ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.60 by root, Wed Nov 28 18:29:30 2007 UTC vs.
Revision 1.133 by root, Sun Feb 24 06:50:16 2008 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
9	=head1 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	#include <ev.h>	11	#include <ev.h>
12		12
13	ev_io stdin_watcher;	13	ev_io stdin_watcher;
14	ev_timer timeout_watcher;	14	ev_timer timeout_watcher;
…		…
48	return 0;	48	return 0;
49	}	49	}
50		50
51	=head1 DESCRIPTION	51	=head1 DESCRIPTION
52		52
		53	The newest version of this document is also available as a html-formatted
		54	web page you might find easier to navigate when reading it for the first
		55	time: L<http://cvs.schmorp.de/libev/ev.html>.
		56
53	Libev is an event loop: you register interest in certain events (such as a	57	Libev is an event loop: you register interest in certain events (such as a
54	file descriptor being readable or a timeout occuring), and it will manage	58	file descriptor being readable or a timeout occurring), and it will manage
55	these event sources and provide your program with events.	59	these event sources and provide your program with events.
56		60
57	To do this, it must take more or less complete control over your process	61	To do this, it must take more or less complete control over your process
58	(or thread) by executing the I<event loop> handler, and will then	62	(or thread) by executing the I<event loop> handler, and will then
59	communicate events via a callback mechanism.	63	communicate events via a callback mechanism.
…		…
61	You register interest in certain events by registering so-called I<event	65	You register interest in certain events by registering so-called I<event
62	watchers>, which are relatively small C structures you initialise with the	66	watchers>, which are relatively small C structures you initialise with the
63	details of the event, and then hand it over to libev by I<starting> the	67	details of the event, and then hand it over to libev by I<starting> the
64	watcher.	68	watcher.
65		69
66	=head1 FEATURES	70	=head2 FEATURES
67		71
68	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	72	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
69	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	73	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
70	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	74	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
71	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	75	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
…		…
78		82
79	It also is quite fast (see this	83	It also is quite fast (see this
80	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	84	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
81	for example).	85	for example).
82		86
83	=head1 CONVENTIONS	87	=head2 CONVENTIONS
84		88
85	Libev is very configurable. In this manual the default configuration will	89	Libev is very configurable. In this manual the default configuration will
86	be described, which supports multiple event loops. For more info about	90	be described, which supports multiple event loops. For more info about
87	various configuration options please have a look at B<EMBED> section in	91	various configuration options please have a look at B<EMBED> section in
88	this manual. If libev was configured without support for multiple event	92	this manual. If libev was configured without support for multiple event
89	loops, then all functions taking an initial argument of name C<loop>	93	loops, then all functions taking an initial argument of name C<loop>
90	(which is always of type C<struct ev_loop *>) will not have this argument.	94	(which is always of type C<struct ev_loop *>) will not have this argument.
91		95
92	=head1 TIME REPRESENTATION	96	=head2 TIME REPRESENTATION
93		97
94	Libev represents time as a single floating point number, representing the	98	Libev represents time as a single floating point number, representing the
95	(fractional) number of seconds since the (POSIX) epoch (somewhere near	99	(fractional) number of seconds since the (POSIX) epoch (somewhere near
96	the beginning of 1970, details are complicated, don't ask). This type is	100	the beginning of 1970, details are complicated, don't ask). This type is
97	called C<ev_tstamp>, which is what you should use too. It usually aliases	101	called C<ev_tstamp>, which is what you should use too. It usually aliases
98	to the C<double> type in C, and when you need to do any calculations on	102	to the C<double> type in C, and when you need to do any calculations on
99	it, you should treat it as such.	103	it, you should treat it as some floatingpoint value. Unlike the name
		104	component C<stamp> might indicate, it is also used for time differences
		105	throughout libev.
100		106
101	=head1 GLOBAL FUNCTIONS	107	=head1 GLOBAL FUNCTIONS
102		108
103	These functions can be called anytime, even before initialising the	109	These functions can be called anytime, even before initialising the
104	library in any way.	110	library in any way.
…		…
109		115
110	Returns the current time as libev would use it. Please note that the	116	Returns the current time as libev would use it. Please note that the
111	C<ev_now> function is usually faster and also often returns the timestamp	117	C<ev_now> function is usually faster and also often returns the timestamp
112	you actually want to know.	118	you actually want to know.
113		119
		120	=item ev_sleep (ev_tstamp interval)
		121
		122	Sleep for the given interval: The current thread will be blocked until
		123	either it is interrupted or the given time interval has passed. Basically
		124	this is a subsecond-resolution C<sleep ()>.
		125
114	=item int ev_version_major ()	126	=item int ev_version_major ()
115		127
116	=item int ev_version_minor ()	128	=item int ev_version_minor ()
117		129
118	You can find out the major and minor version numbers of the library	130	You can find out the major and minor ABI version numbers of the library
119	you linked against by calling the functions C<ev_version_major> and	131	you linked against by calling the functions C<ev_version_major> and
120	C<ev_version_minor>. If you want, you can compare against the global	132	C<ev_version_minor>. If you want, you can compare against the global
121	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	133	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
122	version of the library your program was compiled against.	134	version of the library your program was compiled against.
123		135
		136	These version numbers refer to the ABI version of the library, not the
		137	release version.
		138
124	Usually, it's a good idea to terminate if the major versions mismatch,	139	Usually, it's a good idea to terminate if the major versions mismatch,
125	as this indicates an incompatible change. Minor versions are usually	140	as this indicates an incompatible change. Minor versions are usually
126	compatible to older versions, so a larger minor version alone is usually	141	compatible to older versions, so a larger minor version alone is usually
127	not a problem.	142	not a problem.
128		143
129	Example: Make sure we haven't accidentally been linked against the wrong	144	Example: Make sure we haven't accidentally been linked against the wrong
130	version.	145	version.
…		…
245	flags. If that is troubling you, check C<ev_backend ()> afterwards).	260	flags. If that is troubling you, check C<ev_backend ()> afterwards).
246		261
247	If you don't know what event loop to use, use the one returned from this	262	If you don't know what event loop to use, use the one returned from this
248	function.	263	function.
249		264
		265	The default loop is the only loop that can handle C<ev_signal> and
		266	C<ev_child> watchers, and to do this, it always registers a handler
		267	for C<SIGCHLD>. If this is a problem for your app you can either
		268	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		269	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		270	C<ev_default_init>.
		271
250	The flags argument can be used to specify special behaviour or specific	272	The flags argument can be used to specify special behaviour or specific
251	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	273	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
252		274
253	The following flags are supported:	275	The following flags are supported:
254		276
…		…
266	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	288	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
267	override the flags completely if it is found in the environment. This is	289	override the flags completely if it is found in the environment. This is
268	useful to try out specific backends to test their performance, or to work	290	useful to try out specific backends to test their performance, or to work
269	around bugs.	291	around bugs.
270		292
		293	=item C<EVFLAG_FORKCHECK>
		294
		295	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		296	a fork, you can also make libev check for a fork in each iteration by
		297	enabling this flag.
		298
		299	This works by calling C<getpid ()> on every iteration of the loop,
		300	and thus this might slow down your event loop if you do a lot of loop
		301	iterations and little real work, but is usually not noticeable (on my
		302	Linux system for example, C<getpid> is actually a simple 5-insn sequence
		303	without a syscall and thus I<very> fast, but my Linux system also has
		304	C<pthread_atfork> which is even faster).
		305
		306	The big advantage of this flag is that you can forget about fork (and
		307	forget about forgetting to tell libev about forking) when you use this
		308	flag.
		309
		310	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		311	environment variable.
		312
271	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	313	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
272		314
273	This is your standard select(2) backend. Not I<completely> standard, as	315	This is your standard select(2) backend. Not I<completely> standard, as
274	libev tries to roll its own fd_set with no limits on the number of fds,	316	libev tries to roll its own fd_set with no limits on the number of fds,
275	but if that fails, expect a fairly low limit on the number of fds when	317	but if that fails, expect a fairly low limit on the number of fds when
276	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	318	using this backend. It doesn't scale too well (O(highest_fd)), but its
277	the fastest backend for a low number of fds.	319	usually the fastest backend for a low number of (low-numbered :) fds.
		320
		321	To get good performance out of this backend you need a high amount of
		322	parallelity (most of the file descriptors should be busy). If you are
		323	writing a server, you should C<accept ()> in a loop to accept as many
		324	connections as possible during one iteration. You might also want to have
		325	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		326	readyness notifications you get per iteration.
278		327
279	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	328	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
280		329
281	And this is your standard poll(2) backend. It's more complicated than	330	And this is your standard poll(2) backend. It's more complicated
282	select, but handles sparse fds better and has no artificial limit on the	331	than select, but handles sparse fds better and has no artificial
283	number of fds you can use (except it will slow down considerably with a	332	limit on the number of fds you can use (except it will slow down
284	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	333	considerably with a lot of inactive fds). It scales similarly to select,
		334	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		335	performance tips.
285		336
286	=item C<EVBACKEND_EPOLL> (value 4, Linux)	337	=item C<EVBACKEND_EPOLL> (value 4, Linux)
287		338
288	For few fds, this backend is a bit little slower than poll and select,	339	For few fds, this backend is a bit little slower than poll and select,
289	but it scales phenomenally better. While poll and select usually scale like	340	but it scales phenomenally better. While poll and select usually scale
290	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	341	like O(total_fds) where n is the total number of fds (or the highest fd),
291	either O(1) or O(active_fds).	342	epoll scales either O(1) or O(active_fds). The epoll design has a number
		343	of shortcomings, such as silently dropping events in some hard-to-detect
		344	cases and rewiring a syscall per fd change, no fork support and bad
		345	support for dup.
292		346
293	While stopping and starting an I/O watcher in the same iteration will	347	While stopping, setting and starting an I/O watcher in the same iteration
294	result in some caching, there is still a syscall per such incident	348	will result in some caching, there is still a syscall per such incident
295	(because the fd could point to a different file description now), so its	349	(because the fd could point to a different file description now), so its
296	best to avoid that. Also, dup()ed file descriptors might not work very	350	best to avoid that. Also, C<dup ()>'ed file descriptors might not work
297	well if you register events for both fds.	351	very well if you register events for both fds.
298		352
299	Please note that epoll sometimes generates spurious notifications, so you	353	Please note that epoll sometimes generates spurious notifications, so you
300	need to use non-blocking I/O or other means to avoid blocking when no data	354	need to use non-blocking I/O or other means to avoid blocking when no data
301	(or space) is available.	355	(or space) is available.
302		356
		357	Best performance from this backend is achieved by not unregistering all
		358	watchers for a file descriptor until it has been closed, if possible, i.e.
		359	keep at least one watcher active per fd at all times.
		360
		361	While nominally embeddeble in other event loops, this feature is broken in
		362	all kernel versions tested so far.
		363
303	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	364	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
304		365
305	Kqueue deserves special mention, as at the time of this writing, it	366	Kqueue deserves special mention, as at the time of this writing, it
306	was broken on all BSDs except NetBSD (usually it doesn't work with	367	was broken on all BSDs except NetBSD (usually it doesn't work reliably
307	anything but sockets and pipes, except on Darwin, where of course its	368	with anything but sockets and pipes, except on Darwin, where of course
308	completely useless). For this reason its not being "autodetected"	369	it's completely useless). For this reason it's not being "autodetected"
309	unless you explicitly specify it explicitly in the flags (i.e. using	370	unless you explicitly specify it explicitly in the flags (i.e. using
310	C<EVBACKEND_KQUEUE>).	371	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		372	system like NetBSD.
		373
		374	You still can embed kqueue into a normal poll or select backend and use it
		375	only for sockets (after having made sure that sockets work with kqueue on
		376	the target platform). See C<ev_embed> watchers for more info.
311		377
312	It scales in the same way as the epoll backend, but the interface to the	378	It scales in the same way as the epoll backend, but the interface to the
313	kernel is more efficient (which says nothing about its actual speed, of	379	kernel is more efficient (which says nothing about its actual speed, of
314	course). While starting and stopping an I/O watcher does not cause an	380	course). While stopping, setting and starting an I/O watcher does never
315	extra syscall as with epoll, it still adds up to four event changes per	381	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to
316	incident, so its best to avoid that.	382	two event changes per incident, support for C<fork ()> is very bad and it
		383	drops fds silently in similarly hard-to-detect cases.
		384
		385	This backend usually performs well under most conditions.
		386
		387	While nominally embeddable in other event loops, this doesn't work
		388	everywhere, so you might need to test for this. And since it is broken
		389	almost everywhere, you should only use it when you have a lot of sockets
		390	(for which it usually works), by embedding it into another event loop
		391	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for
		392	sockets.
317		393
318	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	394	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
319		395
320	This is not implemented yet (and might never be).	396	This is not implemented yet (and might never be, unless you send me an
		397	implementation). According to reports, C</dev/poll> only supports sockets
		398	and is not embeddable, which would limit the usefulness of this backend
		399	immensely.
