ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.185 by root, Tue Sep 23 09:13:59 2008 UTC vs.
Revision 1.306 by root, Mon Oct 18 07:36:05 2010 UTC

…		…
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	// a single header file is required	11	// a single header file is required
12	#include <ev.h>	12	#include <ev.h>
13		13
		14	#include <stdio.h> // for puts
		15
14	// every watcher type has its own typedef'd struct	16	// every watcher type has its own typedef'd struct
15	// with the name ev_<type>	17	// with the name ev_TYPE
16	ev_io stdin_watcher;	18	ev_io stdin_watcher;
17	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
18		20
19	// all watcher callbacks have a similar signature	21	// all watcher callbacks have a similar signature
20	// this callback is called when data is readable on stdin	22	// this callback is called when data is readable on stdin
21	static void	23	static void
22	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
23	{	25	{
24	puts ("stdin ready");	26	puts ("stdin ready");
25	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
26	// with its corresponding stop function.	28	// with its corresponding stop function.
27	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
…		…
30	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_unloop (EV_A_ EVUNLOOP_ALL);
31	}	33	}
32		34
33	// another callback, this time for a time-out	35	// another callback, this time for a time-out
34	static void	36	static void
35	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
36	{	38	{
37	puts ("timeout");	39	puts ("timeout");
38	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_loop to stop iterating
39	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_unloop (EV_A_ EVUNLOOP_ONE);
40	}	42	}
…		…
60		62
61	// unloop was called, so exit	63	// unloop was called, so exit
62	return 0;	64	return 0;
63	}	65	}
64		66
65	=head1 DESCRIPTION	67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
66		70
67	The newest version of this document is also available as an html-formatted	71	The newest version of this document is also available as an html-formatted
68	web page you might find easier to navigate when reading it for the first	72	web page you might find easier to navigate when reading it for the first
69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.	73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familiarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 ABOUT LIBEV
70		84
71	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
72	file descriptor being readable or a timeout occurring), and it will manage	86	file descriptor being readable or a timeout occurring), and it will manage
73	these event sources and provide your program with events.	87	these event sources and provide your program with events.
74		88
…		…
84	=head2 FEATURES	98	=head2 FEATURES
85		99
86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	100	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	101	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	102	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	103	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
90	with customised rescheduling (C<ev_periodic>), synchronous signals	104	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
91	(C<ev_signal>), process status change events (C<ev_child>), and event	105	timers (C<ev_timer>), absolute timers with customised rescheduling
92	watchers dealing with the event loop mechanism itself (C<ev_idle>,	106	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as	107	change events (C<ev_child>), and event watchers dealing with the event
94	file watchers (C<ev_stat>) and even limited support for fork events	108	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
95	(C<ev_fork>).	109	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		110	limited support for fork events (C<ev_fork>).
96		111
97	It also is quite fast (see this	112	It also is quite fast (see this
98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	113	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
99	for example).	114	for example).
100		115
…		…
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	123	name C<loop> (which is always of type C<struct ev_loop *>) will not have
109	this argument.	124	this argument.
110		125
111	=head2 TIME REPRESENTATION	126	=head2 TIME REPRESENTATION
112		127
113	Libev represents time as a single floating point number, representing the	128	Libev represents time as a single floating point number, representing
114	(fractional) number of seconds since the (POSIX) epoch (somewhere near	129	the (fractional) number of seconds since the (POSIX) epoch (in practise
115	the beginning of 1970, details are complicated, don't ask). This type is	130	somewhere near the beginning of 1970, details are complicated, don't
116	called C<ev_tstamp>, which is what you should use too. It usually aliases	131	ask). This type is called C<ev_tstamp>, which is what you should use
117	to the C<double> type in C, and when you need to do any calculations on	132	too. It usually aliases to the C<double> type in C. When you need to do
118	it, you should treat it as some floating point value. Unlike the name	133	any calculations on it, you should treat it as some floating point value.
		134
119	component C<stamp> might indicate, it is also used for time differences	135	Unlike the name component C<stamp> might indicate, it is also used for
120	throughout libev.	136	time differences (e.g. delays) throughout libev.
121		137
122	=head1 ERROR HANDLING	138	=head1 ERROR HANDLING
123		139
124	Libev knows three classes of errors: operating system errors, usage errors	140	Libev knows three classes of errors: operating system errors, usage errors
125	and internal errors (bugs).	141	and internal errors (bugs).
…		…
176	as this indicates an incompatible change. Minor versions are usually	192	as this indicates an incompatible change. Minor versions are usually
177	compatible to older versions, so a larger minor version alone is usually	193	compatible to older versions, so a larger minor version alone is usually
178	not a problem.	194	not a problem.
179		195
180	Example: Make sure we haven't accidentally been linked against the wrong	196	Example: Make sure we haven't accidentally been linked against the wrong
181	version.	197	version (note, however, that this will not detect ABI mismatches :).
182		198
183	assert (("libev version mismatch",	199	assert (("libev version mismatch",
184	ev_version_major () == EV_VERSION_MAJOR	200	ev_version_major () == EV_VERSION_MAJOR
185	&& ev_version_minor () >= EV_VERSION_MINOR));	201	&& ev_version_minor () >= EV_VERSION_MINOR));
186		202
…		…
214	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	230	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
215	recommended ones.	231	recommended ones.
216		232
217	See the description of C<ev_embed> watchers for more info.	233	See the description of C<ev_embed> watchers for more info.
218		234
219	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	235	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
220		236
221	Sets the allocation function to use (the prototype is similar - the	237	Sets the allocation function to use (the prototype is similar - the
222	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	238	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
223	used to allocate and free memory (no surprises here). If it returns zero	239	used to allocate and free memory (no surprises here). If it returns zero
224	when memory needs to be allocated (C<size != 0>), the library might abort	240	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
250	}	266	}
251		267
252	...	268	...
253	ev_set_allocator (persistent_realloc);	269	ev_set_allocator (persistent_realloc);
254		270
255	=item ev_set_syserr_cb (void (*cb)(const char *msg));	271	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
256		272
257	Set the callback function to call on a retryable system call error (such	273	Set the callback function to call on a retryable system call error (such
258	as failed select, poll, epoll_wait). The message is a printable string	274	as failed select, poll, epoll_wait). The message is a printable string
259	indicating the system call or subsystem causing the problem. If this	275	indicating the system call or subsystem causing the problem. If this
260	callback is set, then libev will expect it to remedy the situation, no	276	callback is set, then libev will expect it to remedy the situation, no
…		…
276		292
277	=back	293	=back
278		294
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	295	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
280		296
281	An event loop is described by a C<struct ev_loop *>. The library knows two	297	An event loop is described by a C<struct ev_loop *> (the C<struct>
282	types of such loops, the I<default> loop, which supports signals and child	298	is I<not> optional in this case, as there is also an C<ev_loop>
283	events, and dynamically created loops which do not.	299	I<function>).
		300
		301	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
		303	not.
284		304
285	=over 4	305	=over 4
286		306
287	=item struct ev_loop *ev_default_loop (unsigned int flags)	307	=item struct ev_loop *ev_default_loop (unsigned int flags)
288		308
…		…
294	If you don't know what event loop to use, use the one returned from this	314	If you don't know what event loop to use, use the one returned from this
295	function.	315	function.
296		316
297	Note that this function is I<not> thread-safe, so if you want to use it	317	Note that this function is I<not> thread-safe, so if you want to use it
298	from multiple threads, you have to lock (note also that this is unlikely,	318	from multiple threads, you have to lock (note also that this is unlikely,
299	as loops cannot bes hared easily between threads anyway).	319	as loops cannot be shared easily between threads anyway).
300		320
301	The default loop is the only loop that can handle C<ev_signal> and	321	The default loop is the only loop that can handle C<ev_signal> and
302	C<ev_child> watchers, and to do this, it always registers a handler	322	C<ev_child> watchers, and to do this, it always registers a handler
303	for C<SIGCHLD>. If this is a problem for your application you can either	323	for C<SIGCHLD>. If this is a problem for your application you can either
304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	324	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
326	useful to try out specific backends to test their performance, or to work	346	useful to try out specific backends to test their performance, or to work
327	around bugs.	347	around bugs.
328		348
329	=item C<EVFLAG_FORKCHECK>	349	=item C<EVFLAG_FORKCHECK>
330		350
331	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	351	Instead of calling C<ev_loop_fork> manually after a fork, you can also
332	a fork, you can also make libev check for a fork in each iteration by	352	make libev check for a fork in each iteration by enabling this flag.
333	enabling this flag.
334		353
335	This works by calling C<getpid ()> on every iteration of the loop,	354	This works by calling C<getpid ()> on every iteration of the loop,
336	and thus this might slow down your event loop if you do a lot of loop	355	and thus this might slow down your event loop if you do a lot of loop
337	iterations and little real work, but is usually not noticeable (on my	356	iterations and little real work, but is usually not noticeable (on my
338	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	357	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
…		…
344	flag.	363	flag.
345		364
346	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	365	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
347	environment variable.	366	environment variable.
348		367
		368	=item C<EVFLAG_NOINOTIFY>
		369
		370	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
		372	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.
		374
		375	=item C<EVFLAG_SIGNALFD>
		376
		377	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
		379	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
		381	handling with threads, as long as you properly block signals in your
		382	threads that are not interested in handling them.
		383
		384	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
		386	example) that can't properly initialise their signal masks.
		387
349	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	388	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
350		389
351	This is your standard select(2) backend. Not I<completely> standard, as	390	This is your standard select(2) backend. Not I<completely> standard, as
352	libev tries to roll its own fd_set with no limits on the number of fds,	391	libev tries to roll its own fd_set with no limits on the number of fds,
353	but if that fails, expect a fairly low limit on the number of fds when	392	but if that fails, expect a fairly low limit on the number of fds when
…		…
377	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and	416	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
378	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.	417	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
379		418
380	=item C<EVBACKEND_EPOLL> (value 4, Linux)	419	=item C<EVBACKEND_EPOLL> (value 4, Linux)
381		420
		421	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		422	kernels).
		423
382	For few fds, this backend is a bit little slower than poll and select,	424	For few fds, this backend is a bit little slower than poll and select,
383	but it scales phenomenally better. While poll and select usually scale	425	but it scales phenomenally better. While poll and select usually scale
384	like O(total_fds) where n is the total number of fds (or the highest fd),	426	like O(total_fds) where n is the total number of fds (or the highest fd),
385	epoll scales either O(1) or O(active_fds). The epoll design has a number	427	epoll scales either O(1) or O(active_fds).
386	of shortcomings, such as silently dropping events in some hard-to-detect	428
387	cases and requiring a system call per fd change, no fork support and bad	429	The epoll mechanism deserves honorable mention as the most misdesigned
388	support for dup.	430	of the more advanced event mechanisms: mere annoyances include silently
		431	dropping file descriptors, requiring a system call per change per file
		432	descriptor (and unnecessary guessing of parameters), problems with dup and
		433	so on. The biggest issue is fork races, however - if a program forks then
		434	I<both> parent and child process have to recreate the epoll set, which can
		435	take considerable time (one syscall per file descriptor) and is of course
		436	hard to detect.
		437
		438	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		439	of course I<doesn't>, and epoll just loves to report events for totally
		440	I<different> file descriptors (even already closed ones, so one cannot
		441	even remove them from the set) than registered in the set (especially
		442	on SMP systems). Libev tries to counter these spurious notifications by
		443	employing an additional generation counter and comparing that against the
		444	events to filter out spurious ones, recreating the set when required. Last
		445	not least, it also refuses to work with some file descriptors which work
		446	perfectly fine with C<select> (files, many character devices...).
389		447
390	While stopping, setting and starting an I/O watcher in the same iteration	448	While stopping, setting and starting an I/O watcher in the same iteration
391	will result in some caching, there is still a system call per such incident	449	will result in some caching, there is still a system call per such
392	(because the fd could point to a different file description now), so its	450	incident (because the same I<file descriptor> could point to a different
393	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	451	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
394	very well if you register events for both fds.	452	file descriptors might not work very well if you register events for both
395		453	file descriptors.
396	Please note that epoll sometimes generates spurious notifications, so you
397	need to use non-blocking I/O or other means to avoid blocking when no data
398	(or space) is available.
399		454
400	Best performance from this backend is achieved by not unregistering all	455	Best performance from this backend is achieved by not unregistering all
401	watchers for a file descriptor until it has been closed, if possible,	456	watchers for a file descriptor until it has been closed, if possible,
402	i.e. keep at least one watcher active per fd at all times. Stopping and	457	i.e. keep at least one watcher active per fd at all times. Stopping and
403	starting a watcher (without re-setting it) also usually doesn't cause	458	starting a watcher (without re-setting it) also usually doesn't cause
404	extra overhead.	459	extra overhead. A fork can both result in spurious notifications as well
		460	as in libev having to destroy and recreate the epoll object, which can
		461	take considerable time and thus should be avoided.
		462
		463	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		464	faster than epoll for maybe up to a hundred file descriptors, depending on
		465	the usage. So sad.
405		466
406	While nominally embeddable in other event loops, this feature is broken in	467	While nominally embeddable in other event loops, this feature is broken in
407	all kernel versions tested so far.	468	all kernel versions tested so far.
408		469
409	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	470	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
410	C<EVBACKEND_POLL>.	471	C<EVBACKEND_POLL>.
411		472
412	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	473	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
413		474
414	Kqueue deserves special mention, as at the time of this writing, it was	475	Kqueue deserves special mention, as at the time of this writing, it
415	broken on all BSDs except NetBSD (usually it doesn't work reliably with	476	was broken on all BSDs except NetBSD (usually it doesn't work reliably
416	anything but sockets and pipes, except on Darwin, where of course it's	477	with anything but sockets and pipes, except on Darwin, where of course
417	completely useless). For this reason it's not being "auto-detected" unless	478	it's completely useless). Unlike epoll, however, whose brokenness
418	you explicitly specify it in the flags (i.e. using C<EVBACKEND_KQUEUE>) or	479	is by design, these kqueue bugs can (and eventually will) be fixed
419	libev was compiled on a known-to-be-good (-enough) system like NetBSD.	480	without API changes to existing programs. For this reason it's not being
		481	"auto-detected" unless you explicitly specify it in the flags (i.e. using
		482	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		483	system like NetBSD.
420		484
421	You still can embed kqueue into a normal poll or select backend and use it	485	You still can embed kqueue into a normal poll or select backend and use it
422	only for sockets (after having made sure that sockets work with kqueue on	486	only for sockets (after having made sure that sockets work with kqueue on
423	the target platform). See C<ev_embed> watchers for more info.	487	the target platform). See C<ev_embed> watchers for more info.
424		488
425	It scales in the same way as the epoll backend, but the interface to the	489	It scales in the same way as the epoll backend, but the interface to the
426	kernel is more efficient (which says nothing about its actual speed, of	490	kernel is more efficient (which says nothing about its actual speed, of
427	course). While stopping, setting and starting an I/O watcher does never	491	course). While stopping, setting and starting an I/O watcher does never
428	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	492	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
429	two event changes per incident. Support for C<fork ()> is very bad and it	493	two event changes per incident. Support for C<fork ()> is very bad (but
430	drops fds silently in similarly hard-to-detect cases.	494	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		495	cases
431		496
432	This backend usually performs well under most conditions.	497	This backend usually performs well under most conditions.
433		498
434	While nominally embeddable in other event loops, this doesn't work	499	While nominally embeddable in other event loops, this doesn't work
435	everywhere, so you might need to test for this. And since it is broken	500	everywhere, so you might need to test for this. And since it is broken
436	almost everywhere, you should only use it when you have a lot of sockets	501	almost everywhere, you should only use it when you have a lot of sockets
437	(for which it usually works), by embedding it into another event loop	502	(for which it usually works), by embedding it into another event loop
438	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,	503	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
439	using it only for sockets.	504	also broken on OS X)) and, did I mention it, using it only for sockets.
440		505
441	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with	506	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
442	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with	507	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
443	C<NOTE_EOF>.	508	C<NOTE_EOF>.
444		509
…		…
464	might perform better.	529	might perform better.
465		530
466	On the positive side, with the exception of the spurious readiness	531	On the positive side, with the exception of the spurious readiness
467	notifications, this backend actually performed fully to specification	532	notifications, this backend actually performed fully to specification
468	in all tests and is fully embeddable, which is a rare feat among the	533	in all tests and is fully embeddable, which is a rare feat among the
469	OS-specific backends.	534	OS-specific backends (I vastly prefer correctness over speed hacks).
470		535
471	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	536	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
472	C<EVBACKEND_POLL>.	537	C<EVBACKEND_POLL>.
473		538
474	=item C<EVBACKEND_ALL>	539	=item C<EVBACKEND_ALL>
…		…
479		544
480	It is definitely not recommended to use this flag.	545	It is definitely not recommended to use this flag.
481		546
482	=back	547	=back
483		548
484	If one or more of these are or'ed into the flags value, then only these	549	If one or more of the backend flags are or'ed into the flags value,
485	backends will be tried (in the reverse order as listed here). If none are	550	then only these backends will be tried (in the reverse order as listed
486	specified, all backends in C<ev_recommended_backends ()> will be tried.	551	here). If none are specified, all backends in C<ev_recommended_backends
		552	()> will be tried.
487		553
488	Example: This is the most typical usage.	554	Example: This is the most typical usage.
489		555
490	if (!ev_default_loop (0))	556	if (!ev_default_loop (0))
491	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	557	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
…		…
503	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	569	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
504		570
505	=item struct ev_loop *ev_loop_new (unsigned int flags)	571	=item struct ev_loop *ev_loop_new (unsigned int flags)
506		572
507	Similar to C<ev_default_loop>, but always creates a new event loop that is	573	Similar to C<ev_default_loop>, but always creates a new event loop that is
508	always distinct from the default loop. Unlike the default loop, it cannot	574	always distinct from the default loop.
509	handle signal and child watchers, and attempts to do so will be greeted by
510	undefined behaviour (or a failed assertion if assertions are enabled).
511		575
512	Note that this function I<is> thread-safe, and the recommended way to use	576	Note that this function I<is> thread-safe, and one common way to use
513	libev with threads is indeed to create one loop per thread, and using the	577	libev with threads is indeed to create one loop per thread, and using the
514	default loop in the "main" or "initial" thread.	578	default loop in the "main" or "initial" thread.
515		579
516	Example: Try to create a event loop that uses epoll and nothing else.	580	Example: Try to create a event loop that uses epoll and nothing else.
517		581
…		…
519	if (!epoller)	583	if (!epoller)
520	fatal ("no epoll found here, maybe it hides under your chair");	584	fatal ("no epoll found here, maybe it hides under your chair");
521		585
522	=item ev_default_destroy ()	586	=item ev_default_destroy ()
523		587
524	Destroys the default loop again (frees all memory and kernel state	588	Destroys the default loop (frees all memory and kernel state etc.). None
525	etc.). None of the active event watchers will be stopped in the normal	589	of the active event watchers will be stopped in the normal sense, so
526	sense, so e.g. C<ev_is_active> might still return true. It is your	590	e.g. C<ev_is_active> might still return true. It is your responsibility to
527	responsibility to either stop all watchers cleanly yourself I<before>	591	either stop all watchers cleanly yourself I<before> calling this function,
528	calling this function, or cope with the fact afterwards (which is usually	592	or cope with the fact afterwards (which is usually the easiest thing, you
529	the easiest thing, you can just ignore the watchers and/or C<free ()> them	593	can just ignore the watchers and/or C<free ()> them for example).
530	for example).
531		594
532	Note that certain global state, such as signal state, will not be freed by	595	Note that certain global state, such as signal state (and installed signal
533	this function, and related watchers (such as signal and child watchers)	596	handlers), will not be freed by this function, and related watchers (such
534	would need to be stopped manually.	597	as signal and child watchers) would need to be stopped manually.
535		598
536	In general it is not advisable to call this function except in the	599	In general it is not advisable to call this function except in the
537	rare occasion where you really need to free e.g. the signal handling	600	rare occasion where you really need to free e.g. the signal handling
538	pipe fds. If you need dynamically allocated loops it is better to use	601	pipe fds. If you need dynamically allocated loops it is better to use
539	C<ev_loop_new> and C<ev_loop_destroy>).	602	C<ev_loop_new> and C<ev_loop_destroy>.
540		603
541	=item ev_loop_destroy (loop)	604	=item ev_loop_destroy (loop)
542		605
543	Like C<ev_default_destroy>, but destroys an event loop created by an	606	Like C<ev_default_destroy>, but destroys an event loop created by an
544	earlier call to C<ev_loop_new>.	607	earlier call to C<ev_loop_new>.
…		…
550	name, you can call it anytime, but it makes most sense after forking, in	613	name, you can call it anytime, but it makes most sense after forking, in
551	the child process (or both child and parent, but that again makes little	614	the child process (or both child and parent, but that again makes little
552	sense). You I<must> call it in the child before using any of the libev	615	sense). You I<must> call it in the child before using any of the libev
553	functions, and it will only take effect at the next C<ev_loop> iteration.	616	functions, and it will only take effect at the next C<ev_loop> iteration.
554		617
		618	Again, you I<have> to call it on I<any> loop that you want to re-use after
		619	a fork, I<even if you do not plan to use the loop in the parent>. This is
		620	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		621	during fork.
		622
555	On the other hand, you only need to call this function in the child	623	On the other hand, you only need to call this function in the child
556	process if and only if you want to use the event library in the child. If	624	process if and only if you want to use the event loop in the child. If you
557	you just fork+exec, you don't have to call it at all.	625	just fork+exec or create a new loop in the child, you don't have to call
		626	it at all.
558		627
559	The function itself is quite fast and it's usually not a problem to call	628	The function itself is quite fast and it's usually not a problem to call
560	it just in case after a fork. To make this easy, the function will fit in	629	it just in case after a fork. To make this easy, the function will fit in
561	quite nicely into a call to C<pthread_atfork>:	630	quite nicely into a call to C<pthread_atfork>:
562		631
…		…
564		633
565	=item ev_loop_fork (loop)	634	=item ev_loop_fork (loop)
566		635
567	Like C<ev_default_fork>, but acts on an event loop created by	636	Like C<ev_default_fork>, but acts on an event loop created by
568	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	637	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
569	after fork that you want to re-use in the child, and how you do this is	638	after fork that you want to re-use in the child, and how you keep track of
570	entirely your own problem.	639	them is entirely your own problem.
