734		819
		820	This function is I<mostly> exception-safe - you can break out of a
		821	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		822	exception and so on. This does not decrement the C<ev_depth> value, nor
		823	will it clear any outstanding C<EVBREAK_ONE> breaks.
		824
735	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	825	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
736	those events and any already outstanding ones, but will not block your	826	those events and any already outstanding ones, but will not wait and
737	process in case there are no events and will return after one iteration of	827	block your process in case there are no events and will return after one
738	the loop.	828	iteration of the loop. This is sometimes useful to poll and handle new
		829	events while doing lengthy calculations, to keep the program responsive.
739		830
740	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	831	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
741	necessary) and will handle those and any already outstanding ones. It	832	necessary) and will handle those and any already outstanding ones. It
742	will block your process until at least one new event arrives (which could	833	will block your process until at least one new event arrives (which could
743	be an event internal to libev itself, so there is no guarantee that a	834	be an event internal to libev itself, so there is no guarantee that a
744	user-registered callback will be called), and will return after one	835	user-registered callback will be called), and will return after one
745	iteration of the loop.	836	iteration of the loop.
746		837
747	This is useful if you are waiting for some external event in conjunction	838	This is useful if you are waiting for some external event in conjunction
748	with something not expressible using other libev watchers (i.e. "roll your	839	with something not expressible using other libev watchers (i.e. "roll your
749	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	840	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
750	usually a better approach for this kind of thing.	841	usually a better approach for this kind of thing.
751		842
752	Here are the gory details of what C<ev_loop> does:	843	Here are the gory details of what C<ev_run> does (this is for your
		844	understanding, not a guarantee that things will work exactly like this in
		845	future versions):
753		846
		847	- Increment loop depth.
		848	- Reset the ev_break status.
754	- Before the first iteration, call any pending watchers.	849	- Before the first iteration, call any pending watchers.
		850	LOOP:
755	* If EVFLAG_FORKCHECK was used, check for a fork.	851	- If EVFLAG_FORKCHECK was used, check for a fork.
756	- If a fork was detected (by any means), queue and call all fork watchers.	852	- If a fork was detected (by any means), queue and call all fork watchers.
757	- Queue and call all prepare watchers.	853	- Queue and call all prepare watchers.
		854	- If ev_break was called, goto FINISH.
758	- If we have been forked, detach and recreate the kernel state	855	- If we have been forked, detach and recreate the kernel state
759	as to not disturb the other process.	856	as to not disturb the other process.
760	- Update the kernel state with all outstanding changes.	857	- Update the kernel state with all outstanding changes.
761	- Update the "event loop time" (ev_now ()).	858	- Update the "event loop time" (ev_now ()).
762	- Calculate for how long to sleep or block, if at all	859	- Calculate for how long to sleep or block, if at all
763	(active idle watchers, EVLOOP_NONBLOCK or not having	860	(active idle watchers, EVRUN_NOWAIT or not having
764	any active watchers at all will result in not sleeping).	861	any active watchers at all will result in not sleeping).
765	- Sleep if the I/O and timer collect interval say so.	862	- Sleep if the I/O and timer collect interval say so.
		863	- Increment loop iteration counter.
766	- Block the process, waiting for any events.	864	- Block the process, waiting for any events.
767	- Queue all outstanding I/O (fd) events.	865	- Queue all outstanding I/O (fd) events.
768	- Update the "event loop time" (ev_now ()), and do time jump adjustments.	866	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
769	- Queue all expired timers.	867	- Queue all expired timers.
770	- Queue all expired periodics.	868	- Queue all expired periodics.
771	- Unless any events are pending now, queue all idle watchers.	869	- Queue all idle watchers with priority higher than that of pending events.
772	- Queue all check watchers.	870	- Queue all check watchers.
773	- Call all queued watchers in reverse order (i.e. check watchers first).	871	- Call all queued watchers in reverse order (i.e. check watchers first).
774	Signals and child watchers are implemented as I/O watchers, and will	872	Signals and child watchers are implemented as I/O watchers, and will
775	be handled here by queueing them when their watcher gets executed.	873	be handled here by queueing them when their watcher gets executed.
776	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	874	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
777	were used, or there are no active watchers, return, otherwise	875	were used, or there are no active watchers, goto FINISH, otherwise
778	continue with step *.	876	continue with step LOOP.
		877	FINISH:
		878	- Reset the ev_break status iff it was EVBREAK_ONE.
		879	- Decrement the loop depth.
		880	- Return.
779		881
780	Example: Queue some jobs and then loop until no events are outstanding	882	Example: Queue some jobs and then loop until no events are outstanding
781	anymore.	883	anymore.
782		884
783	... queue jobs here, make sure they register event watchers as long	885	... queue jobs here, make sure they register event watchers as long
784	... as they still have work to do (even an idle watcher will do..)	886	... as they still have work to do (even an idle watcher will do..)
785	ev_loop (my_loop, 0);	887	ev_run (my_loop, 0);
786	... jobs done or somebody called unloop. yeah!	888	... jobs done or somebody called break. yeah!
787		889
788	=item ev_unloop (loop, how)	890	=item ev_break (loop, how)
789		891
790	Can be used to make a call to C<ev_loop> return early (but only after it	892	Can be used to make a call to C<ev_run> return early (but only after it
791	has processed all outstanding events). The C<how> argument must be either	893	has processed all outstanding events). The C<how> argument must be either
792	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	894	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
793	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	895	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
794		896
795	This "unloop state" will be cleared when entering C<ev_loop> again.	897	This "break state" will be cleared on the next call to C<ev_run>.
796		898
797	It is safe to call C<ev_unloop> from outside any C<ev_loop> calls.	899	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		900	which case it will have no effect.
798		901
799	=item ev_ref (loop)	902	=item ev_ref (loop)
800		903
801	=item ev_unref (loop)	904	=item ev_unref (loop)
802		905
803	Ref/unref can be used to add or remove a reference count on the event	906	Ref/unref can be used to add or remove a reference count on the event
804	loop: Every watcher keeps one reference, and as long as the reference	907	loop: Every watcher keeps one reference, and as long as the reference
805	count is nonzero, C<ev_loop> will not return on its own.	908	count is nonzero, C<ev_run> will not return on its own.
806		909
807	This is useful when you have a watcher that you never intend to	910	This is useful when you have a watcher that you never intend to
808	unregister, but that nevertheless should not keep C<ev_loop> from	911	unregister, but that nevertheless should not keep C<ev_run> from
809	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>	912	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
810	before stopping it.	913	before stopping it.
811		914
812	As an example, libev itself uses this for its internal signal pipe: It	915	As an example, libev itself uses this for its internal signal pipe: It
813	is not visible to the libev user and should not keep C<ev_loop> from	916	is not visible to the libev user and should not keep C<ev_run> from
814	exiting if no event watchers registered by it are active. It is also an	917	exiting if no event watchers registered by it are active. It is also an
815	excellent way to do this for generic recurring timers or from within	918	excellent way to do this for generic recurring timers or from within
816	third-party libraries. Just remember to I<unref after start> and I<ref	919	third-party libraries. Just remember to I<unref after start> and I<ref
817	before stop> (but only if the watcher wasn't active before, or was active	920	before stop> (but only if the watcher wasn't active before, or was active
818	before, respectively. Note also that libev might stop watchers itself	921	before, respectively. Note also that libev might stop watchers itself
819	(e.g. non-repeating timers) in which case you have to C<ev_ref>	922	(e.g. non-repeating timers) in which case you have to C<ev_ref>
820	in the callback).	923	in the callback).
821		924
822	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	925	Example: Create a signal watcher, but keep it from keeping C<ev_run>
823	running when nothing else is active.	926	running when nothing else is active.
824		927
825	ev_signal exitsig;	928	ev_signal exitsig;
826	ev_signal_init (&exitsig, sig_cb, SIGINT);	929	ev_signal_init (&exitsig, sig_cb, SIGINT);
827	ev_signal_start (loop, &exitsig);	930	ev_signal_start (loop, &exitsig);
828	evf_unref (loop);	931	ev_unref (loop);
829		932
830	Example: For some weird reason, unregister the above signal handler again.	933	Example: For some weird reason, unregister the above signal handler again.
831		934
832	ev_ref (loop);	935	ev_ref (loop);
833	ev_signal_stop (loop, &exitsig);	936	ev_signal_stop (loop, &exitsig);
…		…
853	overhead for the actual polling but can deliver many events at once.	956	overhead for the actual polling but can deliver many events at once.
854		957
855	By setting a higher I<io collect interval> you allow libev to spend more	958	By setting a higher I<io collect interval> you allow libev to spend more
856	time collecting I/O events, so you can handle more events per iteration,	959	time collecting I/O events, so you can handle more events per iteration,
857	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	960	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
858	C<ev_timer>) will be not affected. Setting this to a non-null value will	961	C<ev_timer>) will not be affected. Setting this to a non-null value will
859	introduce an additional C<ev_sleep ()> call into most loop iterations. The	962	introduce an additional C<ev_sleep ()> call into most loop iterations. The
860	sleep time ensures that libev will not poll for I/O events more often then	963	sleep time ensures that libev will not poll for I/O events more often then
861	once per this interval, on average.	964	once per this interval, on average (as long as the host time resolution is
		965	good enough).
862		966
863	Likewise, by setting a higher I<timeout collect interval> you allow libev	967	Likewise, by setting a higher I<timeout collect interval> you allow libev
864	to spend more time collecting timeouts, at the expense of increased	968	to spend more time collecting timeouts, at the expense of increased
865	latency/jitter/inexactness (the watcher callback will be called	969	latency/jitter/inexactness (the watcher callback will be called
866	later). C<ev_io> watchers will not be affected. Setting this to a non-null	970	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
890	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);	994	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
891		995
892	=item ev_invoke_pending (loop)	996	=item ev_invoke_pending (loop)
893		997
894	This call will simply invoke all pending watchers while resetting their	998	This call will simply invoke all pending watchers while resetting their
895	pending state. Normally, C<ev_loop> does this automatically when required,	999	pending state. Normally, C<ev_run> does this automatically when required,
896	but when overriding the invoke callback this call comes handy.	1000	but when overriding the invoke callback this call comes handy. This
		1001	function can be invoked from a watcher - this can be useful for example
		1002	when you want to do some lengthy calculation and want to pass further
		1003	event handling to another thread (you still have to make sure only one
		1004	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
897		1005
898	=item int ev_pending_count (loop)	1006	=item int ev_pending_count (loop)
899		1007
900	Returns the number of pending watchers - zero indicates that no watchers	1008	Returns the number of pending watchers - zero indicates that no watchers
901	are pending.	1009	are pending.
902		1010
903	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))	1011	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
904		1012
905	This overrides the invoke pending functionality of the loop: Instead of	1013	This overrides the invoke pending functionality of the loop: Instead of
906	invoking all pending watchers when there are any, C<ev_loop> will call	1014	invoking all pending watchers when there are any, C<ev_run> will call
907	this callback instead. This is useful, for example, when you want to	1015	this callback instead. This is useful, for example, when you want to
908	invoke the actual watchers inside another context (another thread etc.).	1016	invoke the actual watchers inside another context (another thread etc.).
909		1017
910	If you want to reset the callback, use C<ev_invoke_pending> as new	1018	If you want to reset the callback, use C<ev_invoke_pending> as new
911	callback.	1019	callback.
912		1020
913	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))	1021	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
914		1022
915	Sometimes you want to share the same loop between multiple threads. This	1023	Sometimes you want to share the same loop between multiple threads. This
916	can be done relatively simply by putting mutex_lock/unlock calls around	1024	can be done relatively simply by putting mutex_lock/unlock calls around
917	each call to a libev function.	1025	each call to a libev function.
918		1026
919	However, C<ev_loop> can run an indefinite time, so it is not feasible to	1027	However, C<ev_run> can run an indefinite time, so it is not feasible
920	wait for it to return. One way around this is to wake up the loop via	1028	to wait for it to return. One way around this is to wake up the event
921	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>	1029	loop via C<ev_break> and C<ev_async_send>, another way is to set these
922	and I<acquire> callbacks on the loop.	1030	I<release> and I<acquire> callbacks on the loop.
923		1031
924	When set, then C<release> will be called just before the thread is	1032	When set, then C<release> will be called just before the thread is
925	suspended waiting for new events, and C<acquire> is called just	1033	suspended waiting for new events, and C<acquire> is called just
926	afterwards.	1034	afterwards.
927		1035
…		…
930		1038
931	While event loop modifications are allowed between invocations of	1039	While event loop modifications are allowed between invocations of
932	C<release> and C<acquire> (that's their only purpose after all), no	1040	C<release> and C<acquire> (that's their only purpose after all), no
933	modifications done will affect the event loop, i.e. adding watchers will	1041	modifications done will affect the event loop, i.e. adding watchers will
934	have no effect on the set of file descriptors being watched, or the time	1042	have no effect on the set of file descriptors being watched, or the time
935	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it	1043	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
936	to take note of any changes you made.	1044	to take note of any changes you made.
937		1045
938	In theory, threads executing C<ev_loop> will be async-cancel safe between	1046	In theory, threads executing C<ev_run> will be async-cancel safe between
939	invocations of C<release> and C<acquire>.	1047	invocations of C<release> and C<acquire>.
940		1048
941	See also the locking example in the C<THREADS> section later in this	1049	See also the locking example in the C<THREADS> section later in this
942	document.	1050	document.
943		1051
944	=item ev_set_userdata (loop, void *data)	1052	=item ev_set_userdata (loop, void *data)
945		1053
946	=item ev_userdata (loop)	1054	=item void *ev_userdata (loop)
947		1055
948	Set and retrieve a single C<void *> associated with a loop. When	1056	Set and retrieve a single C<void *> associated with a loop. When
949	C<ev_set_userdata> has never been called, then C<ev_userdata> returns	1057	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
950	C<0.>	1058	C<0>.
951		1059
952	These two functions can be used to associate arbitrary data with a loop,	1060	These two functions can be used to associate arbitrary data with a loop,
953	and are intended solely for the C<invoke_pending_cb>, C<release> and	1061	and are intended solely for the C<invoke_pending_cb>, C<release> and
954	C<acquire> callbacks described above, but of course can be (ab-)used for	1062	C<acquire> callbacks described above, but of course can be (ab-)used for
955	any other purpose as well.	1063	any other purpose as well.
956		1064
957	=item ev_loop_verify (loop)	1065	=item ev_verify (loop)
958		1066
959	This function only does something when C<EV_VERIFY> support has been	1067	This function only does something when C<EV_VERIFY> support has been
960	compiled in, which is the default for non-minimal builds. It tries to go	1068	compiled in, which is the default for non-minimal builds. It tries to go
961	through all internal structures and checks them for validity. If anything	1069	through all internal structures and checks them for validity. If anything
962	is found to be inconsistent, it will print an error message to standard	1070	is found to be inconsistent, it will print an error message to standard
…		…
973		1081
974	In the following description, uppercase C<TYPE> in names stands for the	1082	In the following description, uppercase C<TYPE> in names stands for the
975	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer	1083	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
976	watchers and C<ev_io_start> for I/O watchers.	1084	watchers and C<ev_io_start> for I/O watchers.