321		400
322	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	401	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
323		402
324	This uses the Solaris 10 port mechanism. As with everything on Solaris,	403	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
325	it's really slow, but it still scales very well (O(active_fds)).	404	it's really slow, but it still scales very well (O(active_fds)).
326		405
327	Please note that solaris ports can result in a lot of spurious	406	Please note that solaris event ports can deliver a lot of spurious
328	notifications, so you need to use non-blocking I/O or other means to avoid	407	notifications, so you need to use non-blocking I/O or other means to avoid
329	blocking when no data (or space) is available.	408	blocking when no data (or space) is available.
		409
		410	While this backend scales well, it requires one system call per active
		411	file descriptor per loop iteration. For small and medium numbers of file
		412	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		413	might perform better.
		414
		415	On the positive side, ignoring the spurious readyness notifications, this
		416	backend actually performed to specification in all tests and is fully
		417	embeddable, which is a rare feat among the OS-specific backends.
330		418
331	=item C<EVBACKEND_ALL>	419	=item C<EVBACKEND_ALL>
332		420
333	Try all backends (even potentially broken ones that wouldn't be tried	421	Try all backends (even potentially broken ones that wouldn't be tried
334	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	422	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
335	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	423	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
336		424
		425	It is definitely not recommended to use this flag.
		426
337	=back	427	=back
338		428
339	If one or more of these are ored into the flags value, then only these	429	If one or more of these are ored into the flags value, then only these
340	backends will be tried (in the reverse order as given here). If none are	430	backends will be tried (in the reverse order as listed here). If none are
341	specified, most compiled-in backend will be tried, usually in reverse	431	specified, all backends in C<ev_recommended_backends ()> will be tried.
342	order of their flag values :)
343		432
344	The most typical usage is like this:	433	The most typical usage is like this:
345		434
346	if (!ev_default_loop (0))	435	if (!ev_default_loop (0))
347	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	436	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
…		…
375	Destroys the default loop again (frees all memory and kernel state	464	Destroys the default loop again (frees all memory and kernel state
376	etc.). None of the active event watchers will be stopped in the normal	465	etc.). None of the active event watchers will be stopped in the normal
377	sense, so e.g. C<ev_is_active> might still return true. It is your	466	sense, so e.g. C<ev_is_active> might still return true. It is your
378	responsibility to either stop all watchers cleanly yoursef I<before>	467	responsibility to either stop all watchers cleanly yoursef I<before>
379	calling this function, or cope with the fact afterwards (which is usually	468	calling this function, or cope with the fact afterwards (which is usually
380	the easiest thing, youc na just ignore the watchers and/or C<free ()> them	469	the easiest thing, you can just ignore the watchers and/or C<free ()> them
381	for example).	470	for example).
		471
		472	Note that certain global state, such as signal state, will not be freed by
		473	this function, and related watchers (such as signal and child watchers)
		474	would need to be stopped manually.
		475
		476	In general it is not advisable to call this function except in the
		477	rare occasion where you really need to free e.g. the signal handling
		478	pipe fds. If you need dynamically allocated loops it is better to use
		479	C<ev_loop_new> and C<ev_loop_destroy>).
382		480
383	=item ev_loop_destroy (loop)	481	=item ev_loop_destroy (loop)
384		482
385	Like C<ev_default_destroy>, but destroys an event loop created by an	483	Like C<ev_default_destroy>, but destroys an event loop created by an
386	earlier call to C<ev_loop_new>.	484	earlier call to C<ev_loop_new>.
387		485
388	=item ev_default_fork ()	486	=item ev_default_fork ()
389		487
		488	This function sets a flag that causes subsequent C<ev_loop> iterations
390	This function reinitialises the kernel state for backends that have	489	to reinitialise the kernel state for backends that have one. Despite the
391	one. Despite the name, you can call it anytime, but it makes most sense	490	name, you can call it anytime, but it makes most sense after forking, in
392	after forking, in either the parent or child process (or both, but that	491	the child process (or both child and parent, but that again makes little
393	again makes little sense).	492	sense). You I<must> call it in the child before using any of the libev
		493	functions, and it will only take effect at the next C<ev_loop> iteration.
394		494
395	You I<must> call this function in the child process after forking if and	495	On the other hand, you only need to call this function in the child
396	only if you want to use the event library in both processes. If you just	496	process if and only if you want to use the event library in the child. If
397	fork+exec, you don't have to call it.	497	you just fork+exec, you don't have to call it at all.
398		498
399	The function itself is quite fast and it's usually not a problem to call	499	The function itself is quite fast and it's usually not a problem to call
400	it just in case after a fork. To make this easy, the function will fit in	500	it just in case after a fork. To make this easy, the function will fit in
401	quite nicely into a call to C<pthread_atfork>:	501	quite nicely into a call to C<pthread_atfork>:
402		502
403	pthread_atfork (0, 0, ev_default_fork);	503	pthread_atfork (0, 0, ev_default_fork);
404		504
405	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
406	without calling this function, so if you force one of those backends you
407	do not need to care.
408
409	=item ev_loop_fork (loop)	505	=item ev_loop_fork (loop)
410		506
411	Like C<ev_default_fork>, but acts on an event loop created by	507	Like C<ev_default_fork>, but acts on an event loop created by
412	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	508	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
413	after fork, and how you do this is entirely your own problem.	509	after fork, and how you do this is entirely your own problem.
		510
		511	=item int ev_is_default_loop (loop)
		512
		513	Returns true when the given loop actually is the default loop, false otherwise.
		514
		515	=item unsigned int ev_loop_count (loop)
		516
		517	Returns the count of loop iterations for the loop, which is identical to
		518	the number of times libev did poll for new events. It starts at C<0> and
		519	happily wraps around with enough iterations.
		520
		521	This value can sometimes be useful as a generation counter of sorts (it
		522	"ticks" the number of loop iterations), as it roughly corresponds with
		523	C<ev_prepare> and C<ev_check> calls.
414		524
415	=item unsigned int ev_backend (loop)	525	=item unsigned int ev_backend (loop)
416		526
417	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	527	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
418	use.	528	use.
…		…
421		531
422	Returns the current "event loop time", which is the time the event loop	532	Returns the current "event loop time", which is the time the event loop
423	received events and started processing them. This timestamp does not	533	received events and started processing them. This timestamp does not
424	change as long as callbacks are being processed, and this is also the base	534	change as long as callbacks are being processed, and this is also the base
425	time used for relative timers. You can treat it as the timestamp of the	535	time used for relative timers. You can treat it as the timestamp of the
426	event occuring (or more correctly, libev finding out about it).	536	event occurring (or more correctly, libev finding out about it).
427		537
428	=item ev_loop (loop, int flags)	538	=item ev_loop (loop, int flags)
429		539
430	Finally, this is it, the event handler. This function usually is called	540	Finally, this is it, the event handler. This function usually is called
431	after you initialised all your watchers and you want to start handling	541	after you initialised all your watchers and you want to start handling
…		…
452	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	562	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
453	usually a better approach for this kind of thing.	563	usually a better approach for this kind of thing.
454		564
455	Here are the gory details of what C<ev_loop> does:	565	Here are the gory details of what C<ev_loop> does:
456		566
457	* If there are no active watchers (reference count is zero), return.	567	- Before the first iteration, call any pending watchers.
458	- Queue prepare watchers and then call all outstanding watchers.	568	* If EVFLAG_FORKCHECK was used, check for a fork.
		569	- If a fork was detected, queue and call all fork watchers.
		570	- Queue and call all prepare watchers.
459	- If we have been forked, recreate the kernel state.	571	- If we have been forked, recreate the kernel state.
460	- Update the kernel state with all outstanding changes.	572	- Update the kernel state with all outstanding changes.
461	- Update the "event loop time".	573	- Update the "event loop time".
462	- Calculate for how long to block.	574	- Calculate for how long to sleep or block, if at all
		575	(active idle watchers, EVLOOP_NONBLOCK or not having
		576	any active watchers at all will result in not sleeping).
		577	- Sleep if the I/O and timer collect interval say so.
463	- Block the process, waiting for any events.	578	- Block the process, waiting for any events.
464	- Queue all outstanding I/O (fd) events.	579	- Queue all outstanding I/O (fd) events.
465	- Update the "event loop time" and do time jump handling.	580	- Update the "event loop time" and do time jump handling.
466	- Queue all outstanding timers.	581	- Queue all outstanding timers.
467	- Queue all outstanding periodics.	582	- Queue all outstanding periodics.
468	- If no events are pending now, queue all idle watchers.	583	- If no events are pending now, queue all idle watchers.
469	- Queue all check watchers.	584	- Queue all check watchers.
470	- Call all queued watchers in reverse order (i.e. check watchers first).	585	- Call all queued watchers in reverse order (i.e. check watchers first).
471	Signals and child watchers are implemented as I/O watchers, and will	586	Signals and child watchers are implemented as I/O watchers, and will
472	be handled here by queueing them when their watcher gets executed.	587	be handled here by queueing them when their watcher gets executed.
473	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	588	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
474	were used, return, otherwise continue with step *.	589	were used, or there are no active watchers, return, otherwise
		590	continue with step *.
475		591
476	Example: Queue some jobs and then loop until no events are outsanding	592	Example: Queue some jobs and then loop until no events are outstanding
477	anymore.	593	anymore.
478		594
479	... queue jobs here, make sure they register event watchers as long	595	... queue jobs here, make sure they register event watchers as long
480	... as they still have work to do (even an idle watcher will do..)	596	... as they still have work to do (even an idle watcher will do..)
481	ev_loop (my_loop, 0);	597	ev_loop (my_loop, 0);
…		…
485		601
486	Can be used to make a call to C<ev_loop> return early (but only after it	602	Can be used to make a call to C<ev_loop> return early (but only after it
487	has processed all outstanding events). The C<how> argument must be either	603	has processed all outstanding events). The C<how> argument must be either
488	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	604	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
489	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	605	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
		606
		607	This "unloop state" will be cleared when entering C<ev_loop> again.
490		608
491	=item ev_ref (loop)	609	=item ev_ref (loop)
492		610
493	=item ev_unref (loop)	611	=item ev_unref (loop)
494		612
…		…
499	returning, ev_unref() after starting, and ev_ref() before stopping it. For	617	returning, ev_unref() after starting, and ev_ref() before stopping it. For
500	example, libev itself uses this for its internal signal pipe: It is not	618	example, libev itself uses this for its internal signal pipe: It is not
501	visible to the libev user and should not keep C<ev_loop> from exiting if	619	visible to the libev user and should not keep C<ev_loop> from exiting if
502	no event watchers registered by it are active. It is also an excellent	620	no event watchers registered by it are active. It is also an excellent
503	way to do this for generic recurring timers or from within third-party	621	way to do this for generic recurring timers or from within third-party
504	libraries. Just remember to I<unref after start> and I<ref before stop>.	622	libraries. Just remember to I<unref after start> and I<ref before stop>
		623	(but only if the watcher wasn't active before, or was active before,
		624	respectively).
505		625
506	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	626	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
507	running when nothing else is active.	627	running when nothing else is active.
508		628
509	struct ev_signal exitsig;	629	struct ev_signal exitsig;
…		…
513		633
514	Example: For some weird reason, unregister the above signal handler again.	634	Example: For some weird reason, unregister the above signal handler again.
515		635
516	ev_ref (loop);	636	ev_ref (loop);
517	ev_signal_stop (loop, &exitsig);	637	ev_signal_stop (loop, &exitsig);
		638
		639	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		640
		641	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		642
		643	These advanced functions influence the time that libev will spend waiting
		644	for events. Both are by default C<0>, meaning that libev will try to
		645	invoke timer/periodic callbacks and I/O callbacks with minimum latency.
		646
		647	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		648	allows libev to delay invocation of I/O and timer/periodic callbacks to
		649	increase efficiency of loop iterations.
		650
		651	The background is that sometimes your program runs just fast enough to
		652	handle one (or very few) event(s) per loop iteration. While this makes
		653	the program responsive, it also wastes a lot of CPU time to poll for new
		654	events, especially with backends like C<select ()> which have a high
		655	overhead for the actual polling but can deliver many events at once.
		656
		657	By setting a higher I<io collect interval> you allow libev to spend more
		658	time collecting I/O events, so you can handle more events per iteration,
		659	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		660	C<ev_timer>) will be not affected. Setting this to a non-null value will
		661	introduce an additional C<ev_sleep ()> call into most loop iterations.
		662
		663	Likewise, by setting a higher I<timeout collect interval> you allow libev
		664	to spend more time collecting timeouts, at the expense of increased
		665	latency (the watcher callback will be called later). C<ev_io> watchers
		666	will not be affected. Setting this to a non-null value will not introduce
		667	any overhead in libev.