571		640
572	=item int ev_is_default_loop (loop)	641	=item int ev_is_default_loop (loop)
573		642
574	Returns true when the given loop is, in fact, the default loop, and false	643	Returns true when the given loop is, in fact, the default loop, and false
575	otherwise.	644	otherwise.
576		645
577	=item unsigned int ev_loop_count (loop)	646	=item unsigned int ev_iteration (loop)
578		647
579	Returns the count of loop iterations for the loop, which is identical to	648	Returns the current iteration count for the loop, which is identical to
580	the number of times libev did poll for new events. It starts at C<0> and	649	the number of times libev did poll for new events. It starts at C<0> and
581	happily wraps around with enough iterations.	650	happily wraps around with enough iterations.
582		651
583	This value can sometimes be useful as a generation counter of sorts (it	652	This value can sometimes be useful as a generation counter of sorts (it
584	"ticks" the number of loop iterations), as it roughly corresponds with	653	"ticks" the number of loop iterations), as it roughly corresponds with
585	C<ev_prepare> and C<ev_check> calls.	654	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		655	prepare and check phases.
		656
		657	=item unsigned int ev_depth (loop)
		658
		659	Returns the number of times C<ev_loop> was entered minus the number of
		660	times C<ev_loop> was exited, in other words, the recursion depth.
		661
		662	Outside C<ev_loop>, this number is zero. In a callback, this number is
		663	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		664	in which case it is higher.
		665
		666	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		667	etc.), doesn't count as "exit" - consider this as a hint to avoid such
		668	ungentleman behaviour unless it's really convenient.
586		669
587	=item unsigned int ev_backend (loop)	670	=item unsigned int ev_backend (loop)
588		671
589	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	672	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
590	use.	673	use.
…		…
605		688
606	This function is rarely useful, but when some event callback runs for a	689	This function is rarely useful, but when some event callback runs for a
607	very long time without entering the event loop, updating libev's idea of	690	very long time without entering the event loop, updating libev's idea of
608	the current time is a good idea.	691	the current time is a good idea.
609		692
610	See also "The special problem of time updates" in the C<ev_timer> section.	693	See also L<The special problem of time updates> in the C<ev_timer> section.
		694
		695	=item ev_suspend (loop)
		696
		697	=item ev_resume (loop)
		698
		699	These two functions suspend and resume a loop, for use when the loop is
		700	not used for a while and timeouts should not be processed.
		701
		702	A typical use case would be an interactive program such as a game: When
		703	the user presses C<^Z> to suspend the game and resumes it an hour later it
		704	would be best to handle timeouts as if no time had actually passed while
		705	the program was suspended. This can be achieved by calling C<ev_suspend>
		706	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		707	C<ev_resume> directly afterwards to resume timer processing.
		708
		709	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		710	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		711	will be rescheduled (that is, they will lose any events that would have
		712	occurred while suspended).
		713
		714	After calling C<ev_suspend> you B<must not> call I<any> function on the
		715	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		716	without a previous call to C<ev_suspend>.
		717
		718	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		719	event loop time (see C<ev_now_update>).
611		720
612	=item ev_loop (loop, int flags)	721	=item ev_loop (loop, int flags)
613		722
614	Finally, this is it, the event handler. This function usually is called	723	Finally, this is it, the event handler. This function usually is called
615	after you initialised all your watchers and you want to start handling	724	after you have initialised all your watchers and you want to start
616	events.	725	handling events.
617		726
618	If the flags argument is specified as C<0>, it will not return until	727	If the flags argument is specified as C<0>, it will not return until
619	either no event watchers are active anymore or C<ev_unloop> was called.	728	either no event watchers are active anymore or C<ev_unloop> was called.
620		729
621	Please note that an explicit C<ev_unloop> is usually better than	730	Please note that an explicit C<ev_unloop> is usually better than
…		…
631	the loop.	740	the loop.
632		741
633	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	742	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
634	necessary) and will handle those and any already outstanding ones. It	743	necessary) and will handle those and any already outstanding ones. It
635	will block your process until at least one new event arrives (which could	744	will block your process until at least one new event arrives (which could
636	be an event internal to libev itself, so there is no guarentee that a	745	be an event internal to libev itself, so there is no guarantee that a
637	user-registered callback will be called), and will return after one	746	user-registered callback will be called), and will return after one
638	iteration of the loop.	747	iteration of the loop.
639		748
640	This is useful if you are waiting for some external event in conjunction	749	This is useful if you are waiting for some external event in conjunction
641	with something not expressible using other libev watchers (i.e. "roll your	750	with something not expressible using other libev watchers (i.e. "roll your
…		…
685	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	794	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
686	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	795	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
687		796
688	This "unloop state" will be cleared when entering C<ev_loop> again.	797	This "unloop state" will be cleared when entering C<ev_loop> again.
689		798
		799	It is safe to call C<ev_unloop> from outside any C<ev_loop> calls.
		800
690	=item ev_ref (loop)	801	=item ev_ref (loop)
691		802
692	=item ev_unref (loop)	803	=item ev_unref (loop)
693		804
694	Ref/unref can be used to add or remove a reference count on the event	805	Ref/unref can be used to add or remove a reference count on the event
695	loop: Every watcher keeps one reference, and as long as the reference	806	loop: Every watcher keeps one reference, and as long as the reference
696	count is nonzero, C<ev_loop> will not return on its own.	807	count is nonzero, C<ev_loop> will not return on its own.
697		808
698	If you have a watcher you never unregister that should not keep C<ev_loop>	809	This is useful when you have a watcher that you never intend to
699	from returning, call ev_unref() after starting, and ev_ref() before	810	unregister, but that nevertheless should not keep C<ev_loop> from
		811	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
700	stopping it.	812	before stopping it.
701		813
702	As an example, libev itself uses this for its internal signal pipe: It is	814	As an example, libev itself uses this for its internal signal pipe: It
703	not visible to the libev user and should not keep C<ev_loop> from exiting	815	is not visible to the libev user and should not keep C<ev_loop> from
704	if no event watchers registered by it are active. It is also an excellent	816	exiting if no event watchers registered by it are active. It is also an
705	way to do this for generic recurring timers or from within third-party	817	excellent way to do this for generic recurring timers or from within
706	libraries. Just remember to I<unref after start> and I<ref before stop>	818	third-party libraries. Just remember to I<unref after start> and I<ref
707	(but only if the watcher wasn't active before, or was active before,	819	before stop> (but only if the watcher wasn't active before, or was active
708	respectively).	820	before, respectively. Note also that libev might stop watchers itself
		821	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		822	in the callback).
709		823
710	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	824	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
711	running when nothing else is active.	825	running when nothing else is active.
712		826
713	struct ev_signal exitsig;	827	ev_signal exitsig;
714	ev_signal_init (&exitsig, sig_cb, SIGINT);	828	ev_signal_init (&exitsig, sig_cb, SIGINT);
715	ev_signal_start (loop, &exitsig);	829	ev_signal_start (loop, &exitsig);
716	evf_unref (loop);	830	evf_unref (loop);
717		831
718	Example: For some weird reason, unregister the above signal handler again.	832	Example: For some weird reason, unregister the above signal handler again.
…		…
742		856
743	By setting a higher I<io collect interval> you allow libev to spend more	857	By setting a higher I<io collect interval> you allow libev to spend more
744	time collecting I/O events, so you can handle more events per iteration,	858	time collecting I/O events, so you can handle more events per iteration,
745	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	859	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
746	C<ev_timer>) will be not affected. Setting this to a non-null value will	860	C<ev_timer>) will be not affected. Setting this to a non-null value will
747	introduce an additional C<ev_sleep ()> call into most loop iterations.	861	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		862	sleep time ensures that libev will not poll for I/O events more often then
		863	once per this interval, on average.
748		864
749	Likewise, by setting a higher I<timeout collect interval> you allow libev	865	Likewise, by setting a higher I<timeout collect interval> you allow libev
750	to spend more time collecting timeouts, at the expense of increased	866	to spend more time collecting timeouts, at the expense of increased
751	latency/jitter/inexactness (the watcher callback will be called	867	latency/jitter/inexactness (the watcher callback will be called
752	later). C<ev_io> watchers will not be affected. Setting this to a non-null	868	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
754		870
755	Many (busy) programs can usually benefit by setting the I/O collect	871	Many (busy) programs can usually benefit by setting the I/O collect
756	interval to a value near C<0.1> or so, which is often enough for	872	interval to a value near C<0.1> or so, which is often enough for
757	interactive servers (of course not for games), likewise for timeouts. It	873	interactive servers (of course not for games), likewise for timeouts. It
758	usually doesn't make much sense to set it to a lower value than C<0.01>,	874	usually doesn't make much sense to set it to a lower value than C<0.01>,
759	as this approaches the timing granularity of most systems.	875	as this approaches the timing granularity of most systems. Note that if
		876	you do transactions with the outside world and you can't increase the
		877	parallelity, then this setting will limit your transaction rate (if you
		878	need to poll once per transaction and the I/O collect interval is 0.01,
		879	then you can't do more than 100 transactions per second).
760		880
761	Setting the I<timeout collect interval> can improve the opportunity for	881	Setting the I<timeout collect interval> can improve the opportunity for
762	saving power, as the program will "bundle" timer callback invocations that	882	saving power, as the program will "bundle" timer callback invocations that
763	are "near" in time together, by delaying some, thus reducing the number of	883	are "near" in time together, by delaying some, thus reducing the number of
764	times the process sleeps and wakes up again. Another useful technique to	884	times the process sleeps and wakes up again. Another useful technique to
765	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure	885	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
766	they fire on, say, one-second boundaries only.	886	they fire on, say, one-second boundaries only.
767		887
		888	Example: we only need 0.1s timeout granularity, and we wish not to poll
		889	more often than 100 times per second:
		890
		891	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		892	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		893
		894	=item ev_invoke_pending (loop)
		895
		896	This call will simply invoke all pending watchers while resetting their
		897	pending state. Normally, C<ev_loop> does this automatically when required,
		898	but when overriding the invoke callback this call comes handy.
		899
		900	=item int ev_pending_count (loop)
		901
		902	Returns the number of pending watchers - zero indicates that no watchers
		903	are pending.
		904
		905	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		906
		907	This overrides the invoke pending functionality of the loop: Instead of
		908	invoking all pending watchers when there are any, C<ev_loop> will call
		909	this callback instead. This is useful, for example, when you want to
		910	invoke the actual watchers inside another context (another thread etc.).
		911
		912	If you want to reset the callback, use C<ev_invoke_pending> as new
		913	callback.
		914
		915	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		916
		917	Sometimes you want to share the same loop between multiple threads. This
		918	can be done relatively simply by putting mutex_lock/unlock calls around
		919	each call to a libev function.
		920
		921	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		922	wait for it to return. One way around this is to wake up the loop via
		923	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		924	and I<acquire> callbacks on the loop.
		925
		926	When set, then C<release> will be called just before the thread is
		927	suspended waiting for new events, and C<acquire> is called just
		928	afterwards.
		929
		930	Ideally, C<release> will just call your mutex_unlock function, and
		931	C<acquire> will just call the mutex_lock function again.
		932
		933	While event loop modifications are allowed between invocations of
		934	C<release> and C<acquire> (that's their only purpose after all), no
		935	modifications done will affect the event loop, i.e. adding watchers will
		936	have no effect on the set of file descriptors being watched, or the time
		937	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it
		938	to take note of any changes you made.
		939
		940	In theory, threads executing C<ev_loop> will be async-cancel safe between
		941	invocations of C<release> and C<acquire>.
		942
		943	See also the locking example in the C<THREADS> section later in this
		944	document.
		945
		946	=item ev_set_userdata (loop, void *data)
		947
		948	=item ev_userdata (loop)
		949
		950	Set and retrieve a single C<void *> associated with a loop. When
		951	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		952	C<0.>
		953
		954	These two functions can be used to associate arbitrary data with a loop,
		955	and are intended solely for the C<invoke_pending_cb>, C<release> and
		956	C<acquire> callbacks described above, but of course can be (ab-)used for
		957	any other purpose as well.
		958
768	=item ev_loop_verify (loop)	959	=item ev_loop_verify (loop)
769		960
770	This function only does something when C<EV_VERIFY> support has been	961	This function only does something when C<EV_VERIFY> support has been
771	compiled in. which is the default for non-minimal builds. It tries to go	962	compiled in, which is the default for non-minimal builds. It tries to go
772	through all internal structures and checks them for validity. If anything	963	through all internal structures and checks them for validity. If anything
773	is found to be inconsistent, it will print an error message to standard	964	is found to be inconsistent, it will print an error message to standard
774	error and call C<abort ()>.	965	error and call C<abort ()>.
775		966
776	This can be used to catch bugs inside libev itself: under normal	967	This can be used to catch bugs inside libev itself: under normal
…		…
780	=back	971	=back
781		972
782		973
783	=head1 ANATOMY OF A WATCHER	974	=head1 ANATOMY OF A WATCHER
784		975
		976	In the following description, uppercase C<TYPE> in names stands for the
		977	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		978	watchers and C<ev_io_start> for I/O watchers.
		979
785	A watcher is a structure that you create and register to record your	980	A watcher is a structure that you create and register to record your
786	interest in some event. For instance, if you want to wait for STDIN to	981	interest in some event. For instance, if you want to wait for STDIN to
787	become readable, you would create an C<ev_io> watcher for that:	982	become readable, you would create an C<ev_io> watcher for that:
788		983
789	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	984	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
790	{	985	{
791	ev_io_stop (w);	986	ev_io_stop (w);
792	ev_unloop (loop, EVUNLOOP_ALL);	987	ev_unloop (loop, EVUNLOOP_ALL);
793	}	988	}
794		989
795	struct ev_loop *loop = ev_default_loop (0);	990	struct ev_loop *loop = ev_default_loop (0);
		991
796	struct ev_io stdin_watcher;	992	ev_io stdin_watcher;
		993
797	ev_init (&stdin_watcher, my_cb);	994	ev_init (&stdin_watcher, my_cb);
798	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	995	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
799	ev_io_start (loop, &stdin_watcher);	996	ev_io_start (loop, &stdin_watcher);
		997
800	ev_loop (loop, 0);	998	ev_loop (loop, 0);
801		999
802	As you can see, you are responsible for allocating the memory for your	1000	As you can see, you are responsible for allocating the memory for your
803	watcher structures (and it is usually a bad idea to do this on the stack,	1001	watcher structures (and it is I<usually> a bad idea to do this on the
804	although this can sometimes be quite valid).	1002	stack).
		1003
		1004	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1005	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
805		1006
806	Each watcher structure must be initialised by a call to C<ev_init	1007	Each watcher structure must be initialised by a call to C<ev_init
807	(watcher *, callback)>, which expects a callback to be provided. This	1008	(watcher *, callback)>, which expects a callback to be provided. This
808	callback gets invoked each time the event occurs (or, in the case of I/O	1009	callback gets invoked each time the event occurs (or, in the case of I/O
809	watchers, each time the event loop detects that the file descriptor given	1010	watchers, each time the event loop detects that the file descriptor given
810	is readable and/or writable).	1011	is readable and/or writable).
811		1012
812	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1013	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
813	with arguments specific to this watcher type. There is also a macro	1014	macro to configure it, with arguments specific to the watcher type. There
814	to combine initialisation and setting in one call: C<< ev_<type>_init	1015	is also a macro to combine initialisation and setting in one call: C<<
815	(watcher *, callback, ...) >>.	1016	ev_TYPE_init (watcher *, callback, ...) >>.
816		1017
817	To make the watcher actually watch out for events, you have to start it	1018	To make the watcher actually watch out for events, you have to start it
818	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1019	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
819	*) >>), and you can stop watching for events at any time by calling the	1020	*) >>), and you can stop watching for events at any time by calling the
820	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1021	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
821		1022
822	As long as your watcher is active (has been started but not stopped) you	1023	As long as your watcher is active (has been started but not stopped) you
823	must not touch the values stored in it. Most specifically you must never	1024	must not touch the values stored in it. Most specifically you must never
824	reinitialise it or call its C<set> macro.	1025	reinitialise it or call its C<ev_TYPE_set> macro.
825		1026
826	Each and every callback receives the event loop pointer as first, the	1027	Each and every callback receives the event loop pointer as first, the
827	registered watcher structure as second, and a bitset of received events as	1028	registered watcher structure as second, and a bitset of received events as
828	third argument.	1029	third argument.
829		1030
…		…
838	=item C<EV_WRITE>	1039	=item C<EV_WRITE>
839		1040
840	The file descriptor in the C<ev_io> watcher has become readable and/or	1041	The file descriptor in the C<ev_io> watcher has become readable and/or
841	writable.	1042	writable.
842		1043
843	=item C<EV_TIMEOUT>	1044	=item C<EV_TIMER>
844		1045
845	The C<ev_timer> watcher has timed out.	1046	The C<ev_timer> watcher has timed out.
846		1047
847	=item C<EV_PERIODIC>	1048	=item C<EV_PERIODIC>
848		1049
…		…
887		1088
888	=item C<EV_ASYNC>	1089	=item C<EV_ASYNC>
889		1090
890	The given async watcher has been asynchronously notified (see C<ev_async>).	1091	The given async watcher has been asynchronously notified (see C<ev_async>).
891		1092
		1093	=item C<EV_CUSTOM>
		1094
		1095	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1096	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1097
892	=item C<EV_ERROR>	1098	=item C<EV_ERROR>
893		1099
894	An unspecified error has occurred, the watcher has been stopped. This might	1100	An unspecified error has occurred, the watcher has been stopped. This might
895	happen because the watcher could not be properly started because libev	1101	happen because the watcher could not be properly started because libev
896	ran out of memory, a file descriptor was found to be closed or any other	1102	ran out of memory, a file descriptor was found to be closed or any other
		1103	problem. Libev considers these application bugs.
		1104
897	problem. You best act on it by reporting the problem and somehow coping	1105	You best act on it by reporting the problem and somehow coping with the
898	with the watcher being stopped.	1106	watcher being stopped. Note that well-written programs should not receive
		1107	an error ever, so when your watcher receives it, this usually indicates a
		1108	bug in your program.
899		1109
900	Libev will usually signal a few "dummy" events together with an error, for	1110	Libev will usually signal a few "dummy" events together with an error, for
901	example it might indicate that a fd is readable or writable, and if your	1111	example it might indicate that a fd is readable or writable, and if your
902	callbacks is well-written it can just attempt the operation and cope with	1112	callbacks is well-written it can just attempt the operation and cope with
903	the error from read() or write(). This will not work in multi-threaded	1113	the error from read() or write(). This will not work in multi-threaded
…		…
906		1116
907	=back	1117	=back
908		1118
909	=head2 GENERIC WATCHER FUNCTIONS	1119	=head2 GENERIC WATCHER FUNCTIONS
910		1120
911	In the following description, C<TYPE> stands for the watcher type,
912	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
913
914	=over 4	1121	=over 4
915		1122
916	=item C<ev_init> (ev_TYPE *watcher, callback)	1123	=item C<ev_init> (ev_TYPE *watcher, callback)
917		1124
918	This macro initialises the generic portion of a watcher. The contents	1125	This macro initialises the generic portion of a watcher. The contents
…		…
923	which rolls both calls into one.	1130	which rolls both calls into one.
924		1131
925	You can reinitialise a watcher at any time as long as it has been stopped	1132	You can reinitialise a watcher at any time as long as it has been stopped
926	(or never started) and there are no pending events outstanding.	1133	(or never started) and there are no pending events outstanding.
927		1134
928	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1135	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
929	int revents)>.	1136	int revents)>.
930		1137
931	Example: Initialise an C<ev_io> watcher in two steps.	1138	Example: Initialise an C<ev_io> watcher in two steps.
932		1139
933	ev_io w;	1140	ev_io w;
934	ev_init (&w, my_cb);	1141	ev_init (&w, my_cb);
935	ev_io_set (&w, STDIN_FILENO, EV_READ);	1142	ev_io_set (&w, STDIN_FILENO, EV_READ);
936		1143
937	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1144	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
938		1145
939	This macro initialises the type-specific parts of a watcher. You need to	1146	This macro initialises the type-specific parts of a watcher. You need to
940	call C<ev_init> at least once before you call this macro, but you can	1147	call C<ev_init> at least once before you call this macro, but you can
941	call C<ev_TYPE_set> any number of times. You must not, however, call this	1148	call C<ev_TYPE_set> any number of times. You must not, however, call this
942	macro on a watcher that is active (it can be pending, however, which is a	1149	macro on a watcher that is active (it can be pending, however, which is a
…		…
955		1162
956	Example: Initialise and set an C<ev_io> watcher in one step.	1163	Example: Initialise and set an C<ev_io> watcher in one step.
957		1164
958	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);	1165	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
959		1166
960	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1167	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
961		1168
962	Starts (activates) the given watcher. Only active watchers will receive	1169	Starts (activates) the given watcher. Only active watchers will receive
963	events. If the watcher is already active nothing will happen.	1170	events. If the watcher is already active nothing will happen.
964		1171
965	Example: Start the C<ev_io> watcher that is being abused as example in this	1172	Example: Start the C<ev_io> watcher that is being abused as example in this
966	whole section.	1173	whole section.
967		1174
968	ev_io_start (EV_DEFAULT_UC, &w);	1175	ev_io_start (EV_DEFAULT_UC, &w);
969		1176
970	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1177	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
971		1178
972	Stops the given watcher again (if active) and clears the pending	1179	Stops the given watcher if active, and clears the pending status (whether
		1180	the watcher was active or not).