977		1085
978	A watcher is a structure that you create and register to record your	1086	A watcher is an opaque structure that you allocate and register to record
979	interest in some event. For instance, if you want to wait for STDIN to	1087	your interest in some event. To make a concrete example, imagine you want
980	become readable, you would create an C<ev_io> watcher for that:	1088	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1089	for that:
981		1090
982	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)	1091	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
983	{	1092	{
984	ev_io_stop (w);	1093	ev_io_stop (w);
985	ev_unloop (loop, EVUNLOOP_ALL);	1094	ev_break (loop, EVBREAK_ALL);
986	}	1095	}
987		1096
988	struct ev_loop *loop = ev_default_loop (0);	1097	struct ev_loop *loop = ev_default_loop (0);
989		1098
990	ev_io stdin_watcher;	1099	ev_io stdin_watcher;
991		1100
992	ev_init (&stdin_watcher, my_cb);	1101	ev_init (&stdin_watcher, my_cb);
993	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1102	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
994	ev_io_start (loop, &stdin_watcher);	1103	ev_io_start (loop, &stdin_watcher);
995		1104
996	ev_loop (loop, 0);	1105	ev_run (loop, 0);
997		1106
998	As you can see, you are responsible for allocating the memory for your	1107	As you can see, you are responsible for allocating the memory for your
999	watcher structures (and it is I<usually> a bad idea to do this on the	1108	watcher structures (and it is I<usually> a bad idea to do this on the
1000	stack).	1109	stack).
1001		1110
1002	Each watcher has an associated watcher structure (called C<struct ev_TYPE>	1111	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
1003	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).	1112	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
1004		1113
1005	Each watcher structure must be initialised by a call to C<ev_init	1114	Each watcher structure must be initialised by a call to C<ev_init (watcher
1006	(watcher *, callback)>, which expects a callback to be provided. This	1115	*, callback)>, which expects a callback to be provided. This callback is
1007	callback gets invoked each time the event occurs (or, in the case of I/O	1116	invoked each time the event occurs (or, in the case of I/O watchers, each
1008	watchers, each time the event loop detects that the file descriptor given	1117	time the event loop detects that the file descriptor given is readable
1009	is readable and/or writable).	1118	and/or writable).
1010		1119
1011	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>	1120	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
1012	macro to configure it, with arguments specific to the watcher type. There	1121	macro to configure it, with arguments specific to the watcher type. There
1013	is also a macro to combine initialisation and setting in one call: C<<	1122	is also a macro to combine initialisation and setting in one call: C<<
1014	ev_TYPE_init (watcher *, callback, ...) >>.	1123	ev_TYPE_init (watcher *, callback, ...) >>.
…		…
1065		1174
1066	=item C<EV_PREPARE>	1175	=item C<EV_PREPARE>
1067		1176
1068	=item C<EV_CHECK>	1177	=item C<EV_CHECK>
1069		1178
1070	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1179	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1071	to gather new events, and all C<ev_check> watchers are invoked just after	1180	gather new events, and all C<ev_check> watchers are queued (not invoked)
1072	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1181	just after C<ev_run> has gathered them, but before it queues any callbacks
		1182	for any received events. That means C<ev_prepare> watchers are the last
		1183	watchers invoked before the event loop sleeps or polls for new events, and
		1184	C<ev_check> watchers will be invoked before any other watchers of the same
		1185	or lower priority within an event loop iteration.
		1186
1073	received events. Callbacks of both watcher types can start and stop as	1187	Callbacks of both watcher types can start and stop as many watchers as
1074	many watchers as they want, and all of them will be taken into account	1188	they want, and all of them will be taken into account (for example, a
1075	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1189	C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1076	C<ev_loop> from blocking).	1190	blocking).
1077		1191
1078	=item C<EV_EMBED>	1192	=item C<EV_EMBED>
1079		1193
1080	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1194	The embedded event loop specified in the C<ev_embed> watcher needs attention.
1081		1195
1082	=item C<EV_FORK>	1196	=item C<EV_FORK>
1083		1197
1084	The event loop has been resumed in the child process after fork (see	1198	The event loop has been resumed in the child process after fork (see
1085	C<ev_fork>).	1199	C<ev_fork>).
		1200
		1201	=item C<EV_CLEANUP>
		1202
		1203	The event loop is about to be destroyed (see C<ev_cleanup>).
1086		1204
1087	=item C<EV_ASYNC>	1205	=item C<EV_ASYNC>
1088		1206
1089	The given async watcher has been asynchronously notified (see C<ev_async>).	1207	The given async watcher has been asynchronously notified (see C<ev_async>).
1090		1208
…		…
1263	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related	1381	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1264	functions that do not need a watcher.	1382	functions that do not need a watcher.
1265		1383
1266	=back	1384	=back
1267		1385
		1386	See also the L<ASSOCIATING CUSTOM DATA WITH A WATCHER> and L<BUILDING YOUR
		1387	OWN COMPOSITE WATCHERS> idioms.
1268		1388
1269	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1389	=head2 WATCHER STATES
1270		1390
1271	Each watcher has, by default, a member C<void *data> that you can change	1391	There are various watcher states mentioned throughout this manual -
1272	and read at any time: libev will completely ignore it. This can be used	1392	active, pending and so on. In this section these states and the rules to
1273	to associate arbitrary data with your watcher. If you need more data and	1393	transition between them will be described in more detail - and while these
1274	don't want to allocate memory and store a pointer to it in that data	1394	rules might look complicated, they usually do "the right thing".
1275	member, you can also "subclass" the watcher type and provide your own
1276	data:
1277		1395
1278	struct my_io	1396	=over 4
1279	{
1280	ev_io io;
1281	int otherfd;
1282	void *somedata;
1283	struct whatever *mostinteresting;
1284	};
1285		1397
1286	...	1398	=item initialiased
1287	struct my_io w;
1288	ev_io_init (&w.io, my_cb, fd, EV_READ);
1289		1399
1290	And since your callback will be called with a pointer to the watcher, you	1400	Before a watcher can be registered with the event loop it has to be
1291	can cast it back to your own type:	1401	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1402	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1292		1403
1293	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)	1404	In this state it is simply some block of memory that is suitable for
1294	{	1405	use in an event loop. It can be moved around, freed, reused etc. at
1295	struct my_io *w = (struct my_io *)w_;	1406	will - as long as you either keep the memory contents intact, or call
1296	...	1407	C<ev_TYPE_init> again.
1297	}
1298		1408
1299	More interesting and less C-conformant ways of casting your callback type	1409	=item started/running/active
1300	instead have been omitted.
1301		1410
1302	Another common scenario is to use some data structure with multiple	1411	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1303	embedded watchers:	1412	property of the event loop, and is actively waiting for events. While in
		1413	this state it cannot be accessed (except in a few documented ways), moved,
		1414	freed or anything else - the only legal thing is to keep a pointer to it,
		1415	and call libev functions on it that are documented to work on active watchers.
1304		1416
1305	struct my_biggy	1417	=item pending
1306	{
1307	int some_data;
1308	ev_timer t1;
1309	ev_timer t2;
1310	}
1311		1418
1312	In this case getting the pointer to C<my_biggy> is a bit more	1419	If a watcher is active and libev determines that an event it is interested
1313	complicated: Either you store the address of your C<my_biggy> struct	1420	in has occurred (such as a timer expiring), it will become pending. It will
1314	in the C<data> member of the watcher (for woozies), or you need to use	1421	stay in this pending state until either it is stopped or its callback is
1315	some pointer arithmetic using C<offsetof> inside your watchers (for real	1422	about to be invoked, so it is not normally pending inside the watcher
1316	programmers):	1423	callback.
1317		1424
1318	#include <stddef.h>	1425	The watcher might or might not be active while it is pending (for example,
		1426	an expired non-repeating timer can be pending but no longer active). If it
		1427	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1428	but it is still property of the event loop at this time, so cannot be
		1429	moved, freed or reused. And if it is active the rules described in the
		1430	previous item still apply.
1319		1431
1320	static void	1432	It is also possible to feed an event on a watcher that is not active (e.g.
1321	t1_cb (EV_P_ ev_timer *w, int revents)	1433	via C<ev_feed_event>), in which case it becomes pending without being
1322	{	1434	active.
1323	struct my_biggy big = (struct my_biggy *)
1324	(((char *)w) - offsetof (struct my_biggy, t1));
1325	}
1326		1435
1327	static void	1436	=item stopped
1328	t2_cb (EV_P_ ev_timer *w, int revents)	1437
1329	{	1438	A watcher can be stopped implicitly by libev (in which case it might still
1330	struct my_biggy big = (struct my_biggy *)	1439	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
1331	(((char *)w) - offsetof (struct my_biggy, t2));	1440	latter will clear any pending state the watcher might be in, regardless
1332	}	1441	of whether it was active or not, so stopping a watcher explicitly before
		1442	freeing it is often a good idea.
		1443
		1444	While stopped (and not pending) the watcher is essentially in the
		1445	initialised state, that is, it can be reused, moved, modified in any way
		1446	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1447	it again).
		1448
		1449	=back
1333		1450
1334	=head2 WATCHER PRIORITY MODELS	1451	=head2 WATCHER PRIORITY MODELS
1335		1452
1336	Many event loops support I<watcher priorities>, which are usually small	1453	Many event loops support I<watcher priorities>, which are usually small
1337	integers that influence the ordering of event callback invocation	1454	integers that influence the ordering of event callback invocation
…		…
1464	In general you can register as many read and/or write event watchers per	1581	In general you can register as many read and/or write event watchers per
1465	fd as you want (as long as you don't confuse yourself). Setting all file	1582	fd as you want (as long as you don't confuse yourself). Setting all file
1466	descriptors to non-blocking mode is also usually a good idea (but not	1583	descriptors to non-blocking mode is also usually a good idea (but not
1467	required if you know what you are doing).	1584	required if you know what you are doing).
1468		1585
1469	If you cannot use non-blocking mode, then force the use of a
1470	known-to-be-good backend (at the time of this writing, this includes only
1471	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1472	descriptors for which non-blocking operation makes no sense (such as
1473	files) - libev doesn't guarantee any specific behaviour in that case.
1474
1475	Another thing you have to watch out for is that it is quite easy to	1586	Another thing you have to watch out for is that it is quite easy to
1476	receive "spurious" readiness notifications, that is your callback might	1587	receive "spurious" readiness notifications, that is, your callback might
1477	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1588	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1478	because there is no data. Not only are some backends known to create a	1589	because there is no data. It is very easy to get into this situation even
1479	lot of those (for example Solaris ports), it is very easy to get into	1590	with a relatively standard program structure. Thus it is best to always
1480	this situation even with a relatively standard program structure. Thus	1591	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1481	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1482	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1592	preferable to a program hanging until some data arrives.
1483		1593
1484	If you cannot run the fd in non-blocking mode (for example you should	1594	If you cannot run the fd in non-blocking mode (for example you should
1485	not play around with an Xlib connection), then you have to separately	1595	not play around with an Xlib connection), then you have to separately
1486	re-test whether a file descriptor is really ready with a known-to-be good	1596	re-test whether a file descriptor is really ready with a known-to-be good
1487	interface such as poll (fortunately in our Xlib example, Xlib already	1597	interface such as poll (fortunately in the case of Xlib, it already does
1488	does this on its own, so its quite safe to use). Some people additionally	1598	this on its own, so its quite safe to use). Some people additionally
1489	use C<SIGALRM> and an interval timer, just to be sure you won't block	1599	use C<SIGALRM> and an interval timer, just to be sure you won't block
1490	indefinitely.	1600	indefinitely.
1491		1601
1492	But really, best use non-blocking mode.	1602	But really, best use non-blocking mode.
1493		1603
…		…
1521		1631
1522	There is no workaround possible except not registering events	1632	There is no workaround possible except not registering events
1523	for potentially C<dup ()>'ed file descriptors, or to resort to	1633	for potentially C<dup ()>'ed file descriptors, or to resort to
1524	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1634	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1525		1635
		1636	=head3 The special problem of files
		1637
		1638	Many people try to use C<select> (or libev) on file descriptors
		1639	representing files, and expect it to become ready when their program
		1640	doesn't block on disk accesses (which can take a long time on their own).
		1641
		1642	However, this cannot ever work in the "expected" way - you get a readiness
		1643	notification as soon as the kernel knows whether and how much data is
		1644	there, and in the case of open files, that's always the case, so you
		1645	always get a readiness notification instantly, and your read (or possibly
		1646	write) will still block on the disk I/O.
		1647
		1648	Another way to view it is that in the case of sockets, pipes, character
		1649	devices and so on, there is another party (the sender) that delivers data
		1650	on its own, but in the case of files, there is no such thing: the disk
		1651	will not send data on its own, simply because it doesn't know what you
		1652	wish to read - you would first have to request some data.
		1653
		1654	Since files are typically not-so-well supported by advanced notification
		1655	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1656	to files, even though you should not use it. The reason for this is
		1657	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1658	usually a tty, often a pipe, but also sometimes files or special devices
		1659	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1660	F</dev/urandom>), and even though the file might better be served with
		1661	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1662	it "just works" instead of freezing.
		1663
		1664	So avoid file descriptors pointing to files when you know it (e.g. use
		1665	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1666	when you rarely read from a file instead of from a socket, and want to
		1667	reuse the same code path.
		1668
1526	=head3 The special problem of fork	1669	=head3 The special problem of fork
1527		1670
1528	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1671	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1529	useless behaviour. Libev fully supports fork, but needs to be told about	1672	useless behaviour. Libev fully supports fork, but needs to be told about
1530	it in the child.	1673	it in the child if you want to continue to use it in the child.
1531		1674
1532	To support fork in your programs, you either have to call	1675	To support fork in your child processes, you have to call C<ev_loop_fork
1533	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1676	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1534	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1677	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1535	C<EVBACKEND_POLL>.
1536		1678
1537	=head3 The special problem of SIGPIPE	1679	=head3 The special problem of SIGPIPE
1538		1680
1539	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1681	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1540	when writing to a pipe whose other end has been closed, your program gets	1682	when writing to a pipe whose other end has been closed, your program gets
…		…
1622	...	1764	...
1623	struct ev_loop *loop = ev_default_init (0);	1765	struct ev_loop *loop = ev_default_init (0);
1624	ev_io stdin_readable;	1766	ev_io stdin_readable;
1625	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1767	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1626	ev_io_start (loop, &stdin_readable);	1768	ev_io_start (loop, &stdin_readable);
1627	ev_loop (loop, 0);	1769	ev_run (loop, 0);
1628		1770
1629		1771
1630	=head2 C<ev_timer> - relative and optionally repeating timeouts	1772	=head2 C<ev_timer> - relative and optionally repeating timeouts
1631		1773
1632	Timer watchers are simple relative timers that generate an event after a	1774	Timer watchers are simple relative timers that generate an event after a
…		…
1638	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1780	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1639	monotonic clock option helps a lot here).	1781	monotonic clock option helps a lot here).
1640		1782
1641	The callback is guaranteed to be invoked only I<after> its timeout has	1783	The callback is guaranteed to be invoked only I<after> its timeout has
1642	passed (not I<at>, so on systems with very low-resolution clocks this	1784	passed (not I<at>, so on systems with very low-resolution clocks this
1643	might introduce a small delay). If multiple timers become ready during the	1785	might introduce a small delay, see "the special problem of being too
		1786	early", below). If multiple timers become ready during the same loop
1644	same loop iteration then the ones with earlier time-out values are invoked	1787	iteration then the ones with earlier time-out values are invoked before
1645	before ones of the same priority with later time-out values (but this is	1788	ones of the same priority with later time-out values (but this is no
1646	no longer true when a callback calls C<ev_loop> recursively).	1789	longer true when a callback calls C<ev_run> recursively).