		668
		669	Many (busy) programs can usually benefit by setting the io collect
		670	interval to a value near C<0.1> or so, which is often enough for
		671	interactive servers (of course not for games), likewise for timeouts. It
		672	usually doesn't make much sense to set it to a lower value than C<0.01>,
		673	as this approsaches the timing granularity of most systems.
518		674
519	=back	675	=back
520		676
521		677
522	=head1 ANATOMY OF A WATCHER	678	=head1 ANATOMY OF A WATCHER
…		…
622	=item C<EV_FORK>	778	=item C<EV_FORK>
623		779
624	The event loop has been resumed in the child process after fork (see	780	The event loop has been resumed in the child process after fork (see
625	C<ev_fork>).	781	C<ev_fork>).
626		782
		783	=item C<EV_ASYNC>
		784
		785	The given async watcher has been asynchronously notified (see C<ev_async>).
		786
627	=item C<EV_ERROR>	787	=item C<EV_ERROR>
628		788
629	An unspecified error has occured, the watcher has been stopped. This might	789	An unspecified error has occured, the watcher has been stopped. This might
630	happen because the watcher could not be properly started because libev	790	happen because the watcher could not be properly started because libev
631	ran out of memory, a file descriptor was found to be closed or any other	791	ran out of memory, a file descriptor was found to be closed or any other
…		…
702	=item bool ev_is_pending (ev_TYPE *watcher)	862	=item bool ev_is_pending (ev_TYPE *watcher)
703		863
704	Returns a true value iff the watcher is pending, (i.e. it has outstanding	864	Returns a true value iff the watcher is pending, (i.e. it has outstanding
705	events but its callback has not yet been invoked). As long as a watcher	865	events but its callback has not yet been invoked). As long as a watcher
706	is pending (but not active) you must not call an init function on it (but	866	is pending (but not active) you must not call an init function on it (but
707	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	867	C<ev_TYPE_set> is safe), you must not change its priority, and you must
708	libev (e.g. you cnanot C<free ()> it).	868	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		869	it).
709		870
710	=item callback ev_cb (ev_TYPE *watcher)	871	=item callback ev_cb (ev_TYPE *watcher)
711		872
712	Returns the callback currently set on the watcher.	873	Returns the callback currently set on the watcher.
713		874
714	=item ev_cb_set (ev_TYPE *watcher, callback)	875	=item ev_cb_set (ev_TYPE *watcher, callback)
715		876
716	Change the callback. You can change the callback at virtually any time	877	Change the callback. You can change the callback at virtually any time
717	(modulo threads).	878	(modulo threads).
		879
		880	=item ev_set_priority (ev_TYPE *watcher, priority)
		881
		882	=item int ev_priority (ev_TYPE *watcher)
		883
		884	Set and query the priority of the watcher. The priority is a small
		885	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		886	(default: C<-2>). Pending watchers with higher priority will be invoked
		887	before watchers with lower priority, but priority will not keep watchers
		888	from being executed (except for C<ev_idle> watchers).
		889
		890	This means that priorities are I<only> used for ordering callback
		891	invocation after new events have been received. This is useful, for
		892	example, to reduce latency after idling, or more often, to bind two
		893	watchers on the same event and make sure one is called first.
		894
		895	If you need to suppress invocation when higher priority events are pending
		896	you need to look at C<ev_idle> watchers, which provide this functionality.
		897
		898	You I<must not> change the priority of a watcher as long as it is active or
		899	pending.
		900
		901	The default priority used by watchers when no priority has been set is
		902	always C<0>, which is supposed to not be too high and not be too low :).
		903
		904	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		905	fine, as long as you do not mind that the priority value you query might
		906	or might not have been adjusted to be within valid range.
		907
		908	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		909
		910	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		911	C<loop> nor C<revents> need to be valid as long as the watcher callback
		912	can deal with that fact.
		913
		914	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		915
		916	If the watcher is pending, this function returns clears its pending status
		917	and returns its C<revents> bitset (as if its callback was invoked). If the
		918	watcher isn't pending it does nothing and returns C<0>.
718		919
719	=back	920	=back
720		921
721		922
722	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	923	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
…		…
807	In general you can register as many read and/or write event watchers per	1008	In general you can register as many read and/or write event watchers per
808	fd as you want (as long as you don't confuse yourself). Setting all file	1009	fd as you want (as long as you don't confuse yourself). Setting all file
809	descriptors to non-blocking mode is also usually a good idea (but not	1010	descriptors to non-blocking mode is also usually a good idea (but not
810	required if you know what you are doing).	1011	required if you know what you are doing).
811		1012
812	You have to be careful with dup'ed file descriptors, though. Some backends
813	(the linux epoll backend is a notable example) cannot handle dup'ed file
814	descriptors correctly if you register interest in two or more fds pointing
815	to the same underlying file/socket/etc. description (that is, they share
816	the same underlying "file open").
817
818	If you must do this, then force the use of a known-to-be-good backend	1013	If you must do this, then force the use of a known-to-be-good backend
819	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1014	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
820	C<EVBACKEND_POLL>).	1015	C<EVBACKEND_POLL>).
821		1016
822	Another thing you have to watch out for is that it is quite easy to	1017	Another thing you have to watch out for is that it is quite easy to
…		…
828	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1023	it is best to always use non-blocking I/O: An extra C<read>(2) returning
829	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1024	C<EAGAIN> is far preferable to a program hanging until some data arrives.
830		1025
831	If you cannot run the fd in non-blocking mode (for example you should not	1026	If you cannot run the fd in non-blocking mode (for example you should not
832	play around with an Xlib connection), then you have to seperately re-test	1027	play around with an Xlib connection), then you have to seperately re-test
833	wether a file descriptor is really ready with a known-to-be good interface	1028	whether a file descriptor is really ready with a known-to-be good interface
834	such as poll (fortunately in our Xlib example, Xlib already does this on	1029	such as poll (fortunately in our Xlib example, Xlib already does this on
835	its own, so its quite safe to use).	1030	its own, so its quite safe to use).
		1031
		1032	=head3 The special problem of disappearing file descriptors
		1033
		1034	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1035	descriptor (either by calling C<close> explicitly or by any other means,
		1036	such as C<dup>). The reason is that you register interest in some file
		1037	descriptor, but when it goes away, the operating system will silently drop
		1038	this interest. If another file descriptor with the same number then is
		1039	registered with libev, there is no efficient way to see that this is, in
		1040	fact, a different file descriptor.
		1041
		1042	To avoid having to explicitly tell libev about such cases, libev follows
		1043	the following policy: Each time C<ev_io_set> is being called, libev
		1044	will assume that this is potentially a new file descriptor, otherwise
		1045	it is assumed that the file descriptor stays the same. That means that
		1046	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1047	descriptor even if the file descriptor number itself did not change.
		1048
		1049	This is how one would do it normally anyway, the important point is that
		1050	the libev application should not optimise around libev but should leave
		1051	optimisations to libev.
		1052
		1053	=head3 The special problem of dup'ed file descriptors
		1054
		1055	Some backends (e.g. epoll), cannot register events for file descriptors,
		1056	but only events for the underlying file descriptions. That means when you
		1057	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1058	events for them, only one file descriptor might actually receive events.
		1059
		1060	There is no workaround possible except not registering events
		1061	for potentially C<dup ()>'ed file descriptors, or to resort to
		1062	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1063
		1064	=head3 The special problem of fork
		1065
		1066	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1067	useless behaviour. Libev fully supports fork, but needs to be told about
		1068	it in the child.
		1069
		1070	To support fork in your programs, you either have to call
		1071	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
		1072	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
		1073	C<EVBACKEND_POLL>.
		1074
		1075
		1076	=head3 Watcher-Specific Functions
836		1077
837	=over 4	1078	=over 4
838		1079
839	=item ev_io_init (ev_io *, callback, int fd, int events)	1080	=item ev_io_init (ev_io *, callback, int fd, int events)
840		1081
…		…
851	=item int events [read-only]	1092	=item int events [read-only]
852		1093
853	The events being watched.	1094	The events being watched.
854		1095
855	=back	1096	=back
		1097
		1098	=head3 Examples
856		1099
857	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1100	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
858	readable, but only once. Since it is likely line-buffered, you could	1101	readable, but only once. Since it is likely line-buffered, you could
859	attempt to read a whole line in the callback.	1102	attempt to read a whole line in the callback.
860		1103
…		…
894		1137
895	The callback is guarenteed to be invoked only when its timeout has passed,	1138	The callback is guarenteed to be invoked only when its timeout has passed,
896	but if multiple timers become ready during the same loop iteration then	1139	but if multiple timers become ready during the same loop iteration then
897	order of execution is undefined.	1140	order of execution is undefined.
898		1141
		1142	=head3 Watcher-Specific Functions and Data Members
		1143
899	=over 4	1144	=over 4
900		1145
901	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1146	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
902		1147
903	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1148	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
…		…
911	configure a timer to trigger every 10 seconds, then it will trigger at	1156	configure a timer to trigger every 10 seconds, then it will trigger at
912	exactly 10 second intervals. If, however, your program cannot keep up with	1157	exactly 10 second intervals. If, however, your program cannot keep up with
913	the timer (because it takes longer than those 10 seconds to do stuff) the	1158	the timer (because it takes longer than those 10 seconds to do stuff) the
914	timer will not fire more than once per event loop iteration.	1159	timer will not fire more than once per event loop iteration.
915		1160
916	=item ev_timer_again (loop)	1161	=item ev_timer_again (loop, ev_timer *)
917		1162
918	This will act as if the timer timed out and restart it again if it is	1163	This will act as if the timer timed out and restart it again if it is
919	repeating. The exact semantics are:	1164	repeating. The exact semantics are:
920		1165
		1166	If the timer is pending, its pending status is cleared.
		1167
921	If the timer is started but nonrepeating, stop it.	1168	If the timer is started but nonrepeating, stop it (as if it timed out).
922		1169
923	If the timer is repeating, either start it if necessary (with the repeat	1170	If the timer is repeating, either start it if necessary (with the
924	value), or reset the running timer to the repeat value.	1171	C<repeat> value), or reset the running timer to the C<repeat> value.
925		1172
926	This sounds a bit complicated, but here is a useful and typical	1173	This sounds a bit complicated, but here is a useful and typical
927	example: Imagine you have a tcp connection and you want a so-called	1174	example: Imagine you have a tcp connection and you want a so-called idle
928	idle timeout, that is, you want to be called when there have been,	1175	timeout, that is, you want to be called when there have been, say, 60
929	say, 60 seconds of inactivity on the socket. The easiest way to do	1176	seconds of inactivity on the socket. The easiest way to do this is to
930	this is to configure an C<ev_timer> with C<after>=C<repeat>=C<60> and calling	1177	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
931	C<ev_timer_again> each time you successfully read or write some data. If	1178	C<ev_timer_again> each time you successfully read or write some data. If
932	you go into an idle state where you do not expect data to travel on the	1179	you go into an idle state where you do not expect data to travel on the
933	socket, you can stop the timer, and again will automatically restart it if	1180	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
934	need be.	1181	automatically restart it if need be.
935		1182
936	You can also ignore the C<after> value and C<ev_timer_start> altogether	1183	That means you can ignore the C<after> value and C<ev_timer_start>
937	and only ever use the C<repeat> value:	1184	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
938		1185
939	ev_timer_init (timer, callback, 0., 5.);	1186	ev_timer_init (timer, callback, 0., 5.);
940	ev_timer_again (loop, timer);	1187	ev_timer_again (loop, timer);
941	...	1188	...
942	timer->again = 17.;	1189	timer->again = 17.;
943	ev_timer_again (loop, timer);	1190	ev_timer_again (loop, timer);
944	...	1191	...
945	timer->again = 10.;	1192	timer->again = 10.;
946	ev_timer_again (loop, timer);	1193	ev_timer_again (loop, timer);
947		1194
948	This is more efficient then stopping/starting the timer eahc time you want	1195	This is more slightly efficient then stopping/starting the timer each time
949	to modify its timeout value.	1196	you want to modify its timeout value.
950		1197
951	=item ev_tstamp repeat [read-write]	1198	=item ev_tstamp repeat [read-write]
952		1199
953	The current C<repeat> value. Will be used each time the watcher times out	1200	The current C<repeat> value. Will be used each time the watcher times out
954	or C<ev_timer_again> is called and determines the next timeout (if any),	1201	or C<ev_timer_again> is called and determines the next timeout (if any),
955	which is also when any modifications are taken into account.	1202	which is also when any modifications are taken into account.
956		1203
957	=back	1204	=back
		1205
		1206	=head3 Examples
958		1207
959	Example: Create a timer that fires after 60 seconds.	1208	Example: Create a timer that fires after 60 seconds.
960		1209
961	static void	1210	static void
962	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1211	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
…		…
996	but on wallclock time (absolute time). You can tell a periodic watcher	1245	but on wallclock time (absolute time). You can tell a periodic watcher
997	to trigger "at" some specific point in time. For example, if you tell a	1246	to trigger "at" some specific point in time. For example, if you tell a
998	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1247	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
999	+ 10.>) and then reset your system clock to the last year, then it will	1248	+ 10.>) and then reset your system clock to the last year, then it will
1000	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1249	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
1001	roughly 10 seconds later and of course not if you reset your system time	1250	roughly 10 seconds later).
