[ViewVC] Diff of: cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.53 by root, Tue Nov 27 20:15:02 2007 UTC vs.
Revision 1.88 by ayin, Tue Dec 18 13:06:18 2007 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	/* this is the only header you need */
8	#include <ev.h>	7	#include <ev.h>
9		8
10	/* what follows is a fully working example program */	9	=head1 EXAMPLE PROGRAM
		10
		11	#include <ev.h>
		12
11	ev_io stdin_watcher;	13	ev_io stdin_watcher;
12	ev_timer timeout_watcher;	14	ev_timer timeout_watcher;
13		15
14	/* called when data readable on stdin */	16	/* called when data readable on stdin */
15	static void	17	static void
…		…
45		47
46	return 0;	48	return 0;
47	}	49	}
48		50
49	=head1 DESCRIPTION	51	=head1 DESCRIPTION
		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>.
50		56
51	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
52	file descriptor being readable or a timeout occuring), and it will manage	58	file descriptor being readable or a timeout occuring), and it will manage
53	these event sources and provide your program with events.	59	these event sources and provide your program with events.
54		60
…		…
61	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
62	watcher.	68	watcher.
63		69
64	=head1 FEATURES	70	=head1 FEATURES
65		71
66	Libev supports select, poll, the linux-specific epoll and the bsd-specific	72	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
67	kqueue mechanisms for file descriptor events, relative timers, absolute	73	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
68	timers with customised rescheduling, signal events, process status change	74	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
69	events (related to SIGCHLD), and event watchers dealing with the event	75	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
70	loop mechanism itself (idle, prepare and check watchers). It also is quite	76	with customised rescheduling (C<ev_periodic>), synchronous signals
		77	(C<ev_signal>), process status change events (C<ev_child>), and event
		78	watchers dealing with the event loop mechanism itself (C<ev_idle>,
		79	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as
		80	file watchers (C<ev_stat>) and even limited support for fork events
		81	(C<ev_fork>).
		82
		83	It also is quite fast (see this
71	fast (see this L<benchmark\|http://libev.schmorp.de/bench.html> comparing	84	L<benchmark\|http://libev.schmorp.de/bench.html> comparing it to libevent
72	it to libevent for example).	85	for example).
73		86
74	=head1 CONVENTIONS	87	=head1 CONVENTIONS
75		88
76	Libev is very configurable. In this manual the default configuration	89	Libev is very configurable. In this manual the default configuration will
77	will be described, which supports multiple event loops. For more info	90	be described, which supports multiple event loops. For more info about
78	about various configuration options please have a look at the file	91	various configuration options please have a look at B<EMBED> section in
79	F<README.embed> in the libev distribution. If libev was configured without	92	this manual. If libev was configured without support for multiple event
80	support for multiple event loops, then all functions taking an initial	93	loops, then all functions taking an initial argument of name C<loop>
81	argument of name C<loop> (which is always of type C<struct ev_loop *>)	94	(which is always of type C<struct ev_loop *>) will not have this argument.
82	will not have this argument.
83		95
84	=head1 TIME REPRESENTATION	96	=head1 TIME REPRESENTATION
85		97
86	Libev represents time as a single floating point number, representing the	98	Libev represents time as a single floating point number, representing the
87	(fractional) number of seconds since the (POSIX) epoch (somewhere near	99	(fractional) number of seconds since the (POSIX) epoch (somewhere near
88	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
89	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
90	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
91	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.
92		106
93	=head1 GLOBAL FUNCTIONS	107	=head1 GLOBAL FUNCTIONS
94		108
95	These functions can be called anytime, even before initialising the	109	These functions can be called anytime, even before initialising the
96	library in any way.	110	library in any way.
…		…
105		119
106	=item int ev_version_major ()	120	=item int ev_version_major ()
107		121
108	=item int ev_version_minor ()	122	=item int ev_version_minor ()
109		123
110	You can find out the major and minor version numbers of the library	124	You can find out the major and minor ABI version numbers of the library
111	you linked against by calling the functions C<ev_version_major> and	125	you linked against by calling the functions C<ev_version_major> and
112	C<ev_version_minor>. If you want, you can compare against the global	126	C<ev_version_minor>. If you want, you can compare against the global
113	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	127	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
114	version of the library your program was compiled against.	128	version of the library your program was compiled against.