1647		1790
1648	=head3 Be smart about timeouts	1791	=head3 Be smart about timeouts
1649		1792
1650	Many real-world problems involve some kind of timeout, usually for error	1793	Many real-world problems involve some kind of timeout, usually for error
1651	recovery. A typical example is an HTTP request - if the other side hangs,	1794	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1726		1869
1727	In this case, it would be more efficient to leave the C<ev_timer> alone,	1870	In this case, it would be more efficient to leave the C<ev_timer> alone,
1728	but remember the time of last activity, and check for a real timeout only	1871	but remember the time of last activity, and check for a real timeout only
1729	within the callback:	1872	within the callback:
1730		1873
		1874	ev_tstamp timeout = 60.;
1731	ev_tstamp last_activity; // time of last activity	1875	ev_tstamp last_activity; // time of last activity
		1876	ev_timer timer;
1732		1877
1733	static void	1878	static void
1734	callback (EV_P_ ev_timer *w, int revents)	1879	callback (EV_P_ ev_timer *w, int revents)
1735	{	1880	{
1736	ev_tstamp now = ev_now (EV_A);	1881	// calculate when the timeout would happen
1737	ev_tstamp timeout = last_activity + 60.;	1882	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1738		1883
1739	// if last_activity + 60. is older than now, we did time out	1884	// if negative, it means we the timeout already occurred
1740	if (timeout < now)	1885	if (after < 0.)
1741	{	1886	{
1742	// timeout occurred, take action	1887	// timeout occurred, take action
1743	}	1888	}
1744	else	1889	else
1745	{	1890	{
1746	// callback was invoked, but there was some activity, re-arm	1891	// callback was invoked, but there was some recent
1747	// the watcher to fire in last_activity + 60, which is	1892	// activity. simply restart the timer to time out
1748	// guaranteed to be in the future, so "again" is positive:	1893	// after "after" seconds, which is the earliest time
1749	w->repeat = timeout - now;	1894	// the timeout can occur.
		1895	ev_timer_set (w, after, 0.);
1750	ev_timer_again (EV_A_ w);	1896	ev_timer_start (EV_A_ w);
1751	}	1897	}
1752	}	1898	}
1753		1899
1754	To summarise the callback: first calculate the real timeout (defined	1900	To summarise the callback: first calculate in how many seconds the
1755	as "60 seconds after the last activity"), then check if that time has	1901	timeout will occur (by calculating the absolute time when it would occur,
1756	been reached, which means something I<did>, in fact, time out. Otherwise	1902	C<last_activity + timeout>, and subtracting the current time, C<ev_now
1757	the callback was invoked too early (C<timeout> is in the future), so	1903	(EV_A)> from that).
1758	re-schedule the timer to fire at that future time, to see if maybe we have
1759	a timeout then.
1760		1904
1761	Note how C<ev_timer_again> is used, taking advantage of the	1905	If this value is negative, then we are already past the timeout, i.e. we
1762	C<ev_timer_again> optimisation when the timer is already running.	1906	timed out, and need to do whatever is needed in this case.
		1907
		1908	Otherwise, we now the earliest time at which the timeout would trigger,
		1909	and simply start the timer with this timeout value.
		1910
		1911	In other words, each time the callback is invoked it will check whether
		1912	the timeout occurred. If not, it will simply reschedule itself to check
		1913	again at the earliest time it could time out. Rinse. Repeat.
1763		1914
1764	This scheme causes more callback invocations (about one every 60 seconds	1915	This scheme causes more callback invocations (about one every 60 seconds
1765	minus half the average time between activity), but virtually no calls to	1916	minus half the average time between activity), but virtually no calls to
1766	libev to change the timeout.	1917	libev to change the timeout.
1767		1918
1768	To start the timer, simply initialise the watcher and set C<last_activity>	1919	To start the machinery, simply initialise the watcher and set
1769	to the current time (meaning we just have some activity :), then call the	1920	C<last_activity> to the current time (meaning there was some activity just
1770	callback, which will "do the right thing" and start the timer:	1921	now), then call the callback, which will "do the right thing" and start
		1922	the timer:
1771		1923
		1924	last_activity = ev_now (EV_A);
1772	ev_init (timer, callback);	1925	ev_init (&timer, callback);
1773	last_activity = ev_now (loop);	1926	callback (EV_A_ &timer, 0);
1774	callback (loop, timer, EV_TIMER);
1775		1927
1776	And when there is some activity, simply store the current time in	1928	When there is some activity, simply store the current time in
1777	C<last_activity>, no libev calls at all:	1929	C<last_activity>, no libev calls at all:
1778		1930
		1931	if (activity detected)
1779	last_activity = ev_now (loop);	1932	last_activity = ev_now (EV_A);
		1933
		1934	When your timeout value changes, then the timeout can be changed by simply
		1935	providing a new value, stopping the timer and calling the callback, which
		1936	will again do the right thing (for example, time out immediately :).
		1937
		1938	timeout = new_value;
		1939	ev_timer_stop (EV_A_ &timer);
		1940	callback (EV_A_ &timer, 0);
1780		1941
1781	This technique is slightly more complex, but in most cases where the	1942	This technique is slightly more complex, but in most cases where the
1782	time-out is unlikely to be triggered, much more efficient.	1943	time-out is unlikely to be triggered, much more efficient.
1783
1784	Changing the timeout is trivial as well (if it isn't hard-coded in the
1785	callback :) - just change the timeout and invoke the callback, which will
1786	fix things for you.
1787		1944
1788	=item 4. Wee, just use a double-linked list for your timeouts.	1945	=item 4. Wee, just use a double-linked list for your timeouts.
1789		1946
1790	If there is not one request, but many thousands (millions...), all	1947	If there is not one request, but many thousands (millions...), all
1791	employing some kind of timeout with the same timeout value, then one can	1948	employing some kind of timeout with the same timeout value, then one can
…		…
1818	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is	1975	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1819	rather complicated, but extremely efficient, something that really pays	1976	rather complicated, but extremely efficient, something that really pays
1820	off after the first million or so of active timers, i.e. it's usually	1977	off after the first million or so of active timers, i.e. it's usually
1821	overkill :)	1978	overkill :)
1822		1979
		1980	=head3 The special problem of being too early
		1981
		1982	If you ask a timer to call your callback after three seconds, then
		1983	you expect it to be invoked after three seconds - but of course, this
		1984	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		1985	guaranteed to any precision by libev - imagine somebody suspending the
		1986	process with a STOP signal for a few hours for example.
		1987
		1988	So, libev tries to invoke your callback as soon as possible I<after> the
		1989	delay has occurred, but cannot guarantee this.
		1990
		1991	A less obvious failure mode is calling your callback too early: many event
		1992	loops compare timestamps with a "elapsed delay >= requested delay", but
		1993	this can cause your callback to be invoked much earlier than you would
		1994	expect.
		1995
		1996	To see why, imagine a system with a clock that only offers full second
		1997	resolution (think windows if you can't come up with a broken enough OS
		1998	yourself). If you schedule a one-second timer at the time 500.9, then the
		1999	event loop will schedule your timeout to elapse at a system time of 500
		2000	(500.9 truncated to the resolution) + 1, or 501.
		2001
		2002	If an event library looks at the timeout 0.1s later, it will see "501 >=
		2003	501" and invoke the callback 0.1s after it was started, even though a
		2004	one-second delay was requested - this is being "too early", despite best
		2005	intentions.
		2006
		2007	This is the reason why libev will never invoke the callback if the elapsed
		2008	delay equals the requested delay, but only when the elapsed delay is
		2009	larger than the requested delay. In the example above, libev would only invoke
		2010	the callback at system time 502, or 1.1s after the timer was started.
		2011
		2012	So, while libev cannot guarantee that your callback will be invoked
		2013	exactly when requested, it I<can> and I<does> guarantee that the requested
		2014	delay has actually elapsed, or in other words, it always errs on the "too
		2015	late" side of things.
		2016
1823	=head3 The special problem of time updates	2017	=head3 The special problem of time updates
1824		2018
1825	Establishing the current time is a costly operation (it usually takes at	2019	Establishing the current time is a costly operation (it usually takes
1826	least two system calls): EV therefore updates its idea of the current	2020	at least one system call): EV therefore updates its idea of the current
1827	time only before and after C<ev_loop> collects new events, which causes a	2021	time only before and after C<ev_run> collects new events, which causes a
1828	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2022	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1829	lots of events in one iteration.	2023	lots of events in one iteration.
1830		2024
1831	The relative timeouts are calculated relative to the C<ev_now ()>	2025	The relative timeouts are calculated relative to the C<ev_now ()>
1832	time. This is usually the right thing as this timestamp refers to the time	2026	time. This is usually the right thing as this timestamp refers to the time
…		…
1837	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2031	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1838		2032
1839	If the event loop is suspended for a long time, you can also force an	2033	If the event loop is suspended for a long time, you can also force an
1840	update of the time returned by C<ev_now ()> by calling C<ev_now_update	2034	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1841	()>.	2035	()>.
		2036
		2037	=head3 The special problem of unsynchronised clocks
		2038
		2039	Modern systems have a variety of clocks - libev itself uses the normal
		2040	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2041	jumps).
		2042
		2043	Neither of these clocks is synchronised with each other or any other clock
		2044	on the system, so C<ev_time ()> might return a considerably different time
		2045	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2046	a call to C<gettimeofday> might return a second count that is one higher
		2047	than a directly following call to C<time>.
		2048
		2049	The moral of this is to only compare libev-related timestamps with
		2050	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2051	a second or so.
		2052
		2053	One more problem arises due to this lack of synchronisation: if libev uses
		2054	the system monotonic clock and you compare timestamps from C<ev_time>
		2055	or C<ev_now> from when you started your timer and when your callback is
		2056	invoked, you will find that sometimes the callback is a bit "early".
		2057
		2058	This is because C<ev_timer>s work in real time, not wall clock time, so
		2059	libev makes sure your callback is not invoked before the delay happened,
		2060	I<measured according to the real time>, not the system clock.
		2061
		2062	If your timeouts are based on a physical timescale (e.g. "time out this
		2063	connection after 100 seconds") then this shouldn't bother you as it is
		2064	exactly the right behaviour.
		2065
		2066	If you want to compare wall clock/system timestamps to your timers, then
		2067	you need to use C<ev_periodic>s, as these are based on the wall clock
		2068	time, where your comparisons will always generate correct results.
1842		2069
1843	=head3 The special problems of suspended animation	2070	=head3 The special problems of suspended animation
1844		2071
1845	When you leave the server world it is quite customary to hit machines that	2072	When you leave the server world it is quite customary to hit machines that
1846	can suspend/hibernate - what happens to the clocks during such a suspend?	2073	can suspend/hibernate - what happens to the clocks during such a suspend?
…		…
1890	keep up with the timer (because it takes longer than those 10 seconds to	2117	keep up with the timer (because it takes longer than those 10 seconds to
1891	do stuff) the timer will not fire more than once per event loop iteration.	2118	do stuff) the timer will not fire more than once per event loop iteration.
1892		2119
1893	=item ev_timer_again (loop, ev_timer *)	2120	=item ev_timer_again (loop, ev_timer *)
1894		2121
1895	This will act as if the timer timed out and restart it again if it is	2122	This will act as if the timer timed out, and restarts it again if it is
1896	repeating. The exact semantics are:	2123	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2124	timeout to the C<repeat> value and calling C<ev_timer_start>.
1897		2125
		2126	The exact semantics are as in the following rules, all of which will be
		2127	applied to the watcher:
		2128
		2129	=over 4
		2130
1898	If the timer is pending, its pending status is cleared.	2131	=item If the timer is pending, the pending status is always cleared.
1899		2132
1900	If the timer is started but non-repeating, stop it (as if it timed out).	2133	=item If the timer is started but non-repeating, stop it (as if it timed
		2134	out, without invoking it).
1901		2135
1902	If the timer is repeating, either start it if necessary (with the	2136	=item If the timer is repeating, make the C<repeat> value the new timeout
1903	C<repeat> value), or reset the running timer to the C<repeat> value.	2137	and start the timer, if necessary.
		2138
		2139	=back
1904		2140
1905	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a	2141	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1906	usage example.	2142	usage example.
1907		2143
1908	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)	2144	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
…		…
1949	}	2185	}
1950		2186
1951	ev_timer mytimer;	2187	ev_timer mytimer;
1952	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2188	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1953	ev_timer_again (&mytimer); /* start timer */	2189	ev_timer_again (&mytimer); /* start timer */
1954	ev_loop (loop, 0);	2190	ev_run (loop, 0);
1955		2191
1956	// and in some piece of code that gets executed on any "activity":	2192	// and in some piece of code that gets executed on any "activity":
1957	// reset the timeout to start ticking again at 10 seconds	2193	// reset the timeout to start ticking again at 10 seconds
1958	ev_timer_again (&mytimer);	2194	ev_timer_again (&mytimer);
1959		2195
…		…
1985		2221
1986	As with timers, the callback is guaranteed to be invoked only when the	2222	As with timers, the callback is guaranteed to be invoked only when the
1987	point in time where it is supposed to trigger has passed. If multiple	2223	point in time where it is supposed to trigger has passed. If multiple
1988	timers become ready during the same loop iteration then the ones with	2224	timers become ready during the same loop iteration then the ones with
1989	earlier time-out values are invoked before ones with later time-out values	2225	earlier time-out values are invoked before ones with later time-out values
1990	(but this is no longer true when a callback calls C<ev_loop> recursively).	2226	(but this is no longer true when a callback calls C<ev_run> recursively).
1991		2227
1992	=head3 Watcher-Specific Functions and Data Members	2228	=head3 Watcher-Specific Functions and Data Members
1993		2229
1994	=over 4	2230	=over 4
1995		2231
…		…
2030		2266
2031	Another way to think about it (for the mathematically inclined) is that	2267	Another way to think about it (for the mathematically inclined) is that
2032	C<ev_periodic> will try to run the callback in this mode at the next possible	2268	C<ev_periodic> will try to run the callback in this mode at the next possible
2033	time where C<time = offset (mod interval)>, regardless of any time jumps.	2269	time where C<time = offset (mod interval)>, regardless of any time jumps.
2034		2270
2035	For numerical stability it is preferable that the C<offset> value is near	2271	The C<interval> I<MUST> be positive, and for numerical stability, the
2036	C<ev_now ()> (the current time), but there is no range requirement for	2272	interval value should be higher than C<1/8192> (which is around 100
2037	this value, and in fact is often specified as zero.	2273	microseconds) and C<offset> should be higher than C<0> and should have
		2274	at most a similar magnitude as the current time (say, within a factor of
		2275	ten). Typical values for offset are, in fact, C<0> or something between
		2276	C<0> and C<interval>, which is also the recommended range.
2038		2277
2039	Note also that there is an upper limit to how often a timer can fire (CPU	2278	Note also that there is an upper limit to how often a timer can fire (CPU
2040	speed for example), so if C<interval> is very small then timing stability	2279	speed for example), so if C<interval> is very small then timing stability
2041	will of course deteriorate. Libev itself tries to be exact to be about one	2280	will of course deteriorate. Libev itself tries to be exact to be about one
2042	millisecond (if the OS supports it and the machine is fast enough).	2281	millisecond (if the OS supports it and the machine is fast enough).
…		…
2123	Example: Call a callback every hour, or, more precisely, whenever the	2362	Example: Call a callback every hour, or, more precisely, whenever the
2124	system time is divisible by 3600. The callback invocation times have	2363	system time is divisible by 3600. The callback invocation times have
2125	potentially a lot of jitter, but good long-term stability.	2364	potentially a lot of jitter, but good long-term stability.
2126		2365
2127	static void	2366	static void
2128	clock_cb (struct ev_loop *loop, ev_io *w, int revents)	2367	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
2129	{	2368	{
2130	... its now a full hour (UTC, or TAI or whatever your clock follows)	2369	... its now a full hour (UTC, or TAI or whatever your clock follows)
2131	}	2370	}
2132		2371
2133	ev_periodic hourly_tick;	2372	ev_periodic hourly_tick;
…		…
2156		2395
2157	=head2 C<ev_signal> - signal me when a signal gets signalled!	2396	=head2 C<ev_signal> - signal me when a signal gets signalled!