1002	again).
1003		1251
1004	They can also be used to implement vastly more complex timers, such as	1252	They can also be used to implement vastly more complex timers, such as
1005	triggering an event on eahc midnight, local time.	1253	triggering an event on each midnight, local time or other, complicated,
		1254	rules.
1006		1255
1007	As with timers, the callback is guarenteed to be invoked only when the	1256	As with timers, the callback is guarenteed to be invoked only when the
1008	time (C<at>) has been passed, but if multiple periodic timers become ready	1257	time (C<at>) has been passed, but if multiple periodic timers become ready
1009	during the same loop iteration then order of execution is undefined.	1258	during the same loop iteration then order of execution is undefined.
1010		1259
		1260	=head3 Watcher-Specific Functions and Data Members
		1261
1011	=over 4	1262	=over 4
1012		1263
1013	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1264	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
1014		1265
1015	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1266	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
…		…
1017	Lots of arguments, lets sort it out... There are basically three modes of	1268	Lots of arguments, lets sort it out... There are basically three modes of
1018	operation, and we will explain them from simplest to complex:	1269	operation, and we will explain them from simplest to complex:
1019		1270
1020	=over 4	1271	=over 4
1021		1272
1022	=item * absolute timer (interval = reschedule_cb = 0)	1273	=item * absolute timer (at = time, interval = reschedule_cb = 0)
1023		1274
1024	In this configuration the watcher triggers an event at the wallclock time	1275	In this configuration the watcher triggers an event at the wallclock time
1025	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1276	C<at> and doesn't repeat. It will not adjust when a time jump occurs,
1026	that is, if it is to be run at January 1st 2011 then it will run when the	1277	that is, if it is to be run at January 1st 2011 then it will run when the
1027	system time reaches or surpasses this time.	1278	system time reaches or surpasses this time.
1028		1279
1029	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1280	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
1030		1281
1031	In this mode the watcher will always be scheduled to time out at the next	1282	In this mode the watcher will always be scheduled to time out at the next
1032	C<at + N * interval> time (for some integer N) and then repeat, regardless	1283	C<at + N * interval> time (for some integer N, which can also be negative)
1033	of any time jumps.	1284	and then repeat, regardless of any time jumps.
1034		1285
1035	This can be used to create timers that do not drift with respect to system	1286	This can be used to create timers that do not drift with respect to system
1036	time:	1287	time:
1037		1288
1038	ev_periodic_set (&periodic, 0., 3600., 0);	1289	ev_periodic_set (&periodic, 0., 3600., 0);
…		…
1044		1295
1045	Another way to think about it (for the mathematically inclined) is that	1296	Another way to think about it (for the mathematically inclined) is that
1046	C<ev_periodic> will try to run the callback in this mode at the next possible	1297	C<ev_periodic> will try to run the callback in this mode at the next possible
1047	time where C<time = at (mod interval)>, regardless of any time jumps.	1298	time where C<time = at (mod interval)>, regardless of any time jumps.
1048		1299
		1300	For numerical stability it is preferable that the C<at> value is near
		1301	C<ev_now ()> (the current time), but there is no range requirement for
		1302	this value.
		1303
1049	=item * manual reschedule mode (reschedule_cb = callback)	1304	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1050		1305
1051	In this mode the values for C<interval> and C<at> are both being	1306	In this mode the values for C<interval> and C<at> are both being
1052	ignored. Instead, each time the periodic watcher gets scheduled, the	1307	ignored. Instead, each time the periodic watcher gets scheduled, the
1053	reschedule callback will be called with the watcher as first, and the	1308	reschedule callback will be called with the watcher as first, and the
1054	current time as second argument.	1309	current time as second argument.
1055		1310
1056	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1311	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
1057	ever, or make any event loop modifications>. If you need to stop it,	1312	ever, or make any event loop modifications>. If you need to stop it,
1058	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	1313	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
1059	starting a prepare watcher).	1314	starting an C<ev_prepare> watcher, which is legal).
1060		1315
1061	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1316	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
1062	ev_tstamp now)>, e.g.:	1317	ev_tstamp now)>, e.g.:
1063		1318
1064	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1319	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
…		…
1087	Simply stops and restarts the periodic watcher again. This is only useful	1342	Simply stops and restarts the periodic watcher again. This is only useful
1088	when you changed some parameters or the reschedule callback would return	1343	when you changed some parameters or the reschedule callback would return
1089	a different time than the last time it was called (e.g. in a crond like	1344	a different time than the last time it was called (e.g. in a crond like
1090	program when the crontabs have changed).	1345	program when the crontabs have changed).
1091		1346
		1347	=item ev_tstamp offset [read-write]
		1348
		1349	When repeating, this contains the offset value, otherwise this is the
		1350	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1351
		1352	Can be modified any time, but changes only take effect when the periodic
		1353	timer fires or C<ev_periodic_again> is being called.
		1354
1092	=item ev_tstamp interval [read-write]	1355	=item ev_tstamp interval [read-write]
1093		1356
1094	The current interval value. Can be modified any time, but changes only	1357	The current interval value. Can be modified any time, but changes only
1095	take effect when the periodic timer fires or C<ev_periodic_again> is being	1358	take effect when the periodic timer fires or C<ev_periodic_again> is being
1096	called.	1359	called.
…		…
1099		1362
1100	The current reschedule callback, or C<0>, if this functionality is	1363	The current reschedule callback, or C<0>, if this functionality is
1101	switched off. Can be changed any time, but changes only take effect when	1364	switched off. Can be changed any time, but changes only take effect when
1102	the periodic timer fires or C<ev_periodic_again> is being called.	1365	the periodic timer fires or C<ev_periodic_again> is being called.
1103		1366
		1367	=item ev_tstamp at [read-only]
		1368
		1369	When active, contains the absolute time that the watcher is supposed to
		1370	trigger next.
		1371
1104	=back	1372	=back
		1373
		1374	=head3 Examples
1105		1375
1106	Example: Call a callback every hour, or, more precisely, whenever the	1376	Example: Call a callback every hour, or, more precisely, whenever the
1107	system clock is divisible by 3600. The callback invocation times have	1377	system clock is divisible by 3600. The callback invocation times have
1108	potentially a lot of jittering, but good long-term stability.	1378	potentially a lot of jittering, but good long-term stability.
1109		1379
…		…
1149	with the kernel (thus it coexists with your own signal handlers as long	1419	with the kernel (thus it coexists with your own signal handlers as long
1150	as you don't register any with libev). Similarly, when the last signal	1420	as you don't register any with libev). Similarly, when the last signal
1151	watcher for a signal is stopped libev will reset the signal handler to	1421	watcher for a signal is stopped libev will reset the signal handler to
1152	SIG_DFL (regardless of what it was set to before).	1422	SIG_DFL (regardless of what it was set to before).
1153		1423
		1424	=head3 Watcher-Specific Functions and Data Members
		1425
1154	=over 4	1426	=over 4
1155		1427
1156	=item ev_signal_init (ev_signal *, callback, int signum)	1428	=item ev_signal_init (ev_signal *, callback, int signum)
1157		1429
1158	=item ev_signal_set (ev_signal *, int signum)	1430	=item ev_signal_set (ev_signal *, int signum)
…		…
1164		1436
1165	The signal the watcher watches out for.	1437	The signal the watcher watches out for.
1166		1438
1167	=back	1439	=back
1168		1440
		1441	=head3 Examples
		1442
		1443	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1444
		1445	static void
		1446	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1447	{
		1448	ev_unloop (loop, EVUNLOOP_ALL);
		1449	}
		1450
		1451	struct ev_signal signal_watcher;
		1452	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1453	ev_signal_start (loop, &sigint_cb);
		1454
1169		1455
1170	=head2 C<ev_child> - watch out for process status changes	1456	=head2 C<ev_child> - watch out for process status changes
1171		1457
1172	Child watchers trigger when your process receives a SIGCHLD in response to	1458	Child watchers trigger when your process receives a SIGCHLD in response to
1173	some child status changes (most typically when a child of yours dies).	1459	some child status changes (most typically when a child of yours dies).
1174		1460
		1461	=head3 Watcher-Specific Functions and Data Members
		1462
1175	=over 4	1463	=over 4
1176		1464
1177	=item ev_child_init (ev_child *, callback, int pid)	1465	=item ev_child_init (ev_child *, callback, int pid, int trace)
1178		1466
1179	=item ev_child_set (ev_child *, int pid)	1467	=item ev_child_set (ev_child *, int pid, int trace)
1180		1468
1181	Configures the watcher to wait for status changes of process C<pid> (or	1469	Configures the watcher to wait for status changes of process C<pid> (or
1182	I<any> process if C<pid> is specified as C<0>). The callback can look	1470	I<any> process if C<pid> is specified as C<0>). The callback can look
1183	at the C<rstatus> member of the C<ev_child> watcher structure to see	1471	at the C<rstatus> member of the C<ev_child> watcher structure to see
1184	the status word (use the macros from C<sys/wait.h> and see your systems	1472	the status word (use the macros from C<sys/wait.h> and see your systems
1185	C<waitpid> documentation). The C<rpid> member contains the pid of the	1473	C<waitpid> documentation). The C<rpid> member contains the pid of the
1186	process causing the status change.	1474	process causing the status change. C<trace> must be either C<0> (only
		1475	activate the watcher when the process terminates) or C<1> (additionally
		1476	activate the watcher when the process is stopped or continued).
1187		1477
1188	=item int pid [read-only]	1478	=item int pid [read-only]
1189		1479
1190	The process id this watcher watches out for, or C<0>, meaning any process id.	1480	The process id this watcher watches out for, or C<0>, meaning any process id.
1191		1481
…		…
1197		1487
1198	The process exit/trace status caused by C<rpid> (see your systems	1488	The process exit/trace status caused by C<rpid> (see your systems
1199	C<waitpid> and C<sys/wait.h> documentation for details).	1489	C<waitpid> and C<sys/wait.h> documentation for details).
1200		1490
1201	=back	1491	=back
1202
1203	Example: Try to exit cleanly on SIGINT and SIGTERM.
1204
1205	static void
1206	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
1207	{
1208	ev_unloop (loop, EVUNLOOP_ALL);
1209	}
1210
1211	struct ev_signal signal_watcher;
1212	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1213	ev_signal_start (loop, &sigint_cb);
1214		1492
1215		1493
1216	=head2 C<ev_stat> - did the file attributes just change?	1494	=head2 C<ev_stat> - did the file attributes just change?
1217		1495
1218	This watches a filesystem path for attribute changes. That is, it calls	1496	This watches a filesystem path for attribute changes. That is, it calls
…		…
1247	semantics of C<ev_stat> watchers, which means that libev sometimes needs	1525	semantics of C<ev_stat> watchers, which means that libev sometimes needs
1248	to fall back to regular polling again even with inotify, but changes are	1526	to fall back to regular polling again even with inotify, but changes are
1249	usually detected immediately, and if the file exists there will be no	1527	usually detected immediately, and if the file exists there will be no
1250	polling.	1528	polling.
1251		1529
		1530	=head3 Inotify
		1531
		1532	When C<inotify (7)> support has been compiled into libev (generally only
		1533	available on Linux) and present at runtime, it will be used to speed up
		1534	change detection where possible. The inotify descriptor will be created lazily
		1535	when the first C<ev_stat> watcher is being started.
		1536
		1537	Inotify presense does not change the semantics of C<ev_stat> watchers
		1538	except that changes might be detected earlier, and in some cases, to avoid
		1539	making regular C<stat> calls. Even in the presense of inotify support
		1540	there are many cases where libev has to resort to regular C<stat> polling.
		1541
		1542	(There is no support for kqueue, as apparently it cannot be used to
		1543	implement this functionality, due to the requirement of having a file
		1544	descriptor open on the object at all times).
		1545
		1546	=head3 The special problem of stat time resolution
		1547
		1548	The C<stat ()> syscall only supports full-second resolution portably, and
		1549	even on systems where the resolution is higher, many filesystems still
		1550	only support whole seconds.
		1551
		1552	That means that, if the time is the only thing that changes, you might
		1553	miss updates: on the first update, C<ev_stat> detects a change and calls
		1554	your callback, which does something. When there is another update within
		1555	the same second, C<ev_stat> will be unable to detect it.
		1556
		1557	The solution to this is to delay acting on a change for a second (or till
		1558	the next second boundary), using a roughly one-second delay C<ev_timer>
		1559	(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>
		1560	is added to work around small timing inconsistencies of some operating
		1561	systems.