		1181
973	status. It is possible that stopped watchers are pending (for example,	1182	It is possible that stopped watchers are pending - for example,
974	non-repeating timers are being stopped when they become pending), but	1183	non-repeating timers are being stopped when they become pending - but
975	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1184	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
976	you want to free or reuse the memory used by the watcher it is therefore a	1185	pending. If you want to free or reuse the memory used by the watcher it is
977	good idea to always call its C<ev_TYPE_stop> function.	1186	therefore a good idea to always call its C<ev_TYPE_stop> function.
978		1187
979	=item bool ev_is_active (ev_TYPE *watcher)	1188	=item bool ev_is_active (ev_TYPE *watcher)
980		1189
981	Returns a true value iff the watcher is active (i.e. it has been started	1190	Returns a true value iff the watcher is active (i.e. it has been started
982	and not yet been stopped). As long as a watcher is active you must not modify	1191	and not yet been stopped). As long as a watcher is active you must not modify
…		…
998	=item ev_cb_set (ev_TYPE *watcher, callback)	1207	=item ev_cb_set (ev_TYPE *watcher, callback)
999		1208
1000	Change the callback. You can change the callback at virtually any time	1209	Change the callback. You can change the callback at virtually any time
1001	(modulo threads).	1210	(modulo threads).
1002		1211
1003	=item ev_set_priority (ev_TYPE *watcher, priority)	1212	=item ev_set_priority (ev_TYPE *watcher, int priority)
1004		1213
1005	=item int ev_priority (ev_TYPE *watcher)	1214	=item int ev_priority (ev_TYPE *watcher)
1006		1215
1007	Set and query the priority of the watcher. The priority is a small	1216	Set and query the priority of the watcher. The priority is a small
1008	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1217	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1009	(default: C<-2>). Pending watchers with higher priority will be invoked	1218	(default: C<-2>). Pending watchers with higher priority will be invoked
1010	before watchers with lower priority, but priority will not keep watchers	1219	before watchers with lower priority, but priority will not keep watchers
1011	from being executed (except for C<ev_idle> watchers).	1220	from being executed (except for C<ev_idle> watchers).
1012		1221
1013	This means that priorities are I<only> used for ordering callback
1014	invocation after new events have been received. This is useful, for
1015	example, to reduce latency after idling, or more often, to bind two
1016	watchers on the same event and make sure one is called first.
1017
1018	If you need to suppress invocation when higher priority events are pending	1222	If you need to suppress invocation when higher priority events are pending
1019	you need to look at C<ev_idle> watchers, which provide this functionality.	1223	you need to look at C<ev_idle> watchers, which provide this functionality.
1020		1224
1021	You I<must not> change the priority of a watcher as long as it is active or	1225	You I<must not> change the priority of a watcher as long as it is active or
1022	pending.	1226	pending.
1023		1227
		1228	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1229	fine, as long as you do not mind that the priority value you query might
		1230	or might not have been clamped to the valid range.
		1231
1024	The default priority used by watchers when no priority has been set is	1232	The default priority used by watchers when no priority has been set is
1025	always C<0>, which is supposed to not be too high and not be too low :).	1233	always C<0>, which is supposed to not be too high and not be too low :).
1026		1234
1027	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1235	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1028	fine, as long as you do not mind that the priority value you query might	1236	priorities.
1029	or might not have been adjusted to be within valid range.
1030		1237
1031	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1238	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1032		1239
1033	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1240	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1034	C<loop> nor C<revents> need to be valid as long as the watcher callback	1241	C<loop> nor C<revents> need to be valid as long as the watcher callback
…		…
1041	returns its C<revents> bitset (as if its callback was invoked). If the	1248	returns its C<revents> bitset (as if its callback was invoked). If the
1042	watcher isn't pending it does nothing and returns C<0>.	1249	watcher isn't pending it does nothing and returns C<0>.
1043		1250
1044	Sometimes it can be useful to "poll" a watcher instead of waiting for its	1251	Sometimes it can be useful to "poll" a watcher instead of waiting for its
1045	callback to be invoked, which can be accomplished with this function.	1252	callback to be invoked, which can be accomplished with this function.
		1253
		1254	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1255
		1256	Feeds the given event set into the event loop, as if the specified event
		1257	had happened for the specified watcher (which must be a pointer to an
		1258	initialised but not necessarily started event watcher). Obviously you must
		1259	not free the watcher as long as it has pending events.
		1260
		1261	Stopping the watcher, letting libev invoke it, or calling
		1262	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1263	not started in the first place.
		1264
		1265	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1266	functions that do not need a watcher.
1046		1267
1047	=back	1268	=back
1048		1269
1049		1270
1050	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1271	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
…		…
1056	member, you can also "subclass" the watcher type and provide your own	1277	member, you can also "subclass" the watcher type and provide your own
1057	data:	1278	data:
1058		1279
1059	struct my_io	1280	struct my_io
1060	{	1281	{
1061	struct ev_io io;	1282	ev_io io;
1062	int otherfd;	1283	int otherfd;
1063	void *somedata;	1284	void *somedata;
1064	struct whatever *mostinteresting;	1285	struct whatever *mostinteresting;
1065	};	1286	};
1066		1287
…		…
1069	ev_io_init (&w.io, my_cb, fd, EV_READ);	1290	ev_io_init (&w.io, my_cb, fd, EV_READ);
1070		1291
1071	And since your callback will be called with a pointer to the watcher, you	1292	And since your callback will be called with a pointer to the watcher, you
1072	can cast it back to your own type:	1293	can cast it back to your own type:
1073		1294
1074	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1295	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1075	{	1296	{
1076	struct my_io *w = (struct my_io *)w_;	1297	struct my_io *w = (struct my_io *)w_;
1077	...	1298	...
1078	}	1299	}
1079		1300
…		…
1097	programmers):	1318	programmers):
1098		1319
1099	#include <stddef.h>	1320	#include <stddef.h>
1100		1321
1101	static void	1322	static void
1102	t1_cb (EV_P_ struct ev_timer *w, int revents)	1323	t1_cb (EV_P_ ev_timer *w, int revents)
1103	{	1324	{
1104	struct my_biggy big = (struct my_biggy *	1325	struct my_biggy big = (struct my_biggy *)
1105	(((char *)w) - offsetof (struct my_biggy, t1));	1326	(((char *)w) - offsetof (struct my_biggy, t1));
1106	}	1327	}
1107		1328
1108	static void	1329	static void
1109	t2_cb (EV_P_ struct ev_timer *w, int revents)	1330	t2_cb (EV_P_ ev_timer *w, int revents)
1110	{	1331	{
1111	struct my_biggy big = (struct my_biggy *	1332	struct my_biggy big = (struct my_biggy *)
1112	(((char *)w) - offsetof (struct my_biggy, t2));	1333	(((char *)w) - offsetof (struct my_biggy, t2));
1113	}	1334	}
		1335
		1336	=head2 WATCHER PRIORITY MODELS
		1337
		1338	Many event loops support I<watcher priorities>, which are usually small
		1339	integers that influence the ordering of event callback invocation
		1340	between watchers in some way, all else being equal.
		1341
		1342	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1343	description for the more technical details such as the actual priority
		1344	range.
		1345
		1346	There are two common ways how these these priorities are being interpreted
		1347	by event loops:
		1348
		1349	In the more common lock-out model, higher priorities "lock out" invocation
		1350	of lower priority watchers, which means as long as higher priority
		1351	watchers receive events, lower priority watchers are not being invoked.
		1352
		1353	The less common only-for-ordering model uses priorities solely to order
		1354	callback invocation within a single event loop iteration: Higher priority
		1355	watchers are invoked before lower priority ones, but they all get invoked
		1356	before polling for new events.
		1357
		1358	Libev uses the second (only-for-ordering) model for all its watchers
		1359	except for idle watchers (which use the lock-out model).
		1360
		1361	The rationale behind this is that implementing the lock-out model for
		1362	watchers is not well supported by most kernel interfaces, and most event
		1363	libraries will just poll for the same events again and again as long as
		1364	their callbacks have not been executed, which is very inefficient in the
		1365	common case of one high-priority watcher locking out a mass of lower
		1366	priority ones.
		1367
		1368	Static (ordering) priorities are most useful when you have two or more
		1369	watchers handling the same resource: a typical usage example is having an
		1370	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1371	timeouts. Under load, data might be received while the program handles
		1372	other jobs, but since timers normally get invoked first, the timeout
		1373	handler will be executed before checking for data. In that case, giving
		1374	the timer a lower priority than the I/O watcher ensures that I/O will be
		1375	handled first even under adverse conditions (which is usually, but not
		1376	always, what you want).
		1377
		1378	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1379	will only be executed when no same or higher priority watchers have
		1380	received events, they can be used to implement the "lock-out" model when
		1381	required.
		1382
		1383	For example, to emulate how many other event libraries handle priorities,
		1384	you can associate an C<ev_idle> watcher to each such watcher, and in
		1385	the normal watcher callback, you just start the idle watcher. The real
		1386	processing is done in the idle watcher callback. This causes libev to
		1387	continuously poll and process kernel event data for the watcher, but when
		1388	the lock-out case is known to be rare (which in turn is rare :), this is
		1389	workable.
		1390
		1391	Usually, however, the lock-out model implemented that way will perform
		1392	miserably under the type of load it was designed to handle. In that case,
		1393	it might be preferable to stop the real watcher before starting the
		1394	idle watcher, so the kernel will not have to process the event in case
		1395	the actual processing will be delayed for considerable time.
		1396
		1397	Here is an example of an I/O watcher that should run at a strictly lower
		1398	priority than the default, and which should only process data when no
		1399	other events are pending:
		1400
		1401	ev_idle idle; // actual processing watcher
		1402	ev_io io; // actual event watcher
		1403
		1404	static void
		1405	io_cb (EV_P_ ev_io *w, int revents)
		1406	{
		1407	// stop the I/O watcher, we received the event, but
		1408	// are not yet ready to handle it.
		1409	ev_io_stop (EV_A_ w);
		1410
		1411	// start the idle watcher to handle the actual event.
		1412	// it will not be executed as long as other watchers
		1413	// with the default priority are receiving events.
		1414	ev_idle_start (EV_A_ &idle);
		1415	}
		1416
		1417	static void
		1418	idle_cb (EV_P_ ev_idle *w, int revents)
		1419	{
		1420	// actual processing
		1421	read (STDIN_FILENO, ...);
		1422
		1423	// have to start the I/O watcher again, as
		1424	// we have handled the event
		1425	ev_io_start (EV_P_ &io);
		1426	}
		1427
		1428	// initialisation
		1429	ev_idle_init (&idle, idle_cb);
		1430	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1431	ev_io_start (EV_DEFAULT_ &io);
		1432
		1433	In the "real" world, it might also be beneficial to start a timer, so that
		1434	low-priority connections can not be locked out forever under load. This
		1435	enables your program to keep a lower latency for important connections
		1436	during short periods of high load, while not completely locking out less
		1437	important ones.
1114		1438
1115		1439
1116	=head1 WATCHER TYPES	1440	=head1 WATCHER TYPES
1117		1441
1118	This section describes each watcher in detail, but will not repeat	1442	This section describes each watcher in detail, but will not repeat
…		…
1144	descriptors to non-blocking mode is also usually a good idea (but not	1468	descriptors to non-blocking mode is also usually a good idea (but not
1145	required if you know what you are doing).	1469	required if you know what you are doing).
1146		1470
1147	If you cannot use non-blocking mode, then force the use of a	1471	If you cannot use non-blocking mode, then force the use of a
1148	known-to-be-good backend (at the time of this writing, this includes only	1472	known-to-be-good backend (at the time of this writing, this includes only
1149	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>).	1473	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1474	descriptors for which non-blocking operation makes no sense (such as
		1475	files) - libev doesn't guarantee any specific behaviour in that case.
1150		1476
1151	Another thing you have to watch out for is that it is quite easy to	1477	Another thing you have to watch out for is that it is quite easy to
1152	receive "spurious" readiness notifications, that is your callback might	1478	receive "spurious" readiness notifications, that is your callback might
1153	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1479	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1154	because there is no data. Not only are some backends known to create a	1480	because there is no data. Not only are some backends known to create a
…		…
1219		1545
1220	So when you encounter spurious, unexplained daemon exits, make sure you	1546	So when you encounter spurious, unexplained daemon exits, make sure you
1221	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1547	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1222	somewhere, as that would have given you a big clue).	1548	somewhere, as that would have given you a big clue).
1223		1549
		1550	=head3 The special problem of accept()ing when you can't
		1551
		1552	Many implementations of the POSIX C<accept> function (for example,
		1553	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1554	connection from the pending queue in all error cases.
		1555
		1556	For example, larger servers often run out of file descriptors (because
		1557	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1558	rejecting the connection, leading to libev signalling readiness on
		1559	the next iteration again (the connection still exists after all), and
		1560	typically causing the program to loop at 100% CPU usage.
		1561
		1562	Unfortunately, the set of errors that cause this issue differs between
		1563	operating systems, there is usually little the app can do to remedy the
		1564	situation, and no known thread-safe method of removing the connection to
		1565	cope with overload is known (to me).
		1566
		1567	One of the easiest ways to handle this situation is to just ignore it
		1568	- when the program encounters an overload, it will just loop until the
		1569	situation is over. While this is a form of busy waiting, no OS offers an
		1570	event-based way to handle this situation, so it's the best one can do.
		1571
		1572	A better way to handle the situation is to log any errors other than
		1573	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1574	messages, and continue as usual, which at least gives the user an idea of
		1575	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1576	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1577	usage.
		1578
		1579	If your program is single-threaded, then you could also keep a dummy file
		1580	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1581	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1582	close that fd, and create a new dummy fd. This will gracefully refuse
		1583	clients under typical overload conditions.
		1584
		1585	The last way to handle it is to simply log the error and C<exit>, as
		1586	is often done with C<malloc> failures, but this results in an easy
		1587	opportunity for a DoS attack.
1224		1588
1225	=head3 Watcher-Specific Functions	1589	=head3 Watcher-Specific Functions
1226		1590
1227	=over 4	1591	=over 4
1228		1592
…		…
1249	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1613	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1250	readable, but only once. Since it is likely line-buffered, you could	1614	readable, but only once. Since it is likely line-buffered, you could
1251	attempt to read a whole line in the callback.	1615	attempt to read a whole line in the callback.
1252		1616
1253	static void	1617	static void
1254	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1618	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1255	{	1619	{
1256	ev_io_stop (loop, w);	1620	ev_io_stop (loop, w);
1257	.. read from stdin here (or from w->fd) and handle any I/O errors	1621	.. read from stdin here (or from w->fd) and handle any I/O errors
1258	}	1622	}
1259		1623
1260	...	1624	...
1261	struct ev_loop *loop = ev_default_init (0);	1625	struct ev_loop *loop = ev_default_init (0);
1262	struct ev_io stdin_readable;	1626	ev_io stdin_readable;
1263	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1627	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1264	ev_io_start (loop, &stdin_readable);	1628	ev_io_start (loop, &stdin_readable);
1265	ev_loop (loop, 0);	1629	ev_loop (loop, 0);
1266		1630
1267		1631
…		…
1275	year, it will still time out after (roughly) one hour. "Roughly" because	1639	year, it will still time out after (roughly) one hour. "Roughly" because
1276	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1640	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1277	monotonic clock option helps a lot here).	1641	monotonic clock option helps a lot here).
1278		1642
1279	The callback is guaranteed to be invoked only I<after> its timeout has	1643	The callback is guaranteed to be invoked only I<after> its timeout has
1280	passed, but if multiple timers become ready during the same loop iteration	1644	passed (not I<at>, so on systems with very low-resolution clocks this
1281	then order of execution is undefined.	1645	might introduce a small delay). If multiple timers become ready during the
		1646	same loop iteration then the ones with earlier time-out values are invoked
		1647	before ones of the same priority with later time-out values (but this is
		1648	no longer true when a callback calls C<ev_loop> recursively).
		1649
		1650	=head3 Be smart about timeouts
		1651
		1652	Many real-world problems involve some kind of timeout, usually for error
		1653	recovery. A typical example is an HTTP request - if the other side hangs,
		1654	you want to raise some error after a while.
		1655
		1656	What follows are some ways to handle this problem, from obvious and
		1657	inefficient to smart and efficient.
		1658
		1659	In the following, a 60 second activity timeout is assumed - a timeout that
		1660	gets reset to 60 seconds each time there is activity (e.g. each time some
		1661	data or other life sign was received).
		1662
		1663	=over 4
		1664
		1665	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1666
		1667	This is the most obvious, but not the most simple way: In the beginning,
		1668	start the watcher:
		1669
		1670	ev_timer_init (timer, callback, 60., 0.);
		1671	ev_timer_start (loop, timer);
		1672
		1673	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1674	and start it again:
		1675
		1676	ev_timer_stop (loop, timer);
		1677	ev_timer_set (timer, 60., 0.);
		1678	ev_timer_start (loop, timer);
		1679
		1680	This is relatively simple to implement, but means that each time there is
		1681	some activity, libev will first have to remove the timer from its internal
		1682	data structure and then add it again. Libev tries to be fast, but it's
		1683	still not a constant-time operation.
		1684
		1685	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1686
		1687	This is the easiest way, and involves using C<ev_timer_again> instead of
		1688	C<ev_timer_start>.
		1689
		1690	To implement this, configure an C<ev_timer> with a C<repeat> value
		1691	of C<60> and then call C<ev_timer_again> at start and each time you
		1692	successfully read or write some data. If you go into an idle state where
		1693	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1694	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1695
		1696	That means you can ignore both the C<ev_timer_start> function and the
		1697	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1698	member and C<ev_timer_again>.
		1699
		1700	At start:
		1701
		1702	ev_init (timer, callback);
		1703	timer->repeat = 60.;
		1704	ev_timer_again (loop, timer);
		1705
		1706	Each time there is some activity:
		1707
		1708	ev_timer_again (loop, timer);
		1709
		1710	It is even possible to change the time-out on the fly, regardless of
		1711	whether the watcher is active or not:
		1712
		1713	timer->repeat = 30.;
		1714	ev_timer_again (loop, timer);
		1715
		1716	This is slightly more efficient then stopping/starting the timer each time
		1717	you want to modify its timeout value, as libev does not have to completely
		1718	remove and re-insert the timer from/into its internal data structure.
		1719
		1720	It is, however, even simpler than the "obvious" way to do it.
		1721
		1722	=item 3. Let the timer time out, but then re-arm it as required.
		1723
		1724	This method is more tricky, but usually most efficient: Most timeouts are
		1725	relatively long compared to the intervals between other activity - in
		1726	our example, within 60 seconds, there are usually many I/O events with
		1727	associated activity resets.
		1728
		1729	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1730	but remember the time of last activity, and check for a real timeout only
		1731	within the callback:
		1732
		1733	ev_tstamp last_activity; // time of last activity
		1734
		1735	static void
		1736	callback (EV_P_ ev_timer *w, int revents)
		1737	{
		1738	ev_tstamp now = ev_now (EV_A);
		1739	ev_tstamp timeout = last_activity + 60.;
		1740
		1741	// if last_activity + 60. is older than now, we did time out
		1742	if (timeout < now)
		1743	{
		1744	// timeout occurred, take action
		1745	}
		1746	else
		1747	{
		1748	// callback was invoked, but there was some activity, re-arm
		1749	// the watcher to fire in last_activity + 60, which is
		1750	// guaranteed to be in the future, so "again" is positive:
		1751	w->repeat = timeout - now;
		1752	ev_timer_again (EV_A_ w);
		1753	}
		1754	}
		1755
		1756	To summarise the callback: first calculate the real timeout (defined
		1757	as "60 seconds after the last activity"), then check if that time has
		1758	been reached, which means something I<did>, in fact, time out. Otherwise
		1759	the callback was invoked too early (C<timeout> is in the future), so
		1760	re-schedule the timer to fire at that future time, to see if maybe we have
		1761	a timeout then.
		1762
		1763	Note how C<ev_timer_again> is used, taking advantage of the
		1764	C<ev_timer_again> optimisation when the timer is already running.
		1765
		1766	This scheme causes more callback invocations (about one every 60 seconds
		1767	minus half the average time between activity), but virtually no calls to
		1768	libev to change the timeout.
		1769
		1770	To start the timer, simply initialise the watcher and set C<last_activity>
		1771	to the current time (meaning we just have some activity :), then call the
		1772	callback, which will "do the right thing" and start the timer:
		1773
		1774	ev_init (timer, callback);
		1775	last_activity = ev_now (loop);
		1776	callback (loop, timer, EV_TIMER);
		1777
		1778	And when there is some activity, simply store the current time in
		1779	C<last_activity>, no libev calls at all:
		1780
		1781	last_activity = ev_now (loop);
		1782
		1783	This technique is slightly more complex, but in most cases where the
		1784	time-out is unlikely to be triggered, much more efficient.
		1785
		1786	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1787	callback :) - just change the timeout and invoke the callback, which will
		1788	fix things for you.
		1789
		1790	=item 4. Wee, just use a double-linked list for your timeouts.
		1791
		1792	If there is not one request, but many thousands (millions...), all
		1793	employing some kind of timeout with the same timeout value, then one can
		1794	do even better:
		1795
		1796	When starting the timeout, calculate the timeout value and put the timeout
		1797	at the I<end> of the list.
		1798
		1799	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1800	the list is expected to fire (for example, using the technique #3).
		1801
		1802	When there is some activity, remove the timer from the list, recalculate
		1803	the timeout, append it to the end of the list again, and make sure to
		1804	update the C<ev_timer> if it was taken from the beginning of the list.
		1805
		1806	This way, one can manage an unlimited number of timeouts in O(1) time for
		1807	starting, stopping and updating the timers, at the expense of a major
		1808	complication, and having to use a constant timeout. The constant timeout
		1809	ensures that the list stays sorted.
		1810
		1811	=back
		1812
		1813	So which method the best?
		1814
		1815	Method #2 is a simple no-brain-required solution that is adequate in most
		1816	situations. Method #3 requires a bit more thinking, but handles many cases
		1817	better, and isn't very complicated either. In most case, choosing either
		1818	one is fine, with #3 being better in typical situations.