2158		2397
2159	Signal watchers will trigger an event when the process receives a specific	2398	Signal watchers will trigger an event when the process receives a specific
2160	signal one or more times. Even though signals are very asynchronous, libev	2399	signal one or more times. Even though signals are very asynchronous, libev
2161	will try it's best to deliver signals synchronously, i.e. as part of the	2400	will try its best to deliver signals synchronously, i.e. as part of the
2162	normal event processing, like any other event.	2401	normal event processing, like any other event.
2163		2402
2164	If you want signals to be delivered truly asynchronously, just use	2403	If you want signals to be delivered truly asynchronously, just use
2165	C<sigaction> as you would do without libev and forget about sharing	2404	C<sigaction> as you would do without libev and forget about sharing
2166	the signal. You can even use C<ev_async> from a signal handler to	2405	the signal. You can even use C<ev_async> from a signal handler to
…		…
2185	=head3 The special problem of inheritance over fork/execve/pthread_create	2424	=head3 The special problem of inheritance over fork/execve/pthread_create
2186		2425
2187	Both the signal mask (C<sigprocmask>) and the signal disposition	2426	Both the signal mask (C<sigprocmask>) and the signal disposition
2188	(C<sigaction>) are unspecified after starting a signal watcher (and after	2427	(C<sigaction>) are unspecified after starting a signal watcher (and after
2189	stopping it again), that is, libev might or might not block the signal,	2428	stopping it again), that is, libev might or might not block the signal,
2190	and might or might not set or restore the installed signal handler.	2429	and might or might not set or restore the installed signal handler (but
		2430	see C<EVFLAG_NOSIGMASK>).
2191		2431
2192	While this does not matter for the signal disposition (libev never	2432	While this does not matter for the signal disposition (libev never
2193	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on	2433	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2194	C<execve>), this matters for the signal mask: many programs do not expect	2434	C<execve>), this matters for the signal mask: many programs do not expect
2195	certain signals to be blocked.	2435	certain signals to be blocked.
…		…
2209		2449
2210	So I can't stress this enough: I<If you do not reset your signal mask when	2450	So I can't stress this enough: I<If you do not reset your signal mask when
2211	you expect it to be empty, you have a race condition in your code>. This	2451	you expect it to be empty, you have a race condition in your code>. This
2212	is not a libev-specific thing, this is true for most event libraries.	2452	is not a libev-specific thing, this is true for most event libraries.
2213		2453
		2454	=head3 The special problem of threads signal handling
		2455
		2456	POSIX threads has problematic signal handling semantics, specifically,
		2457	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2458	threads in a process block signals, which is hard to achieve.
		2459
		2460	When you want to use sigwait (or mix libev signal handling with your own
		2461	for the same signals), you can tackle this problem by globally blocking
		2462	all signals before creating any threads (or creating them with a fully set
		2463	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2464	loops. Then designate one thread as "signal receiver thread" which handles
		2465	these signals. You can pass on any signals that libev might be interested
		2466	in by calling C<ev_feed_signal>.
		2467
2214	=head3 Watcher-Specific Functions and Data Members	2468	=head3 Watcher-Specific Functions and Data Members
2215		2469
2216	=over 4	2470	=over 4
2217		2471
2218	=item ev_signal_init (ev_signal *, callback, int signum)	2472	=item ev_signal_init (ev_signal *, callback, int signum)
…		…
2233	Example: Try to exit cleanly on SIGINT.	2487	Example: Try to exit cleanly on SIGINT.
2234		2488
2235	static void	2489	static void
2236	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)	2490	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
2237	{	2491	{
2238	ev_unloop (loop, EVUNLOOP_ALL);	2492	ev_break (loop, EVBREAK_ALL);
2239	}	2493	}
2240		2494
2241	ev_signal signal_watcher;	2495	ev_signal signal_watcher;
2242	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2496	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
2243	ev_signal_start (loop, &signal_watcher);	2497	ev_signal_start (loop, &signal_watcher);
…		…
2593	Apart from keeping your process non-blocking (which is a useful	2847	Apart from keeping your process non-blocking (which is a useful
2594	effect on its own sometimes), idle watchers are a good place to do	2848	effect on its own sometimes), idle watchers are a good place to do
2595	"pseudo-background processing", or delay processing stuff to after the	2849	"pseudo-background processing", or delay processing stuff to after the
2596	event loop has handled all outstanding events.	2850	event loop has handled all outstanding events.
2597		2851
		2852	=head3 Abusing an C<ev_idle> watcher for its side-effect
		2853
		2854	As long as there is at least one active idle watcher, libev will never
		2855	sleep unnecessarily. Or in other words, it will loop as fast as possible.
		2856	For this to work, the idle watcher doesn't need to be invoked at all - the
		2857	lowest priority will do.
		2858
		2859	This mode of operation can be useful together with an C<ev_check> watcher,
		2860	to do something on each event loop iteration - for example to balance load
		2861	between different connections.
		2862
		2863	See L<Abusing an C<ev_check> watcher for its side-effect> for a longer
		2864	example.
		2865
2598	=head3 Watcher-Specific Functions and Data Members	2866	=head3 Watcher-Specific Functions and Data Members
2599		2867
2600	=over 4	2868	=over 4
2601		2869
2602	=item ev_idle_init (ev_idle *, callback)	2870	=item ev_idle_init (ev_idle *, callback)
…		…
2625	ev_idle_start (loop, idle_watcher);	2893	ev_idle_start (loop, idle_watcher);
2626		2894
2627		2895
2628	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2896	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
2629		2897
2630	Prepare and check watchers are usually (but not always) used in pairs:	2898	Prepare and check watchers are often (but not always) used in pairs:
2631	prepare watchers get invoked before the process blocks and check watchers	2899	prepare watchers get invoked before the process blocks and check watchers
2632	afterwards.	2900	afterwards.
2633		2901
2634	You I<must not> call C<ev_loop> or similar functions that enter	2902	You I<must not> call C<ev_run> or similar functions that enter
2635	the current event loop from either C<ev_prepare> or C<ev_check>	2903	the current event loop from either C<ev_prepare> or C<ev_check>
2636	watchers. Other loops than the current one are fine, however. The	2904	watchers. Other loops than the current one are fine, however. The
2637	rationale behind this is that you do not need to check for recursion in	2905	rationale behind this is that you do not need to check for recursion in
2638	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2906	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
2639	C<ev_check> so if you have one watcher of each kind they will always be	2907	C<ev_check> so if you have one watcher of each kind they will always be
…		…
2663	with priority higher than or equal to the event loop and one coroutine	2931	with priority higher than or equal to the event loop and one coroutine
2664	of lower priority, but only once, using idle watchers to keep the event	2932	of lower priority, but only once, using idle watchers to keep the event
2665	loop from blocking if lower-priority coroutines are active, thus mapping	2933	loop from blocking if lower-priority coroutines are active, thus mapping
2666	low-priority coroutines to idle/background tasks).	2934	low-priority coroutines to idle/background tasks).
2667		2935
2668	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2936	When used for this purpose, it is recommended to give C<ev_check> watchers
2669	priority, to ensure that they are being run before any other watchers	2937	highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
2670	after the poll (this doesn't matter for C<ev_prepare> watchers).	2938	any other watchers after the poll (this doesn't matter for C<ev_prepare>
		2939	watchers).
2671		2940
2672	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not	2941	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
2673	activate ("feed") events into libev. While libev fully supports this, they	2942	activate ("feed") events into libev. While libev fully supports this, they
2674	might get executed before other C<ev_check> watchers did their job. As	2943	might get executed before other C<ev_check> watchers did their job. As
2675	C<ev_check> watchers are often used to embed other (non-libev) event	2944	C<ev_check> watchers are often used to embed other (non-libev) event
2676	loops those other event loops might be in an unusable state until their	2945	loops those other event loops might be in an unusable state until their
2677	C<ev_check> watcher ran (always remind yourself to coexist peacefully with	2946	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
2678	others).	2947	others).
		2948
		2949	=head3 Abusing an C<ev_check> watcher for its side-effect
		2950
		2951	C<ev_check> (and less often also C<ev_prepare>) watchers can also be
		2952	useful because they are called once per event loop iteration. For
		2953	example, if you want to handle a large number of connections fairly, you
		2954	normally only do a bit of work for each active connection, and if there
		2955	is more work to do, you wait for the next event loop iteration, so other
		2956	connections have a chance of making progress.
		2957
		2958	Using an C<ev_check> watcher is almost enough: it will be called on the
		2959	next event loop iteration. However, that isn't as soon as possible -
		2960	without external events, your C<ev_check> watcher will not be invoked.
		2961
		2962
		2963	This is where C<ev_idle> watchers come in handy - all you need is a
		2964	single global idle watcher that is active as long as you have one active
		2965	C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
		2966	will not sleep, and the C<ev_check> watcher makes sure a callback gets
		2967	invoked. Neither watcher alone can do that.
2679		2968
2680	=head3 Watcher-Specific Functions and Data Members	2969	=head3 Watcher-Specific Functions and Data Members
2681		2970
2682	=over 4	2971	=over 4
2683		2972
…		…
2807		3096
2808	if (timeout >= 0)	3097	if (timeout >= 0)
2809	// create/start timer	3098	// create/start timer
2810		3099
2811	// poll	3100	// poll
2812	ev_loop (EV_A_ 0);	3101	ev_run (EV_A_ 0);
2813		3102
2814	// stop timer again	3103	// stop timer again
2815	if (timeout >= 0)	3104	if (timeout >= 0)
2816	ev_timer_stop (EV_A_ &to);	3105	ev_timer_stop (EV_A_ &to);
2817		3106
…		…
2895	if you do not want that, you need to temporarily stop the embed watcher).	3184	if you do not want that, you need to temporarily stop the embed watcher).
2896		3185
2897	=item ev_embed_sweep (loop, ev_embed *)	3186	=item ev_embed_sweep (loop, ev_embed *)
2898		3187
2899	Make a single, non-blocking sweep over the embedded loop. This works	3188	Make a single, non-blocking sweep over the embedded loop. This works
2900	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3189	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
2901	appropriate way for embedded loops.	3190	appropriate way for embedded loops.
2902		3191
2903	=item struct ev_loop *other [read-only]	3192	=item struct ev_loop *other [read-only]
2904		3193
2905	The embedded event loop.	3194	The embedded event loop.
…		…
2965	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3254	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2966	handlers will be invoked, too, of course.	3255	handlers will be invoked, too, of course.
2967		3256
2968	=head3 The special problem of life after fork - how is it possible?	3257	=head3 The special problem of life after fork - how is it possible?
2969		3258
2970	Most uses of C<fork()> consist of forking, then some simple calls to ste	3259	Most uses of C<fork()> consist of forking, then some simple calls to set
2971	up/change the process environment, followed by a call to C<exec()>. This	3260	up/change the process environment, followed by a call to C<exec()>. This
2972	sequence should be handled by libev without any problems.	3261	sequence should be handled by libev without any problems.
2973		3262
2974	This changes when the application actually wants to do event handling	3263	This changes when the application actually wants to do event handling
2975	in the child, or both parent in child, in effect "continuing" after the	3264	in the child, or both parent in child, in effect "continuing" after the
…		…
2991	disadvantage of having to use multiple event loops (which do not support	3280	disadvantage of having to use multiple event loops (which do not support
2992	signal watchers).	3281	signal watchers).
2993		3282
2994	When this is not possible, or you want to use the default loop for	3283	When this is not possible, or you want to use the default loop for
2995	other reasons, then in the process that wants to start "fresh", call	3284	other reasons, then in the process that wants to start "fresh", call
2996	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying	3285	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
2997	the default loop will "orphan" (not stop) all registered watchers, so you	3286	Destroying the default loop will "orphan" (not stop) all registered
2998	have to be careful not to execute code that modifies those watchers. Note	3287	watchers, so you have to be careful not to execute code that modifies
2999	also that in that case, you have to re-register any signal watchers.	3288	those watchers. Note also that in that case, you have to re-register any
		3289	signal watchers.
3000		3290
3001	=head3 Watcher-Specific Functions and Data Members	3291	=head3 Watcher-Specific Functions and Data Members
3002		3292
3003	=over 4	3293	=over 4
3004		3294
3005	=item ev_fork_init (ev_signal *, callback)	3295	=item ev_fork_init (ev_fork *, callback)
3006		3296
3007	Initialises and configures the fork watcher - it has no parameters of any	3297	Initialises and configures the fork watcher - it has no parameters of any
3008	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3298	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
3009	believe me.	3299	really.
3010		3300
3011	=back	3301	=back
3012		3302
3013		3303
		3304	=head2 C<ev_cleanup> - even the best things end
		3305
		3306	Cleanup watchers are called just before the event loop is being destroyed
		3307	by a call to C<ev_loop_destroy>.
		3308
		3309	While there is no guarantee that the event loop gets destroyed, cleanup
		3310	watchers provide a convenient method to install cleanup hooks for your
		3311	program, worker threads and so on - you just to make sure to destroy the
		3312	loop when you want them to be invoked.
		3313
		3314	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3315	all other watchers, they do not keep a reference to the event loop (which
		3316	makes a lot of sense if you think about it). Like all other watchers, you
		3317	can call libev functions in the callback, except C<ev_cleanup_start>.
		3318
		3319	=head3 Watcher-Specific Functions and Data Members
		3320
		3321	=over 4
		3322
		3323	=item ev_cleanup_init (ev_cleanup *, callback)
		3324
		3325	Initialises and configures the cleanup watcher - it has no parameters of
		3326	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3327	pointless, I assure you.
		3328
		3329	=back
		3330
		3331	Example: Register an atexit handler to destroy the default loop, so any
		3332	cleanup functions are called.
		3333
		3334	static void
		3335	program_exits (void)
		3336	{
		3337	ev_loop_destroy (EV_DEFAULT_UC);
		3338	}
		3339
		3340	...
		3341	atexit (program_exits);
		3342
		3343
3014	=head2 C<ev_async> - how to wake up another event loop	3344	=head2 C<ev_async> - how to wake up an event loop
3015		3345
3016	In general, you cannot use an C<ev_loop> from multiple threads or other	3346	In general, you cannot use an C<ev_loop> from multiple threads or other
3017	asynchronous sources such as signal handlers (as opposed to multiple event	3347	asynchronous sources such as signal handlers (as opposed to multiple event
3018	loops - those are of course safe to use in different threads).	3348	loops - those are of course safe to use in different threads).
3019		3349
3020	Sometimes, however, you need to wake up another event loop you do not	3350	Sometimes, however, you need to wake up an event loop you do not control,
3021	control, for example because it belongs to another thread. This is what	3351	for example because it belongs to another thread. This is what C<ev_async>
3022	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3352	watchers do: as long as the C<ev_async> watcher is active, you can signal
3023	can signal it by calling C<ev_async_send>, which is thread- and signal	3353	it by calling C<ev_async_send>, which is thread- and signal safe.
3024	safe.
3025		3354
3026	This functionality is very similar to C<ev_signal> watchers, as signals,	3355	This functionality is very similar to C<ev_signal> watchers, as signals,
3027	too, are asynchronous in nature, and signals, too, will be compressed	3356	too, are asynchronous in nature, and signals, too, will be compressed
3028	(i.e. the number of callback invocations may be less than the number of	3357	(i.e. the number of callback invocations may be less than the number of
3029	C<ev_async_sent> calls).	3358	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
3030		3359	of "global async watchers" by using a watcher on an otherwise unused
3031	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not	3360	signal, and C<ev_feed_signal> to signal this watcher from another thread,
3032	just the default loop.	3361	even without knowing which loop owns the signal.