115		129
		130	These version numbers refer to the ABI version of the library, not the
		131	release version.
		132
116	Usually, it's a good idea to terminate if the major versions mismatch,	133	Usually, it's a good idea to terminate if the major versions mismatch,
117	as this indicates an incompatible change. Minor versions are usually	134	as this indicates an incompatible change. Minor versions are usually
118	compatible to older versions, so a larger minor version alone is usually	135	compatible to older versions, so a larger minor version alone is usually
119	not a problem.	136	not a problem.
120		137
121	Example: make sure we haven't accidentally been linked against the wrong	138	Example: Make sure we haven't accidentally been linked against the wrong
122	version:	139	version.
123		140
124	assert (("libev version mismatch",	141	assert (("libev version mismatch",
125	ev_version_major () == EV_VERSION_MAJOR	142	ev_version_major () == EV_VERSION_MAJOR
126	&& ev_version_minor () >= EV_VERSION_MINOR));	143	&& ev_version_minor () >= EV_VERSION_MINOR));
127		144
…		…
155	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	172	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
156	recommended ones.	173	recommended ones.
157		174
158	See the description of C<ev_embed> watchers for more info.	175	See the description of C<ev_embed> watchers for more info.
159		176
160	=item ev_set_allocator (void (cb)(void *ptr, size_t size))	177	=item ev_set_allocator (void (cb)(void *ptr, long size))
161		178
162	Sets the allocation function to use (the prototype and semantics are	179	Sets the allocation function to use (the prototype is similar - the
163	identical to the realloc C function). It is used to allocate and free	180	semantics is identical - to the realloc C function). It is used to
164	memory (no surprises here). If it returns zero when memory needs to be	181	allocate and free memory (no surprises here). If it returns zero when
165	allocated, the library might abort or take some potentially destructive	182	memory needs to be allocated, the library might abort or take some
166	action. The default is your system realloc function.	183	potentially destructive action. The default is your system realloc
		184	function.
167		185
168	You could override this function in high-availability programs to, say,	186	You could override this function in high-availability programs to, say,
169	free some memory if it cannot allocate memory, to use a special allocator,	187	free some memory if it cannot allocate memory, to use a special allocator,
170	or even to sleep a while and retry until some memory is available.	188	or even to sleep a while and retry until some memory is available.
171		189
172	Example: replace the libev allocator with one that waits a bit and then	190	Example: Replace the libev allocator with one that waits a bit and then
173	retries: better than mine).	191	retries).
174		192
175	static void *	193	static void *
176	persistent_realloc (void *ptr, size_t size)	194	persistent_realloc (void *ptr, size_t size)
177	{	195	{
178	for (;;)	196	for (;;)
…		…
197	callback is set, then libev will expect it to remedy the sitution, no	215	callback is set, then libev will expect it to remedy the sitution, no
198	matter what, when it returns. That is, libev will generally retry the	216	matter what, when it returns. That is, libev will generally retry the
199	requested operation, or, if the condition doesn't go away, do bad stuff	217	requested operation, or, if the condition doesn't go away, do bad stuff
200	(such as abort).	218	(such as abort).
201		219
202	Example: do the same thing as libev does internally:	220	Example: This is basically the same thing that libev does internally, too.
203		221
204	static void	222	static void
205	fatal_error (const char *msg)	223	fatal_error (const char *msg)
206	{	224	{
207	perror (msg);	225	perror (msg);
…		…
257	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	275	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
258	override the flags completely if it is found in the environment. This is	276	override the flags completely if it is found in the environment. This is
259	useful to try out specific backends to test their performance, or to work	277	useful to try out specific backends to test their performance, or to work
260	around bugs.	278	around bugs.
261		279
		280	=item C<EVFLAG_FORKCHECK>
		281
		282	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		283	a fork, you can also make libev check for a fork in each iteration by
		284	enabling this flag.