		1819
		1820	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1821	rather complicated, but extremely efficient, something that really pays
		1822	off after the first million or so of active timers, i.e. it's usually
		1823	overkill :)
1282		1824
1283	=head3 The special problem of time updates	1825	=head3 The special problem of time updates
1284		1826
1285	Establishing the current time is a costly operation (it usually takes at	1827	Establishing the current time is a costly operation (it usually takes at
1286	least two system calls): EV therefore updates its idea of the current	1828	least two system calls): EV therefore updates its idea of the current
…		…
1298		1840
1299	If the event loop is suspended for a long time, you can also force an	1841	If the event loop is suspended for a long time, you can also force an
1300	update of the time returned by C<ev_now ()> by calling C<ev_now_update	1842	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1301	()>.	1843	()>.
1302		1844
		1845	=head3 The special problems of suspended animation
		1846
		1847	When you leave the server world it is quite customary to hit machines that
		1848	can suspend/hibernate - what happens to the clocks during such a suspend?
		1849
		1850	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1851	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1852	to run until the system is suspended, but they will not advance while the
		1853	system is suspended. That means, on resume, it will be as if the program
		1854	was frozen for a few seconds, but the suspend time will not be counted
		1855	towards C<ev_timer> when a monotonic clock source is used. The real time
		1856	clock advanced as expected, but if it is used as sole clocksource, then a
		1857	long suspend would be detected as a time jump by libev, and timers would
		1858	be adjusted accordingly.
		1859
		1860	I would not be surprised to see different behaviour in different between
		1861	operating systems, OS versions or even different hardware.
		1862
		1863	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1864	time jump in the monotonic clocks and the realtime clock. If the program
		1865	is suspended for a very long time, and monotonic clock sources are in use,
		1866	then you can expect C<ev_timer>s to expire as the full suspension time
		1867	will be counted towards the timers. When no monotonic clock source is in
		1868	use, then libev will again assume a timejump and adjust accordingly.
		1869
		1870	It might be beneficial for this latter case to call C<ev_suspend>
		1871	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1872	deterministic behaviour in this case (you can do nothing against
		1873	C<SIGSTOP>).
		1874
1303	=head3 Watcher-Specific Functions and Data Members	1875	=head3 Watcher-Specific Functions and Data Members
1304		1876
1305	=over 4	1877	=over 4
1306		1878
1307	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1879	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
…		…
1330	If the timer is started but non-repeating, stop it (as if it timed out).	1902	If the timer is started but non-repeating, stop it (as if it timed out).
1331		1903
1332	If the timer is repeating, either start it if necessary (with the	1904	If the timer is repeating, either start it if necessary (with the
1333	C<repeat> value), or reset the running timer to the C<repeat> value.	1905	C<repeat> value), or reset the running timer to the C<repeat> value.
1334		1906
1335	This sounds a bit complicated, but here is a useful and typical	1907	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1336	example: Imagine you have a TCP connection and you want a so-called idle	1908	usage example.
1337	timeout, that is, you want to be called when there have been, say, 60
1338	seconds of inactivity on the socket. The easiest way to do this is to
1339	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1340	C<ev_timer_again> each time you successfully read or write some data. If
1341	you go into an idle state where you do not expect data to travel on the
1342	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1343	automatically restart it if need be.
1344		1909
1345	That means you can ignore the C<after> value and C<ev_timer_start>	1910	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1346	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1347		1911
1348	ev_timer_init (timer, callback, 0., 5.);	1912	Returns the remaining time until a timer fires. If the timer is active,
1349	ev_timer_again (loop, timer);	1913	then this time is relative to the current event loop time, otherwise it's
1350	...	1914	the timeout value currently configured.
1351	timer->again = 17.;
1352	ev_timer_again (loop, timer);
1353	...
1354	timer->again = 10.;
1355	ev_timer_again (loop, timer);
1356		1915
1357	This is more slightly efficient then stopping/starting the timer each time	1916	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1358	you want to modify its timeout value.	1917	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
1359		1918	will return C<4>. When the timer expires and is restarted, it will return
1360	Note, however, that it is often even more efficient to remember the	1919	roughly C<7> (likely slightly less as callback invocation takes some time,
1361	time of the last activity and let the timer time-out naturally. In the	1920	too), and so on.
1362	callback, you then check whether the time-out is real, or, if there was
1363	some activity, you reschedule the watcher to time-out in "last_activity +
1364	timeout - ev_now ()" seconds.
1365		1921
1366	=item ev_tstamp repeat [read-write]	1922	=item ev_tstamp repeat [read-write]
1367		1923
1368	The current C<repeat> value. Will be used each time the watcher times out	1924	The current C<repeat> value. Will be used each time the watcher times out
1369	or C<ev_timer_again> is called, and determines the next timeout (if any),	1925	or C<ev_timer_again> is called, and determines the next timeout (if any),
…		…
1374	=head3 Examples	1930	=head3 Examples
1375		1931
1376	Example: Create a timer that fires after 60 seconds.	1932	Example: Create a timer that fires after 60 seconds.
1377		1933
1378	static void	1934	static void
1379	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1935	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1380	{	1936	{
1381	.. one minute over, w is actually stopped right here	1937	.. one minute over, w is actually stopped right here
1382	}	1938	}
1383		1939
1384	struct ev_timer mytimer;	1940	ev_timer mytimer;
1385	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1941	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1386	ev_timer_start (loop, &mytimer);	1942	ev_timer_start (loop, &mytimer);
1387		1943
1388	Example: Create a timeout timer that times out after 10 seconds of	1944	Example: Create a timeout timer that times out after 10 seconds of
1389	inactivity.	1945	inactivity.
1390		1946
1391	static void	1947	static void
1392	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1948	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1393	{	1949	{
1394	.. ten seconds without any activity	1950	.. ten seconds without any activity
1395	}	1951	}
1396		1952
1397	struct ev_timer mytimer;	1953	ev_timer mytimer;
1398	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1954	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1399	ev_timer_again (&mytimer); /* start timer */	1955	ev_timer_again (&mytimer); /* start timer */
1400	ev_loop (loop, 0);	1956	ev_loop (loop, 0);
1401		1957
1402	// and in some piece of code that gets executed on any "activity":	1958	// and in some piece of code that gets executed on any "activity":
…		…
1407	=head2 C<ev_periodic> - to cron or not to cron?	1963	=head2 C<ev_periodic> - to cron or not to cron?
1408		1964
1409	Periodic watchers are also timers of a kind, but they are very versatile	1965	Periodic watchers are also timers of a kind, but they are very versatile
1410	(and unfortunately a bit complex).	1966	(and unfortunately a bit complex).
1411		1967
1412	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1968	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1413	but on wall clock time (absolute time). You can tell a periodic watcher	1969	relative time, the physical time that passes) but on wall clock time
1414	to trigger after some specific point in time. For example, if you tell a	1970	(absolute time, the thing you can read on your calender or clock). The
1415	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1971	difference is that wall clock time can run faster or slower than real
1416	+ 10.>, that is, an absolute time not a delay) and then reset your system	1972	time, and time jumps are not uncommon (e.g. when you adjust your
1417	clock to January of the previous year, then it will take more than year	1973	wrist-watch).
1418	to trigger the event (unlike an C<ev_timer>, which would still trigger
1419	roughly 10 seconds later as it uses a relative timeout).
1420		1974
		1975	You can tell a periodic watcher to trigger after some specific point
		1976	in time: for example, if you tell a periodic watcher to trigger "in 10
		1977	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1978	not a delay) and then reset your system clock to January of the previous
		1979	year, then it will take a year or more to trigger the event (unlike an
		1980	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1981	it, as it uses a relative timeout).
		1982
1421	C<ev_periodic>s can also be used to implement vastly more complex timers,	1983	C<ev_periodic> watchers can also be used to implement vastly more complex
1422	such as triggering an event on each "midnight, local time", or other	1984	timers, such as triggering an event on each "midnight, local time", or
1423	complicated rules.	1985	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1986	those cannot react to time jumps.
1424		1987
1425	As with timers, the callback is guaranteed to be invoked only when the	1988	As with timers, the callback is guaranteed to be invoked only when the
1426	time (C<at>) has passed, but if multiple periodic timers become ready	1989	point in time where it is supposed to trigger has passed. If multiple
1427	during the same loop iteration, then order of execution is undefined.	1990	timers become ready during the same loop iteration then the ones with
		1991	earlier time-out values are invoked before ones with later time-out values
		1992	(but this is no longer true when a callback calls C<ev_loop> recursively).
1428		1993
1429	=head3 Watcher-Specific Functions and Data Members	1994	=head3 Watcher-Specific Functions and Data Members
1430		1995
1431	=over 4	1996	=over 4
1432		1997
1433	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1998	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1434		1999
1435	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2000	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1436		2001
1437	Lots of arguments, lets sort it out... There are basically three modes of	2002	Lots of arguments, let's sort it out... There are basically three modes of
1438	operation, and we will explain them from simplest to most complex:	2003	operation, and we will explain them from simplest to most complex:
1439		2004
1440	=over 4	2005	=over 4
1441		2006
1442	=item * absolute timer (at = time, interval = reschedule_cb = 0)	2007	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1443		2008
1444	In this configuration the watcher triggers an event after the wall clock	2009	In this configuration the watcher triggers an event after the wall clock
1445	time C<at> has passed. It will not repeat and will not adjust when a time	2010	time C<offset> has passed. It will not repeat and will not adjust when a
1446	jump occurs, that is, if it is to be run at January 1st 2011 then it will	2011	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1447	only run when the system clock reaches or surpasses this time.	2012	will be stopped and invoked when the system clock reaches or surpasses
		2013	this point in time.
1448		2014
1449	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	2015	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1450		2016
1451	In this mode the watcher will always be scheduled to time out at the next	2017	In this mode the watcher will always be scheduled to time out at the next
1452	C<at + N * interval> time (for some integer N, which can also be negative)	2018	C<offset + N * interval> time (for some integer N, which can also be
1453	and then repeat, regardless of any time jumps.	2019	negative) and then repeat, regardless of any time jumps. The C<offset>
		2020	argument is merely an offset into the C<interval> periods.
1454		2021
1455	This can be used to create timers that do not drift with respect to the	2022	This can be used to create timers that do not drift with respect to the
1456	system clock, for example, here is a C<ev_periodic> that triggers each	2023	system clock, for example, here is an C<ev_periodic> that triggers each
1457	hour, on the hour:	2024	hour, on the hour (with respect to UTC):
1458		2025
1459	ev_periodic_set (&periodic, 0., 3600., 0);	2026	ev_periodic_set (&periodic, 0., 3600., 0);
1460		2027
1461	This doesn't mean there will always be 3600 seconds in between triggers,	2028	This doesn't mean there will always be 3600 seconds in between triggers,
1462	but only that the callback will be called when the system time shows a	2029	but only that the callback will be called when the system time shows a
1463	full hour (UTC), or more correctly, when the system time is evenly divisible	2030	full hour (UTC), or more correctly, when the system time is evenly divisible
1464	by 3600.	2031	by 3600.
1465		2032
1466	Another way to think about it (for the mathematically inclined) is that	2033	Another way to think about it (for the mathematically inclined) is that
1467	C<ev_periodic> will try to run the callback in this mode at the next possible	2034	C<ev_periodic> will try to run the callback in this mode at the next possible
1468	time where C<time = at (mod interval)>, regardless of any time jumps.	2035	time where C<time = offset (mod interval)>, regardless of any time jumps.
1469		2036
1470	For numerical stability it is preferable that the C<at> value is near	2037	For numerical stability it is preferable that the C<offset> value is near
1471	C<ev_now ()> (the current time), but there is no range requirement for	2038	C<ev_now ()> (the current time), but there is no range requirement for
1472	this value, and in fact is often specified as zero.	2039	this value, and in fact is often specified as zero.
1473		2040
1474	Note also that there is an upper limit to how often a timer can fire (CPU	2041	Note also that there is an upper limit to how often a timer can fire (CPU
1475	speed for example), so if C<interval> is very small then timing stability	2042	speed for example), so if C<interval> is very small then timing stability
1476	will of course deteriorate. Libev itself tries to be exact to be about one	2043	will of course deteriorate. Libev itself tries to be exact to be about one
1477	millisecond (if the OS supports it and the machine is fast enough).	2044	millisecond (if the OS supports it and the machine is fast enough).
1478		2045
1479	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2046	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1480		2047
1481	In this mode the values for C<interval> and C<at> are both being	2048	In this mode the values for C<interval> and C<offset> are both being
1482	ignored. Instead, each time the periodic watcher gets scheduled, the	2049	ignored. Instead, each time the periodic watcher gets scheduled, the
1483	reschedule callback will be called with the watcher as first, and the	2050	reschedule callback will be called with the watcher as first, and the
1484	current time as second argument.	2051	current time as second argument.
1485		2052
1486	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2053	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1487	ever, or make ANY event loop modifications whatsoever>.	2054	or make ANY other event loop modifications whatsoever, unless explicitly
		2055	allowed by documentation here>.
1488		2056
1489	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2057	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1490	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2058	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1491	only event loop modification you are allowed to do).	2059	only event loop modification you are allowed to do).
1492		2060
1493	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	2061	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1494	*w, ev_tstamp now)>, e.g.:	2062	*w, ev_tstamp now)>, e.g.:
1495		2063
		2064	static ev_tstamp
1496	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2065	my_rescheduler (ev_periodic *w, ev_tstamp now)
1497	{	2066	{
1498	return now + 60.;	2067	return now + 60.;
1499	}	2068	}
1500		2069
1501	It must return the next time to trigger, based on the passed time value	2070	It must return the next time to trigger, based on the passed time value
…		…
1521	a different time than the last time it was called (e.g. in a crond like	2090	a different time than the last time it was called (e.g. in a crond like
1522	program when the crontabs have changed).	2091	program when the crontabs have changed).
1523		2092
1524	=item ev_tstamp ev_periodic_at (ev_periodic *)	2093	=item ev_tstamp ev_periodic_at (ev_periodic *)
1525		2094
1526	When active, returns the absolute time that the watcher is supposed to	2095	When active, returns the absolute time that the watcher is supposed
1527	trigger next.	2096	to trigger next. This is not the same as the C<offset> argument to
		2097	C<ev_periodic_set>, but indeed works even in interval and manual
		2098	rescheduling modes.
1528		2099
1529	=item ev_tstamp offset [read-write]	2100	=item ev_tstamp offset [read-write]
1530		2101
1531	When repeating, this contains the offset value, otherwise this is the	2102	When repeating, this contains the offset value, otherwise this is the
1532	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2103	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2104	although libev might modify this value for better numerical stability).
1533		2105
1534	Can be modified any time, but changes only take effect when the periodic	2106	Can be modified any time, but changes only take effect when the periodic
1535	timer fires or C<ev_periodic_again> is being called.	2107	timer fires or C<ev_periodic_again> is being called.
1536		2108
1537	=item ev_tstamp interval [read-write]	2109	=item ev_tstamp interval [read-write]
1538		2110
1539	The current interval value. Can be modified any time, but changes only	2111	The current interval value. Can be modified any time, but changes only
1540	take effect when the periodic timer fires or C<ev_periodic_again> is being	2112	take effect when the periodic timer fires or C<ev_periodic_again> is being
1541	called.	2113	called.
1542		2114
1543	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2115	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1544		2116
1545	The current reschedule callback, or C<0>, if this functionality is	2117	The current reschedule callback, or C<0>, if this functionality is
1546	switched off. Can be changed any time, but changes only take effect when	2118	switched off. Can be changed any time, but changes only take effect when
1547	the periodic timer fires or C<ev_periodic_again> is being called.	2119	the periodic timer fires or C<ev_periodic_again> is being called.
1548		2120
…		…
1553	Example: Call a callback every hour, or, more precisely, whenever the	2125	Example: Call a callback every hour, or, more precisely, whenever the
1554	system time is divisible by 3600. The callback invocation times have	2126	system time is divisible by 3600. The callback invocation times have
1555	potentially a lot of jitter, but good long-term stability.	2127	potentially a lot of jitter, but good long-term stability.
1556		2128
1557	static void	2129	static void
1558	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2130	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
1559	{	2131	{
1560	... its now a full hour (UTC, or TAI or whatever your clock follows)	2132	... its now a full hour (UTC, or TAI or whatever your clock follows)
1561	}	2133	}
1562		2134
1563	struct ev_periodic hourly_tick;	2135	ev_periodic hourly_tick;
1564	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2136	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1565	ev_periodic_start (loop, &hourly_tick);	2137	ev_periodic_start (loop, &hourly_tick);
1566		2138
1567	Example: The same as above, but use a reschedule callback to do it:	2139	Example: The same as above, but use a reschedule callback to do it:
1568		2140
1569	#include <math.h>	2141	#include <math.h>
1570		2142
1571	static ev_tstamp	2143	static ev_tstamp
1572	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2144	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1573	{	2145	{
1574	return now + (3600. - fmod (now, 3600.));	2146	return now + (3600. - fmod (now, 3600.));
1575	}	2147	}
1576		2148
1577	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2149	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1578		2150
1579	Example: Call a callback every hour, starting now:	2151	Example: Call a callback every hour, starting now:
1580		2152
1581	struct ev_periodic hourly_tick;	2153	ev_periodic hourly_tick;
1582	ev_periodic_init (&hourly_tick, clock_cb,	2154	ev_periodic_init (&hourly_tick, clock_cb,
1583	fmod (ev_now (loop), 3600.), 3600., 0);	2155	fmod (ev_now (loop), 3600.), 3600., 0);
1584	ev_periodic_start (loop, &hourly_tick);	2156	ev_periodic_start (loop, &hourly_tick);
1585		2157
1586		2158
…		…
1589	Signal watchers will trigger an event when the process receives a specific	2161	Signal watchers will trigger an event when the process receives a specific
1590	signal one or more times. Even though signals are very asynchronous, libev	2162	signal one or more times. Even though signals are very asynchronous, libev
1591	will try it's best to deliver signals synchronously, i.e. as part of the	2163	will try it's best to deliver signals synchronously, i.e. as part of the
1592	normal event processing, like any other event.	2164	normal event processing, like any other event.
1593		2165
1594	If you want signals asynchronously, just use C<sigaction> as you would	2166	If you want signals to be delivered truly asynchronously, just use
1595	do without libev and forget about sharing the signal. You can even use	2167	C<sigaction> as you would do without libev and forget about sharing
1596	C<ev_async> from a signal handler to synchronously wake up an event loop.	2168	the signal. You can even use C<ev_async> from a signal handler to
		2169	synchronously wake up an event loop.
1597		2170
1598	You can configure as many watchers as you like per signal. Only when the	2171	You can configure as many watchers as you like for the same signal, but
		2172	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2173	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2174	C<SIGINT> in both the default loop and another loop at the same time. At
		2175	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2176
1599	first watcher gets started will libev actually register a signal handler	2177	When the first watcher gets started will libev actually register something
1600	with the kernel (thus it coexists with your own signal handlers as long as	2178	with the kernel (thus it coexists with your own signal handlers as long as
1601	you don't register any with libev for the same signal). Similarly, when	2179	you don't register any with libev for the same signal).
1602	the last signal watcher for a signal is stopped, libev will reset the
1603	signal handler to SIG_DFL (regardless of what it was set to before).
1604		2180
1605	If possible and supported, libev will install its handlers with	2181	If possible and supported, libev will install its handlers with
1606	C<SA_RESTART> behaviour enabled, so system calls should not be unduly	2182	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1607	interrupted. If you have a problem with system calls getting interrupted by	2183	not be unduly interrupted. If you have a problem with system calls getting
1608	signals you can block all signals in an C<ev_check> watcher and unblock	2184	interrupted by signals you can block all signals in an C<ev_check> watcher
1609	them in an C<ev_prepare> watcher.	2185	and unblock them in an C<ev_prepare> watcher.
		2186
		2187	=head3 The special problem of inheritance over fork/execve/pthread_create
		2188
		2189	Both the signal mask (C<sigprocmask>) and the signal disposition
		2190	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2191	stopping it again), that is, libev might or might not block the signal,
		2192	and might or might not set or restore the installed signal handler.
		2193
		2194	While this does not matter for the signal disposition (libev never
		2195	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2196	C<execve>), this matters for the signal mask: many programs do not expect
		2197	certain signals to be blocked.
		2198
		2199	This means that before calling C<exec> (from the child) you should reset
		2200	the signal mask to whatever "default" you expect (all clear is a good
		2201	choice usually).
		2202
		2203	The simplest way to ensure that the signal mask is reset in the child is
		2204	to install a fork handler with C<pthread_atfork> that resets it. That will
		2205	catch fork calls done by libraries (such as the libc) as well.
		2206
		2207	In current versions of libev, the signal will not be blocked indefinitely
		2208	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2209	the window of opportunity for problems, it will not go away, as libev
		2210	I<has> to modify the signal mask, at least temporarily.
		2211
		2212	So I can't stress this enough: I<If you do not reset your signal mask when
		2213	you expect it to be empty, you have a race condition in your code>. This
		2214	is not a libev-specific thing, this is true for most event libraries.
1610		2215
1611	=head3 Watcher-Specific Functions and Data Members	2216	=head3 Watcher-Specific Functions and Data Members
1612		2217
1613	=over 4	2218	=over 4
1614		2219
…		…
1625		2230
1626	=back	2231	=back
1627		2232
1628	=head3 Examples	2233	=head3 Examples
1629		2234
1630	Example: Try to exit cleanly on SIGINT and SIGTERM.	2235	Example: Try to exit cleanly on SIGINT.
1631		2236
1632	static void	2237	static void
1633	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2238	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1634	{	2239	{
1635	ev_unloop (loop, EVUNLOOP_ALL);	2240	ev_unloop (loop, EVUNLOOP_ALL);
1636	}	2241	}
1637		2242
1638	struct ev_signal signal_watcher;	2243	ev_signal signal_watcher;
1639	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2244	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1640	ev_signal_start (loop, &sigint_cb);	2245	ev_signal_start (loop, &signal_watcher);
1641		2246
1642		2247
1643	=head2 C<ev_child> - watch out for process status changes	2248	=head2 C<ev_child> - watch out for process status changes
1644		2249
1645	Child watchers trigger when your process receives a SIGCHLD in response to	2250	Child watchers trigger when your process receives a SIGCHLD in response to
1646	some child status changes (most typically when a child of yours dies or	2251	some child status changes (most typically when a child of yours dies or
1647	exits). It is permissible to install a child watcher I<after> the child	2252	exits). It is permissible to install a child watcher I<after> the child
1648	has been forked (which implies it might have already exited), as long	2253	has been forked (which implies it might have already exited), as long
1649	as the event loop isn't entered (or is continued from a watcher), i.e.,	2254	as the event loop isn't entered (or is continued from a watcher), i.e.,
1650	forking and then immediately registering a watcher for the child is fine,	2255	forking and then immediately registering a watcher for the child is fine,
1651	but forking and registering a watcher a few event loop iterations later is	2256	but forking and registering a watcher a few event loop iterations later or
1652	not.	2257	in the next callback invocation is not.