3033		3362
3034	=head3 Queueing	3363	=head3 Queueing
3035		3364
3036	C<ev_async> does not support queueing of data in any way. The reason	3365	C<ev_async> does not support queueing of data in any way. The reason
3037	is that the author does not know of a simple (or any) algorithm for a	3366	is that the author does not know of a simple (or any) algorithm for a
…		…
3129	trust me.	3458	trust me.
3130		3459
3131	=item ev_async_send (loop, ev_async *)	3460	=item ev_async_send (loop, ev_async *)
3132		3461
3133	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3462	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3134	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3463	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3464	returns.
		3465
3135	C<ev_feed_event>, this call is safe to do from other threads, signal or	3466	Unlike C<ev_feed_event>, this call is safe to do from other threads,
3136	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3467	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3137	section below on what exactly this means).	3468	embedding section below on what exactly this means).
3138		3469
3139	Note that, as with other watchers in libev, multiple events might get	3470	Note that, as with other watchers in libev, multiple events might get
3140	compressed into a single callback invocation (another way to look at this	3471	compressed into a single callback invocation (another way to look at
3141	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,	3472	this is that C<ev_async> watchers are level-triggered: they are set on
3142	reset when the event loop detects that).	3473	C<ev_async_send>, reset when the event loop detects that).
3143		3474
3144	This call incurs the overhead of a system call only once per event loop	3475	This call incurs the overhead of at most one extra system call per event
3145	iteration, so while the overhead might be noticeable, it doesn't apply to	3476	loop iteration, if the event loop is blocked, and no syscall at all if
3146	repeated calls to C<ev_async_send> for the same event loop.	3477	the event loop (or your program) is processing events. That means that
		3478	repeated calls are basically free (there is no need to avoid calls for
		3479	performance reasons) and that the overhead becomes smaller (typically
		3480	zero) under load.
3147		3481
3148	=item bool = ev_async_pending (ev_async *)	3482	=item bool = ev_async_pending (ev_async *)
3149		3483
3150	Returns a non-zero value when C<ev_async_send> has been called on the	3484	Returns a non-zero value when C<ev_async_send> has been called on the
3151	watcher but the event has not yet been processed (or even noted) by the	3485	watcher but the event has not yet been processed (or even noted) by the
…		…
3206	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3540	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3207		3541
3208	=item ev_feed_fd_event (loop, int fd, int revents)	3542	=item ev_feed_fd_event (loop, int fd, int revents)
3209		3543
3210	Feed an event on the given fd, as if a file descriptor backend detected	3544	Feed an event on the given fd, as if a file descriptor backend detected
3211	the given events it.	3545	the given events.
3212		3546
3213	=item ev_feed_signal_event (loop, int signum)	3547	=item ev_feed_signal_event (loop, int signum)
3214		3548
3215	Feed an event as if the given signal occurred (C<loop> must be the default	3549	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3216	loop!).	3550	which is async-safe.
3217		3551
3218	=back	3552	=back
		3553
		3554
		3555	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3556
		3557	This section explains some common idioms that are not immediately
		3558	obvious. Note that examples are sprinkled over the whole manual, and this
		3559	section only contains stuff that wouldn't fit anywhere else.
		3560
		3561	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3562
		3563	Each watcher has, by default, a C<void *data> member that you can read
		3564	or modify at any time: libev will completely ignore it. This can be used
		3565	to associate arbitrary data with your watcher. If you need more data and
		3566	don't want to allocate memory separately and store a pointer to it in that
		3567	data member, you can also "subclass" the watcher type and provide your own
		3568	data:
		3569
		3570	struct my_io
		3571	{
		3572	ev_io io;
		3573	int otherfd;
		3574	void *somedata;
		3575	struct whatever *mostinteresting;
		3576	};
		3577
		3578	...
		3579	struct my_io w;
		3580	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3581
		3582	And since your callback will be called with a pointer to the watcher, you
		3583	can cast it back to your own type:
		3584
		3585	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3586	{
		3587	struct my_io *w = (struct my_io *)w_;
		3588	...
		3589	}
		3590
		3591	More interesting and less C-conformant ways of casting your callback
		3592	function type instead have been omitted.
		3593
		3594	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3595
		3596	Another common scenario is to use some data structure with multiple
		3597	embedded watchers, in effect creating your own watcher that combines
		3598	multiple libev event sources into one "super-watcher":
		3599
		3600	struct my_biggy
		3601	{
		3602	int some_data;
		3603	ev_timer t1;
		3604	ev_timer t2;
		3605	}
		3606
		3607	In this case getting the pointer to C<my_biggy> is a bit more
		3608	complicated: Either you store the address of your C<my_biggy> struct in
		3609	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3610	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3611	real programmers):
		3612
		3613	#include <stddef.h>
		3614
		3615	static void
		3616	t1_cb (EV_P_ ev_timer *w, int revents)
		3617	{
		3618	struct my_biggy big = (struct my_biggy *)
		3619	(((char *)w) - offsetof (struct my_biggy, t1));
		3620	}
		3621
		3622	static void
		3623	t2_cb (EV_P_ ev_timer *w, int revents)
		3624	{
		3625	struct my_biggy big = (struct my_biggy *)
		3626	(((char *)w) - offsetof (struct my_biggy, t2));
		3627	}
		3628
		3629	=head2 AVOIDING FINISHING BEFORE RETURNING
		3630
		3631	Often you have structures like this in event-based programs:
		3632
		3633	callback ()
		3634	{
		3635	free (request);
		3636	}
		3637
		3638	request = start_new_request (..., callback);
		3639
		3640	The intent is to start some "lengthy" operation. The C<request> could be
		3641	used to cancel the operation, or do other things with it.
		3642
		3643	It's not uncommon to have code paths in C<start_new_request> that
		3644	immediately invoke the callback, for example, to report errors. Or you add
		3645	some caching layer that finds that it can skip the lengthy aspects of the
		3646	operation and simply invoke the callback with the result.
		3647
		3648	The problem here is that this will happen I<before> C<start_new_request>
		3649	has returned, so C<request> is not set.
		3650
		3651	Even if you pass the request by some safer means to the callback, you
		3652	might want to do something to the request after starting it, such as
		3653	canceling it, which probably isn't working so well when the callback has
		3654	already been invoked.
		3655
		3656	A common way around all these issues is to make sure that
		3657	C<start_new_request> I<always> returns before the callback is invoked. If
		3658	C<start_new_request> immediately knows the result, it can artificially
		3659	delay invoking the callback by e.g. using a C<prepare> or C<idle> watcher
		3660	for example, or more sneakily, by reusing an existing (stopped) watcher
		3661	and pushing it into the pending queue:
		3662
		3663	ev_set_cb (watcher, callback);
		3664	ev_feed_event (EV_A_ watcher, 0);
		3665
		3666	This way, C<start_new_request> can safely return before the callback is
		3667	invoked, while not delaying callback invocation too much.
		3668
		3669	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3670
		3671	Often (especially in GUI toolkits) there are places where you have
		3672	I<modal> interaction, which is most easily implemented by recursively
		3673	invoking C<ev_run>.
		3674
		3675	This brings the problem of exiting - a callback might want to finish the
		3676	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3677	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3678	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3679	other combination: In these cases, C<ev_break> will not work alone.
		3680
		3681	The solution is to maintain "break this loop" variable for each C<ev_run>
		3682	invocation, and use a loop around C<ev_run> until the condition is
		3683	triggered, using C<EVRUN_ONCE>:
		3684
		3685	// main loop
		3686	int exit_main_loop = 0;
		3687
		3688	while (!exit_main_loop)
		3689	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3690
		3691	// in a modal watcher
		3692	int exit_nested_loop = 0;
		3693
		3694	while (!exit_nested_loop)
		3695	ev_run (EV_A_ EVRUN_ONCE);
		3696
		3697	To exit from any of these loops, just set the corresponding exit variable:
		3698
		3699	// exit modal loop
		3700	exit_nested_loop = 1;
		3701
		3702	// exit main program, after modal loop is finished
		3703	exit_main_loop = 1;
		3704
		3705	// exit both
		3706	exit_main_loop = exit_nested_loop = 1;
		3707
		3708	=head2 THREAD LOCKING EXAMPLE
		3709
		3710	Here is a fictitious example of how to run an event loop in a different
		3711	thread from where callbacks are being invoked and watchers are
		3712	created/added/removed.
		3713
		3714	For a real-world example, see the C<EV::Loop::Async> perl module,
		3715	which uses exactly this technique (which is suited for many high-level
		3716	languages).
		3717
		3718	The example uses a pthread mutex to protect the loop data, a condition
		3719	variable to wait for callback invocations, an async watcher to notify the
		3720	event loop thread and an unspecified mechanism to wake up the main thread.
		3721
		3722	First, you need to associate some data with the event loop:
		3723
		3724	typedef struct {
		3725	mutex_t lock; /* global loop lock */
		3726	ev_async async_w;
		3727	thread_t tid;
		3728	cond_t invoke_cv;
		3729	} userdata;
		3730
		3731	void prepare_loop (EV_P)
		3732	{
		3733	// for simplicity, we use a static userdata struct.
		3734	static userdata u;
		3735
		3736	ev_async_init (&u->async_w, async_cb);
		3737	ev_async_start (EV_A_ &u->async_w);
		3738
		3739	pthread_mutex_init (&u->lock, 0);
		3740	pthread_cond_init (&u->invoke_cv, 0);
		3741
		3742	// now associate this with the loop
		3743	ev_set_userdata (EV_A_ u);
		3744	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3745	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3746
		3747	// then create the thread running ev_run
		3748	pthread_create (&u->tid, 0, l_run, EV_A);
		3749	}
		3750
		3751	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3752	solely to wake up the event loop so it takes notice of any new watchers
		3753	that might have been added:
		3754
		3755	static void
		3756	async_cb (EV_P_ ev_async *w, int revents)
		3757	{
		3758	// just used for the side effects
		3759	}
		3760
		3761	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3762	protecting the loop data, respectively.
		3763
		3764	static void
		3765	l_release (EV_P)
		3766	{
		3767	userdata *u = ev_userdata (EV_A);
		3768	pthread_mutex_unlock (&u->lock);
		3769	}
		3770
		3771	static void
		3772	l_acquire (EV_P)
		3773	{
		3774	userdata *u = ev_userdata (EV_A);
		3775	pthread_mutex_lock (&u->lock);
		3776	}
		3777
		3778	The event loop thread first acquires the mutex, and then jumps straight
		3779	into C<ev_run>:
		3780
		3781	void *
		3782	l_run (void *thr_arg)
		3783	{
		3784	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3785
		3786	l_acquire (EV_A);
		3787	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3788	ev_run (EV_A_ 0);
		3789	l_release (EV_A);
		3790
		3791	return 0;
		3792	}
		3793
		3794	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3795	signal the main thread via some unspecified mechanism (signals? pipe
		3796	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3797	have been called (in a while loop because a) spurious wakeups are possible
		3798	and b) skipping inter-thread-communication when there are no pending
		3799	watchers is very beneficial):
		3800
		3801	static void
		3802	l_invoke (EV_P)
		3803	{
		3804	userdata *u = ev_userdata (EV_A);
		3805
		3806	while (ev_pending_count (EV_A))
		3807	{
		3808	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3809	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3810	}
		3811	}
		3812
		3813	Now, whenever the main thread gets told to invoke pending watchers, it
		3814	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3815	thread to continue:
		3816
		3817	static void
		3818	real_invoke_pending (EV_P)
		3819	{
		3820	userdata *u = ev_userdata (EV_A);
		3821
		3822	pthread_mutex_lock (&u->lock);
		3823	ev_invoke_pending (EV_A);
		3824	pthread_cond_signal (&u->invoke_cv);
		3825	pthread_mutex_unlock (&u->lock);
		3826	}
		3827
		3828	Whenever you want to start/stop a watcher or do other modifications to an
		3829	event loop, you will now have to lock:
		3830
		3831	ev_timer timeout_watcher;
		3832	userdata *u = ev_userdata (EV_A);
		3833
		3834	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3835
		3836	pthread_mutex_lock (&u->lock);
		3837	ev_timer_start (EV_A_ &timeout_watcher);
		3838	ev_async_send (EV_A_ &u->async_w);
		3839	pthread_mutex_unlock (&u->lock);
		3840
		3841	Note that sending the C<ev_async> watcher is required because otherwise
		3842	an event loop currently blocking in the kernel will have no knowledge
		3843	about the newly added timer. By waking up the loop it will pick up any new
		3844	watchers in the next event loop iteration.
		3845
		3846	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3847
		3848	While the overhead of a callback that e.g. schedules a thread is small, it
		3849	is still an overhead. If you embed libev, and your main usage is with some
		3850	kind of threads or coroutines, you might want to customise libev so that
		3851	doesn't need callbacks anymore.
		3852
		3853	Imagine you have coroutines that you can switch to using a function
		3854	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3855	and that due to some magic, the currently active coroutine is stored in a
		3856	global called C<current_coro>. Then you can build your own "wait for libev
		3857	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3858	the differing C<;> conventions):
		3859
		3860	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3861	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3862
		3863	That means instead of having a C callback function, you store the
		3864	coroutine to switch to in each watcher, and instead of having libev call
		3865	your callback, you instead have it switch to that coroutine.
		3866
		3867	A coroutine might now wait for an event with a function called
		3868	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3869	matter when, or whether the watcher is active or not when this function is
		3870	called):
		3871
		3872	void
		3873	wait_for_event (ev_watcher *w)
		3874	{
		3875	ev_cb_set (w) = current_coro;
		3876	switch_to (libev_coro);
		3877	}
		3878
		3879	That basically suspends the coroutine inside C<wait_for_event> and
		3880	continues the libev coroutine, which, when appropriate, switches back to
		3881	this or any other coroutine.
		3882
		3883	You can do similar tricks if you have, say, threads with an event queue -
		3884	instead of storing a coroutine, you store the queue object and instead of
		3885	switching to a coroutine, you push the watcher onto the queue and notify
		3886	any waiters.
		3887
		3888	To embed libev, see L<EMBEDDING>, but in short, it's easiest to create two
		3889	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3890
		3891	// my_ev.h
		3892	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3893	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb);
		3894	#include "../libev/ev.h"
		3895
		3896	// my_ev.c
		3897	#define EV_H "my_ev.h"
		3898	#include "../libev/ev.c"
		3899
		3900	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3901	F<my_ev.c> into your project. When properly specifying include paths, you
		3902	can even use F<ev.h> as header file name directly.
3219		3903
3220		3904
3221	=head1 LIBEVENT EMULATION	3905	=head1 LIBEVENT EMULATION
3222		3906
3223	Libev offers a compatibility emulation layer for libevent. It cannot	3907	Libev offers a compatibility emulation layer for libevent. It cannot
3224	emulate the internals of libevent, so here are some usage hints:	3908	emulate the internals of libevent, so here are some usage hints:
3225		3909
3226	=over 4	3910	=over 4
		3911
		3912	=item * Only the libevent-1.4.1-beta API is being emulated.
		3913
		3914	This was the newest libevent version available when libev was implemented,
		3915	and is still mostly unchanged in 2010.
3227		3916
3228	=item * Use it by including <event.h>, as usual.	3917	=item * Use it by including <event.h>, as usual.
3229		3918
3230	=item * The following members are fully supported: ev_base, ev_callback,	3919	=item * The following members are fully supported: ev_base, ev_callback,
3231	ev_arg, ev_fd, ev_res, ev_events.	3920	ev_arg, ev_fd, ev_res, ev_events.