		285
		286	This works by calling C<getpid ()> on every iteration of the loop,
		287	and thus this might slow down your event loop if you do a lot of loop
		288	iterations and little real work, but is usually not noticeable (on my
		289	Linux system for example, C<getpid> is actually a simple 5-insn sequence
		290	without a syscall and thus I<very> fast, but my Linux system also has
		291	C<pthread_atfork> which is even faster).
		292
		293	The big advantage of this flag is that you can forget about fork (and
		294	forget about forgetting to tell libev about forking) when you use this
		295	flag.
		296
		297	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		298	environment variable.
		299
262	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	300	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
263		301
264	This is your standard select(2) backend. Not I<completely> standard, as	302	This is your standard select(2) backend. Not I<completely> standard, as
265	libev tries to roll its own fd_set with no limits on the number of fds,	303	libev tries to roll its own fd_set with no limits on the number of fds,
266	but if that fails, expect a fairly low limit on the number of fds when	304	but if that fails, expect a fairly low limit on the number of fds when
…		…
353	Similar to C<ev_default_loop>, but always creates a new event loop that is	391	Similar to C<ev_default_loop>, but always creates a new event loop that is
354	always distinct from the default loop. Unlike the default loop, it cannot	392	always distinct from the default loop. Unlike the default loop, it cannot
355	handle signal and child watchers, and attempts to do so will be greeted by	393	handle signal and child watchers, and attempts to do so will be greeted by
356	undefined behaviour (or a failed assertion if assertions are enabled).	394	undefined behaviour (or a failed assertion if assertions are enabled).
357		395
358	Example: try to create a event loop that uses epoll and nothing else.	396	Example: Try to create a event loop that uses epoll and nothing else.
359		397
360	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);	398	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);
361	if (!epoller)	399	if (!epoller)
362	fatal ("no epoll found here, maybe it hides under your chair");	400	fatal ("no epoll found here, maybe it hides under your chair");
363		401
…		…
366	Destroys the default loop again (frees all memory and kernel state	404	Destroys the default loop again (frees all memory and kernel state
367	etc.). None of the active event watchers will be stopped in the normal	405	etc.). None of the active event watchers will be stopped in the normal
368	sense, so e.g. C<ev_is_active> might still return true. It is your	406	sense, so e.g. C<ev_is_active> might still return true. It is your
369	responsibility to either stop all watchers cleanly yoursef I<before>	407	responsibility to either stop all watchers cleanly yoursef I<before>
370	calling this function, or cope with the fact afterwards (which is usually	408	calling this function, or cope with the fact afterwards (which is usually
371	the easiest thing, youc na just ignore the watchers and/or C<free ()> them	409	the easiest thing, you can just ignore the watchers and/or C<free ()> them
372	for example).	410	for example).
		411
		412	Note that certain global state, such as signal state, will not be freed by
		413	this function, and related watchers (such as signal and child watchers)
		414	would need to be stopped manually.
		415
		416	In general it is not advisable to call this function except in the
		417	rare occasion where you really need to free e.g. the signal handling
		418	pipe fds. If you need dynamically allocated loops it is better to use
		419	C<ev_loop_new> and C<ev_loop_destroy>).
373		420
374	=item ev_loop_destroy (loop)	421	=item ev_loop_destroy (loop)
375		422
376	Like C<ev_default_destroy>, but destroys an event loop created by an	423	Like C<ev_default_destroy>, but destroys an event loop created by an
377	earlier call to C<ev_loop_new>.	424	earlier call to C<ev_loop_new>.
…		…
400	=item ev_loop_fork (loop)	447	=item ev_loop_fork (loop)
401		448
402	Like C<ev_default_fork>, but acts on an event loop created by	449	Like C<ev_default_fork>, but acts on an event loop created by
403	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	450	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
404	after fork, and how you do this is entirely your own problem.	451	after fork, and how you do this is entirely your own problem.
		452
		453	=item unsigned int ev_loop_count (loop)
		454
		455	Returns the count of loop iterations for the loop, which is identical to
		456	the number of times libev did poll for new events. It starts at C<0> and
		457	happily wraps around with enough iterations.
		458
		459	This value can sometimes be useful as a generation counter of sorts (it
		460	"ticks" the number of loop iterations), as it roughly corresponds with
		461	C<ev_prepare> and C<ev_check> calls.
405		462
406	=item unsigned int ev_backend (loop)	463	=item unsigned int ev_backend (loop)
407		464
408	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	465	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
409	use.	466	use.