		1562
		1563	=head3 Watcher-Specific Functions and Data Members
		1564
1252	=over 4	1565	=over 4
1253		1566
1254	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)	1567	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
1255		1568
1256	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)	1569	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
…		…
1263		1576
1264	The callback will be receive C<EV_STAT> when a change was detected,	1577	The callback will be receive C<EV_STAT> when a change was detected,
1265	relative to the attributes at the time the watcher was started (or the	1578	relative to the attributes at the time the watcher was started (or the
1266	last change was detected).	1579	last change was detected).
1267		1580
1268	=item ev_stat_stat (ev_stat *)	1581	=item ev_stat_stat (loop, ev_stat *)
1269		1582
1270	Updates the stat buffer immediately with new values. If you change the	1583	Updates the stat buffer immediately with new values. If you change the
1271	watched path in your callback, you could call this fucntion to avoid	1584	watched path in your callback, you could call this fucntion to avoid
1272	detecting this change (while introducing a race condition). Can also be	1585	detecting this change (while introducing a race condition). Can also be
1273	useful simply to find out the new values.	1586	useful simply to find out the new values.
…		…
1291	=item const char *path [read-only]	1604	=item const char *path [read-only]
1292		1605
1293	The filesystem path that is being watched.	1606	The filesystem path that is being watched.
1294		1607
1295	=back	1608	=back
		1609
		1610	=head3 Examples
1296		1611
1297	Example: Watch C</etc/passwd> for attribute changes.	1612	Example: Watch C</etc/passwd> for attribute changes.
1298		1613
1299	static void	1614	static void
1300	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	1615	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
…		…
1313	}	1628	}
1314		1629
1315	...	1630	...
1316	ev_stat passwd;	1631	ev_stat passwd;
1317		1632
1318	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");	1633	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1319	ev_stat_start (loop, &passwd);	1634	ev_stat_start (loop, &passwd);
1320		1635
		1636	Example: Like above, but additionally use a one-second delay so we do not
		1637	miss updates (however, frequent updates will delay processing, too, so
		1638	one might do the work both on C<ev_stat> callback invocation I<and> on
		1639	C<ev_timer> callback invocation).
		1640
		1641	static ev_stat passwd;
		1642	static ev_timer timer;
		1643
		1644	static void
		1645	timer_cb (EV_P_ ev_timer *w, int revents)
		1646	{
		1647	ev_timer_stop (EV_A_ w);
		1648
		1649	/* now it's one second after the most recent passwd change */
		1650	}
		1651
		1652	static void
		1653	stat_cb (EV_P_ ev_stat *w, int revents)
		1654	{
		1655	/* reset the one-second timer */
		1656	ev_timer_again (EV_A_ &timer);
		1657	}
		1658
		1659	...
		1660	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		1661	ev_stat_start (loop, &passwd);
		1662	ev_timer_init (&timer, timer_cb, 0., 1.01);
		1663
1321		1664
1322	=head2 C<ev_idle> - when you've got nothing better to do...	1665	=head2 C<ev_idle> - when you've got nothing better to do...
1323		1666
1324	Idle watchers trigger events when there are no other events are pending	1667	Idle watchers trigger events when no other events of the same or higher
1325	(prepare, check and other idle watchers do not count). That is, as long	1668	priority are pending (prepare, check and other idle watchers do not
1326	as your process is busy handling sockets or timeouts (or even signals,	1669	count).
1327	imagine) it will not be triggered. But when your process is idle all idle	1670
1328	watchers are being called again and again, once per event loop iteration -	1671	That is, as long as your process is busy handling sockets or timeouts
		1672	(or even signals, imagine) of the same or higher priority it will not be
		1673	triggered. But when your process is idle (or only lower-priority watchers
		1674	are pending), the idle watchers are being called once per event loop
1329	until stopped, that is, or your process receives more events and becomes	1675	iteration - until stopped, that is, or your process receives more events
1330	busy.	1676	and becomes busy again with higher priority stuff.
1331		1677
1332	The most noteworthy effect is that as long as any idle watchers are	1678	The most noteworthy effect is that as long as any idle watchers are
1333	active, the process will not block when waiting for new events.	1679	active, the process will not block when waiting for new events.
1334		1680
1335	Apart from keeping your process non-blocking (which is a useful	1681	Apart from keeping your process non-blocking (which is a useful
1336	effect on its own sometimes), idle watchers are a good place to do	1682	effect on its own sometimes), idle watchers are a good place to do
1337	"pseudo-background processing", or delay processing stuff to after the	1683	"pseudo-background processing", or delay processing stuff to after the
1338	event loop has handled all outstanding events.	1684	event loop has handled all outstanding events.
1339		1685
		1686	=head3 Watcher-Specific Functions and Data Members
		1687
1340	=over 4	1688	=over 4
1341		1689
1342	=item ev_idle_init (ev_signal *, callback)	1690	=item ev_idle_init (ev_signal *, callback)
1343		1691
1344	Initialises and configures the idle watcher - it has no parameters of any	1692	Initialises and configures the idle watcher - it has no parameters of any
1345	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1693	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1346	believe me.	1694	believe me.
1347		1695
1348	=back	1696	=back
		1697
		1698	=head3 Examples
1349		1699
1350	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	1700	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1351	callback, free it. Also, use no error checking, as usual.	1701	callback, free it. Also, use no error checking, as usual.
1352		1702
1353	static void	1703	static void
1354	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	1704	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1355	{	1705	{
1356	free (w);	1706	free (w);
1357	// now do something you wanted to do when the program has	1707	// now do something you wanted to do when the program has
1358	// no longer asnything immediate to do.	1708	// no longer anything immediate to do.
1359	}	1709	}
1360		1710
1361	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	1711	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1362	ev_idle_init (idle_watcher, idle_cb);	1712	ev_idle_init (idle_watcher, idle_cb);
1363	ev_idle_start (loop, idle_cb);	1713	ev_idle_start (loop, idle_cb);
…		…
1401	with priority higher than or equal to the event loop and one coroutine	1751	with priority higher than or equal to the event loop and one coroutine
1402	of lower priority, but only once, using idle watchers to keep the event	1752	of lower priority, but only once, using idle watchers to keep the event
1403	loop from blocking if lower-priority coroutines are active, thus mapping	1753	loop from blocking if lower-priority coroutines are active, thus mapping
1404	low-priority coroutines to idle/background tasks).	1754	low-priority coroutines to idle/background tasks).
1405		1755
		1756	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1757	priority, to ensure that they are being run before any other watchers
		1758	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1759	too) should not activate ("feed") events into libev. While libev fully
		1760	supports this, they will be called before other C<ev_check> watchers
		1761	did their job. As C<ev_check> watchers are often used to embed other
		1762	(non-libev) event loops those other event loops might be in an unusable
		1763	state until their C<ev_check> watcher ran (always remind yourself to
		1764	coexist peacefully with others).
		1765
		1766	=head3 Watcher-Specific Functions and Data Members
		1767
1406	=over 4	1768	=over 4
1407		1769
1408	=item ev_prepare_init (ev_prepare *, callback)	1770	=item ev_prepare_init (ev_prepare *, callback)
1409		1771
1410	=item ev_check_init (ev_check *, callback)	1772	=item ev_check_init (ev_check *, callback)
…		…
1413	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1775	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1414	macros, but using them is utterly, utterly and completely pointless.	1776	macros, but using them is utterly, utterly and completely pointless.
1415		1777
1416	=back	1778	=back
1417		1779
1418	Example: To include a library such as adns, you would add IO watchers	1780	=head3 Examples
1419	and a timeout watcher in a prepare handler, as required by libadns, and	1781
		1782	There are a number of principal ways to embed other event loops or modules
		1783	into libev. Here are some ideas on how to include libadns into libev
		1784	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1785	use for an actually working example. Another Perl module named C<EV::Glib>
		1786	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV
		1787	into the Glib event loop).
		1788
		1789	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1420	in a check watcher, destroy them and call into libadns. What follows is	1790	and in a check watcher, destroy them and call into libadns. What follows
1421	pseudo-code only of course:	1791	is pseudo-code only of course. This requires you to either use a low
		1792	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1793	the callbacks for the IO/timeout watchers might not have been called yet.
1422		1794
1423	static ev_io iow [nfd];	1795	static ev_io iow [nfd];
1424	static ev_timer tw;	1796	static ev_timer tw;
1425		1797
1426	static void	1798	static void
1427	io_cb (ev_loop *loop, ev_io *w, int revents)	1799	io_cb (ev_loop *loop, ev_io *w, int revents)
1428	{	1800	{
1429	// set the relevant poll flags
1430	// could also call adns_processreadable etc. here
1431	struct pollfd *fd = (struct pollfd *)w->data;
1432	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1433	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1434	}	1801	}
1435		1802
1436	// create io watchers for each fd and a timer before blocking	1803	// create io watchers for each fd and a timer before blocking
1437	static void	1804	static void
1438	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	1805	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
1439	{	1806	{
1440	int timeout = 3600000;truct pollfd fds [nfd];	1807	int timeout = 3600000;
		1808	struct pollfd fds [nfd];
1441	// actual code will need to loop here and realloc etc.	1809	// actual code will need to loop here and realloc etc.
1442	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	1810	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1443		1811
1444	/* the callback is illegal, but won't be called as we stop during check */	1812	/* the callback is illegal, but won't be called as we stop during check */
1445	ev_timer_init (&tw, 0, timeout * 1e-3);	1813	ev_timer_init (&tw, 0, timeout * 1e-3);
1446	ev_timer_start (loop, &tw);	1814	ev_timer_start (loop, &tw);
1447		1815
1448	// create on ev_io per pollfd	1816	// create one ev_io per pollfd
1449	for (int i = 0; i < nfd; ++i)	1817	for (int i = 0; i < nfd; ++i)
1450	{	1818	{
1451	ev_io_init (iow + i, io_cb, fds [i].fd,	1819	ev_io_init (iow + i, io_cb, fds [i].fd,
1452	((fds [i].events & POLLIN ? EV_READ : 0)	1820	((fds [i].events & POLLIN ? EV_READ : 0)
1453	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	1821	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1454		1822
1455	fds [i].revents = 0;	1823	fds [i].revents = 0;
1456	iow [i].data = fds + i;
1457	ev_io_start (loop, iow + i);	1824	ev_io_start (loop, iow + i);
1458	}	1825	}
1459	}	1826	}
1460		1827
1461	// stop all watchers after blocking	1828	// stop all watchers after blocking
…		…
1463	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	1830	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
1464	{	1831	{
1465	ev_timer_stop (loop, &tw);	1832	ev_timer_stop (loop, &tw);
1466		1833
1467	for (int i = 0; i < nfd; ++i)	1834	for (int i = 0; i < nfd; ++i)
		1835	{
		1836	// set the relevant poll flags
		1837	// could also call adns_processreadable etc. here
		1838	struct pollfd *fd = fds + i;
		1839	int revents = ev_clear_pending (iow + i);
		1840	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1841	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1842
		1843	// now stop the watcher
1468	ev_io_stop (loop, iow + i);	1844	ev_io_stop (loop, iow + i);
		1845	}
1469		1846
1470	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	1847	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1848	}
		1849
		1850	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		1851	in the prepare watcher and would dispose of the check watcher.
		1852
		1853	Method 3: If the module to be embedded supports explicit event
		1854	notification (adns does), you can also make use of the actual watcher
		1855	callbacks, and only destroy/create the watchers in the prepare watcher.
		1856
		1857	static void
		1858	timer_cb (EV_P_ ev_timer *w, int revents)
		1859	{
		1860	adns_state ads = (adns_state)w->data;
		1861	update_now (EV_A);
		1862
		1863	adns_processtimeouts (ads, &tv_now);
		1864	}
		1865
		1866	static void
		1867	io_cb (EV_P_ ev_io *w, int revents)
		1868	{
		1869	adns_state ads = (adns_state)w->data;
		1870	update_now (EV_A);
		1871
		1872	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		1873	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		1874	}
		1875
		1876	// do not ever call adns_afterpoll
		1877
		1878	Method 4: Do not use a prepare or check watcher because the module you
		1879	want to embed is too inflexible to support it. Instead, youc na override
		1880	their poll function. The drawback with this solution is that the main
		1881	loop is now no longer controllable by EV. The C<Glib::EV> module does
		1882	this.
		1883
		1884	static gint
		1885	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		1886	{
		1887	int got_events = 0;
		1888
		1889	for (n = 0; n < nfds; ++n)
		1890	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		1891
		1892	if (timeout >= 0)
		1893	// create/start timer
		1894
		1895	// poll
		1896	ev_loop (EV_A_ 0);
		1897
		1898	// stop timer again
		1899	if (timeout >= 0)
		1900	ev_timer_stop (EV_A_ &to);
		1901
		1902	// stop io watchers again - their callbacks should have set
		1903	for (n = 0; n < nfds; ++n)
		1904	ev_io_stop (EV_A_ iow [n]);
		1905
		1906	return got_events;
1471	}	1907	}
1472		1908
1473		1909
1474	=head2 C<ev_embed> - when one backend isn't enough...	1910	=head2 C<ev_embed> - when one backend isn't enough...