…		…
3237	=item * Priorities are not currently supported. Initialising priorities	3926	=item * Priorities are not currently supported. Initialising priorities
3238	will fail and all watchers will have the same priority, even though there	3927	will fail and all watchers will have the same priority, even though there
3239	is an ev_pri field.	3928	is an ev_pri field.
3240		3929
3241	=item * In libevent, the last base created gets the signals, in libev, the	3930	=item * In libevent, the last base created gets the signals, in libev, the
3242	first base created (== the default loop) gets the signals.	3931	base that registered the signal gets the signals.
3243		3932
3244	=item * Other members are not supported.	3933	=item * Other members are not supported.
3245		3934
3246	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3935	=item * The libev emulation is I<not> ABI compatible to libevent, you need
3247	to use the libev header file and library.	3936	to use the libev header file and library.
3248		3937
3249	=back	3938	=back
3250		3939
3251	=head1 C++ SUPPORT	3940	=head1 C++ SUPPORT
		3941
		3942	=head2 C API
		3943
		3944	The normal C API should work fine when used from C++: both ev.h and the
		3945	libev sources can be compiled as C++. Therefore, code that uses the C API
		3946	will work fine.
		3947
		3948	Proper exception specifications might have to be added to callbacks passed
		3949	to libev: exceptions may be thrown only from watcher callbacks, all
		3950	other callbacks (allocator, syserr, loop acquire/release and periodioc
		3951	reschedule callbacks) must not throw exceptions, and might need a C<throw
		3952	()> specification. If you have code that needs to be compiled as both C
		3953	and C++ you can use the C<EV_THROW> macro for this:
		3954
		3955	static void
		3956	fatal_error (const char *msg) EV_THROW
		3957	{
		3958	perror (msg);
		3959	abort ();
		3960	}
		3961
		3962	...
		3963	ev_set_syserr_cb (fatal_error);
		3964
		3965	The only API functions that can currently throw exceptions are C<ev_run>,
		3966	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		3967	because it runs cleanup watchers).
		3968
		3969	Throwing exceptions in watcher callbacks is only supported if libev itself
		3970	is compiled with a C++ compiler or your C and C++ environments allow
		3971	throwing exceptions through C libraries (most do).
		3972
		3973	=head2 C++ API
3252		3974
3253	Libev comes with some simplistic wrapper classes for C++ that mainly allow	3975	Libev comes with some simplistic wrapper classes for C++ that mainly allow
3254	you to use some convenience methods to start/stop watchers and also change	3976	you to use some convenience methods to start/stop watchers and also change
3255	the callback model to a model using method callbacks on objects.	3977	the callback model to a model using method callbacks on objects.
3256		3978
…		…
3266	Care has been taken to keep the overhead low. The only data member the C++	3988	Care has been taken to keep the overhead low. The only data member the C++
3267	classes add (compared to plain C-style watchers) is the event loop pointer	3989	classes add (compared to plain C-style watchers) is the event loop pointer
3268	that the watcher is associated with (or no additional members at all if	3990	that the watcher is associated with (or no additional members at all if
3269	you disable C<EV_MULTIPLICITY> when embedding libev).	3991	you disable C<EV_MULTIPLICITY> when embedding libev).
3270		3992
3271	Currently, functions, and static and non-static member functions can be	3993	Currently, functions, static and non-static member functions and classes
3272	used as callbacks. Other types should be easy to add as long as they only	3994	with C<operator ()> can be used as callbacks. Other types should be easy
3273	need one additional pointer for context. If you need support for other	3995	to add as long as they only need one additional pointer for context. If
3274	types of functors please contact the author (preferably after implementing	3996	you need support for other types of functors please contact the author
3275	it).	3997	(preferably after implementing it).
		3998
		3999	For all this to work, your C++ compiler either has to use the same calling
		4000	conventions as your C compiler (for static member functions), or you have
		4001	to embed libev and compile libev itself as C++.
3276		4002
3277	Here is a list of things available in the C<ev> namespace:	4003	Here is a list of things available in the C<ev> namespace:
3278		4004
3279	=over 4	4005	=over 4
3280		4006
…		…
3290	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.	4016	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
3291		4017
3292	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of	4018	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
3293	the same name in the C<ev> namespace, with the exception of C<ev_signal>	4019	the same name in the C<ev> namespace, with the exception of C<ev_signal>
3294	which is called C<ev::sig> to avoid clashes with the C<signal> macro	4020	which is called C<ev::sig> to avoid clashes with the C<signal> macro
3295	defines by many implementations.	4021	defined by many implementations.
3296		4022
3297	All of those classes have these methods:	4023	All of those classes have these methods:
3298		4024
3299	=over 4	4025	=over 4
3300		4026
…		…
3390	Associates a different C<struct ev_loop> with this watcher. You can only	4116	Associates a different C<struct ev_loop> with this watcher. You can only
3391	do this when the watcher is inactive (and not pending either).	4117	do this when the watcher is inactive (and not pending either).
3392		4118
3393	=item w->set ([arguments])	4119	=item w->set ([arguments])
3394		4120
3395	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be	4121	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
3396	called at least once. Unlike the C counterpart, an active watcher gets	4122	method or a suitable start method must be called at least once. Unlike the
3397	automatically stopped and restarted when reconfiguring it with this	4123	C counterpart, an active watcher gets automatically stopped and restarted
3398	method.	4124	when reconfiguring it with this method.
3399		4125
3400	=item w->start ()	4126	=item w->start ()
3401		4127
3402	Starts the watcher. Note that there is no C<loop> argument, as the	4128	Starts the watcher. Note that there is no C<loop> argument, as the
3403	constructor already stores the event loop.	4129	constructor already stores the event loop.
3404		4130
		4131	=item w->start ([arguments])
		4132
		4133	Instead of calling C<set> and C<start> methods separately, it is often
		4134	convenient to wrap them in one call. Uses the same type of arguments as
		4135	the configure C<set> method of the watcher.
		4136
3405	=item w->stop ()	4137	=item w->stop ()
3406		4138
3407	Stops the watcher if it is active. Again, no C<loop> argument.	4139	Stops the watcher if it is active. Again, no C<loop> argument.
3408		4140
3409	=item w->again () (C<ev::timer>, C<ev::periodic> only)	4141	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
3421		4153
3422	=back	4154	=back
3423		4155
3424	=back	4156	=back
3425		4157
3426	Example: Define a class with an IO and idle watcher, start one of them in	4158	Example: Define a class with two I/O and idle watchers, start the I/O
3427	the constructor.	4159	watchers in the constructor.
3428		4160
3429	class myclass	4161	class myclass
3430	{	4162	{
3431	ev::io io ; void io_cb (ev::io &w, int revents);	4163	ev::io io ; void io_cb (ev::io &w, int revents);
		4164	ev::io io2 ; void io2_cb (ev::io &w, int revents);
3432	ev::idle idle; void idle_cb (ev::idle &w, int revents);	4165	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3433		4166
3434	myclass (int fd)	4167	myclass (int fd)
3435	{	4168	{
3436	io .set <myclass, &myclass::io_cb > (this);	4169	io .set <myclass, &myclass::io_cb > (this);
		4170	io2 .set <myclass, &myclass::io2_cb > (this);
3437	idle.set <myclass, &myclass::idle_cb> (this);	4171	idle.set <myclass, &myclass::idle_cb> (this);
3438		4172
3439	io.start (fd, ev::READ);	4173	io.set (fd, ev::WRITE); // configure the watcher
		4174	io.start (); // start it whenever convenient
		4175
		4176	io2.start (fd, ev::READ); // set + start in one call
3440	}	4177	}
3441	};	4178	};
3442		4179
3443		4180
3444	=head1 OTHER LANGUAGE BINDINGS	4181	=head1 OTHER LANGUAGE BINDINGS
…		…
3483	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.	4220	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
3484		4221
3485	=item D	4222	=item D
3486		4223
3487	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	4224	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
3488	be found at L<http://proj.llucax.com.ar/wiki/evd>.	4225	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
3489		4226
3490	=item Ocaml	4227	=item Ocaml
3491		4228
3492	Erkki Seppala has written Ocaml bindings for libev, to be found at	4229	Erkki Seppala has written Ocaml bindings for libev, to be found at
3493	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	4230	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
…		…
3518	loop argument"). The C<EV_A> form is used when this is the sole argument,	4255	loop argument"). The C<EV_A> form is used when this is the sole argument,
3519	C<EV_A_> is used when other arguments are following. Example:	4256	C<EV_A_> is used when other arguments are following. Example:
3520		4257
3521	ev_unref (EV_A);	4258	ev_unref (EV_A);
3522	ev_timer_add (EV_A_ watcher);	4259	ev_timer_add (EV_A_ watcher);
3523	ev_loop (EV_A_ 0);	4260	ev_run (EV_A_ 0);
3524		4261
3525	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,	4262	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
3526	which is often provided by the following macro.	4263	which is often provided by the following macro.
3527		4264
3528	=item C<EV_P>, C<EV_P_>	4265	=item C<EV_P>, C<EV_P_>
…		…
3541	suitable for use with C<EV_A>.	4278	suitable for use with C<EV_A>.
3542		4279
3543	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4280	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3544		4281
3545	Similar to the other two macros, this gives you the value of the default	4282	Similar to the other two macros, this gives you the value of the default
3546	loop, if multiple loops are supported ("ev loop default").	4283	loop, if multiple loops are supported ("ev loop default"). The default loop
		4284	will be initialised if it isn't already initialised.
		4285
		4286	For non-multiplicity builds, these macros do nothing, so you always have
		4287	to initialise the loop somewhere.
3547		4288
3548	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>	4289	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3549		4290
3550	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the	4291	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3551	default loop has been initialised (C<UC> == unchecked). Their behaviour	4292	default loop has been initialised (C<UC> == unchecked). Their behaviour
…		…
3568	}	4309	}
3569		4310
3570	ev_check check;	4311	ev_check check;
3571	ev_check_init (&check, check_cb);	4312	ev_check_init (&check, check_cb);
3572	ev_check_start (EV_DEFAULT_ &check);	4313	ev_check_start (EV_DEFAULT_ &check);
3573	ev_loop (EV_DEFAULT_ 0);	4314	ev_run (EV_DEFAULT_ 0);
3574		4315
3575	=head1 EMBEDDING	4316	=head1 EMBEDDING
3576		4317
3577	Libev can (and often is) directly embedded into host	4318	Libev can (and often is) directly embedded into host
3578	applications. Examples of applications that embed it include the Deliantra	4319	applications. Examples of applications that embed it include the Deliantra
…		…
3670	users of libev and the libev code itself must be compiled with compatible	4411	users of libev and the libev code itself must be compiled with compatible
3671	settings.	4412	settings.
3672		4413
3673	=over 4	4414	=over 4
3674		4415
		4416	=item EV_COMPAT3 (h)
		4417
		4418	Backwards compatibility is a major concern for libev. This is why this
		4419	release of libev comes with wrappers for the functions and symbols that
		4420	have been renamed between libev version 3 and 4.
		4421
		4422	You can disable these wrappers (to test compatibility with future
		4423	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4424	sources. This has the additional advantage that you can drop the C<struct>
		4425	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4426	typedef in that case.
		4427
		4428	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4429	and in some even more future version the compatibility code will be
		4430	removed completely.
		4431
3675	=item EV_STANDALONE (h)	4432	=item EV_STANDALONE (h)
3676		4433
3677	Must always be C<1> if you do not use autoconf configuration, which	4434	Must always be C<1> if you do not use autoconf configuration, which
3678	keeps libev from including F<config.h>, and it also defines dummy	4435	keeps libev from including F<config.h>, and it also defines dummy
3679	implementations for some libevent functions (such as logging, which is not	4436	implementations for some libevent functions (such as logging, which is not
3680	supported). It will also not define any of the structs usually found in	4437	supported). It will also not define any of the structs usually found in
3681	F<event.h> that are not directly supported by the libev core alone.	4438	F<event.h> that are not directly supported by the libev core alone.
3682		4439
3683	In standalone mode, libev will still try to automatically deduce the	4440	In standalone mode, libev will still try to automatically deduce the
3684	configuration, but has to be more conservative.	4441	configuration, but has to be more conservative.
		4442
		4443	=item EV_USE_FLOOR
		4444
		4445	If defined to be C<1>, libev will use the C<floor ()> function for its
		4446	periodic reschedule calculations, otherwise libev will fall back on a
		4447	portable (slower) implementation. If you enable this, you usually have to
		4448	link against libm or something equivalent. Enabling this when the C<floor>
		4449	function is not available will fail, so the safe default is to not enable
		4450	this.
3685		4451
3686	=item EV_USE_MONOTONIC	4452	=item EV_USE_MONOTONIC
3687		4453
3688	If defined to be C<1>, libev will try to detect the availability of the	4454	If defined to be C<1>, libev will try to detect the availability of the
3689	monotonic clock option at both compile time and runtime. Otherwise no	4455	monotonic clock option at both compile time and runtime. Otherwise no
…		…
3819	If defined to be C<1>, libev will compile in support for the Linux inotify	4585	If defined to be C<1>, libev will compile in support for the Linux inotify
3820	interface to speed up C<ev_stat> watchers. Its actual availability will	4586	interface to speed up C<ev_stat> watchers. Its actual availability will
3821	be detected at runtime. If undefined, it will be enabled if the headers	4587	be detected at runtime. If undefined, it will be enabled if the headers
3822	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4588	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3823		4589
		4590	=item EV_NO_SMP
		4591
		4592	If defined to be C<1>, libev will assume that memory is always coherent
		4593	between threads, that is, threads can be used, but threads never run on
		4594	different cpus (or different cpu cores). This reduces dependencies
		4595	and makes libev faster.
		4596
		4597	=item EV_NO_THREADS
		4598
		4599	If defined to be C<1>, libev will assume that it will never be called
		4600	from different threads, which is a stronger assumption than C<EV_NO_SMP>,
		4601	above. This reduces dependencies and makes libev faster.
		4602
3824	=item EV_ATOMIC_T	4603	=item EV_ATOMIC_T
3825		4604
3826	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	4605	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
3827	access is atomic with respect to other threads or signal contexts. No such	4606	access is atomic and serialised with respect to other threads or signal
3828	type is easily found in the C language, so you can provide your own type	4607	contexts. No such type is easily found in the C language, so you can
3829	that you know is safe for your purposes. It is used both for signal handler "locking"	4608	provide your own type that you know is safe for your purposes. It is used
3830	as well as for signal and thread safety in C<ev_async> watchers.	4609	both for signal handler "locking" as well as for signal and thread safety
		4610	in C<ev_async> watchers.
3831		4611
3832	In the absence of this define, libev will use C<sig_atomic_t volatile>	4612	In the absence of this define, libev will use C<sig_atomic_t volatile>
3833	(from F<signal.h>), which is usually good enough on most platforms.	4613	(from F<signal.h>), which is usually good enough on most platforms,
		4614	although strictly speaking using a type that also implies a memory fence
		4615	is required.
3834		4616
3835	=item EV_H (h)	4617	=item EV_H (h)
3836		4618
3837	The name of the F<ev.h> header file used to include it. The default if	4619	The name of the F<ev.h> header file used to include it. The default if
3838	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	4620	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
…		…
3862	will have the C<struct ev_loop *> as first argument, and you can create	4644	will have the C<struct ev_loop *> as first argument, and you can create
3863	additional independent event loops. Otherwise there will be no support	4645	additional independent event loops. Otherwise there will be no support
3864	for multiple event loops and there is no first event loop pointer	4646	for multiple event loops and there is no first event loop pointer
3865	argument. Instead, all functions act on the single default loop.	4647	argument. Instead, all functions act on the single default loop.