1653		2258
1654	Only the default event loop is capable of handling signals, and therefore	2259	Only the default event loop is capable of handling signals, and therefore
1655	you can only register child watchers in the default event loop.	2260	you can only register child watchers in the default event loop.
1656		2261
		2262	Due to some design glitches inside libev, child watchers will always be
		2263	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2264	libev)
		2265
1657	=head3 Process Interaction	2266	=head3 Process Interaction
1658		2267
1659	Libev grabs C<SIGCHLD> as soon as the default event loop is	2268	Libev grabs C<SIGCHLD> as soon as the default event loop is
1660	initialised. This is necessary to guarantee proper behaviour even if	2269	initialised. This is necessary to guarantee proper behaviour even if the
1661	the first child watcher is started after the child exits. The occurrence	2270	first child watcher is started after the child exits. The occurrence
1662	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2271	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1663	synchronously as part of the event loop processing. Libev always reaps all	2272	synchronously as part of the event loop processing. Libev always reaps all
1664	children, even ones not watched.	2273	children, even ones not watched.
1665		2274
1666	=head3 Overriding the Built-In Processing	2275	=head3 Overriding the Built-In Processing
…		…
1676	=head3 Stopping the Child Watcher	2285	=head3 Stopping the Child Watcher
1677		2286
1678	Currently, the child watcher never gets stopped, even when the	2287	Currently, the child watcher never gets stopped, even when the
1679	child terminates, so normally one needs to stop the watcher in the	2288	child terminates, so normally one needs to stop the watcher in the
1680	callback. Future versions of libev might stop the watcher automatically	2289	callback. Future versions of libev might stop the watcher automatically
1681	when a child exit is detected.	2290	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2291	problem).
1682		2292
1683	=head3 Watcher-Specific Functions and Data Members	2293	=head3 Watcher-Specific Functions and Data Members
1684		2294
1685	=over 4	2295	=over 4
1686		2296
…		…
1718	its completion.	2328	its completion.
1719		2329
1720	ev_child cw;	2330	ev_child cw;
1721		2331
1722	static void	2332	static void
1723	child_cb (EV_P_ struct ev_child *w, int revents)	2333	child_cb (EV_P_ ev_child *w, int revents)
1724	{	2334	{
1725	ev_child_stop (EV_A_ w);	2335	ev_child_stop (EV_A_ w);
1726	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2336	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1727	}	2337	}
1728		2338
…		…
1743		2353
1744		2354
1745	=head2 C<ev_stat> - did the file attributes just change?	2355	=head2 C<ev_stat> - did the file attributes just change?
1746		2356
1747	This watches a file system path for attribute changes. That is, it calls	2357	This watches a file system path for attribute changes. That is, it calls
1748	C<stat> regularly (or when the OS says it changed) and sees if it changed	2358	C<stat> on that path in regular intervals (or when the OS says it changed)
1749	compared to the last time, invoking the callback if it did.	2359	and sees if it changed compared to the last time, invoking the callback if
		2360	it did.
1750		2361
1751	The path does not need to exist: changing from "path exists" to "path does	2362	The path does not need to exist: changing from "path exists" to "path does
1752	not exist" is a status change like any other. The condition "path does	2363	not exist" is a status change like any other. The condition "path does not
1753	not exist" is signified by the C<st_nlink> field being zero (which is	2364	exist" (or more correctly "path cannot be stat'ed") is signified by the
1754	otherwise always forced to be at least one) and all the other fields of	2365	C<st_nlink> field being zero (which is otherwise always forced to be at
1755	the stat buffer having unspecified contents.	2366	least one) and all the other fields of the stat buffer having unspecified
		2367	contents.
1756		2368
1757	The path I<should> be absolute and I<must not> end in a slash. If it is	2369	The path I<must not> end in a slash or contain special components such as
		2370	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1758	relative and your working directory changes, the behaviour is undefined.	2371	your working directory changes, then the behaviour is undefined.
1759		2372
1760	Since there is no standard kernel interface to do this, the portable	2373	Since there is no portable change notification interface available, the
1761	implementation simply calls C<stat (2)> regularly on the path to see if	2374	portable implementation simply calls C<stat(2)> regularly on the path
1762	it changed somehow. You can specify a recommended polling interval for	2375	to see if it changed somehow. You can specify a recommended polling
1763	this case. If you specify a polling interval of C<0> (highly recommended!)	2376	interval for this case. If you specify a polling interval of C<0> (highly
1764	then a I<suitable, unspecified default> value will be used (which	2377	recommended!) then a I<suitable, unspecified default> value will be used
1765	you can expect to be around five seconds, although this might change	2378	(which you can expect to be around five seconds, although this might
1766	dynamically). Libev will also impose a minimum interval which is currently	2379	change dynamically). Libev will also impose a minimum interval which is
1767	around C<0.1>, but thats usually overkill.	2380	currently around C<0.1>, but that's usually overkill.
1768		2381
1769	This watcher type is not meant for massive numbers of stat watchers,	2382	This watcher type is not meant for massive numbers of stat watchers,
1770	as even with OS-supported change notifications, this can be	2383	as even with OS-supported change notifications, this can be
1771	resource-intensive.	2384	resource-intensive.
1772		2385
1773	At the time of this writing, the only OS-specific interface implemented	2386	At the time of this writing, the only OS-specific interface implemented
1774	is the Linux inotify interface (implementing kqueue support is left as	2387	is the Linux inotify interface (implementing kqueue support is left as an
1775	an exercise for the reader. Note, however, that the author sees no way	2388	exercise for the reader. Note, however, that the author sees no way of
1776	of implementing C<ev_stat> semantics with kqueue).	2389	implementing C<ev_stat> semantics with kqueue, except as a hint).
1777		2390
1778	=head3 ABI Issues (Largefile Support)	2391	=head3 ABI Issues (Largefile Support)
1779		2392
1780	Libev by default (unless the user overrides this) uses the default	2393	Libev by default (unless the user overrides this) uses the default
1781	compilation environment, which means that on systems with large file	2394	compilation environment, which means that on systems with large file
1782	support disabled by default, you get the 32 bit version of the stat	2395	support disabled by default, you get the 32 bit version of the stat
1783	structure. When using the library from programs that change the ABI to	2396	structure. When using the library from programs that change the ABI to
1784	use 64 bit file offsets the programs will fail. In that case you have to	2397	use 64 bit file offsets the programs will fail. In that case you have to
1785	compile libev with the same flags to get binary compatibility. This is	2398	compile libev with the same flags to get binary compatibility. This is
1786	obviously the case with any flags that change the ABI, but the problem is	2399	obviously the case with any flags that change the ABI, but the problem is
1787	most noticeably disabled with ev_stat and large file support.	2400	most noticeably displayed with ev_stat and large file support.
1788		2401
1789	The solution for this is to lobby your distribution maker to make large	2402	The solution for this is to lobby your distribution maker to make large
1790	file interfaces available by default (as e.g. FreeBSD does) and not	2403	file interfaces available by default (as e.g. FreeBSD does) and not
1791	optional. Libev cannot simply switch on large file support because it has	2404	optional. Libev cannot simply switch on large file support because it has
1792	to exchange stat structures with application programs compiled using the	2405	to exchange stat structures with application programs compiled using the
1793	default compilation environment.	2406	default compilation environment.
1794		2407
1795	=head3 Inotify and Kqueue	2408	=head3 Inotify and Kqueue
1796		2409
1797	When C<inotify (7)> support has been compiled into libev (generally only	2410	When C<inotify (7)> support has been compiled into libev and present at
1798	available with Linux) and present at runtime, it will be used to speed up	2411	runtime, it will be used to speed up change detection where possible. The
1799	change detection where possible. The inotify descriptor will be created lazily	2412	inotify descriptor will be created lazily when the first C<ev_stat>
1800	when the first C<ev_stat> watcher is being started.	2413	watcher is being started.
1801		2414
1802	Inotify presence does not change the semantics of C<ev_stat> watchers	2415	Inotify presence does not change the semantics of C<ev_stat> watchers
1803	except that changes might be detected earlier, and in some cases, to avoid	2416	except that changes might be detected earlier, and in some cases, to avoid
1804	making regular C<stat> calls. Even in the presence of inotify support	2417	making regular C<stat> calls. Even in the presence of inotify support
1805	there are many cases where libev has to resort to regular C<stat> polling,	2418	there are many cases where libev has to resort to regular C<stat> polling,
1806	but as long as the path exists, libev usually gets away without polling.	2419	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2420	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2421	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2422	xfs are fully working) libev usually gets away without polling.
1807		2423
1808	There is no support for kqueue, as apparently it cannot be used to	2424	There is no support for kqueue, as apparently it cannot be used to
1809	implement this functionality, due to the requirement of having a file	2425	implement this functionality, due to the requirement of having a file
1810	descriptor open on the object at all times, and detecting renames, unlinks	2426	descriptor open on the object at all times, and detecting renames, unlinks
1811	etc. is difficult.	2427	etc. is difficult.
1812		2428
		2429	=head3 C<stat ()> is a synchronous operation
		2430
		2431	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2432	the process. The exception are C<ev_stat> watchers - those call C<stat
		2433	()>, which is a synchronous operation.
		2434
		2435	For local paths, this usually doesn't matter: unless the system is very
		2436	busy or the intervals between stat's are large, a stat call will be fast,
		2437	as the path data is usually in memory already (except when starting the
		2438	watcher).
		2439
		2440	For networked file systems, calling C<stat ()> can block an indefinite
		2441	time due to network issues, and even under good conditions, a stat call
		2442	often takes multiple milliseconds.
		2443
		2444	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2445	paths, although this is fully supported by libev.
		2446
1813	=head3 The special problem of stat time resolution	2447	=head3 The special problem of stat time resolution
1814		2448
1815	The C<stat ()> system call only supports full-second resolution portably, and	2449	The C<stat ()> system call only supports full-second resolution portably,
1816	even on systems where the resolution is higher, most file systems still	2450	and even on systems where the resolution is higher, most file systems
1817	only support whole seconds.	2451	still only support whole seconds.
1818		2452
1819	That means that, if the time is the only thing that changes, you can	2453	That means that, if the time is the only thing that changes, you can
1820	easily miss updates: on the first update, C<ev_stat> detects a change and	2454	easily miss updates: on the first update, C<ev_stat> detects a change and
1821	calls your callback, which does something. When there is another update	2455	calls your callback, which does something. When there is another update
1822	within the same second, C<ev_stat> will be unable to detect unless the	2456	within the same second, C<ev_stat> will be unable to detect unless the
…		…
1965		2599
1966	=head3 Watcher-Specific Functions and Data Members	2600	=head3 Watcher-Specific Functions and Data Members
1967		2601
1968	=over 4	2602	=over 4
1969		2603
1970	=item ev_idle_init (ev_signal *, callback)	2604	=item ev_idle_init (ev_idle *, callback)
1971		2605
1972	Initialises and configures the idle watcher - it has no parameters of any	2606	Initialises and configures the idle watcher - it has no parameters of any
1973	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2607	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1974	believe me.	2608	believe me.
1975		2609
…		…
1979		2613
1980	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2614	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1981	callback, free it. Also, use no error checking, as usual.	2615	callback, free it. Also, use no error checking, as usual.
1982		2616
1983	static void	2617	static void
1984	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2618	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1985	{	2619	{
1986	free (w);	2620	free (w);
1987	// now do something you wanted to do when the program has	2621	// now do something you wanted to do when the program has
1988	// no longer anything immediate to do.	2622	// no longer anything immediate to do.
1989	}	2623	}
1990		2624
1991	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2625	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1992	ev_idle_init (idle_watcher, idle_cb);	2626	ev_idle_init (idle_watcher, idle_cb);
1993	ev_idle_start (loop, idle_cb);	2627	ev_idle_start (loop, idle_watcher);
1994		2628
1995		2629
1996	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2630	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1997		2631
1998	Prepare and check watchers are usually (but not always) used in pairs:	2632	Prepare and check watchers are usually (but not always) used in pairs:
…		…
2077		2711
2078	static ev_io iow [nfd];	2712	static ev_io iow [nfd];
2079	static ev_timer tw;	2713	static ev_timer tw;
2080		2714
2081	static void	2715	static void
2082	io_cb (ev_loop *loop, ev_io *w, int revents)	2716	io_cb (struct ev_loop *loop, ev_io *w, int revents)
2083	{	2717	{
2084	}	2718	}
2085		2719
2086	// create io watchers for each fd and a timer before blocking	2720	// create io watchers for each fd and a timer before blocking
2087	static void	2721	static void
2088	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2722	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
2089	{	2723	{
2090	int timeout = 3600000;	2724	int timeout = 3600000;
2091	struct pollfd fds [nfd];	2725	struct pollfd fds [nfd];
2092	// actual code will need to loop here and realloc etc.	2726	// actual code will need to loop here and realloc etc.
2093	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2727	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
2094		2728
2095	/* the callback is illegal, but won't be called as we stop during check */	2729	/* the callback is illegal, but won't be called as we stop during check */
2096	ev_timer_init (&tw, 0, timeout * 1e-3);	2730	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
2097	ev_timer_start (loop, &tw);	2731	ev_timer_start (loop, &tw);
2098		2732
2099	// create one ev_io per pollfd	2733	// create one ev_io per pollfd
2100	for (int i = 0; i < nfd; ++i)	2734	for (int i = 0; i < nfd; ++i)
2101	{	2735	{
…		…
2108	}	2742	}
2109	}	2743	}
2110		2744
2111	// stop all watchers after blocking	2745	// stop all watchers after blocking
2112	static void	2746	static void
2113	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2747	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
2114	{	2748	{
2115	ev_timer_stop (loop, &tw);	2749	ev_timer_stop (loop, &tw);
2116		2750
2117	for (int i = 0; i < nfd; ++i)	2751	for (int i = 0; i < nfd; ++i)
2118	{	2752	{
…		…
2214	some fds have to be watched and handled very quickly (with low latency),	2848	some fds have to be watched and handled very quickly (with low latency),
2215	and even priorities and idle watchers might have too much overhead. In	2849	and even priorities and idle watchers might have too much overhead. In
2216	this case you would put all the high priority stuff in one loop and all	2850	this case you would put all the high priority stuff in one loop and all
2217	the rest in a second one, and embed the second one in the first.	2851	the rest in a second one, and embed the second one in the first.
2218		2852
2219	As long as the watcher is active, the callback will be invoked every time	2853	As long as the watcher is active, the callback will be invoked every
2220	there might be events pending in the embedded loop. The callback must then	2854	time there might be events pending in the embedded loop. The callback
2221	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2855	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2222	their callbacks (you could also start an idle watcher to give the embedded	2856	sweep and invoke their callbacks (the callback doesn't need to invoke the
2223	loop strictly lower priority for example). You can also set the callback	2857	C<ev_embed_sweep> function directly, it could also start an idle watcher
2224	to C<0>, in which case the embed watcher will automatically execute the	2858	to give the embedded loop strictly lower priority for example).
2225	embedded loop sweep.
2226		2859
2227	As long as the watcher is started it will automatically handle events. The	2860	You can also set the callback to C<0>, in which case the embed watcher
2228	callback will be invoked whenever some events have been handled. You can	2861	will automatically execute the embedded loop sweep whenever necessary.
2229	set the callback to C<0> to avoid having to specify one if you are not
2230	interested in that.
2231		2862
2232	Also, there have not currently been made special provisions for forking:	2863	Fork detection will be handled transparently while the C<ev_embed> watcher
2233	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2864	is active, i.e., the embedded loop will automatically be forked when the
2234	but you will also have to stop and restart any C<ev_embed> watchers	2865	embedding loop forks. In other cases, the user is responsible for calling
2235	yourself - but you can use a fork watcher to handle this automatically,	2866	C<ev_loop_fork> on the embedded loop.
2236	and future versions of libev might do just that.
2237		2867
2238	Unfortunately, not all backends are embeddable: only the ones returned by	2868	Unfortunately, not all backends are embeddable: only the ones returned by
2239	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2869	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2240	portable one.	2870	portable one.
2241		2871
2242	So when you want to use this feature you will always have to be prepared	2872	So when you want to use this feature you will always have to be prepared
2243	that you cannot get an embeddable loop. The recommended way to get around	2873	that you cannot get an embeddable loop. The recommended way to get around
2244	this is to have a separate variables for your embeddable loop, try to	2874	this is to have a separate variables for your embeddable loop, try to
2245	create it, and if that fails, use the normal loop for everything.	2875	create it, and if that fails, use the normal loop for everything.
		2876
		2877	=head3 C<ev_embed> and fork
		2878
		2879	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2880	automatically be applied to the embedded loop as well, so no special
		2881	fork handling is required in that case. When the watcher is not running,
		2882	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2883	as applicable.
2246		2884
2247	=head3 Watcher-Specific Functions and Data Members	2885	=head3 Watcher-Specific Functions and Data Members
2248		2886
2249	=over 4	2887	=over 4
2250		2888
…		…
2278	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2916	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2279	used).	2917	used).
2280		2918
2281	struct ev_loop *loop_hi = ev_default_init (0);	2919	struct ev_loop *loop_hi = ev_default_init (0);
2282	struct ev_loop *loop_lo = 0;	2920	struct ev_loop *loop_lo = 0;
2283	struct ev_embed embed;	2921	ev_embed embed;
2284		2922
2285	// see if there is a chance of getting one that works	2923	// see if there is a chance of getting one that works
2286	// (remember that a flags value of 0 means autodetection)	2924	// (remember that a flags value of 0 means autodetection)
2287	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2925	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2288	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2926	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2302	kqueue implementation). Store the kqueue/socket-only event loop in	2940	kqueue implementation). Store the kqueue/socket-only event loop in
2303	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2941	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2304		2942
2305	struct ev_loop *loop = ev_default_init (0);	2943	struct ev_loop *loop = ev_default_init (0);
2306	struct ev_loop *loop_socket = 0;	2944	struct ev_loop *loop_socket = 0;
2307	struct ev_embed embed;	2945	ev_embed embed;
2308		2946
2309	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2947	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2310	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2948	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2311	{	2949	{
2312	ev_embed_init (&embed, 0, loop_socket);	2950	ev_embed_init (&embed, 0, loop_socket);
…		…
2327	event loop blocks next and before C<ev_check> watchers are being called,	2965	event loop blocks next and before C<ev_check> watchers are being called,
2328	and only in the child after the fork. If whoever good citizen calling	2966	and only in the child after the fork. If whoever good citizen calling
2329	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2967	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2330	handlers will be invoked, too, of course.	2968	handlers will be invoked, too, of course.
2331		2969
		2970	=head3 The special problem of life after fork - how is it possible?
		2971
		2972	Most uses of C<fork()> consist of forking, then some simple calls to set
		2973	up/change the process environment, followed by a call to C<exec()>. This
		2974	sequence should be handled by libev without any problems.
		2975
		2976	This changes when the application actually wants to do event handling
		2977	in the child, or both parent in child, in effect "continuing" after the
		2978	fork.
		2979
		2980	The default mode of operation (for libev, with application help to detect
		2981	forks) is to duplicate all the state in the child, as would be expected
		2982	when I<either> the parent I<or> the child process continues.
		2983
		2984	When both processes want to continue using libev, then this is usually the
		2985	wrong result. In that case, usually one process (typically the parent) is
		2986	supposed to continue with all watchers in place as before, while the other
		2987	process typically wants to start fresh, i.e. without any active watchers.
		2988
		2989	The cleanest and most efficient way to achieve that with libev is to
		2990	simply create a new event loop, which of course will be "empty", and
		2991	use that for new watchers. This has the advantage of not touching more
		2992	memory than necessary, and thus avoiding the copy-on-write, and the
		2993	disadvantage of having to use multiple event loops (which do not support
		2994	signal watchers).
		2995
		2996	When this is not possible, or you want to use the default loop for
		2997	other reasons, then in the process that wants to start "fresh", call
		2998	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2999	the default loop will "orphan" (not stop) all registered watchers, so you
		3000	have to be careful not to execute code that modifies those watchers. Note
		3001	also that in that case, you have to re-register any signal watchers.
		3002
2332	=head3 Watcher-Specific Functions and Data Members	3003	=head3 Watcher-Specific Functions and Data Members
2333		3004
2334	=over 4	3005	=over 4
2335		3006
2336	=item ev_fork_init (ev_signal *, callback)	3007	=item ev_fork_init (ev_signal *, callback)
…		…
2340	believe me.	3011	believe me.
2341		3012
2342	=back	3013	=back
2343		3014
2344		3015
2345	=head2 C<ev_async> - how to wake up another event loop	3016	=head2 C<ev_async> - how to wake up an event loop
2346		3017
2347	In general, you cannot use an C<ev_loop> from multiple threads or other	3018	In general, you cannot use an C<ev_loop> from multiple threads or other
2348	asynchronous sources such as signal handlers (as opposed to multiple event	3019	asynchronous sources such as signal handlers (as opposed to multiple event
2349	loops - those are of course safe to use in different threads).	3020	loops - those are of course safe to use in different threads).