1475		1911
…		…
1518	portable one.	1954	portable one.
1519		1955
1520	So when you want to use this feature you will always have to be prepared	1956	So when you want to use this feature you will always have to be prepared
1521	that you cannot get an embeddable loop. The recommended way to get around	1957	that you cannot get an embeddable loop. The recommended way to get around
1522	this is to have a separate variables for your embeddable loop, try to	1958	this is to have a separate variables for your embeddable loop, try to
1523	create it, and if that fails, use the normal loop for everything:	1959	create it, and if that fails, use the normal loop for everything.
		1960
		1961	=head3 Watcher-Specific Functions and Data Members
		1962
		1963	=over 4
		1964
		1965	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		1966
		1967	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		1968
		1969	Configures the watcher to embed the given loop, which must be
		1970	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		1971	invoked automatically, otherwise it is the responsibility of the callback
		1972	to invoke it (it will continue to be called until the sweep has been done,
		1973	if you do not want thta, you need to temporarily stop the embed watcher).
		1974
		1975	=item ev_embed_sweep (loop, ev_embed *)
		1976
		1977	Make a single, non-blocking sweep over the embedded loop. This works
		1978	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		1979	apropriate way for embedded loops.
		1980
		1981	=item struct ev_loop *other [read-only]
		1982
		1983	The embedded event loop.
		1984
		1985	=back
		1986
		1987	=head3 Examples
		1988
		1989	Example: Try to get an embeddable event loop and embed it into the default
		1990	event loop. If that is not possible, use the default loop. The default
		1991	loop is stored in C<loop_hi>, while the mebeddable loop is stored in
		1992	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be
		1993	used).
1524		1994
1525	struct ev_loop *loop_hi = ev_default_init (0);	1995	struct ev_loop *loop_hi = ev_default_init (0);
1526	struct ev_loop *loop_lo = 0;	1996	struct ev_loop *loop_lo = 0;
1527	struct ev_embed embed;	1997	struct ev_embed embed;
1528		1998
…		…
1539	ev_embed_start (loop_hi, &embed);	2009	ev_embed_start (loop_hi, &embed);
1540	}	2010	}
1541	else	2011	else
1542	loop_lo = loop_hi;	2012	loop_lo = loop_hi;
1543		2013
1544	=over 4	2014	Example: Check if kqueue is available but not recommended and create
		2015	a kqueue backend for use with sockets (which usually work with any
		2016	kqueue implementation). Store the kqueue/socket-only event loop in
		2017	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
1545		2018
1546	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2019	struct ev_loop *loop = ev_default_init (0);
		2020	struct ev_loop *loop_socket = 0;
		2021	struct ev_embed embed;
		2022
		2023	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2024	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2025	{
		2026	ev_embed_init (&embed, 0, loop_socket);
		2027	ev_embed_start (loop, &embed);
		2028	}
1547		2029
1548	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	2030	if (!loop_socket)
		2031	loop_socket = loop;
1549		2032
1550	Configures the watcher to embed the given loop, which must be	2033	// now use loop_socket for all sockets, and loop for everything else
1551	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1552	invoked automatically, otherwise it is the responsibility of the callback
1553	to invoke it (it will continue to be called until the sweep has been done,
1554	if you do not want thta, you need to temporarily stop the embed watcher).
1555
1556	=item ev_embed_sweep (loop, ev_embed *)
1557
1558	Make a single, non-blocking sweep over the embedded loop. This works
1559	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1560	apropriate way for embedded loops.
1561
1562	=item struct ev_loop *loop [read-only]
1563
1564	The embedded event loop.
1565
1566	=back
1567		2034
1568		2035
1569	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	2036	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
1570		2037
1571	Fork watchers are called when a C<fork ()> was detected (usually because	2038	Fork watchers are called when a C<fork ()> was detected (usually because
…		…
1574	event loop blocks next and before C<ev_check> watchers are being called,	2041	event loop blocks next and before C<ev_check> watchers are being called,
1575	and only in the child after the fork. If whoever good citizen calling	2042	and only in the child after the fork. If whoever good citizen calling
1576	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2043	C<ev_default_fork> cheats and calls it in the wrong process, the fork
1577	handlers will be invoked, too, of course.	2044	handlers will be invoked, too, of course.
1578		2045
		2046	=head3 Watcher-Specific Functions and Data Members
		2047
1579	=over 4	2048	=over 4
1580		2049
1581	=item ev_fork_init (ev_signal *, callback)	2050	=item ev_fork_init (ev_signal *, callback)
1582		2051
1583	Initialises and configures the fork watcher - it has no parameters of any	2052	Initialises and configures the fork watcher - it has no parameters of any
1584	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	2053	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
1585	believe me.	2054	believe me.
		2055
		2056	=back
		2057
		2058
		2059	=head2 C<ev_async> - how to wake up another event loop
		2060
		2061	In general, you cannot use an C<ev_loop> from multiple threads or other
		2062	asynchronous sources such as signal handlers (as opposed to multiple event
		2063	loops - those are of course safe to use in different threads).
		2064
		2065	Sometimes, however, you need to wake up another event loop you do not
		2066	control, for example because it belongs to another thread. This is what
		2067	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		2068	can signal it by calling C<ev_async_send>, which is thread- and signal
		2069	safe.
		2070
		2071	This functionality is very similar to C<ev_signal> watchers, as signals,
		2072	too, are asynchronous in nature, and signals, too, will be compressed
		2073	(i.e. the number of callback invocations may be less than the number of
		2074	C<ev_async_sent> calls).
		2075
		2076	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		2077	just the default loop.
		2078
		2079	=head3 Queueing
		2080
		2081	C<ev_async> does not support queueing of data in any way. The reason
		2082	is that the author does not know of a simple (or any) algorithm for a
		2083	multiple-writer-single-reader queue that works in all cases and doesn't
		2084	need elaborate support such as pthreads.
		2085
		2086	That means that if you want to queue data, you have to provide your own
		2087	queue. But at least I can tell you would implement locking around your
		2088	queue:
		2089
		2090	=over 4
		2091
		2092	=item queueing from a signal handler context
		2093
		2094	To implement race-free queueing, you simply add to the queue in the signal
		2095	handler but you block the signal handler in the watcher callback. Here is an example that does that for
		2096	some fictitiuous SIGUSR1 handler:
		2097
		2098	static ev_async mysig;
		2099
		2100	static void
		2101	sigusr1_handler (void)
		2102	{
		2103	sometype data;
		2104
		2105	// no locking etc.
		2106	queue_put (data);
		2107	ev_async_send (EV_DEFAULT_ &mysig);
		2108	}
		2109
		2110	static void
		2111	mysig_cb (EV_P_ ev_async *w, int revents)
		2112	{
		2113	sometype data;
		2114	sigset_t block, prev;
		2115
		2116	sigemptyset (&block);
		2117	sigaddset (&block, SIGUSR1);
		2118	sigprocmask (SIG_BLOCK, &block, &prev);
		2119
		2120	while (queue_get (&data))
		2121	process (data);
		2122
		2123	if (sigismember (&prev, SIGUSR1)
		2124	sigprocmask (SIG_UNBLOCK, &block, 0);
		2125	}
		2126
		2127	(Note: pthreads in theory requires you to use C<pthread_setmask>
		2128	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		2129	either...).
		2130
		2131	=item queueing from a thread context
		2132
		2133	The strategy for threads is different, as you cannot (easily) block
		2134	threads but you can easily preempt them, so to queue safely you need to
		2135	employ a traditional mutex lock, such as in this pthread example:
		2136
		2137	static ev_async mysig;
		2138	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		2139
		2140	static void
		2141	otherthread (void)
		2142	{
		2143	// only need to lock the actual queueing operation
		2144	pthread_mutex_lock (&mymutex);
		2145	queue_put (data);
		2146	pthread_mutex_unlock (&mymutex);
		2147
		2148	ev_async_send (EV_DEFAULT_ &mysig);
		2149	}
		2150
		2151	static void
		2152	mysig_cb (EV_P_ ev_async *w, int revents)
		2153	{
		2154	pthread_mutex_lock (&mymutex);
		2155
		2156	while (queue_get (&data))
		2157	process (data);
		2158
		2159	pthread_mutex_unlock (&mymutex);
		2160	}
		2161
		2162	=back
		2163
		2164
		2165	=head3 Watcher-Specific Functions and Data Members
		2166
		2167	=over 4
		2168
		2169	=item ev_async_init (ev_async *, callback)
		2170
		2171	Initialises and configures the async watcher - it has no parameters of any
		2172	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,
		2173	believe me.
		2174
		2175	=item ev_async_send (loop, ev_async *)
		2176
		2177	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		2178	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		2179	C<ev_feed_event>, this call is safe to do in other threads, signal or
		2180	similar contexts (see the dicusssion of C<EV_ATOMIC_T> in the embedding
		2181	section below on what exactly this means).
		2182
		2183	This call incurs the overhead of a syscall only once per loop iteration,
		2184	so while the overhead might be noticable, it doesn't apply to repeated
		2185	calls to C<ev_async_send>.
1586		2186
1587	=back	2187	=back
1588		2188
1589		2189
1590	=head1 OTHER FUNCTIONS	2190	=head1 OTHER FUNCTIONS
…		…
1679		2279
1680	To use it,	2280	To use it,
1681		2281
1682	#include <ev++.h>	2282	#include <ev++.h>
1683		2283
1684	(it is not installed by default). This automatically includes F<ev.h>	2284	This automatically includes F<ev.h> and puts all of its definitions (many
1685	and puts all of its definitions (many of them macros) into the global	2285	of them macros) into the global namespace. All C++ specific things are
1686	namespace. All C++ specific things are put into the C<ev> namespace.	2286	put into the C<ev> namespace. It should support all the same embedding
		2287	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1687		2288
1688	It should support all the same embedding options as F<ev.h>, most notably	2289	Care has been taken to keep the overhead low. The only data member the C++
1689	C<EV_MULTIPLICITY>.	2290	classes add (compared to plain C-style watchers) is the event loop pointer
		2291	that the watcher is associated with (or no additional members at all if
		2292	you disable C<EV_MULTIPLICITY> when embedding libev).
		2293
		2294	Currently, functions, and static and non-static member functions can be
		2295	used as callbacks. Other types should be easy to add as long as they only
		2296	need one additional pointer for context. If you need support for other
		2297	types of functors please contact the author (preferably after implementing
		2298	it).
1690		2299
1691	Here is a list of things available in the C<ev> namespace:	2300	Here is a list of things available in the C<ev> namespace:
1692		2301
1693	=over 4	2302	=over 4
1694		2303
…		…
1710		2319
1711	All of those classes have these methods:	2320	All of those classes have these methods:
1712		2321
1713	=over 4	2322	=over 4
1714		2323
1715	=item ev::TYPE::TYPE (object *, object::method *)	2324	=item ev::TYPE::TYPE ()
1716		2325
1717	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)	2326	=item ev::TYPE::TYPE (struct ev_loop *)
1718		2327
1719	=item ev::TYPE::~TYPE	2328	=item ev::TYPE::~TYPE
1720		2329
1721	The constructor takes a pointer to an object and a method pointer to	2330	The constructor (optionally) takes an event loop to associate the watcher
1722	the event handler callback to call in this class. The constructor calls	2331	with. If it is omitted, it will use C<EV_DEFAULT>.
1723	C<ev_init> for you, which means you have to call the C<set> method	2332
1724	before starting it. If you do not specify a loop then the constructor	2333	The constructor calls C<ev_init> for you, which means you have to call the
1725	automatically associates the default loop with this watcher.	2334	C<set> method before starting it.
		2335
		2336	It will not set a callback, however: You have to call the templated C<set>
		2337	method to set a callback before you can start the watcher.
		2338
		2339	(The reason why you have to use a method is a limitation in C++ which does
		2340	not allow explicit template arguments for constructors).
1726		2341
1727	The destructor automatically stops the watcher if it is active.	2342	The destructor automatically stops the watcher if it is active.
		2343
		2344	=item w->set<class, &class::method> (object *)
		2345
		2346	This method sets the callback method to call. The method has to have a
		2347	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		2348	first argument and the C<revents> as second. The object must be given as
		2349	parameter and is stored in the C<data> member of the watcher.
		2350
		2351	This method synthesizes efficient thunking code to call your method from
		2352	the C callback that libev requires. If your compiler can inline your
		2353	callback (i.e. it is visible to it at the place of the C<set> call and
		2354	your compiler is good :), then the method will be fully inlined into the
		2355	thunking function, making it as fast as a direct C callback.