3866		4648
		4649	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4650	default loop when multiplicity is switched off - you always have to
		4651	initialise the loop manually in this case.
		4652
3867	=item EV_MINPRI	4653	=item EV_MINPRI
3868		4654
3869	=item EV_MAXPRI	4655	=item EV_MAXPRI
3870		4656
3871	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4657	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
3885	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,	4671	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
3886	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.	4672	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3887		4673
3888	If undefined or defined to be C<1> (and the platform supports it), then	4674	If undefined or defined to be C<1> (and the platform supports it), then
3889	the respective watcher type is supported. If defined to be C<0>, then it	4675	the respective watcher type is supported. If defined to be C<0>, then it
3890	is not. Disabling watcher types mainly saves codesize.	4676	is not. Disabling watcher types mainly saves code size.
3891		4677
3892	=item EV_FEATURES	4678	=item EV_FEATURES
3893		4679
3894	If you need to shave off some kilobytes of code at the expense of some	4680	If you need to shave off some kilobytes of code at the expense of some
3895	speed (but with the full API), you can define this symbol to request	4681	speed (but with the full API), you can define this symbol to request
…		…
3907	#define EV_USE_POLL 1	4693	#define EV_USE_POLL 1
3908	#define EV_CHILD_ENABLE 1	4694	#define EV_CHILD_ENABLE 1
3909	#define EV_ASYNC_ENABLE 1	4695	#define EV_ASYNC_ENABLE 1
3910		4696
3911	The actual value is a bitset, it can be a combination of the following	4697	The actual value is a bitset, it can be a combination of the following
3912	values:	4698	values (by default, all of these are enabled):
3913		4699
3914	=over 4	4700	=over 4
3915		4701
3916	=item C<1> - faster/larger code	4702	=item C<1> - faster/larger code
3917		4703
3918	Use larger code to speed up some operations.	4704	Use larger code to speed up some operations.
3919		4705
3920	Currently this is used to override some inlining decisions (enlarging the roughly	4706	Currently this is used to override some inlining decisions (enlarging the
3921	30% code size on amd64.	4707	code size by roughly 30% on amd64).
3922		4708
3923	When optimising for size, use of compiler flags such as C<-Os> with	4709	When optimising for size, use of compiler flags such as C<-Os> with
3924	gcc recommended, as well as C<-DNDEBUG>, as libev contains a number of	4710	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
3925	assertions.	4711	assertions.
3926		4712
		4713	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4714	(e.g. gcc with C<-Os>).
		4715
3927	=item C<2> - faster/larger data structures	4716	=item C<2> - faster/larger data structures
3928		4717
3929	Replaces the small 2-heap for timer management by a faster 4-heap, larger	4718	Replaces the small 2-heap for timer management by a faster 4-heap, larger
3930	hash table sizes and so on. This will usually further increase codesize	4719	hash table sizes and so on. This will usually further increase code size
3931	and can additionally have an effect on the size of data structures at	4720	and can additionally have an effect on the size of data structures at
3932	runtime.	4721	runtime.
		4722
		4723	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4724	(e.g. gcc with C<-Os>).
3933		4725
3934	=item C<4> - full API configuration	4726	=item C<4> - full API configuration
3935		4727
3936	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and	4728	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
3937	enables multiplicity (C<EV_MULTIPLICITY>=1).	4729	enables multiplicity (C<EV_MULTIPLICITY>=1).
…		…
3969	With an intelligent-enough linker (gcc+binutils are intelligent enough	4761	With an intelligent-enough linker (gcc+binutils are intelligent enough
3970	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by	4762	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
3971	your program might be left out as well - a binary starting a timer and an	4763	your program might be left out as well - a binary starting a timer and an
3972	I/O watcher then might come out at only 5Kb.	4764	I/O watcher then might come out at only 5Kb.
3973		4765
		4766	=item EV_API_STATIC
		4767
		4768	If this symbol is defined (by default it is not), then all identifiers
		4769	will have static linkage. This means that libev will not export any
		4770	identifiers, and you cannot link against libev anymore. This can be useful
		4771	when you embed libev, only want to use libev functions in a single file,
		4772	and do not want its identifiers to be visible.
		4773
		4774	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4775	wants to use libev.
		4776
		4777	This option only works when libev is compiled with a C compiler, as C++
		4778	doesn't support the required declaration syntax.
		4779
3974	=item EV_AVOID_STDIO	4780	=item EV_AVOID_STDIO
3975		4781
3976	If this is set to C<1> at compiletime, then libev will avoid using stdio	4782	If this is set to C<1> at compiletime, then libev will avoid using stdio
3977	functions (printf, scanf, perror etc.). This will increase the codesize	4783	functions (printf, scanf, perror etc.). This will increase the code size
3978	somewhat, but if your program doesn't otherwise depend on stdio and your	4784	somewhat, but if your program doesn't otherwise depend on stdio and your
3979	libc allows it, this avoids linking in the stdio library which is quite	4785	libc allows it, this avoids linking in the stdio library which is quite
3980	big.	4786	big.
3981		4787
3982	Note that error messages might become less precise when this option is	4788	Note that error messages might become less precise when this option is
…		…
4028	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it	4834	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
4029	will be C<0>.	4835	will be C<0>.
4030		4836
4031	=item EV_VERIFY	4837	=item EV_VERIFY
4032		4838
4033	Controls how much internal verification (see C<ev_loop_verify ()>) will	4839	Controls how much internal verification (see C<ev_verify ()>) will
4034	be done: If set to C<0>, no internal verification code will be compiled	4840	be done: If set to C<0>, no internal verification code will be compiled
4035	in. If set to C<1>, then verification code will be compiled in, but not	4841	in. If set to C<1>, then verification code will be compiled in, but not
4036	called. If set to C<2>, then the internal verification code will be	4842	called. If set to C<2>, then the internal verification code will be
4037	called once per loop, which can slow down libev. If set to C<3>, then the	4843	called once per loop, which can slow down libev. If set to C<3>, then the
4038	verification code will be called very frequently, which will slow down	4844	verification code will be called very frequently, which will slow down
…		…
4042	will be C<0>.	4848	will be C<0>.
4043		4849
4044	=item EV_COMMON	4850	=item EV_COMMON
4045		4851
4046	By default, all watchers have a C<void *data> member. By redefining	4852	By default, all watchers have a C<void *data> member. By redefining
4047	this macro to a something else you can include more and other types of	4853	this macro to something else you can include more and other types of
4048	members. You have to define it each time you include one of the files,	4854	members. You have to define it each time you include one of the files,
4049	though, and it must be identical each time.	4855	though, and it must be identical each time.
4050		4856
4051	For example, the perl EV module uses something like this:	4857	For example, the perl EV module uses something like this:
4052		4858
…		…
4121	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4927	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4122		4928
4123	#include "ev_cpp.h"	4929	#include "ev_cpp.h"
4124	#include "ev.c"	4930	#include "ev.c"
4125		4931
4126	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES	4932	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
4127		4933
4128	=head2 THREADS AND COROUTINES	4934	=head2 THREADS AND COROUTINES
4129		4935
4130	=head3 THREADS	4936	=head3 THREADS
4131		4937
…		…
4182	default loop and triggering an C<ev_async> watcher from the default loop	4988	default loop and triggering an C<ev_async> watcher from the default loop
4183	watcher callback into the event loop interested in the signal.	4989	watcher callback into the event loop interested in the signal.
4184		4990
4185	=back	4991	=back
4186		4992
4187	=head4 THREAD LOCKING EXAMPLE	4993	See also L<THREAD LOCKING EXAMPLE>.
4188
4189	Here is a fictitious example of how to run an event loop in a different
4190	thread than where callbacks are being invoked and watchers are
4191	created/added/removed.
4192
4193	For a real-world example, see the C<EV::Loop::Async> perl module,
4194	which uses exactly this technique (which is suited for many high-level
4195	languages).
4196
4197	The example uses a pthread mutex to protect the loop data, a condition
4198	variable to wait for callback invocations, an async watcher to notify the
4199	event loop thread and an unspecified mechanism to wake up the main thread.
4200
4201	First, you need to associate some data with the event loop:
4202
4203	typedef struct {
4204	mutex_t lock; /* global loop lock */
4205	ev_async async_w;
4206	thread_t tid;
4207	cond_t invoke_cv;
4208	} userdata;
4209
4210	void prepare_loop (EV_P)
4211	{
4212	// for simplicity, we use a static userdata struct.
4213	static userdata u;
4214
4215	ev_async_init (&u->async_w, async_cb);
4216	ev_async_start (EV_A_ &u->async_w);
4217
4218	pthread_mutex_init (&u->lock, 0);
4219	pthread_cond_init (&u->invoke_cv, 0);
4220
4221	// now associate this with the loop
4222	ev_set_userdata (EV_A_ u);
4223	ev_set_invoke_pending_cb (EV_A_ l_invoke);
4224	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4225
4226	// then create the thread running ev_loop
4227	pthread_create (&u->tid, 0, l_run, EV_A);
4228	}
4229
4230	The callback for the C<ev_async> watcher does nothing: the watcher is used
4231	solely to wake up the event loop so it takes notice of any new watchers
4232	that might have been added:
4233
4234	static void
4235	async_cb (EV_P_ ev_async *w, int revents)
4236	{
4237	// just used for the side effects
4238	}
4239
4240	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4241	protecting the loop data, respectively.
4242
4243	static void
4244	l_release (EV_P)
4245	{
4246	userdata *u = ev_userdata (EV_A);
4247	pthread_mutex_unlock (&u->lock);
4248	}
4249
4250	static void
4251	l_acquire (EV_P)
4252	{
4253	userdata *u = ev_userdata (EV_A);
4254	pthread_mutex_lock (&u->lock);
4255	}
4256
4257	The event loop thread first acquires the mutex, and then jumps straight
4258	into C<ev_loop>:
4259
4260	void *
4261	l_run (void *thr_arg)
4262	{
4263	struct ev_loop *loop = (struct ev_loop *)thr_arg;
4264
4265	l_acquire (EV_A);
4266	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4267	ev_loop (EV_A_ 0);
4268	l_release (EV_A);
4269
4270	return 0;
4271	}
4272
4273	Instead of invoking all pending watchers, the C<l_invoke> callback will
4274	signal the main thread via some unspecified mechanism (signals? pipe
4275	writes? C<Async::Interrupt>?) and then waits until all pending watchers
4276	have been called (in a while loop because a) spurious wakeups are possible
4277	and b) skipping inter-thread-communication when there are no pending
4278	watchers is very beneficial):
4279
4280	static void
4281	l_invoke (EV_P)
4282	{
4283	userdata *u = ev_userdata (EV_A);
4284
4285	while (ev_pending_count (EV_A))
4286	{
4287	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4288	pthread_cond_wait (&u->invoke_cv, &u->lock);
4289	}
4290	}
4291
4292	Now, whenever the main thread gets told to invoke pending watchers, it
4293	will grab the lock, call C<ev_invoke_pending> and then signal the loop
4294	thread to continue:
4295
4296	static void
4297	real_invoke_pending (EV_P)
4298	{
4299	userdata *u = ev_userdata (EV_A);
4300
4301	pthread_mutex_lock (&u->lock);
4302	ev_invoke_pending (EV_A);
4303	pthread_cond_signal (&u->invoke_cv);
4304	pthread_mutex_unlock (&u->lock);
4305	}
4306
4307	Whenever you want to start/stop a watcher or do other modifications to an
4308	event loop, you will now have to lock:
4309
4310	ev_timer timeout_watcher;
4311	userdata *u = ev_userdata (EV_A);
4312
4313	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4314
4315	pthread_mutex_lock (&u->lock);
4316	ev_timer_start (EV_A_ &timeout_watcher);
4317	ev_async_send (EV_A_ &u->async_w);
4318	pthread_mutex_unlock (&u->lock);
4319
4320	Note that sending the C<ev_async> watcher is required because otherwise
4321	an event loop currently blocking in the kernel will have no knowledge
4322	about the newly added timer. By waking up the loop it will pick up any new
4323	watchers in the next event loop iteration.
4324		4994
4325	=head3 COROUTINES	4995	=head3 COROUTINES
4326		4996
4327	Libev is very accommodating to coroutines ("cooperative threads"):	4997	Libev is very accommodating to coroutines ("cooperative threads"):
4328	libev fully supports nesting calls to its functions from different	4998	libev fully supports nesting calls to its functions from different
4329	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4999	coroutines (e.g. you can call C<ev_run> on the same loop from two
4330	different coroutines, and switch freely between both coroutines running	5000	different coroutines, and switch freely between both coroutines running
4331	the loop, as long as you don't confuse yourself). The only exception is	5001	the loop, as long as you don't confuse yourself). The only exception is
4332	that you must not do this from C<ev_periodic> reschedule callbacks.	5002	that you must not do this from C<ev_periodic> reschedule callbacks.
4333		5003
4334	Care has been taken to ensure that libev does not keep local state inside	5004	Care has been taken to ensure that libev does not keep local state inside
4335	C<ev_loop>, and other calls do not usually allow for coroutine switches as	5005	C<ev_run>, and other calls do not usually allow for coroutine switches as
4336	they do not call any callbacks.	5006	they do not call any callbacks.
4337		5007
4338	=head2 COMPILER WARNINGS	5008	=head2 COMPILER WARNINGS
4339		5009
4340	Depending on your compiler and compiler settings, you might get no or a	5010	Depending on your compiler and compiler settings, you might get no or a
…		…
4351	maintainable.	5021	maintainable.
4352		5022
4353	And of course, some compiler warnings are just plain stupid, or simply	5023	And of course, some compiler warnings are just plain stupid, or simply
4354	wrong (because they don't actually warn about the condition their message	5024	wrong (because they don't actually warn about the condition their message
4355	seems to warn about). For example, certain older gcc versions had some	5025	seems to warn about). For example, certain older gcc versions had some
4356	warnings that resulted an extreme number of false positives. These have	5026	warnings that resulted in an extreme number of false positives. These have
4357	been fixed, but some people still insist on making code warn-free with	5027	been fixed, but some people still insist on making code warn-free with
4358	such buggy versions.	5028	such buggy versions.
4359		5029
4360	While libev is written to generate as few warnings as possible,	5030	While libev is written to generate as few warnings as possible,
4361	"warn-free" code is not a goal, and it is recommended not to build libev	5031	"warn-free" code is not a goal, and it is recommended not to build libev
…		…
4397	I suggest using suppression lists.	5067	I suggest using suppression lists.
4398		5068
4399		5069
4400	=head1 PORTABILITY NOTES	5070	=head1 PORTABILITY NOTES
4401		5071
		5072	=head2 GNU/LINUX 32 BIT LIMITATIONS
		5073
		5074	GNU/Linux is the only common platform that supports 64 bit file/large file
		5075	interfaces but I<disables> them by default.
		5076
		5077	That means that libev compiled in the default environment doesn't support
		5078	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		5079
		5080	Unfortunately, many programs try to work around this GNU/Linux issue
		5081	by enabling the large file API, which makes them incompatible with the
		5082	standard libev compiled for their system.
		5083
		5084	Likewise, libev cannot enable the large file API itself as this would
		5085	suddenly make it incompatible to the default compile time environment,
		5086	i.e. all programs not using special compile switches.
		5087
		5088	=head2 OS/X AND DARWIN BUGS
		5089
		5090	The whole thing is a bug if you ask me - basically any system interface
		5091	you touch is broken, whether it is locales, poll, kqueue or even the
		5092	OpenGL drivers.
		5093
		5094	=head3 C<kqueue> is buggy
		5095
		5096	The kqueue syscall is broken in all known versions - most versions support
		5097	only sockets, many support pipes.