2350		3021
2351	Sometimes, however, you need to wake up another event loop you do not	3022	Sometimes, however, you need to wake up an event loop you do not control,
2352	control, for example because it belongs to another thread. This is what	3023	for example because it belongs to another thread. This is what C<ev_async>
2353	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3024	watchers do: as long as the C<ev_async> watcher is active, you can signal
2354	can signal it by calling C<ev_async_send>, which is thread- and signal	3025	it by calling C<ev_async_send>, which is thread- and signal safe.
2355	safe.
2356		3026
2357	This functionality is very similar to C<ev_signal> watchers, as signals,	3027	This functionality is very similar to C<ev_signal> watchers, as signals,
2358	too, are asynchronous in nature, and signals, too, will be compressed	3028	too, are asynchronous in nature, and signals, too, will be compressed
2359	(i.e. the number of callback invocations may be less than the number of	3029	(i.e. the number of callback invocations may be less than the number of
2360	C<ev_async_sent> calls).	3030	C<ev_async_sent> calls).
…		…
2365	=head3 Queueing	3035	=head3 Queueing
2366		3036
2367	C<ev_async> does not support queueing of data in any way. The reason	3037	C<ev_async> does not support queueing of data in any way. The reason
2368	is that the author does not know of a simple (or any) algorithm for a	3038	is that the author does not know of a simple (or any) algorithm for a
2369	multiple-writer-single-reader queue that works in all cases and doesn't	3039	multiple-writer-single-reader queue that works in all cases and doesn't
2370	need elaborate support such as pthreads.	3040	need elaborate support such as pthreads or unportable memory access
		3041	semantics.
2371		3042
2372	That means that if you want to queue data, you have to provide your own	3043	That means that if you want to queue data, you have to provide your own
2373	queue. But at least I can tell you how to implement locking around your	3044	queue. But at least I can tell you how to implement locking around your
2374	queue:	3045	queue:
2375		3046
2376	=over 4	3047	=over 4
2377		3048
2378	=item queueing from a signal handler context	3049	=item queueing from a signal handler context
2379		3050
2380	To implement race-free queueing, you simply add to the queue in the signal	3051	To implement race-free queueing, you simply add to the queue in the signal
2381	handler but you block the signal handler in the watcher callback. Here is an example that does that for	3052	handler but you block the signal handler in the watcher callback. Here is
2382	some fictitious SIGUSR1 handler:	3053	an example that does that for some fictitious SIGUSR1 handler:
2383		3054
2384	static ev_async mysig;	3055	static ev_async mysig;
2385		3056
2386	static void	3057	static void
2387	sigusr1_handler (void)	3058	sigusr1_handler (void)
…		…
2453	=over 4	3124	=over 4
2454		3125
2455	=item ev_async_init (ev_async *, callback)	3126	=item ev_async_init (ev_async *, callback)
2456		3127
2457	Initialises and configures the async watcher - it has no parameters of any	3128	Initialises and configures the async watcher - it has no parameters of any
2458	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3129	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2459	trust me.	3130	trust me.
2460		3131
2461	=item ev_async_send (loop, ev_async *)	3132	=item ev_async_send (loop, ev_async *)
2462		3133
2463	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3134	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2464	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3135	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2465	C<ev_feed_event>, this call is safe to do from other threads, signal or	3136	C<ev_feed_event>, this call is safe to do from other threads, signal or
2466	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3137	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2467	section below on what exactly this means).	3138	section below on what exactly this means).
2468		3139
		3140	Note that, as with other watchers in libev, multiple events might get
		3141	compressed into a single callback invocation (another way to look at this
		3142	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3143	reset when the event loop detects that).
		3144
2469	This call incurs the overhead of a system call only once per loop iteration,	3145	This call incurs the overhead of a system call only once per event loop
2470	so while the overhead might be noticeable, it doesn't apply to repeated	3146	iteration, so while the overhead might be noticeable, it doesn't apply to
2471	calls to C<ev_async_send>.	3147	repeated calls to C<ev_async_send> for the same event loop.
2472		3148
2473	=item bool = ev_async_pending (ev_async *)	3149	=item bool = ev_async_pending (ev_async *)
2474		3150
2475	Returns a non-zero value when C<ev_async_send> has been called on the	3151	Returns a non-zero value when C<ev_async_send> has been called on the
2476	watcher but the event has not yet been processed (or even noted) by the	3152	watcher but the event has not yet been processed (or even noted) by the
…		…
2479	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3155	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2480	the loop iterates next and checks for the watcher to have become active,	3156	the loop iterates next and checks for the watcher to have become active,
2481	it will reset the flag again. C<ev_async_pending> can be used to very	3157	it will reset the flag again. C<ev_async_pending> can be used to very
2482	quickly check whether invoking the loop might be a good idea.	3158	quickly check whether invoking the loop might be a good idea.
2483		3159
2484	Not that this does I<not> check whether the watcher itself is pending, only	3160	Not that this does I<not> check whether the watcher itself is pending,
2485	whether it has been requested to make this watcher pending.	3161	only whether it has been requested to make this watcher pending: there
		3162	is a time window between the event loop checking and resetting the async
		3163	notification, and the callback being invoked.
2486		3164
2487	=back	3165	=back
2488		3166
2489		3167
2490	=head1 OTHER FUNCTIONS	3168	=head1 OTHER FUNCTIONS
…		…
2494	=over 4	3172	=over 4
2495		3173
2496	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3174	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2497		3175
2498	This function combines a simple timer and an I/O watcher, calls your	3176	This function combines a simple timer and an I/O watcher, calls your
2499	callback on whichever event happens first and automatically stop both	3177	callback on whichever event happens first and automatically stops both
2500	watchers. This is useful if you want to wait for a single event on an fd	3178	watchers. This is useful if you want to wait for a single event on an fd
2501	or timeout without having to allocate/configure/start/stop/free one or	3179	or timeout without having to allocate/configure/start/stop/free one or
2502	more watchers yourself.	3180	more watchers yourself.
2503		3181
2504	If C<fd> is less than 0, then no I/O watcher will be started and events	3182	If C<fd> is less than 0, then no I/O watcher will be started and the
2505	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3183	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2506	C<events> set will be created and started.	3184	the given C<fd> and C<events> set will be created and started.
2507		3185
2508	If C<timeout> is less than 0, then no timeout watcher will be	3186	If C<timeout> is less than 0, then no timeout watcher will be
2509	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3187	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2510	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3188	repeat = 0) will be started. C<0> is a valid timeout.
2511	dubious value.
2512		3189
2513	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3190	The callback has the type C<void (*cb)(int revents, void *arg)> and is
2514	passed an C<revents> set like normal event callbacks (a combination of	3191	passed an C<revents> set like normal event callbacks (a combination of
2515	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3192	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
2516	value passed to C<ev_once>:	3193	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3194	a timeout and an io event at the same time - you probably should give io
		3195	events precedence.
		3196
		3197	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2517		3198
2518	static void stdin_ready (int revents, void *arg)	3199	static void stdin_ready (int revents, void *arg)
2519	{	3200	{
		3201	if (revents & EV_READ)
		3202	/* stdin might have data for us, joy! */;
2520	if (revents & EV_TIMEOUT)	3203	else if (revents & EV_TIMER)
2521	/* doh, nothing entered */;	3204	/* doh, nothing entered */;
2522	else if (revents & EV_READ)
2523	/* stdin might have data for us, joy! */;
2524	}	3205	}
2525		3206
2526	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3207	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2527		3208
2528	=item ev_feed_event (ev_loop *, watcher *, int revents)
2529
2530	Feeds the given event set into the event loop, as if the specified event
2531	had happened for the specified watcher (which must be a pointer to an
2532	initialised but not necessarily started event watcher).
2533
2534	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3209	=item ev_feed_fd_event (loop, int fd, int revents)
2535		3210
2536	Feed an event on the given fd, as if a file descriptor backend detected	3211	Feed an event on the given fd, as if a file descriptor backend detected
2537	the given events it.	3212	the given events it.
2538		3213
2539	=item ev_feed_signal_event (ev_loop *loop, int signum)	3214	=item ev_feed_signal_event (loop, int signum)
2540		3215
2541	Feed an event as if the given signal occurred (C<loop> must be the default	3216	Feed an event as if the given signal occurred (C<loop> must be the default
2542	loop!).	3217	loop!).
2543		3218
2544	=back	3219	=back
…		…
2624		3299
2625	=over 4	3300	=over 4
2626		3301
2627	=item ev::TYPE::TYPE ()	3302	=item ev::TYPE::TYPE ()
2628		3303
2629	=item ev::TYPE::TYPE (struct ev_loop *)	3304	=item ev::TYPE::TYPE (loop)
2630		3305
2631	=item ev::TYPE::~TYPE	3306	=item ev::TYPE::~TYPE
2632		3307
2633	The constructor (optionally) takes an event loop to associate the watcher	3308	The constructor (optionally) takes an event loop to associate the watcher
2634	with. If it is omitted, it will use C<EV_DEFAULT>.	3309	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2666		3341
2667	myclass obj;	3342	myclass obj;
2668	ev::io iow;	3343	ev::io iow;
2669	iow.set <myclass, &myclass::io_cb> (&obj);	3344	iow.set <myclass, &myclass::io_cb> (&obj);
2670		3345
		3346	=item w->set (object *)
		3347
		3348	This is a variation of a method callback - leaving out the method to call
		3349	will default the method to C<operator ()>, which makes it possible to use
		3350	functor objects without having to manually specify the C<operator ()> all
		3351	the time. Incidentally, you can then also leave out the template argument
		3352	list.
		3353
		3354	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3355	int revents)>.
		3356
		3357	See the method-C<set> above for more details.
		3358
		3359	Example: use a functor object as callback.
		3360
		3361	struct myfunctor
		3362	{
		3363	void operator() (ev::io &w, int revents)
		3364	{
		3365	...
		3366	}
		3367	}
		3368
		3369	myfunctor f;
		3370
		3371	ev::io w;
		3372	w.set (&f);
		3373
2671	=item w->set<function> (void *data = 0)	3374	=item w->set<function> (void *data = 0)
2672		3375
2673	Also sets a callback, but uses a static method or plain function as	3376	Also sets a callback, but uses a static method or plain function as
2674	callback. The optional C<data> argument will be stored in the watcher's	3377	callback. The optional C<data> argument will be stored in the watcher's
2675	C<data> member and is free for you to use.	3378	C<data> member and is free for you to use.
…		…
2681	Example: Use a plain function as callback.	3384	Example: Use a plain function as callback.
2682		3385
2683	static void io_cb (ev::io &w, int revents) { }	3386	static void io_cb (ev::io &w, int revents) { }
2684	iow.set <io_cb> ();	3387	iow.set <io_cb> ();
2685		3388
2686	=item w->set (struct ev_loop *)	3389	=item w->set (loop)
2687		3390
2688	Associates a different C<struct ev_loop> with this watcher. You can only	3391	Associates a different C<struct ev_loop> with this watcher. You can only
2689	do this when the watcher is inactive (and not pending either).	3392	do this when the watcher is inactive (and not pending either).
2690		3393
2691	=item w->set ([arguments])	3394	=item w->set ([arguments])
…		…
2761	L<http://software.schmorp.de/pkg/EV>.	3464	L<http://software.schmorp.de/pkg/EV>.
2762		3465
2763	=item Python	3466	=item Python
2764		3467
2765	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	3468	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2766	seems to be quite complete and well-documented. Note, however, that the	3469	seems to be quite complete and well-documented.
2767	patch they require for libev is outright dangerous as it breaks the ABI
2768	for everybody else, and therefore, should never be applied in an installed
2769	libev (if python requires an incompatible ABI then it needs to embed
2770	libev).
2771		3470
2772	=item Ruby	3471	=item Ruby
2773		3472
2774	Tony Arcieri has written a ruby extension that offers access to a subset	3473	Tony Arcieri has written a ruby extension that offers access to a subset
2775	of the libev API and adds file handle abstractions, asynchronous DNS and	3474	of the libev API and adds file handle abstractions, asynchronous DNS and
2776	more on top of it. It can be found via gem servers. Its homepage is at	3475	more on top of it. It can be found via gem servers. Its homepage is at
2777	L<http://rev.rubyforge.org/>.	3476	L<http://rev.rubyforge.org/>.
2778		3477
		3478	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3479	makes rev work even on mingw.
		3480
		3481	=item Haskell
		3482
		3483	A haskell binding to libev is available at
		3484	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3485
2779	=item D	3486	=item D
2780		3487
2781	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3488	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2782	be found at L<http://proj.llucax.com.ar/wiki/evd>.	3489	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3490
		3491	=item Ocaml
		3492
		3493	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3494	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3495
		3496	=item Lua
		3497
		3498	Brian Maher has written a partial interface to libev for lua (at the
		3499	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		3500	L<http://github.com/brimworks/lua-ev>.
2783		3501
2784	=back	3502	=back
2785		3503
2786		3504
2787	=head1 MACRO MAGIC	3505	=head1 MACRO MAGIC
…		…
2888		3606
2889	#define EV_STANDALONE 1	3607	#define EV_STANDALONE 1
2890	#include "ev.h"	3608	#include "ev.h"
2891		3609
2892	Both header files and implementation files can be compiled with a C++	3610	Both header files and implementation files can be compiled with a C++
2893	compiler (at least, thats a stated goal, and breakage will be treated	3611	compiler (at least, that's a stated goal, and breakage will be treated
2894	as a bug).	3612	as a bug).
2895		3613
2896	You need the following files in your source tree, or in a directory	3614	You need the following files in your source tree, or in a directory
2897	in your include path (e.g. in libev/ when using -Ilibev):	3615	in your include path (e.g. in libev/ when using -Ilibev):
2898		3616
…		…
2941	libev.m4	3659	libev.m4
2942		3660
2943	=head2 PREPROCESSOR SYMBOLS/MACROS	3661	=head2 PREPROCESSOR SYMBOLS/MACROS
2944		3662
2945	Libev can be configured via a variety of preprocessor symbols you have to	3663	Libev can be configured via a variety of preprocessor symbols you have to
2946	define before including any of its files. The default in the absence of	3664	define before including (or compiling) any of its files. The default in
2947	autoconf is documented for every option.	3665	the absence of autoconf is documented for every option.
		3666
		3667	Symbols marked with "(h)" do not change the ABI, and can have different
		3668	values when compiling libev vs. including F<ev.h>, so it is permissible
		3669	to redefine them before including F<ev.h> without breaking compatibility
		3670	to a compiled library. All other symbols change the ABI, which means all
		3671	users of libev and the libev code itself must be compiled with compatible
		3672	settings.
2948		3673
2949	=over 4	3674	=over 4
2950		3675
2951	=item EV_STANDALONE	3676	=item EV_STANDALONE (h)
2952		3677
2953	Must always be C<1> if you do not use autoconf configuration, which	3678	Must always be C<1> if you do not use autoconf configuration, which
2954	keeps libev from including F<config.h>, and it also defines dummy	3679	keeps libev from including F<config.h>, and it also defines dummy
2955	implementations for some libevent functions (such as logging, which is not	3680	implementations for some libevent functions (such as logging, which is not
2956	supported). It will also not define any of the structs usually found in	3681	supported). It will also not define any of the structs usually found in
2957	F<event.h> that are not directly supported by the libev core alone.	3682	F<event.h> that are not directly supported by the libev core alone.
2958		3683
		3684	In standalone mode, libev will still try to automatically deduce the
		3685	configuration, but has to be more conservative.
		3686
2959	=item EV_USE_MONOTONIC	3687	=item EV_USE_MONOTONIC
2960		3688
2961	If defined to be C<1>, libev will try to detect the availability of the	3689	If defined to be C<1>, libev will try to detect the availability of the
2962	monotonic clock option at both compile time and runtime. Otherwise no use	3690	monotonic clock option at both compile time and runtime. Otherwise no
2963	of the monotonic clock option will be attempted. If you enable this, you	3691	use of the monotonic clock option will be attempted. If you enable this,
2964	usually have to link against librt or something similar. Enabling it when	3692	you usually have to link against librt or something similar. Enabling it
2965	the functionality isn't available is safe, though, although you have	3693	when the functionality isn't available is safe, though, although you have
2966	to make sure you link against any libraries where the C<clock_gettime>	3694	to make sure you link against any libraries where the C<clock_gettime>
2967	function is hiding in (often F<-lrt>).	3695	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2968		3696
2969	=item EV_USE_REALTIME	3697	=item EV_USE_REALTIME
2970		3698
2971	If defined to be C<1>, libev will try to detect the availability of the	3699	If defined to be C<1>, libev will try to detect the availability of the
2972	real-time clock option at compile time (and assume its availability at	3700	real-time clock option at compile time (and assume its availability
2973	runtime if successful). Otherwise no use of the real-time clock option will	3701	at runtime if successful). Otherwise no use of the real-time clock
2974	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3702	option will be attempted. This effectively replaces C<gettimeofday>
2975	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3703	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2976	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3704	correctness. See the note about libraries in the description of
		3705	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3706	C<EV_USE_CLOCK_SYSCALL>.
		3707
		3708	=item EV_USE_CLOCK_SYSCALL
		3709
		3710	If defined to be C<1>, libev will try to use a direct syscall instead
		3711	of calling the system-provided C<clock_gettime> function. This option
		3712	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3713	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3714	programs needlessly. Using a direct syscall is slightly slower (in
		3715	theory), because no optimised vdso implementation can be used, but avoids
		3716	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3717	higher, as it simplifies linking (no need for C<-lrt>).
2977		3718
2978	=item EV_USE_NANOSLEEP	3719	=item EV_USE_NANOSLEEP
2979		3720
2980	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3721	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2981	and will use it for delays. Otherwise it will use C<select ()>.	3722	and will use it for delays. Otherwise it will use C<select ()>.
…		…
2997		3738
2998	=item EV_SELECT_USE_FD_SET	3739	=item EV_SELECT_USE_FD_SET
2999		3740
3000	If defined to C<1>, then the select backend will use the system C<fd_set>	3741	If defined to C<1>, then the select backend will use the system C<fd_set>
3001	structure. This is useful if libev doesn't compile due to a missing	3742	structure. This is useful if libev doesn't compile due to a missing
3002	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3743	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
3003	exotic systems. This usually limits the range of file descriptors to some	3744	on exotic systems. This usually limits the range of file descriptors to
3004	low limit such as 1024 or might have other limitations (winsocket only	3745	some low limit such as 1024 or might have other limitations (winsocket
3005	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3746	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
3006	influence the size of the C<fd_set> used.	3747	configures the maximum size of the C<fd_set>.
3007		3748
3008	=item EV_SELECT_IS_WINSOCKET	3749	=item EV_SELECT_IS_WINSOCKET
3009		3750
3010	When defined to C<1>, the select backend will assume that	3751	When defined to C<1>, the select backend will assume that
3011	select/socket/connect etc. don't understand file descriptors but	3752	select/socket/connect etc. don't understand file descriptors but
…		…
3013	be used is the winsock select). This means that it will call	3754	be used is the winsock select). This means that it will call
3014	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3755	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
3015	it is assumed that all these functions actually work on fds, even	3756	it is assumed that all these functions actually work on fds, even
3016	on win32. Should not be defined on non-win32 platforms.	3757	on win32. Should not be defined on non-win32 platforms.
3017		3758
3018	=item EV_FD_TO_WIN32_HANDLE	3759	=item EV_FD_TO_WIN32_HANDLE(fd)
3019		3760
3020	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	3761	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
3021	file descriptors to socket handles. When not defining this symbol (the	3762	file descriptors to socket handles. When not defining this symbol (the
3022	default), then libev will call C<_get_osfhandle>, which is usually	3763	default), then libev will call C<_get_osfhandle>, which is usually
3023	correct. In some cases, programs use their own file descriptor management,	3764	correct. In some cases, programs use their own file descriptor management,
3024	in which case they can provide this function to map fds to socket handles.	3765	in which case they can provide this function to map fds to socket handles.
		3766
		3767	=item EV_WIN32_HANDLE_TO_FD(handle)
		3768
		3769	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3770	using the standard C<_open_osfhandle> function. For programs implementing
		3771	their own fd to handle mapping, overwriting this function makes it easier
		3772	to do so. This can be done by defining this macro to an appropriate value.
		3773
		3774	=item EV_WIN32_CLOSE_FD(fd)
		3775
		3776	If programs implement their own fd to handle mapping on win32, then this
		3777	macro can be used to override the C<close> function, useful to unregister
		3778	file descriptors again. Note that the replacement function has to close
		3779	the underlying OS handle.
3025		3780
3026	=item EV_USE_POLL	3781	=item EV_USE_POLL
3027		3782
3028	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3783	If defined to be C<1>, libev will compile in support for the C<poll>(2)
3029	backend. Otherwise it will be enabled on non-win32 platforms. It	3784	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
3076	as well as for signal and thread safety in C<ev_async> watchers.	3831	as well as for signal and thread safety in C<ev_async> watchers.
3077		3832
3078	In the absence of this define, libev will use C<sig_atomic_t volatile>	3833	In the absence of this define, libev will use C<sig_atomic_t volatile>
3079	(from F<signal.h>), which is usually good enough on most platforms.	3834	(from F<signal.h>), which is usually good enough on most platforms.
3080		3835
3081	=item EV_H	3836	=item EV_H (h)
3082		3837
3083	The name of the F<ev.h> header file used to include it. The default if	3838	The name of the F<ev.h> header file used to include it. The default if
3084	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	3839	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
3085	used to virtually rename the F<ev.h> header file in case of conflicts.	3840	used to virtually rename the F<ev.h> header file in case of conflicts.
3086		3841
3087	=item EV_CONFIG_H	3842	=item EV_CONFIG_H (h)
3088		3843
3089	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	3844	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
3090	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	3845	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
3091	C<EV_H>, above.	3846	C<EV_H>, above.
3092		3847
3093	=item EV_EVENT_H	3848	=item EV_EVENT_H (h)
3094		3849
3095	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	3850	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
3096	of how the F<event.h> header can be found, the default is C<"event.h">.	3851	of how the F<event.h> header can be found, the default is C<"event.h">.