		2356
		2357	Example: simple class declaration and watcher initialisation
		2358
		2359	struct myclass
		2360	{
		2361	void io_cb (ev::io &w, int revents) { }
		2362	}
		2363
		2364	myclass obj;
		2365	ev::io iow;
		2366	iow.set <myclass, &myclass::io_cb> (&obj);
		2367
		2368	=item w->set<function> (void *data = 0)
		2369
		2370	Also sets a callback, but uses a static method or plain function as
		2371	callback. The optional C<data> argument will be stored in the watcher's
		2372	C<data> member and is free for you to use.
		2373
		2374	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2375
		2376	See the method-C<set> above for more details.
		2377
		2378	Example:
		2379
		2380	static void io_cb (ev::io &w, int revents) { }
		2381	iow.set <io_cb> ();
1728		2382
1729	=item w->set (struct ev_loop *)	2383	=item w->set (struct ev_loop *)
1730		2384
1731	Associates a different C<struct ev_loop> with this watcher. You can only	2385	Associates a different C<struct ev_loop> with this watcher. You can only
1732	do this when the watcher is inactive (and not pending either).	2386	do this when the watcher is inactive (and not pending either).
1733		2387
1734	=item w->set ([args])	2388	=item w->set ([args])
1735		2389
1736	Basically the same as C<ev_TYPE_set>, with the same args. Must be	2390	Basically the same as C<ev_TYPE_set>, with the same args. Must be
1737	called at least once. Unlike the C counterpart, an active watcher gets	2391	called at least once. Unlike the C counterpart, an active watcher gets
1738	automatically stopped and restarted.	2392	automatically stopped and restarted when reconfiguring it with this
		2393	method.
1739		2394
1740	=item w->start ()	2395	=item w->start ()
1741		2396
1742	Starts the watcher. Note that there is no C<loop> argument as the	2397	Starts the watcher. Note that there is no C<loop> argument, as the
1743	constructor already takes the loop.	2398	constructor already stores the event loop.
1744		2399
1745	=item w->stop ()	2400	=item w->stop ()
1746		2401
1747	Stops the watcher if it is active. Again, no C<loop> argument.	2402	Stops the watcher if it is active. Again, no C<loop> argument.
1748		2403
1749	=item w->again () C<ev::timer>, C<ev::periodic> only	2404	=item w->again () (C<ev::timer>, C<ev::periodic> only)
1750		2405
1751	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding	2406	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
1752	C<ev_TYPE_again> function.	2407	C<ev_TYPE_again> function.
1753		2408
1754	=item w->sweep () C<ev::embed> only	2409	=item w->sweep () (C<ev::embed> only)
1755		2410
1756	Invokes C<ev_embed_sweep>.	2411	Invokes C<ev_embed_sweep>.
1757		2412
1758	=item w->update () C<ev::stat> only	2413	=item w->update () (C<ev::stat> only)
1759		2414
1760	Invokes C<ev_stat_stat>.	2415	Invokes C<ev_stat_stat>.
1761		2416
1762	=back	2417	=back
1763		2418
…		…
1766	Example: Define a class with an IO and idle watcher, start one of them in	2421	Example: Define a class with an IO and idle watcher, start one of them in
1767	the constructor.	2422	the constructor.
1768		2423
1769	class myclass	2424	class myclass
1770	{	2425	{
1771	ev_io io; void io_cb (ev::io &w, int revents);	2426	ev::io io; void io_cb (ev::io &w, int revents);
1772	ev_idle idle void idle_cb (ev::idle &w, int revents);	2427	ev:idle idle void idle_cb (ev::idle &w, int revents);
1773		2428
1774	myclass ();	2429	myclass (int fd)
1775	}
1776
1777	myclass::myclass (int fd)
1778	: io (this, &myclass::io_cb),
1779	idle (this, &myclass::idle_cb)
1780	{	2430	{
		2431	io .set <myclass, &myclass::io_cb > (this);
		2432	idle.set <myclass, &myclass::idle_cb> (this);
		2433
1781	io.start (fd, ev::READ);	2434	io.start (fd, ev::READ);
		2435	}
1782	}	2436	};
1783		2437
1784		2438
1785	=head1 MACRO MAGIC	2439	=head1 MACRO MAGIC
1786		2440
1787	Libev can be compiled with a variety of options, the most fundemantal is	2441	Libev can be compiled with a variety of options, the most fundamantal
1788	C<EV_MULTIPLICITY>. This option determines wether (most) functions and	2442	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
1789	callbacks have an initial C<struct ev_loop *> argument.	2443	functions and callbacks have an initial C<struct ev_loop *> argument.
1790		2444
1791	To make it easier to write programs that cope with either variant, the	2445	To make it easier to write programs that cope with either variant, the
1792	following macros are defined:	2446	following macros are defined:
1793		2447
1794	=over 4	2448	=over 4
…		…
1826	Similar to the other two macros, this gives you the value of the default	2480	Similar to the other two macros, this gives you the value of the default
1827	loop, if multiple loops are supported ("ev loop default").	2481	loop, if multiple loops are supported ("ev loop default").
1828		2482
1829	=back	2483	=back
1830		2484
1831	Example: Declare and initialise a check watcher, working regardless of	2485	Example: Declare and initialise a check watcher, utilising the above
1832	wether multiple loops are supported or not.	2486	macros so it will work regardless of whether multiple loops are supported
		2487	or not.
1833		2488
1834	static void	2489	static void
1835	check_cb (EV_P_ ev_timer *w, int revents)	2490	check_cb (EV_P_ ev_timer *w, int revents)
1836	{	2491	{
1837	ev_check_stop (EV_A_ w);	2492	ev_check_stop (EV_A_ w);
…		…
1840	ev_check check;	2495	ev_check check;
1841	ev_check_init (&check, check_cb);	2496	ev_check_init (&check, check_cb);
1842	ev_check_start (EV_DEFAULT_ &check);	2497	ev_check_start (EV_DEFAULT_ &check);
1843	ev_loop (EV_DEFAULT_ 0);	2498	ev_loop (EV_DEFAULT_ 0);
1844		2499
1845
1846	=head1 EMBEDDING	2500	=head1 EMBEDDING
1847		2501
1848	Libev can (and often is) directly embedded into host	2502	Libev can (and often is) directly embedded into host
1849	applications. Examples of applications that embed it include the Deliantra	2503	applications. Examples of applications that embed it include the Deliantra
1850	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)	2504	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
1851	and rxvt-unicode.	2505	and rxvt-unicode.
1852		2506
1853	The goal is to enable you to just copy the neecssary files into your	2507	The goal is to enable you to just copy the necessary files into your
1854	source directory without having to change even a single line in them, so	2508	source directory without having to change even a single line in them, so
1855	you can easily upgrade by simply copying (or having a checked-out copy of	2509	you can easily upgrade by simply copying (or having a checked-out copy of
1856	libev somewhere in your source tree).	2510	libev somewhere in your source tree).
1857		2511
1858	=head2 FILESETS	2512	=head2 FILESETS
…		…
1889	ev_vars.h	2543	ev_vars.h
1890	ev_wrap.h	2544	ev_wrap.h
1891		2545
1892	ev_win32.c required on win32 platforms only	2546	ev_win32.c required on win32 platforms only
1893		2547
1894	ev_select.c only when select backend is enabled (which is by default)	2548	ev_select.c only when select backend is enabled (which is enabled by default)
1895	ev_poll.c only when poll backend is enabled (disabled by default)	2549	ev_poll.c only when poll backend is enabled (disabled by default)
1896	ev_epoll.c only when the epoll backend is enabled (disabled by default)	2550	ev_epoll.c only when the epoll backend is enabled (disabled by default)
1897	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	2551	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1898	ev_port.c only when the solaris port backend is enabled (disabled by default)	2552	ev_port.c only when the solaris port backend is enabled (disabled by default)
1899		2553
…		…
1948		2602
1949	If defined to be C<1>, libev will try to detect the availability of the	2603	If defined to be C<1>, libev will try to detect the availability of the
1950	monotonic clock option at both compiletime and runtime. Otherwise no use	2604	monotonic clock option at both compiletime and runtime. Otherwise no use
1951	of the monotonic clock option will be attempted. If you enable this, you	2605	of the monotonic clock option will be attempted. If you enable this, you
1952	usually have to link against librt or something similar. Enabling it when	2606	usually have to link against librt or something similar. Enabling it when
1953	the functionality isn't available is safe, though, althoguh you have	2607	the functionality isn't available is safe, though, although you have
1954	to make sure you link against any libraries where the C<clock_gettime>	2608	to make sure you link against any libraries where the C<clock_gettime>
1955	function is hiding in (often F<-lrt>).	2609	function is hiding in (often F<-lrt>).
1956		2610
1957	=item EV_USE_REALTIME	2611	=item EV_USE_REALTIME
1958		2612
1959	If defined to be C<1>, libev will try to detect the availability of the	2613	If defined to be C<1>, libev will try to detect the availability of the
1960	realtime clock option at compiletime (and assume its availability at	2614	realtime clock option at compiletime (and assume its availability at
1961	runtime if successful). Otherwise no use of the realtime clock option will	2615	runtime if successful). Otherwise no use of the realtime clock option will
1962	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	2616	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
1963	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries	2617	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
1964	in the description of C<EV_USE_MONOTONIC>, though.	2618	note about libraries in the description of C<EV_USE_MONOTONIC>, though.
		2619
		2620	=item EV_USE_NANOSLEEP
		2621
		2622	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		2623	and will use it for delays. Otherwise it will use C<select ()>.
1965		2624
1966	=item EV_USE_SELECT	2625	=item EV_USE_SELECT
1967		2626
1968	If undefined or defined to be C<1>, libev will compile in support for the	2627	If undefined or defined to be C<1>, libev will compile in support for the
1969	C<select>(2) backend. No attempt at autodetection will be done: if no	2628	C<select>(2) backend. No attempt at autodetection will be done: if no
…		…
1987	wants osf handles on win32 (this is the case when the select to	2646	wants osf handles on win32 (this is the case when the select to
1988	be used is the winsock select). This means that it will call	2647	be used is the winsock select). This means that it will call
1989	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	2648	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
1990	it is assumed that all these functions actually work on fds, even	2649	it is assumed that all these functions actually work on fds, even
1991	on win32. Should not be defined on non-win32 platforms.	2650	on win32. Should not be defined on non-win32 platforms.
		2651
		2652	=item EV_FD_TO_WIN32_HANDLE
		2653
		2654	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		2655	file descriptors to socket handles. When not defining this symbol (the
		2656	default), then libev will call C<_get_osfhandle>, which is usually
		2657	correct. In some cases, programs use their own file descriptor management,
		2658	in which case they can provide this function to map fds to socket handles.
1992		2659
1993	=item EV_USE_POLL	2660	=item EV_USE_POLL
1994		2661
1995	If defined to be C<1>, libev will compile in support for the C<poll>(2)	2662	If defined to be C<1>, libev will compile in support for the C<poll>(2)
1996	backend. Otherwise it will be enabled on non-win32 platforms. It	2663	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
2030		2697
2031	If defined to be C<1>, libev will compile in support for the Linux inotify	2698	If defined to be C<1>, libev will compile in support for the Linux inotify
2032	interface to speed up C<ev_stat> watchers. Its actual availability will	2699	interface to speed up C<ev_stat> watchers. Its actual availability will
2033	be detected at runtime.	2700	be detected at runtime.
2034		2701
		2702	=item EV_ATOMIC_T
		2703
		2704	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		2705	access is atomic with respect to other threads or signal contexts. No such
		2706	type is easily found in the C language, so you can provide your own type
		2707	that you know is safe for your purposes. It is used both for signal handler "locking"
		2708	as well as for signal and thread safety in C<ev_async> watchers.
		2709
		2710	In the absense of this define, libev will use C<sig_atomic_t volatile>
		2711	(from F<signal.h>), which is usually good enough on most platforms.
		2712
2035	=item EV_H	2713	=item EV_H
2036		2714
2037	The name of the F<ev.h> header file used to include it. The default if	2715	The name of the F<ev.h> header file used to include it. The default if
2038	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This	2716	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
2039	can be used to virtually rename the F<ev.h> header file in case of conflicts.	2717	used to virtually rename the F<ev.h> header file in case of conflicts.
2040		2718
2041	=item EV_CONFIG_H	2719	=item EV_CONFIG_H
2042		2720
2043	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	2721	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2044	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	2722	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2045	C<EV_H>, above.	2723	C<EV_H>, above.
2046		2724
2047	=item EV_EVENT_H	2725	=item EV_EVENT_H
2048		2726
2049	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	2727	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2050	of how the F<event.h> header can be found.	2728	of how the F<event.h> header can be found, the default is C<"event.h">.
2051		2729
2052	=item EV_PROTOTYPES	2730	=item EV_PROTOTYPES
2053		2731
2054	If defined to be C<0>, then F<ev.h> will not define any function	2732	If defined to be C<0>, then F<ev.h> will not define any function
2055	prototypes, but still define all the structs and other symbols. This is	2733	prototypes, but still define all the structs and other symbols. This is
…		…
2062	will have the C<struct ev_loop *> as first argument, and you can create	2740	will have the C<struct ev_loop *> as first argument, and you can create
2063	additional independent event loops. Otherwise there will be no support	2741	additional independent event loops. Otherwise there will be no support
2064	for multiple event loops and there is no first event loop pointer	2742	for multiple event loops and there is no first event loop pointer
2065	argument. Instead, all functions act on the single default loop.	2743	argument. Instead, all functions act on the single default loop.