		5098
		5099	Libev tries to work around this by not using C<kqueue> by default on this
		5100	rotten platform, but of course you can still ask for it when creating a
		5101	loop - embedding a socket-only kqueue loop into a select-based one is
		5102	probably going to work well.
		5103
		5104	=head3 C<poll> is buggy
		5105
		5106	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		5107	implementation by something calling C<kqueue> internally around the 10.5.6
		5108	release, so now C<kqueue> I<and> C<poll> are broken.
		5109
		5110	Libev tries to work around this by not using C<poll> by default on
		5111	this rotten platform, but of course you can still ask for it when creating
		5112	a loop.
		5113
		5114	=head3 C<select> is buggy
		5115
		5116	All that's left is C<select>, and of course Apple found a way to fuck this
		5117	one up as well: On OS/X, C<select> actively limits the number of file
		5118	descriptors you can pass in to 1024 - your program suddenly crashes when
		5119	you use more.
		5120
		5121	There is an undocumented "workaround" for this - defining
		5122	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		5123	work on OS/X.
		5124
		5125	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		5126
		5127	=head3 C<errno> reentrancy
		5128
		5129	The default compile environment on Solaris is unfortunately so
		5130	thread-unsafe that you can't even use components/libraries compiled
		5131	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		5132	defined by default. A valid, if stupid, implementation choice.
		5133
		5134	If you want to use libev in threaded environments you have to make sure
		5135	it's compiled with C<_REENTRANT> defined.
		5136
		5137	=head3 Event port backend
		5138
		5139	The scalable event interface for Solaris is called "event
		5140	ports". Unfortunately, this mechanism is very buggy in all major
		5141	releases. If you run into high CPU usage, your program freezes or you get
		5142	a large number of spurious wakeups, make sure you have all the relevant
		5143	and latest kernel patches applied. No, I don't know which ones, but there
		5144	are multiple ones to apply, and afterwards, event ports actually work
		5145	great.
		5146
		5147	If you can't get it to work, you can try running the program by setting
		5148	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		5149	C<select> backends.
		5150
		5151	=head2 AIX POLL BUG
		5152
		5153	AIX unfortunately has a broken C<poll.h> header. Libev works around
		5154	this by trying to avoid the poll backend altogether (i.e. it's not even
		5155	compiled in), which normally isn't a big problem as C<select> works fine
		5156	with large bitsets on AIX, and AIX is dead anyway.
		5157
4402	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	5158	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		5159
		5160	=head3 General issues
4403		5161
4404	Win32 doesn't support any of the standards (e.g. POSIX) that libev	5162	Win32 doesn't support any of the standards (e.g. POSIX) that libev
4405	requires, and its I/O model is fundamentally incompatible with the POSIX	5163	requires, and its I/O model is fundamentally incompatible with the POSIX
4406	model. Libev still offers limited functionality on this platform in	5164	model. Libev still offers limited functionality on this platform in
4407	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5165	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4408	descriptors. This only applies when using Win32 natively, not when using	5166	descriptors. This only applies when using Win32 natively, not when using
4409	e.g. cygwin.	5167	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		5168	as every compiler comes with a slightly differently broken/incompatible
		5169	environment.
4410		5170
4411	Lifting these limitations would basically require the full	5171	Lifting these limitations would basically require the full
4412	re-implementation of the I/O system. If you are into these kinds of	5172	re-implementation of the I/O system. If you are into this kind of thing,
4413	things, then note that glib does exactly that for you in a very portable	5173	then note that glib does exactly that for you in a very portable way (note
4414	way (note also that glib is the slowest event library known to man).	5174	also that glib is the slowest event library known to man).
4415		5175
4416	There is no supported compilation method available on windows except	5176	There is no supported compilation method available on windows except
4417	embedding it into other applications.	5177	embedding it into other applications.
4418		5178
4419	Sensible signal handling is officially unsupported by Microsoft - libev	5179	Sensible signal handling is officially unsupported by Microsoft - libev
…		…
4447	you do I<not> compile the F<ev.c> or any other embedded source files!):	5207	you do I<not> compile the F<ev.c> or any other embedded source files!):
4448		5208
4449	#include "evwrap.h"	5209	#include "evwrap.h"
4450	#include "ev.c"	5210	#include "ev.c"
4451		5211
4452	=over 4
4453
4454	=item The winsocket select function	5212	=head3 The winsocket C<select> function
4455		5213
4456	The winsocket C<select> function doesn't follow POSIX in that it	5214	The winsocket C<select> function doesn't follow POSIX in that it
4457	requires socket I<handles> and not socket I<file descriptors> (it is	5215	requires socket I<handles> and not socket I<file descriptors> (it is
4458	also extremely buggy). This makes select very inefficient, and also	5216	also extremely buggy). This makes select very inefficient, and also
4459	requires a mapping from file descriptors to socket handles (the Microsoft	5217	requires a mapping from file descriptors to socket handles (the Microsoft
…		…
4468	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	5226	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
4469		5227
4470	Note that winsockets handling of fd sets is O(n), so you can easily get a	5228	Note that winsockets handling of fd sets is O(n), so you can easily get a
4471	complexity in the O(n²) range when using win32.	5229	complexity in the O(n²) range when using win32.
4472		5230
4473	=item Limited number of file descriptors	5231	=head3 Limited number of file descriptors
4474		5232
4475	Windows has numerous arbitrary (and low) limits on things.	5233	Windows has numerous arbitrary (and low) limits on things.
4476		5234
4477	Early versions of winsocket's select only supported waiting for a maximum	5235	Early versions of winsocket's select only supported waiting for a maximum
4478	of C<64> handles (probably owning to the fact that all windows kernels	5236	of C<64> handles (probably owning to the fact that all windows kernels
…		…
4493	runtime libraries. This might get you to about C<512> or C<2048> sockets	5251	runtime libraries. This might get you to about C<512> or C<2048> sockets
4494	(depending on windows version and/or the phase of the moon). To get more,	5252	(depending on windows version and/or the phase of the moon). To get more,
4495	you need to wrap all I/O functions and provide your own fd management, but	5253	you need to wrap all I/O functions and provide your own fd management, but
4496	the cost of calling select (O(n²)) will likely make this unworkable.	5254	the cost of calling select (O(n²)) will likely make this unworkable.
4497		5255
4498	=back
4499
4500	=head2 PORTABILITY REQUIREMENTS	5256	=head2 PORTABILITY REQUIREMENTS
4501		5257
4502	In addition to a working ISO-C implementation and of course the	5258	In addition to a working ISO-C implementation and of course the
4503	backend-specific APIs, libev relies on a few additional extensions:	5259	backend-specific APIs, libev relies on a few additional extensions:
4504		5260
…		…
4510	Libev assumes not only that all watcher pointers have the same internal	5266	Libev assumes not only that all watcher pointers have the same internal
4511	structure (guaranteed by POSIX but not by ISO C for example), but it also	5267	structure (guaranteed by POSIX but not by ISO C for example), but it also
4512	assumes that the same (machine) code can be used to call any watcher	5268	assumes that the same (machine) code can be used to call any watcher
4513	callback: The watcher callbacks have different type signatures, but libev	5269	callback: The watcher callbacks have different type signatures, but libev
4514	calls them using an C<ev_watcher *> internally.	5270	calls them using an C<ev_watcher *> internally.
		5271
		5272	=item pointer accesses must be thread-atomic
		5273
		5274	Accessing a pointer value must be atomic, it must both be readable and
		5275	writable in one piece - this is the case on all current architectures.
4515		5276
4516	=item C<sig_atomic_t volatile> must be thread-atomic as well	5277	=item C<sig_atomic_t volatile> must be thread-atomic as well
4517		5278
4518	The type C<sig_atomic_t volatile> (or whatever is defined as	5279	The type C<sig_atomic_t volatile> (or whatever is defined as
4519	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different	5280	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
…		…
4542	watchers.	5303	watchers.
4543		5304
4544	=item C<double> must hold a time value in seconds with enough accuracy	5305	=item C<double> must hold a time value in seconds with enough accuracy
4545		5306
4546	The type C<double> is used to represent timestamps. It is required to	5307	The type C<double> is used to represent timestamps. It is required to
4547	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	5308	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4548	enough for at least into the year 4000. This requirement is fulfilled by	5309	good enough for at least into the year 4000 with millisecond accuracy
		5310	(the design goal for libev). This requirement is overfulfilled by
4549	implementations implementing IEEE 754, which is basically all existing	5311	implementations using IEEE 754, which is basically all existing ones.
		5312
4550	ones. With IEEE 754 doubles, you get microsecond accuracy until at least	5313	With IEEE 754 doubles, you get microsecond accuracy until at least the
4551	2200.	5314	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5315	is either obsolete or somebody patched it to use C<long double> or
		5316	something like that, just kidding).
4552		5317
4553	=back	5318	=back
4554		5319
4555	If you know of other additional requirements drop me a note.	5320	If you know of other additional requirements drop me a note.
4556		5321
…		…
4618	=item Processing ev_async_send: O(number_of_async_watchers)	5383	=item Processing ev_async_send: O(number_of_async_watchers)
4619		5384
4620	=item Processing signals: O(max_signal_number)	5385	=item Processing signals: O(max_signal_number)
4621		5386
4622	Sending involves a system call I<iff> there were no other C<ev_async_send>	5387	Sending involves a system call I<iff> there were no other C<ev_async_send>
4623	calls in the current loop iteration. Checking for async and signal events	5388	calls in the current loop iteration and the loop is currently
		5389	blocked. Checking for async and signal events involves iterating over all
4624	involves iterating over all running async watchers or all signal numbers.	5390	running async watchers or all signal numbers.
4625		5391
4626	=back	5392	=back
4627		5393
4628		5394
4629	=head1 PORTING FROM LIBEV 3.X TO 4.X	5395	=head1 PORTING FROM LIBEV 3.X TO 4.X
4630		5396
4631	The major version 4 introduced some minor incompatible changes to the API.	5397	The major version 4 introduced some incompatible changes to the API.
4632		5398
4633	At the moment, the C<ev.h> header file tries to implement superficial	5399	At the moment, the C<ev.h> header file provides compatibility definitions
4634	compatibility, so most programs should still compile. Those might be	5400	for all changes, so most programs should still compile. The compatibility
4635	removed in later versions of libev, so better update early than late.	5401	layer might be removed in later versions of libev, so better update to the
		5402	new API early than late.
4636		5403
4637	=over 4	5404	=over 4
4638		5405
4639	=item C<ev_loop_count> renamed to C<ev_iteration>	5406	=item C<EV_COMPAT3> backwards compatibility mechanism
4640		5407
4641	=item C<ev_loop_depth> renamed to C<ev_depth>	5408	The backward compatibility mechanism can be controlled by
		5409	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		5410	section.
4642		5411
4643	=item C<ev_loop_verify> renamed to C<ev_verify>	5412	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5413
		5414	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5415
		5416	ev_loop_destroy (EV_DEFAULT_UC);
		5417	ev_loop_fork (EV_DEFAULT);
		5418
		5419	=item function/symbol renames
		5420
		5421	A number of functions and symbols have been renamed:
		5422
		5423	ev_loop => ev_run
		5424	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5425	EVLOOP_ONESHOT => EVRUN_ONCE
		5426
		5427	ev_unloop => ev_break
		5428	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5429	EVUNLOOP_ONE => EVBREAK_ONE
		5430	EVUNLOOP_ALL => EVBREAK_ALL
		5431
		5432	EV_TIMEOUT => EV_TIMER
		5433
		5434	ev_loop_count => ev_iteration
		5435	ev_loop_depth => ev_depth
		5436	ev_loop_verify => ev_verify
4644		5437
4645	Most functions working on C<struct ev_loop> objects don't have an	5438	Most functions working on C<struct ev_loop> objects don't have an
4646	C<ev_loop_> prefix, so it was removed. Note that C<ev_loop_fork> is	5439	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5440	associated constants have been renamed to not collide with the C<struct
		5441	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5442	as all other watcher types. Note that C<ev_loop_fork> is still called
4647	still called C<ev_loop_fork> because it would otherwise clash with the	5443	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
4648	C<ev_fork> typedef.	5444	typedef.
4649
4650	=item C<EV_TIMEOUT> renamed to C<EV_TIMER> in C<revents>
4651
4652	This is a simple rename - all other watcher types use their name
4653	as revents flag, and now C<ev_timer> does, too.
4654
4655	Both C<EV_TIMER> and C<EV_TIMEOUT> symbols were present in 3.x versions
4656	and continue to be present for the foreseeable future, so this is mostly a
4657	documentation change.
4658		5445
4659	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>	5446	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
4660		5447
4661	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different	5448	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
4662	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile	5449	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
…		…
4669		5456
4670	=over 4	5457	=over 4
4671		5458
4672	=item active	5459	=item active
4673		5460
4674	A watcher is active as long as it has been started (has been attached to	5461	A watcher is active as long as it has been started and not yet stopped.
4675	an event loop) but not yet stopped (disassociated from the event loop).	5462	See L<WATCHER STATES> for details.
4676		5463
4677	=item application	5464	=item application
4678		5465
4679	In this document, an application is whatever is using libev.	5466	In this document, an application is whatever is using libev.
		5467
		5468	=item backend
		5469
		5470	The part of the code dealing with the operating system interfaces.
4680		5471
4681	=item callback	5472	=item callback
4682		5473
4683	The address of a function that is called when some event has been	5474	The address of a function that is called when some event has been
4684	detected. Callbacks are being passed the event loop, the watcher that	5475	detected. Callbacks are being passed the event loop, the watcher that
4685	received the event, and the actual event bitset.	5476	received the event, and the actual event bitset.
4686		5477
4687	=item callback invocation	5478	=item callback/watcher invocation
4688		5479
4689	The act of calling the callback associated with a watcher.	5480	The act of calling the callback associated with a watcher.
4690		5481
4691	=item event	5482	=item event
4692		5483
…		…
4711	The model used to describe how an event loop handles and processes	5502	The model used to describe how an event loop handles and processes
4712	watchers and events.	5503	watchers and events.
4713		5504
4714	=item pending	5505	=item pending
4715		5506
4716	A watcher is pending as soon as the corresponding event has been detected,	5507	A watcher is pending as soon as the corresponding event has been
4717	and stops being pending as soon as the watcher will be invoked or its	5508	detected. See L<WATCHER STATES> for details.
4718	pending status is explicitly cleared by the application.
4719
4720	A watcher can be pending, but not active. Stopping a watcher also clears
4721	its pending status.
4722		5509
4723	=item real time	5510	=item real time
4724		5511
4725	The physical time that is observed. It is apparently strictly monotonic :)	5512	The physical time that is observed. It is apparently strictly monotonic :)
4726		5513
4727	=item wall-clock time	5514	=item wall-clock time
4728		5515
4729	The time and date as shown on clocks. Unlike real time, it can actually	5516	The time and date as shown on clocks. Unlike real time, it can actually
4730	be wrong and jump forwards and backwards, e.g. when the you adjust your	5517	be wrong and jump forwards and backwards, e.g. when you adjust your
4731	clock.	5518	clock.
4732		5519
4733	=item watcher	5520	=item watcher
4734		5521
4735	A data structure that describes interest in certain events. Watchers need	5522	A data structure that describes interest in certain events. Watchers need
4736	to be started (attached to an event loop) before they can receive events.	5523	to be started (attached to an event loop) before they can receive events.
4737		5524
4738	=item watcher invocation
4739
4740	The act of calling the callback associated with a watcher.
4741
4742	=back	5525	=back
4743		5526
4744	=head1 AUTHOR	5527	=head1 AUTHOR
4745		5528
4746	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.	5529	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5530	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
4747		5531

Diff Legend

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