…		…
443	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	500	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
444	usually a better approach for this kind of thing.	501	usually a better approach for this kind of thing.
445		502
446	Here are the gory details of what C<ev_loop> does:	503	Here are the gory details of what C<ev_loop> does:
447		504
		505	- Before the first iteration, call any pending watchers.
448	* If there are no active watchers (reference count is zero), return.	506	* If there are no active watchers (reference count is zero), return.
449	- Queue prepare watchers and then call all outstanding watchers.	507	- Queue all prepare watchers and then call all outstanding watchers.
450	- If we have been forked, recreate the kernel state.	508	- If we have been forked, recreate the kernel state.
451	- Update the kernel state with all outstanding changes.	509	- Update the kernel state with all outstanding changes.
452	- Update the "event loop time".	510	- Update the "event loop time".
453	- Calculate for how long to block.	511	- Calculate for how long to block.
454	- Block the process, waiting for any events.	512	- Block the process, waiting for any events.
…		…
462	Signals and child watchers are implemented as I/O watchers, and will	520	Signals and child watchers are implemented as I/O watchers, and will
463	be handled here by queueing them when their watcher gets executed.	521	be handled here by queueing them when their watcher gets executed.
464	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	522	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
465	were used, return, otherwise continue with step *.	523	were used, return, otherwise continue with step *.
466		524
467	Example: queue some jobs and then loop until no events are outsanding	525	Example: Queue some jobs and then loop until no events are outsanding
468	anymore.	526	anymore.
469		527
470	... queue jobs here, make sure they register event watchers as long	528	... queue jobs here, make sure they register event watchers as long
471	... as they still have work to do (even an idle watcher will do..)	529	... as they still have work to do (even an idle watcher will do..)
472	ev_loop (my_loop, 0);	530	ev_loop (my_loop, 0);
…		…
492	visible to the libev user and should not keep C<ev_loop> from exiting if	550	visible to the libev user and should not keep C<ev_loop> from exiting if
493	no event watchers registered by it are active. It is also an excellent	551	no event watchers registered by it are active. It is also an excellent
494	way to do this for generic recurring timers or from within third-party	552	way to do this for generic recurring timers or from within third-party
495	libraries. Just remember to I<unref after start> and I<ref before stop>.	553	libraries. Just remember to I<unref after start> and I<ref before stop>.
496		554
497	Example: create a signal watcher, but keep it from keeping C<ev_loop>	555	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
498	running when nothing else is active.	556	running when nothing else is active.
499		557
500	struct dv_signal exitsig;	558	struct ev_signal exitsig;
501	ev_signal_init (&exitsig, sig_cb, SIGINT);	559	ev_signal_init (&exitsig, sig_cb, SIGINT);
502	ev_signal_start (myloop, &exitsig);	560	ev_signal_start (loop, &exitsig);
503	evf_unref (myloop);	561	evf_unref (loop);
504		562
505	Example: for some weird reason, unregister the above signal handler again.	563	Example: For some weird reason, unregister the above signal handler again.
506		564
507	ev_ref (myloop);	565	ev_ref (loop);
508	ev_signal_stop (myloop, &exitsig);	566	ev_signal_stop (loop, &exitsig);
509		567
510	=back	568	=back
511		569
512		570
513	=head1 ANATOMY OF A WATCHER	571	=head1 ANATOMY OF A WATCHER
…		…
693	=item bool ev_is_pending (ev_TYPE *watcher)	751	=item bool ev_is_pending (ev_TYPE *watcher)
694		752
695	Returns a true value iff the watcher is pending, (i.e. it has outstanding	753	Returns a true value iff the watcher is pending, (i.e. it has outstanding
696	events but its callback has not yet been invoked). As long as a watcher	754	events but its callback has not yet been invoked). As long as a watcher
697	is pending (but not active) you must not call an init function on it (but	755	is pending (but not active) you must not call an init function on it (but
698	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	756	C<ev_TYPE_set> is safe), you must not change its priority, and you must
699	libev (e.g. you cnanot C<free ()> it).	757	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		758	it).