3097		3852
3098	=item EV_PROTOTYPES	3853	=item EV_PROTOTYPES (h)
3099		3854
3100	If defined to be C<0>, then F<ev.h> will not define any function	3855	If defined to be C<0>, then F<ev.h> will not define any function
3101	prototypes, but still define all the structs and other symbols. This is	3856	prototypes, but still define all the structs and other symbols. This is
3102	occasionally useful if you want to provide your own wrapper functions	3857	occasionally useful if you want to provide your own wrapper functions
3103	around libev functions.	3858	around libev functions.
…		…
3125	fine.	3880	fine.
3126		3881
3127	If your embedding application does not need any priorities, defining these	3882	If your embedding application does not need any priorities, defining these
3128	both to C<0> will save some memory and CPU.	3883	both to C<0> will save some memory and CPU.
3129		3884
3130	=item EV_PERIODIC_ENABLE	3885	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		3886	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		3887	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3131		3888
3132	If undefined or defined to be C<1>, then periodic timers are supported. If	3889	If undefined or defined to be C<1> (and the platform supports it), then
3133	defined to be C<0>, then they are not. Disabling them saves a few kB of	3890	the respective watcher type is supported. If defined to be C<0>, then it
3134	code.	3891	is not. Disabling watcher types mainly saves code size.
3135		3892
3136	=item EV_IDLE_ENABLE	3893	=item EV_FEATURES
3137
3138	If undefined or defined to be C<1>, then idle watchers are supported. If
3139	defined to be C<0>, then they are not. Disabling them saves a few kB of
3140	code.
3141
3142	=item EV_EMBED_ENABLE
3143
3144	If undefined or defined to be C<1>, then embed watchers are supported. If
3145	defined to be C<0>, then they are not. Embed watchers rely on most other
3146	watcher types, which therefore must not be disabled.
3147
3148	=item EV_STAT_ENABLE
3149
3150	If undefined or defined to be C<1>, then stat watchers are supported. If
3151	defined to be C<0>, then they are not.
3152
3153	=item EV_FORK_ENABLE
3154
3155	If undefined or defined to be C<1>, then fork watchers are supported. If
3156	defined to be C<0>, then they are not.
3157
3158	=item EV_ASYNC_ENABLE
3159
3160	If undefined or defined to be C<1>, then async watchers are supported. If
3161	defined to be C<0>, then they are not.
3162
3163	=item EV_MINIMAL
3164		3894
3165	If you need to shave off some kilobytes of code at the expense of some	3895	If you need to shave off some kilobytes of code at the expense of some
3166	speed, define this symbol to C<1>. Currently this is used to override some	3896	speed (but with the full API), you can define this symbol to request
3167	inlining decisions, saves roughly 30% code size on amd64. It also selects a	3897	certain subsets of functionality. The default is to enable all features
3168	much smaller 2-heap for timer management over the default 4-heap.	3898	that can be enabled on the platform.
		3899
		3900	A typical way to use this symbol is to define it to C<0> (or to a bitset
		3901	with some broad features you want) and then selectively re-enable
		3902	additional parts you want, for example if you want everything minimal,
		3903	but multiple event loop support, async and child watchers and the poll
		3904	backend, use this:
		3905
		3906	#define EV_FEATURES 0
		3907	#define EV_MULTIPLICITY 1
		3908	#define EV_USE_POLL 1
		3909	#define EV_CHILD_ENABLE 1
		3910	#define EV_ASYNC_ENABLE 1
		3911
		3912	The actual value is a bitset, it can be a combination of the following
		3913	values:
		3914
		3915	=over 4
		3916
		3917	=item C<1> - faster/larger code
		3918
		3919	Use larger code to speed up some operations.
		3920
		3921	Currently this is used to override some inlining decisions (enlarging the
		3922	code size by roughly 30% on amd64).
		3923
		3924	When optimising for size, use of compiler flags such as C<-Os> with
		3925	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		3926	assertions.
		3927
		3928	=item C<2> - faster/larger data structures
		3929
		3930	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		3931	hash table sizes and so on. This will usually further increase code size
		3932	and can additionally have an effect on the size of data structures at
		3933	runtime.
		3934
		3935	=item C<4> - full API configuration
		3936
		3937	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		3938	enables multiplicity (C<EV_MULTIPLICITY>=1).
		3939
		3940	=item C<8> - full API
		3941
		3942	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		3943	details on which parts of the API are still available without this
		3944	feature, and do not complain if this subset changes over time.
		3945
		3946	=item C<16> - enable all optional watcher types
		3947
		3948	Enables all optional watcher types. If you want to selectively enable
		3949	only some watcher types other than I/O and timers (e.g. prepare,
		3950	embed, async, child...) you can enable them manually by defining
		3951	C<EV_watchertype_ENABLE> to C<1> instead.
		3952
		3953	=item C<32> - enable all backends
		3954
		3955	This enables all backends - without this feature, you need to enable at
		3956	least one backend manually (C<EV_USE_SELECT> is a good choice).
		3957
		3958	=item C<64> - enable OS-specific "helper" APIs
		3959
		3960	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		3961	default.
		3962
		3963	=back
		3964
		3965	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		3966	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		3967	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		3968	watchers, timers and monotonic clock support.
		3969
		3970	With an intelligent-enough linker (gcc+binutils are intelligent enough
		3971	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		3972	your program might be left out as well - a binary starting a timer and an
		3973	I/O watcher then might come out at only 5Kb.
		3974
		3975	=item EV_AVOID_STDIO
		3976
		3977	If this is set to C<1> at compiletime, then libev will avoid using stdio
		3978	functions (printf, scanf, perror etc.). This will increase the code size
		3979	somewhat, but if your program doesn't otherwise depend on stdio and your
		3980	libc allows it, this avoids linking in the stdio library which is quite
		3981	big.
		3982
		3983	Note that error messages might become less precise when this option is
		3984	enabled.
		3985
		3986	=item EV_NSIG
		3987
		3988	The highest supported signal number, +1 (or, the number of
		3989	signals): Normally, libev tries to deduce the maximum number of signals
		3990	automatically, but sometimes this fails, in which case it can be
		3991	specified. Also, using a lower number than detected (C<32> should be
		3992	good for about any system in existence) can save some memory, as libev
		3993	statically allocates some 12-24 bytes per signal number.
3169		3994
3170	=item EV_PID_HASHSIZE	3995	=item EV_PID_HASHSIZE
3171		3996
3172	C<ev_child> watchers use a small hash table to distribute workload by	3997	C<ev_child> watchers use a small hash table to distribute workload by
3173	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3998	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3174	than enough. If you need to manage thousands of children you might want to	3999	usually more than enough. If you need to manage thousands of children you
3175	increase this value (I<must> be a power of two).	4000	might want to increase this value (I<must> be a power of two).
3176		4001
3177	=item EV_INOTIFY_HASHSIZE	4002	=item EV_INOTIFY_HASHSIZE
3178		4003
3179	C<ev_stat> watchers use a small hash table to distribute workload by	4004	C<ev_stat> watchers use a small hash table to distribute workload by
3180	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4005	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3181	usually more than enough. If you need to manage thousands of C<ev_stat>	4006	disabled), usually more than enough. If you need to manage thousands of
3182	watchers you might want to increase this value (I<must> be a power of	4007	C<ev_stat> watchers you might want to increase this value (I<must> be a
3183	two).	4008	power of two).
3184		4009
3185	=item EV_USE_4HEAP	4010	=item EV_USE_4HEAP
3186		4011
3187	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4012	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3188	timer and periodics heaps, libev uses a 4-heap when this symbol is defined	4013	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3189	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably	4014	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3190	faster performance with many (thousands) of watchers.	4015	faster performance with many (thousands) of watchers.
3191		4016
3192	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4017	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3193	(disabled).	4018	will be C<0>.
3194		4019
3195	=item EV_HEAP_CACHE_AT	4020	=item EV_HEAP_CACHE_AT
3196		4021
3197	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4022	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3198	timer and periodics heaps, libev can cache the timestamp (I<at>) within	4023	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3199	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4024	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3200	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4025	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3201	but avoids random read accesses on heap changes. This improves performance	4026	but avoids random read accesses on heap changes. This improves performance
3202	noticeably with many (hundreds) of watchers.	4027	noticeably with many (hundreds) of watchers.
3203		4028
3204	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4029	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3205	(disabled).	4030	will be C<0>.
3206		4031
3207	=item EV_VERIFY	4032	=item EV_VERIFY
3208		4033
3209	Controls how much internal verification (see C<ev_loop_verify ()>) will	4034	Controls how much internal verification (see C<ev_loop_verify ()>) will
3210	be done: If set to C<0>, no internal verification code will be compiled	4035	be done: If set to C<0>, no internal verification code will be compiled
…		…
3212	called. If set to C<2>, then the internal verification code will be	4037	called. If set to C<2>, then the internal verification code will be
3213	called once per loop, which can slow down libev. If set to C<3>, then the	4038	called once per loop, which can slow down libev. If set to C<3>, then the
3214	verification code will be called very frequently, which will slow down	4039	verification code will be called very frequently, which will slow down
3215	libev considerably.	4040	libev considerably.
3216		4041
3217	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4042	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3218	C<0>.	4043	will be C<0>.
3219		4044
3220	=item EV_COMMON	4045	=item EV_COMMON
3221		4046
3222	By default, all watchers have a C<void *data> member. By redefining	4047	By default, all watchers have a C<void *data> member. By redefining
3223	this macro to a something else you can include more and other types of	4048	this macro to something else you can include more and other types of
3224	members. You have to define it each time you include one of the files,	4049	members. You have to define it each time you include one of the files,
3225	though, and it must be identical each time.	4050	though, and it must be identical each time.
3226		4051
3227	For example, the perl EV module uses something like this:	4052	For example, the perl EV module uses something like this:
3228		4053
…		…
3281	file.	4106	file.
3282		4107
3283	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4108	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3284	that everybody includes and which overrides some configure choices:	4109	that everybody includes and which overrides some configure choices:
3285		4110
3286	#define EV_MINIMAL 1	4111	#define EV_FEATURES 8
3287	#define EV_USE_POLL 0	4112	#define EV_USE_SELECT 1
3288	#define EV_MULTIPLICITY 0
3289	#define EV_PERIODIC_ENABLE 0	4113	#define EV_PREPARE_ENABLE 1
		4114	#define EV_IDLE_ENABLE 1
3290	#define EV_STAT_ENABLE 0	4115	#define EV_SIGNAL_ENABLE 1
3291	#define EV_FORK_ENABLE 0	4116	#define EV_CHILD_ENABLE 1
		4117	#define EV_USE_STDEXCEPT 0
3292	#define EV_CONFIG_H <config.h>	4118	#define EV_CONFIG_H <config.h>
3293	#define EV_MINPRI 0
3294	#define EV_MAXPRI 0
3295		4119
3296	#include "ev++.h"	4120	#include "ev++.h"
3297		4121
3298	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4122	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3299		4123
3300	#include "ev_cpp.h"	4124	#include "ev_cpp.h"
3301	#include "ev.c"	4125	#include "ev.c"
3302		4126
		4127	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3303		4128
3304	=head1 THREADS AND COROUTINES	4129	=head2 THREADS AND COROUTINES
3305		4130
3306	=head2 THREADS	4131	=head3 THREADS
3307		4132
3308	Libev itself is thread-safe (unless the opposite is specifically	4133	All libev functions are reentrant and thread-safe unless explicitly
3309	documented for a function), but it uses no locking itself. This means that	4134	documented otherwise, but libev implements no locking itself. This means
3310	you can use as many loops as you want in parallel, as long as only one	4135	that you can use as many loops as you want in parallel, as long as there
3311	thread ever calls into one libev function with the same loop parameter:	4136	are no concurrent calls into any libev function with the same loop
		4137	parameter (C<ev_default_*> calls have an implicit default loop parameter,
3312	libev guarantees that different event loops share no data structures that	4138	of course): libev guarantees that different event loops share no data
3313	need locking.	4139	structures that need any locking.
3314		4140
3315	Or to put it differently: calls with different loop parameters can be done	4141	Or to put it differently: calls with different loop parameters can be done
3316	concurrently from multiple threads, calls with the same loop parameter	4142	concurrently from multiple threads, calls with the same loop parameter
3317	must be done serially (but can be done from different threads, as long as	4143	must be done serially (but can be done from different threads, as long as
3318	only one thread ever is inside a call at any point in time, e.g. by using	4144	only one thread ever is inside a call at any point in time, e.g. by using
3319	a mutex per loop).	4145	a mutex per loop).
3320		4146
3321	Specifically to support threads (and signal handlers), libev implements	4147	Specifically to support threads (and signal handlers), libev implements
3322	so-called C<ev_async> watchers, which allow some limited form of	4148	so-called C<ev_async> watchers, which allow some limited form of
3323	concurrency on the same event loop.	4149	concurrency on the same event loop, namely waking it up "from the
		4150	outside".
3324		4151
3325	If you want to know which design (one loop, locking, or multiple loops	4152	If you want to know which design (one loop, locking, or multiple loops
3326	without or something else still) is best for your problem, then I cannot	4153	without or something else still) is best for your problem, then I cannot
3327	help you. I can give some generic advice however:	4154	help you, but here is some generic advice:
3328		4155
3329	=over 4	4156	=over 4
3330		4157
3331	=item * most applications have a main thread: use the default libev loop	4158	=item * most applications have a main thread: use the default libev loop
3332	in that thread, or create a separate thread running only the default loop.	4159	in that thread, or create a separate thread running only the default loop.
…		…
3356	default loop and triggering an C<ev_async> watcher from the default loop	4183	default loop and triggering an C<ev_async> watcher from the default loop
3357	watcher callback into the event loop interested in the signal.	4184	watcher callback into the event loop interested in the signal.
3358		4185
3359	=back	4186	=back
3360		4187
		4188	=head4 THREAD LOCKING EXAMPLE
		4189
		4190	Here is a fictitious example of how to run an event loop in a different
		4191	thread than where callbacks are being invoked and watchers are
		4192	created/added/removed.
		4193
		4194	For a real-world example, see the C<EV::Loop::Async> perl module,
		4195	which uses exactly this technique (which is suited for many high-level
		4196	languages).
		4197
		4198	The example uses a pthread mutex to protect the loop data, a condition
		4199	variable to wait for callback invocations, an async watcher to notify the
		4200	event loop thread and an unspecified mechanism to wake up the main thread.
		4201
		4202	First, you need to associate some data with the event loop:
		4203
		4204	typedef struct {
		4205	mutex_t lock; /* global loop lock */
		4206	ev_async async_w;
		4207	thread_t tid;
		4208	cond_t invoke_cv;
		4209	} userdata;
		4210
		4211	void prepare_loop (EV_P)
		4212	{
		4213	// for simplicity, we use a static userdata struct.
		4214	static userdata u;
		4215
		4216	ev_async_init (&u->async_w, async_cb);
		4217	ev_async_start (EV_A_ &u->async_w);
		4218
		4219	pthread_mutex_init (&u->lock, 0);
		4220	pthread_cond_init (&u->invoke_cv, 0);
		4221
		4222	// now associate this with the loop
		4223	ev_set_userdata (EV_A_ u);
		4224	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4225	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4226
		4227	// then create the thread running ev_loop
		4228	pthread_create (&u->tid, 0, l_run, EV_A);
		4229	}
		4230
		4231	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4232	solely to wake up the event loop so it takes notice of any new watchers
		4233	that might have been added:
		4234
		4235	static void
		4236	async_cb (EV_P_ ev_async *w, int revents)
		4237	{
		4238	// just used for the side effects
		4239	}
		4240
		4241	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4242	protecting the loop data, respectively.
		4243
		4244	static void
		4245	l_release (EV_P)
		4246	{
		4247	userdata *u = ev_userdata (EV_A);
		4248	pthread_mutex_unlock (&u->lock);
		4249	}
		4250
		4251	static void
		4252	l_acquire (EV_P)
		4253	{
		4254	userdata *u = ev_userdata (EV_A);
		4255	pthread_mutex_lock (&u->lock);
		4256	}
		4257
		4258	The event loop thread first acquires the mutex, and then jumps straight
		4259	into C<ev_loop>:
		4260
		4261	void *
		4262	l_run (void *thr_arg)
		4263	{
		4264	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4265
		4266	l_acquire (EV_A);
		4267	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4268	ev_loop (EV_A_ 0);
		4269	l_release (EV_A);
		4270
		4271	return 0;
		4272	}
		4273
		4274	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4275	signal the main thread via some unspecified mechanism (signals? pipe
		4276	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4277	have been called (in a while loop because a) spurious wakeups are possible
		4278	and b) skipping inter-thread-communication when there are no pending
		4279	watchers is very beneficial):
		4280
		4281	static void
		4282	l_invoke (EV_P)
		4283	{
		4284	userdata *u = ev_userdata (EV_A);
		4285
		4286	while (ev_pending_count (EV_A))
		4287	{
		4288	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4289	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4290	}
		4291	}
		4292
		4293	Now, whenever the main thread gets told to invoke pending watchers, it
		4294	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4295	thread to continue:
		4296
		4297	static void
		4298	real_invoke_pending (EV_P)
		4299	{
		4300	userdata *u = ev_userdata (EV_A);
		4301
		4302	pthread_mutex_lock (&u->lock);
		4303	ev_invoke_pending (EV_A);
		4304	pthread_cond_signal (&u->invoke_cv);
		4305	pthread_mutex_unlock (&u->lock);
		4306	}
		4307
		4308	Whenever you want to start/stop a watcher or do other modifications to an
		4309	event loop, you will now have to lock:
		4310
		4311	ev_timer timeout_watcher;
		4312	userdata *u = ev_userdata (EV_A);
		4313
		4314	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4315
		4316	pthread_mutex_lock (&u->lock);
		4317	ev_timer_start (EV_A_ &timeout_watcher);
		4318	ev_async_send (EV_A_ &u->async_w);
		4319	pthread_mutex_unlock (&u->lock);
		4320
		4321	Note that sending the C<ev_async> watcher is required because otherwise
		4322	an event loop currently blocking in the kernel will have no knowledge
		4323	about the newly added timer. By waking up the loop it will pick up any new
		4324	watchers in the next event loop iteration.
		4325
3361	=head2 COROUTINES	4326	=head3 COROUTINES
3362		4327
3363	Libev is much more accommodating to coroutines ("cooperative threads"):	4328	Libev is very accommodating to coroutines ("cooperative threads"):
3364	libev fully supports nesting calls to it's functions from different	4329	libev fully supports nesting calls to its functions from different
3365	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4330	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3366	different coroutines and switch freely between both coroutines running the	4331	different coroutines, and switch freely between both coroutines running
3367	loop, as long as you don't confuse yourself). The only exception is that	4332	the loop, as long as you don't confuse yourself). The only exception is
3368	you must not do this from C<ev_periodic> reschedule callbacks.	4333	that you must not do this from C<ev_periodic> reschedule callbacks.
3369		4334
3370	Care has been taken to ensure that libev does not keep local state inside	4335	Care has been taken to ensure that libev does not keep local state inside
3371	C<ev_loop>, and other calls do not usually allow coroutine switches.	4336	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		4337	they do not call any callbacks.
3372		4338
		4339	=head2 COMPILER WARNINGS
3373		4340
3374	=head1 COMPLEXITIES	4341	Depending on your compiler and compiler settings, you might get no or a
		4342	lot of warnings when compiling libev code. Some people are apparently
		4343	scared by this.
3375		4344
3376	In this section the complexities of (many of) the algorithms used inside	4345	However, these are unavoidable for many reasons. For one, each compiler
3377	libev will be explained. For complexity discussions about backends see the	4346	has different warnings, and each user has different tastes regarding
3378	documentation for C<ev_default_init>.	4347	warning options. "Warn-free" code therefore cannot be a goal except when
		4348	targeting a specific compiler and compiler-version.
3379		4349
3380	All of the following are about amortised time: If an array needs to be	4350	Another reason is that some compiler warnings require elaborate
3381	extended, libev needs to realloc and move the whole array, but this	4351	workarounds, or other changes to the code that make it less clear and less
3382	happens asymptotically never with higher number of elements, so O(1) might	4352	maintainable.
3383	mean it might do a lengthy realloc operation in rare cases, but on average
3384	it is much faster and asymptotically approaches constant time.
3385		4353
3386	=over 4	4354	And of course, some compiler warnings are just plain stupid, or simply
		4355	wrong (because they don't actually warn about the condition their message
		4356	seems to warn about). For example, certain older gcc versions had some
		4357	warnings that resulted in an extreme number of false positives. These have
		4358	been fixed, but some people still insist on making code warn-free with
		4359	such buggy versions.
3387		4360
3388	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4361	While libev is written to generate as few warnings as possible,
		4362	"warn-free" code is not a goal, and it is recommended not to build libev
		4363	with any compiler warnings enabled unless you are prepared to cope with
		4364	them (e.g. by ignoring them). Remember that warnings are just that:
		4365	warnings, not errors, or proof of bugs.
3389		4366
3390	This means that, when you have a watcher that triggers in one hour and
3391	there are 100 watchers that would trigger before that then inserting will
3392	have to skip roughly seven (C<ld 100>) of these watchers.
3393		4367
3394	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4368	=head2 VALGRIND
3395		4369
3396	That means that changing a timer costs less than removing/adding them	4370	Valgrind has a special section here because it is a popular tool that is
3397	as only the relative motion in the event queue has to be paid for.	4371	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3398		4372
3399	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4373	If you think you found a bug (memory leak, uninitialised data access etc.)
		4374	in libev, then check twice: If valgrind reports something like:
3400		4375
3401	These just add the watcher into an array or at the head of a list.	4376	==2274== definitely lost: 0 bytes in 0 blocks.
		4377	==2274== possibly lost: 0 bytes in 0 blocks.
		4378	==2274== still reachable: 256 bytes in 1 blocks.
3402		4379
3403	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4380	Then there is no memory leak, just as memory accounted to global variables
		4381	is not a memleak - the memory is still being referenced, and didn't leak.
3404		4382
3405	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4383	Similarly, under some circumstances, valgrind might report kernel bugs
		4384	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4385	although an acceptable workaround has been found here), or it might be
		4386	confused.