2066		2744
		2745	=item EV_MINPRI
		2746
		2747	=item EV_MAXPRI
		2748
		2749	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2750	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2751	provide for more priorities by overriding those symbols (usually defined
		2752	to be C<-2> and C<2>, respectively).
		2753
		2754	When doing priority-based operations, libev usually has to linearly search
		2755	all the priorities, so having many of them (hundreds) uses a lot of space
		2756	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2757	fine.
		2758
		2759	If your embedding app does not need any priorities, defining these both to
		2760	C<0> will save some memory and cpu.
		2761
2067	=item EV_PERIODIC_ENABLE	2762	=item EV_PERIODIC_ENABLE
2068		2763
2069	If undefined or defined to be C<1>, then periodic timers are supported. If	2764	If undefined or defined to be C<1>, then periodic timers are supported. If
2070	defined to be C<0>, then they are not. Disabling them saves a few kB of	2765	defined to be C<0>, then they are not. Disabling them saves a few kB of
2071	code.	2766	code.
2072		2767
		2768	=item EV_IDLE_ENABLE
		2769
		2770	If undefined or defined to be C<1>, then idle watchers are supported. If
		2771	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2772	code.
		2773
2073	=item EV_EMBED_ENABLE	2774	=item EV_EMBED_ENABLE
2074		2775
2075	If undefined or defined to be C<1>, then embed watchers are supported. If	2776	If undefined or defined to be C<1>, then embed watchers are supported. If
2076	defined to be C<0>, then they are not.	2777	defined to be C<0>, then they are not.
2077		2778
…		…
2081	defined to be C<0>, then they are not.	2782	defined to be C<0>, then they are not.
2082		2783
2083	=item EV_FORK_ENABLE	2784	=item EV_FORK_ENABLE
2084		2785
2085	If undefined or defined to be C<1>, then fork watchers are supported. If	2786	If undefined or defined to be C<1>, then fork watchers are supported. If
		2787	defined to be C<0>, then they are not.
		2788
		2789	=item EV_ASYNC_ENABLE
		2790
		2791	If undefined or defined to be C<1>, then async watchers are supported. If
2086	defined to be C<0>, then they are not.	2792	defined to be C<0>, then they are not.
2087		2793
2088	=item EV_MINIMAL	2794	=item EV_MINIMAL
2089		2795
2090	If you need to shave off some kilobytes of code at the expense of some	2796	If you need to shave off some kilobytes of code at the expense of some
…		…
2098	than enough. If you need to manage thousands of children you might want to	2804	than enough. If you need to manage thousands of children you might want to
2099	increase this value (I<must> be a power of two).	2805	increase this value (I<must> be a power of two).
2100		2806
2101	=item EV_INOTIFY_HASHSIZE	2807	=item EV_INOTIFY_HASHSIZE
2102		2808
2103	C<ev_staz> watchers use a small hash table to distribute workload by	2809	C<ev_stat> watchers use a small hash table to distribute workload by
2104	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	2810	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
2105	usually more than enough. If you need to manage thousands of C<ev_stat>	2811	usually more than enough. If you need to manage thousands of C<ev_stat>
2106	watchers you might want to increase this value (I<must> be a power of	2812	watchers you might want to increase this value (I<must> be a power of
2107	two).	2813	two).
2108		2814
…		…
2125		2831
2126	=item ev_set_cb (ev, cb)	2832	=item ev_set_cb (ev, cb)
2127		2833
2128	Can be used to change the callback member declaration in each watcher,	2834	Can be used to change the callback member declaration in each watcher,
2129	and the way callbacks are invoked and set. Must expand to a struct member	2835	and the way callbacks are invoked and set. Must expand to a struct member
2130	definition and a statement, respectively. See the F<ev.v> header file for	2836	definition and a statement, respectively. See the F<ev.h> header file for
2131	their default definitions. One possible use for overriding these is to	2837	their default definitions. One possible use for overriding these is to
2132	avoid the C<struct ev_loop *> as first argument in all cases, or to use	2838	avoid the C<struct ev_loop *> as first argument in all cases, or to use
2133	method calls instead of plain function calls in C++.	2839	method calls instead of plain function calls in C++.
		2840
		2841	=head2 EXPORTED API SYMBOLS
		2842
		2843	If you need to re-export the API (e.g. via a dll) and you need a list of
		2844	exported symbols, you can use the provided F<Symbol.*> files which list
		2845	all public symbols, one per line:
		2846
		2847	Symbols.ev for libev proper
		2848	Symbols.event for the libevent emulation
		2849
		2850	This can also be used to rename all public symbols to avoid clashes with
		2851	multiple versions of libev linked together (which is obviously bad in
		2852	itself, but sometimes it is inconvinient to avoid this).
		2853
		2854	A sed command like this will create wrapper C<#define>'s that you need to
		2855	include before including F<ev.h>:
		2856
		2857	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		2858
		2859	This would create a file F<wrap.h> which essentially looks like this:
		2860
		2861	#define ev_backend myprefix_ev_backend
		2862	#define ev_check_start myprefix_ev_check_start
		2863	#define ev_check_stop myprefix_ev_check_stop
		2864	...
2134		2865
2135	=head2 EXAMPLES	2866	=head2 EXAMPLES
2136		2867
2137	For a real-world example of a program the includes libev	2868	For a real-world example of a program the includes libev
2138	verbatim, you can have a look at the EV perl module	2869	verbatim, you can have a look at the EV perl module
…		…
2141	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file	2872	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
2142	will be compiled. It is pretty complex because it provides its own header	2873	will be compiled. It is pretty complex because it provides its own header
2143	file.	2874	file.
2144		2875
2145	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	2876	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2146	that everybody includes and which overrides some autoconf choices:	2877	that everybody includes and which overrides some configure choices:
2147		2878
		2879	#define EV_MINIMAL 1
2148	#define EV_USE_POLL 0	2880	#define EV_USE_POLL 0
2149	#define EV_MULTIPLICITY 0	2881	#define EV_MULTIPLICITY 0
2150	#define EV_PERIODICS 0	2882	#define EV_PERIODIC_ENABLE 0
		2883	#define EV_STAT_ENABLE 0
		2884	#define EV_FORK_ENABLE 0
2151	#define EV_CONFIG_H <config.h>	2885	#define EV_CONFIG_H <config.h>
		2886	#define EV_MINPRI 0
		2887	#define EV_MAXPRI 0
2152		2888
2153	#include "ev++.h"	2889	#include "ev++.h"
2154		2890
2155	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	2891	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2156		2892
…		…
2162		2898
2163	In this section the complexities of (many of) the algorithms used inside	2899	In this section the complexities of (many of) the algorithms used inside
2164	libev will be explained. For complexity discussions about backends see the	2900	libev will be explained. For complexity discussions about backends see the
2165	documentation for C<ev_default_init>.	2901	documentation for C<ev_default_init>.
2166		2902
		2903	All of the following are about amortised time: If an array needs to be
		2904	extended, libev needs to realloc and move the whole array, but this
		2905	happens asymptotically never with higher number of elements, so O(1) might
		2906	mean it might do a lengthy realloc operation in rare cases, but on average
		2907	it is much faster and asymptotically approaches constant time.
		2908
2167	=over 4	2909	=over 4
2168		2910
2169	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	2911	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2170		2912
		2913	This means that, when you have a watcher that triggers in one hour and
		2914	there are 100 watchers that would trigger before that then inserting will
		2915	have to skip roughly seven (C<ld 100>) of these watchers.
		2916
2171	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)	2917	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
2172		2918
		2919	That means that changing a timer costs less than removing/adding them
		2920	as only the relative motion in the event queue has to be paid for.
		2921
2173	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	2922	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
2174		2923
		2924	These just add the watcher into an array or at the head of a list.
		2925
2175	=item Stopping check/prepare/idle watchers: O(1)	2926	=item Stopping check/prepare/idle/fork/async watchers: O(1)
2176		2927
2177	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	2928	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
2178		2929
		2930	These watchers are stored in lists then need to be walked to find the
		2931	correct watcher to remove. The lists are usually short (you don't usually
		2932	have many watchers waiting for the same fd or signal).
		2933
2179	=item Finding the next timer per loop iteration: O(1)	2934	=item Finding the next timer in each loop iteration: O(1)
		2935
		2936	By virtue of using a binary heap, the next timer is always found at the
		2937	beginning of the storage array.
2180		2938
2181	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	2939	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2182		2940
2183	=item Activating one watcher: O(1)	2941	A change means an I/O watcher gets started or stopped, which requires
		2942	libev to recalculate its status (and possibly tell the kernel, depending
		2943	on backend and wether C<ev_io_set> was used).
		2944
		2945	=item Activating one watcher (putting it into the pending state): O(1)
		2946
		2947	=item Priority handling: O(number_of_priorities)
		2948
		2949	Priorities are implemented by allocating some space for each
		2950	priority. When doing priority-based operations, libev usually has to
		2951	linearly search all the priorities, but starting/stopping and activating
		2952	watchers becomes O(1) w.r.t. priority handling.
		2953
		2954	=item Sending an ev_async: O(1)
		2955
		2956	=item Processing ev_async_send: O(number_of_async_watchers)
		2957
		2958	=item Processing signals: O(max_signal_number)
		2959
		2960	Sending involves a syscall I<iff> there were no other C<ev_async_send>
		2961	calls in the current loop iteration. Checking for async and signal events
		2962	involves iterating over all running async watchers or all signal numbers.
2184		2963
2185	=back	2964	=back
2186		2965
2187		2966
		2967	=head1 Win32 platform limitations and workarounds
		2968
		2969	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		2970	requires, and its I/O model is fundamentally incompatible with the POSIX
		2971	model. Libev still offers limited functionality on this platform in
		2972	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		2973	descriptors. This only applies when using Win32 natively, not when using
		2974	e.g. cygwin.
		2975
		2976	There is no supported compilation method available on windows except
		2977	embedding it into other applications.
		2978
		2979	Due to the many, low, and arbitrary limits on the win32 platform and the
		2980	abysmal performance of winsockets, using a large number of sockets is not
		2981	recommended (and not reasonable). If your program needs to use more than
		2982	a hundred or so sockets, then likely it needs to use a totally different
		2983	implementation for windows, as libev offers the POSIX model, which cannot
		2984	be implemented efficiently on windows (microsoft monopoly games).
		2985
		2986	=over 4
		2987
		2988	=item The winsocket select function
		2989
		2990	The winsocket C<select> function doesn't follow POSIX in that it requires
		2991	socket I<handles> and not socket I<file descriptors>. This makes select
		2992	very inefficient, and also requires a mapping from file descriptors
		2993	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,
		2994	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor
		2995	symbols for more info.
		2996
		2997	The configuration for a "naked" win32 using the microsoft runtime
		2998	libraries and raw winsocket select is:
		2999
		3000	#define EV_USE_SELECT 1
		3001	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		3002
		3003	Note that winsockets handling of fd sets is O(n), so you can easily get a
		3004	complexity in the O(n²) range when using win32.
		3005
		3006	=item Limited number of file descriptors
		3007
		3008	Windows has numerous arbitrary (and low) limits on things. Early versions
		3009	of winsocket's select only supported waiting for a max. of C<64> handles
		3010	(probably owning to the fact that all windows kernels can only wait for
		3011	C<64> things at the same time internally; microsoft recommends spawning a
		3012	chain of threads and wait for 63 handles and the previous thread in each).
		3013
		3014	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		3015	to some high number (e.g. C<2048>) before compiling the winsocket select
		3016	call (which might be in libev or elsewhere, for example, perl does its own
		3017	select emulation on windows).
		3018
		3019	Another limit is the number of file descriptors in the microsoft runtime
		3020	libraries, which by default is C<64> (there must be a hidden I<64> fetish
		3021	or something like this inside microsoft). You can increase this by calling
		3022	C<_setmaxstdio>, which can increase this limit to C<2048> (another
		3023	arbitrary limit), but is broken in many versions of the microsoft runtime
		3024	libraries.
		3025
		3026	This might get you to about C<512> or C<2048> sockets (depending on
		3027	windows version and/or the phase of the moon). To get more, you need to
		3028	wrap all I/O functions and provide your own fd management, but the cost of
		3029	calling select (O(n²)) will likely make this unworkable.
		3030
		3031	=back
		3032
		3033
2188	=head1 AUTHOR	3034	=head1 AUTHOR
2189		3035
2190	Marc Lehmann <libev@schmorp.de>.	3036	Marc Lehmann <libev@schmorp.de>.
2191		3037

Diff Legend

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