3406		4387
3407	These watchers are stored in lists then need to be walked to find the	4388	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3408	correct watcher to remove. The lists are usually short (you don't usually	4389	make it into some kind of religion.
3409	have many watchers waiting for the same fd or signal).
3410		4390
3411	=item Finding the next timer in each loop iteration: O(1)	4391	If you are unsure about something, feel free to contact the mailing list
		4392	with the full valgrind report and an explanation on why you think this
		4393	is a bug in libev (best check the archives, too :). However, don't be
		4394	annoyed when you get a brisk "this is no bug" answer and take the chance
		4395	of learning how to interpret valgrind properly.
3412		4396
3413	By virtue of using a binary or 4-heap, the next timer is always found at a	4397	If you need, for some reason, empty reports from valgrind for your project
3414	fixed position in the storage array.	4398	I suggest using suppression lists.
3415		4399
3416	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3417		4400
3418	A change means an I/O watcher gets started or stopped, which requires	4401	=head1 PORTABILITY NOTES
3419	libev to recalculate its status (and possibly tell the kernel, depending
3420	on backend and whether C<ev_io_set> was used).
3421		4402
3422	=item Activating one watcher (putting it into the pending state): O(1)	4403	=head2 GNU/LINUX 32 BIT LIMITATIONS
3423		4404
3424	=item Priority handling: O(number_of_priorities)	4405	GNU/Linux is the only common platform that supports 64 bit file/large file
		4406	interfaces but I<disables> them by default.
3425		4407
3426	Priorities are implemented by allocating some space for each	4408	That means that libev compiled in the default environment doesn't support
3427	priority. When doing priority-based operations, libev usually has to	4409	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
3428	linearly search all the priorities, but starting/stopping and activating
3429	watchers becomes O(1) with respect to priority handling.
3430		4410
3431	=item Sending an ev_async: O(1)	4411	Unfortunately, many programs try to work around this GNU/Linux issue
		4412	by enabling the large file API, which makes them incompatible with the
		4413	standard libev compiled for their system.
3432		4414
3433	=item Processing ev_async_send: O(number_of_async_watchers)	4415	Likewise, libev cannot enable the large file API itself as this would
		4416	suddenly make it incompatible to the default compile time environment,
		4417	i.e. all programs not using special compile switches.
3434		4418
3435	=item Processing signals: O(max_signal_number)	4419	=head2 OS/X AND DARWIN BUGS
3436		4420
3437	Sending involves a system call I<iff> there were no other C<ev_async_send>	4421	The whole thing is a bug if you ask me - basically any system interface
3438	calls in the current loop iteration. Checking for async and signal events	4422	you touch is broken, whether it is locales, poll, kqueue or even the
3439	involves iterating over all running async watchers or all signal numbers.	4423	OpenGL drivers.
3440		4424
3441	=back	4425	=head3 C<kqueue> is buggy
3442		4426
		4427	The kqueue syscall is broken in all known versions - most versions support
		4428	only sockets, many support pipes.
3443		4429
		4430	Libev tries to work around this by not using C<kqueue> by default on
		4431	this rotten platform, but of course you can still ask for it when creating
		4432	a loop.
		4433
		4434	=head3 C<poll> is buggy
		4435
		4436	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4437	implementation by something calling C<kqueue> internally around the 10.5.6
		4438	release, so now C<kqueue> I<and> C<poll> are broken.
		4439
		4440	Libev tries to work around this by not using C<poll> by default on
		4441	this rotten platform, but of course you can still ask for it when creating
		4442	a loop.
		4443
		4444	=head3 C<select> is buggy
		4445
		4446	All that's left is C<select>, and of course Apple found a way to fuck this
		4447	one up as well: On OS/X, C<select> actively limits the number of file
		4448	descriptors you can pass in to 1024 - your program suddenly crashes when
		4449	you use more.
		4450
		4451	There is an undocumented "workaround" for this - defining
		4452	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		4453	work on OS/X.
		4454
		4455	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		4456
		4457	=head3 C<errno> reentrancy
		4458
		4459	The default compile environment on Solaris is unfortunately so
		4460	thread-unsafe that you can't even use components/libraries compiled
		4461	without C<-D_REENTRANT> (as long as they use C<errno>), which, of course,
		4462	isn't defined by default.
		4463
		4464	If you want to use libev in threaded environments you have to make sure
		4465	it's compiled with C<_REENTRANT> defined.
		4466
		4467	=head3 Event port backend
		4468
		4469	The scalable event interface for Solaris is called "event ports". Unfortunately,
		4470	this mechanism is very buggy. If you run into high CPU usage, your program
		4471	freezes or you get a large number of spurious wakeups, make sure you have
		4472	all the relevant and latest kernel patches applied. No, I don't know which
		4473	ones, but there are multiple ones.
		4474
		4475	If you can't get it to work, you can try running the program by setting
		4476	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		4477	C<select> backends.
		4478
		4479	=head2 AIX POLL BUG
		4480
		4481	AIX unfortunately has a broken C<poll.h> header. Libev works around
		4482	this by trying to avoid the poll backend altogether (i.e. it's not even
		4483	compiled in), which normally isn't a big problem as C<select> works fine
		4484	with large bitsets, and AIX is dead anyway.
		4485
3444	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4486	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4487
		4488	=head3 General issues
3445		4489
3446	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4490	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3447	requires, and its I/O model is fundamentally incompatible with the POSIX	4491	requires, and its I/O model is fundamentally incompatible with the POSIX
3448	model. Libev still offers limited functionality on this platform in	4492	model. Libev still offers limited functionality on this platform in
3449	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4493	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
3450	descriptors. This only applies when using Win32 natively, not when using	4494	descriptors. This only applies when using Win32 natively, not when using
3451	e.g. cygwin.	4495	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		4496	as every compielr comes with a slightly differently broken/incompatible
		4497	environment.
3452		4498
3453	Lifting these limitations would basically require the full	4499	Lifting these limitations would basically require the full
3454	re-implementation of the I/O system. If you are into these kinds of	4500	re-implementation of the I/O system. If you are into this kind of thing,
3455	things, then note that glib does exactly that for you in a very portable	4501	then note that glib does exactly that for you in a very portable way (note
3456	way (note also that glib is the slowest event library known to man).	4502	also that glib is the slowest event library known to man).
3457		4503
3458	There is no supported compilation method available on windows except	4504	There is no supported compilation method available on windows except
3459	embedding it into other applications.	4505	embedding it into other applications.
		4506
		4507	Sensible signal handling is officially unsupported by Microsoft - libev
		4508	tries its best, but under most conditions, signals will simply not work.
3460		4509
3461	Not a libev limitation but worth mentioning: windows apparently doesn't	4510	Not a libev limitation but worth mentioning: windows apparently doesn't
3462	accept large writes: instead of resulting in a partial write, windows will	4511	accept large writes: instead of resulting in a partial write, windows will
3463	either accept everything or return C<ENOBUFS> if the buffer is too large,	4512	either accept everything or return C<ENOBUFS> if the buffer is too large,
3464	so make sure you only write small amounts into your sockets (less than a	4513	so make sure you only write small amounts into your sockets (less than a
…		…
3469	the abysmal performance of winsockets, using a large number of sockets	4518	the abysmal performance of winsockets, using a large number of sockets
3470	is not recommended (and not reasonable). If your program needs to use	4519	is not recommended (and not reasonable). If your program needs to use
3471	more than a hundred or so sockets, then likely it needs to use a totally	4520	more than a hundred or so sockets, then likely it needs to use a totally
3472	different implementation for windows, as libev offers the POSIX readiness	4521	different implementation for windows, as libev offers the POSIX readiness
3473	notification model, which cannot be implemented efficiently on windows	4522	notification model, which cannot be implemented efficiently on windows
3474	(Microsoft monopoly games).	4523	(due to Microsoft monopoly games).
3475		4524
3476	A typical way to use libev under windows is to embed it (see the embedding	4525	A typical way to use libev under windows is to embed it (see the embedding
3477	section for details) and use the following F<evwrap.h> header file instead	4526	section for details) and use the following F<evwrap.h> header file instead
3478	of F<ev.h>:	4527	of F<ev.h>:
3479		4528
…		…
3486	you do I<not> compile the F<ev.c> or any other embedded source files!):	4535	you do I<not> compile the F<ev.c> or any other embedded source files!):
3487		4536
3488	#include "evwrap.h"	4537	#include "evwrap.h"
3489	#include "ev.c"	4538	#include "ev.c"
3490		4539
3491	=over 4
3492
3493	=item The winsocket select function	4540	=head3 The winsocket C<select> function
3494		4541
3495	The winsocket C<select> function doesn't follow POSIX in that it	4542	The winsocket C<select> function doesn't follow POSIX in that it
3496	requires socket I<handles> and not socket I<file descriptors> (it is	4543	requires socket I<handles> and not socket I<file descriptors> (it is
3497	also extremely buggy). This makes select very inefficient, and also	4544	also extremely buggy). This makes select very inefficient, and also
3498	requires a mapping from file descriptors to socket handles (the Microsoft	4545	requires a mapping from file descriptors to socket handles (the Microsoft
…		…
3507	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	4554	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
3508		4555
3509	Note that winsockets handling of fd sets is O(n), so you can easily get a	4556	Note that winsockets handling of fd sets is O(n), so you can easily get a
3510	complexity in the O(n²) range when using win32.	4557	complexity in the O(n²) range when using win32.
3511		4558
3512	=item Limited number of file descriptors	4559	=head3 Limited number of file descriptors
3513		4560
3514	Windows has numerous arbitrary (and low) limits on things.	4561	Windows has numerous arbitrary (and low) limits on things.
3515		4562
3516	Early versions of winsocket's select only supported waiting for a maximum	4563	Early versions of winsocket's select only supported waiting for a maximum
3517	of C<64> handles (probably owning to the fact that all windows kernels	4564	of C<64> handles (probably owning to the fact that all windows kernels
3518	can only wait for C<64> things at the same time internally; Microsoft	4565	can only wait for C<64> things at the same time internally; Microsoft
3519	recommends spawning a chain of threads and wait for 63 handles and the	4566	recommends spawning a chain of threads and wait for 63 handles and the
3520	previous thread in each. Great).	4567	previous thread in each. Sounds great!).
3521		4568
3522	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4569	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3523	to some high number (e.g. C<2048>) before compiling the winsocket select	4570	to some high number (e.g. C<2048>) before compiling the winsocket select
3524	call (which might be in libev or elsewhere, for example, perl does its own	4571	call (which might be in libev or elsewhere, for example, perl and many
3525	select emulation on windows).	4572	other interpreters do their own select emulation on windows).
3526		4573
3527	Another limit is the number of file descriptors in the Microsoft runtime	4574	Another limit is the number of file descriptors in the Microsoft runtime
3528	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4575	libraries, which by default is C<64> (there must be a hidden I<64>
3529	or something like this inside Microsoft). You can increase this by calling	4576	fetish or something like this inside Microsoft). You can increase this
3530	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4577	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3531	arbitrary limit), but is broken in many versions of the Microsoft runtime	4578	(another arbitrary limit), but is broken in many versions of the Microsoft
3532	libraries.
3533
3534	This might get you to about C<512> or C<2048> sockets (depending on	4579	runtime libraries. This might get you to about C<512> or C<2048> sockets
3535	windows version and/or the phase of the moon). To get more, you need to	4580	(depending on windows version and/or the phase of the moon). To get more,
3536	wrap all I/O functions and provide your own fd management, but the cost of	4581	you need to wrap all I/O functions and provide your own fd management, but
3537	calling select (O(n²)) will likely make this unworkable.	4582	the cost of calling select (O(n²)) will likely make this unworkable.
3538		4583
3539	=back
3540
3541
3542	=head1 PORTABILITY REQUIREMENTS	4584	=head2 PORTABILITY REQUIREMENTS
3543		4585
3544	In addition to a working ISO-C implementation, libev relies on a few	4586	In addition to a working ISO-C implementation and of course the
3545	additional extensions:	4587	backend-specific APIs, libev relies on a few additional extensions:
3546		4588
3547	=over 4	4589	=over 4
3548		4590
3549	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	4591	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3550	calling conventions regardless of C<ev_watcher_type *>.	4592	calling conventions regardless of C<ev_watcher_type *>.
…		…
3575	except the initial one, and run the default loop in the initial thread as	4617	except the initial one, and run the default loop in the initial thread as
3576	well.	4618	well.
3577		4619
3578	=item C<long> must be large enough for common memory allocation sizes	4620	=item C<long> must be large enough for common memory allocation sizes
3579		4621
3580	To improve portability and simplify using libev, libev uses C<long>	4622	To improve portability and simplify its API, libev uses C<long> internally
3581	internally instead of C<size_t> when allocating its data structures. On	4623	instead of C<size_t> when allocating its data structures. On non-POSIX
3582	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4624	systems (Microsoft...) this might be unexpectedly low, but is still at
3583	is still at least 31 bits everywhere, which is enough for hundreds of	4625	least 31 bits everywhere, which is enough for hundreds of millions of
3584	millions of watchers.	4626	watchers.
3585		4627
3586	=item C<double> must hold a time value in seconds with enough accuracy	4628	=item C<double> must hold a time value in seconds with enough accuracy
3587		4629
3588	The type C<double> is used to represent timestamps. It is required to	4630	The type C<double> is used to represent timestamps. It is required to
3589	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4631	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3590	enough for at least into the year 4000. This requirement is fulfilled by	4632	enough for at least into the year 4000. This requirement is fulfilled by
3591	implementations implementing IEEE 754 (basically all existing ones).	4633	implementations implementing IEEE 754, which is basically all existing
		4634	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4635	2200.
3592		4636
3593	=back	4637	=back
3594		4638
3595	If you know of other additional requirements drop me a note.	4639	If you know of other additional requirements drop me a note.
3596		4640
3597		4641
3598	=head1 COMPILER WARNINGS	4642	=head1 ALGORITHMIC COMPLEXITIES
3599		4643
3600	Depending on your compiler and compiler settings, you might get no or a	4644	In this section the complexities of (many of) the algorithms used inside
3601	lot of warnings when compiling libev code. Some people are apparently	4645	libev will be documented. For complexity discussions about backends see
3602	scared by this.	4646	the documentation for C<ev_default_init>.
3603		4647
3604	However, these are unavoidable for many reasons. For one, each compiler	4648	All of the following are about amortised time: If an array needs to be
3605	has different warnings, and each user has different tastes regarding	4649	extended, libev needs to realloc and move the whole array, but this
3606	warning options. "Warn-free" code therefore cannot be a goal except when	4650	happens asymptotically rarer with higher number of elements, so O(1) might
3607	targeting a specific compiler and compiler-version.	4651	mean that libev does a lengthy realloc operation in rare cases, but on
		4652	average it is much faster and asymptotically approaches constant time.
3608		4653
3609	Another reason is that some compiler warnings require elaborate	4654	=over 4
3610	workarounds, or other changes to the code that make it less clear and less
3611	maintainable.
3612		4655
3613	And of course, some compiler warnings are just plain stupid, or simply	4656	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3614	wrong (because they don't actually warn about the condition their message
3615	seems to warn about).
3616		4657
3617	While libev is written to generate as few warnings as possible,	4658	This means that, when you have a watcher that triggers in one hour and
3618	"warn-free" code is not a goal, and it is recommended not to build libev	4659	there are 100 watchers that would trigger before that, then inserting will
3619	with any compiler warnings enabled unless you are prepared to cope with	4660	have to skip roughly seven (C<ld 100>) of these watchers.
3620	them (e.g. by ignoring them). Remember that warnings are just that:
3621	warnings, not errors, or proof of bugs.
3622		4661
		4662	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3623		4663
3624	=head1 VALGRIND	4664	That means that changing a timer costs less than removing/adding them,
		4665	as only the relative motion in the event queue has to be paid for.
3625		4666
3626	Valgrind has a special section here because it is a popular tool that is	4667	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3627	highly useful, but valgrind reports are very hard to interpret.
3628		4668
3629	If you think you found a bug (memory leak, uninitialised data access etc.)	4669	These just add the watcher into an array or at the head of a list.
3630	in libev, then check twice: If valgrind reports something like:
3631		4670
3632	==2274== definitely lost: 0 bytes in 0 blocks.	4671	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3633	==2274== possibly lost: 0 bytes in 0 blocks.
3634	==2274== still reachable: 256 bytes in 1 blocks.
3635		4672
3636	Then there is no memory leak. Similarly, under some circumstances,	4673	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3637	valgrind might report kernel bugs as if it were a bug in libev, or it
3638	might be confused (it is a very good tool, but only a tool).
3639		4674
3640	If you are unsure about something, feel free to contact the mailing list	4675	These watchers are stored in lists, so they need to be walked to find the
3641	with the full valgrind report and an explanation on why you think this is	4676	correct watcher to remove. The lists are usually short (you don't usually
3642	a bug in libev. However, don't be annoyed when you get a brisk "this is	4677	have many watchers waiting for the same fd or signal: one is typical, two
3643	no bug" answer and take the chance of learning how to interpret valgrind	4678	is rare).
3644	properly.
3645		4679
3646	If you need, for some reason, empty reports from valgrind for your project	4680	=item Finding the next timer in each loop iteration: O(1)
3647	I suggest using suppression lists.
3648		4681
		4682	By virtue of using a binary or 4-heap, the next timer is always found at a
		4683	fixed position in the storage array.
		4684
		4685	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4686
		4687	A change means an I/O watcher gets started or stopped, which requires
		4688	libev to recalculate its status (and possibly tell the kernel, depending
		4689	on backend and whether C<ev_io_set> was used).
		4690
		4691	=item Activating one watcher (putting it into the pending state): O(1)
		4692
		4693	=item Priority handling: O(number_of_priorities)
		4694
		4695	Priorities are implemented by allocating some space for each
		4696	priority. When doing priority-based operations, libev usually has to
		4697	linearly search all the priorities, but starting/stopping and activating
		4698	watchers becomes O(1) with respect to priority handling.
		4699
		4700	=item Sending an ev_async: O(1)
		4701
		4702	=item Processing ev_async_send: O(number_of_async_watchers)
		4703
		4704	=item Processing signals: O(max_signal_number)
		4705
		4706	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4707	calls in the current loop iteration. Checking for async and signal events
		4708	involves iterating over all running async watchers or all signal numbers.
		4709
		4710	=back
		4711
		4712
		4713	=head1 PORTING FROM LIBEV 3.X TO 4.X
		4714
		4715	The major version 4 introduced some minor incompatible changes to the API.
		4716
		4717	At the moment, the C<ev.h> header file tries to implement superficial
		4718	compatibility, so most programs should still compile. Those might be
		4719	removed in later versions of libev, so better update early than late.
		4720
		4721	=over 4
		4722
		4723	=item C<ev_loop_count> renamed to C<ev_iteration>
		4724
		4725	=item C<ev_loop_depth> renamed to C<ev_depth>
		4726
		4727	=item C<ev_loop_verify> renamed to C<ev_verify>
		4728
		4729	Most functions working on C<struct ev_loop> objects don't have an
		4730	C<ev_loop_> prefix, so it was removed. Note that C<ev_loop_fork> is
		4731	still called C<ev_loop_fork> because it would otherwise clash with the
		4732	C<ev_fork> typedef.
		4733
		4734	=item C<EV_TIMEOUT> renamed to C<EV_TIMER> in C<revents>
		4735
		4736	This is a simple rename - all other watcher types use their name
		4737	as revents flag, and now C<ev_timer> does, too.
		4738
		4739	Both C<EV_TIMER> and C<EV_TIMEOUT> symbols were present in 3.x versions
		4740	and continue to be present for the foreseeable future, so this is mostly a
		4741	documentation change.
		4742
		4743	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		4744
		4745	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		4746	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		4747	and work, but the library code will of course be larger.
		4748
		4749	=back
		4750
		4751
		4752	=head1 GLOSSARY
		4753
		4754	=over 4
		4755
		4756	=item active
		4757
		4758	A watcher is active as long as it has been started (has been attached to
		4759	an event loop) but not yet stopped (disassociated from the event loop).
		4760
		4761	=item application
		4762
		4763	In this document, an application is whatever is using libev.
		4764
		4765	=item callback
		4766
		4767	The address of a function that is called when some event has been
		4768	detected. Callbacks are being passed the event loop, the watcher that
		4769	received the event, and the actual event bitset.
		4770
		4771	=item callback invocation
		4772
		4773	The act of calling the callback associated with a watcher.
		4774
		4775	=item event
		4776
		4777	A change of state of some external event, such as data now being available
		4778	for reading on a file descriptor, time having passed or simply not having
		4779	any other events happening anymore.
		4780
		4781	In libev, events are represented as single bits (such as C<EV_READ> or
		4782	C<EV_TIMER>).
		4783
		4784	=item event library
		4785
		4786	A software package implementing an event model and loop.
		4787
		4788	=item event loop
		4789
		4790	An entity that handles and processes external events and converts them
		4791	into callback invocations.
		4792
		4793	=item event model
		4794
		4795	The model used to describe how an event loop handles and processes
		4796	watchers and events.
		4797
		4798	=item pending
		4799
		4800	A watcher is pending as soon as the corresponding event has been detected,
		4801	and stops being pending as soon as the watcher will be invoked or its
		4802	pending status is explicitly cleared by the application.
		4803
		4804	A watcher can be pending, but not active. Stopping a watcher also clears
		4805	its pending status.
		4806
		4807	=item real time
		4808
		4809	The physical time that is observed. It is apparently strictly monotonic :)
		4810
		4811	=item wall-clock time
		4812
		4813	The time and date as shown on clocks. Unlike real time, it can actually
		4814	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4815	clock.
		4816
		4817	=item watcher
		4818
		4819	A data structure that describes interest in certain events. Watchers need
		4820	to be started (attached to an event loop) before they can receive events.
		4821
		4822	=item watcher invocation
		4823
		4824	The act of calling the callback associated with a watcher.
		4825
		4826	=back
3649		4827
3650	=head1 AUTHOR	4828	=head1 AUTHOR
3651		4829
3652	Marc Lehmann <libev@schmorp.de>.	4830	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3653		4831

Diff Legend

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