700		759
701	=item callback = ev_cb (ev_TYPE *watcher)	760	=item callback ev_cb (ev_TYPE *watcher)
702		761
703	Returns the callback currently set on the watcher.	762	Returns the callback currently set on the watcher.
704		763
705	=item ev_cb_set (ev_TYPE *watcher, callback)	764	=item ev_cb_set (ev_TYPE *watcher, callback)
706		765
707	Change the callback. You can change the callback at virtually any time	766	Change the callback. You can change the callback at virtually any time
708	(modulo threads).	767	(modulo threads).
		768
		769	=item ev_set_priority (ev_TYPE *watcher, priority)
		770
		771	=item int ev_priority (ev_TYPE *watcher)
		772
		773	Set and query the priority of the watcher. The priority is a small
		774	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		775	(default: C<-2>). Pending watchers with higher priority will be invoked
		776	before watchers with lower priority, but priority will not keep watchers
		777	from being executed (except for C<ev_idle> watchers).
		778
		779	This means that priorities are I<only> used for ordering callback
		780	invocation after new events have been received. This is useful, for
		781	example, to reduce latency after idling, or more often, to bind two
		782	watchers on the same event and make sure one is called first.
		783
		784	If you need to suppress invocation when higher priority events are pending
		785	you need to look at C<ev_idle> watchers, which provide this functionality.
		786
		787	You I<must not> change the priority of a watcher as long as it is active or
		788	pending.
		789
		790	The default priority used by watchers when no priority has been set is
		791	always C<0>, which is supposed to not be too high and not be too low :).
		792
		793	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		794	fine, as long as you do not mind that the priority value you query might
		795	or might not have been adjusted to be within valid range.
		796
		797	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		798
		799	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		800	C<loop> nor C<revents> need to be valid as long as the watcher callback
		801	can deal with that fact.
		802
		803	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		804
		805	If the watcher is pending, this function returns clears its pending status
		806	and returns its C<revents> bitset (as if its callback was invoked). If the
		807	watcher isn't pending it does nothing and returns C<0>.
709		808
710	=back	809	=back
711		810
712		811
713	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	812	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
…		…
734	{	833	{
735	struct my_io w = (struct my_io )w_;	834	struct my_io w = (struct my_io )w_;
736	...	835	...
737	}	836	}
738		837
739	More interesting and less C-conformant ways of catsing your callback type	838	More interesting and less C-conformant ways of casting your callback type
740	have been omitted....	839	instead have been omitted.
		840
		841	Another common scenario is having some data structure with multiple
		842	watchers:
		843
		844	struct my_biggy
		845	{
		846	int some_data;
		847	ev_timer t1;
		848	ev_timer t2;
		849	}
		850
		851	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		852	you need to use C<offsetof>:
		853
		854	#include <stddef.h>
		855
		856	static void
		857	t1_cb (EV_P_ struct ev_timer *w, int revents)
		858	{
		859	struct my_biggy big = (struct my_biggy *
		860	(((char *)w) - offsetof (struct my_biggy, t1));
		861	}
		862
		863	static void
		864	t2_cb (EV_P_ struct ev_timer *w, int revents)
		865	{
		866	struct my_biggy big = (struct my_biggy *
		867	(((char *)w) - offsetof (struct my_biggy, t2));
		868	}
741		869
742		870
743	=head1 WATCHER TYPES	871	=head1 WATCHER TYPES
744		872
745	This section describes each watcher in detail, but will not repeat	873	This section describes each watcher in detail, but will not repeat
…		…
790	it is best to always use non-blocking I/O: An extra C<read>(2) returning	918	it is best to always use non-blocking I/O: An extra C<read>(2) returning
791	C<EAGAIN> is far preferable to a program hanging until some data arrives.	919	C<EAGAIN> is far preferable to a program hanging until some data arrives.
792		920
793	If you cannot run the fd in non-blocking mode (for example you should not	921	If you cannot run the fd in non-blocking mode (for example you should not
794	play around with an Xlib connection), then you have to seperately re-test	922	play around with an Xlib connection), then you have to seperately re-test
795	wether a file descriptor is really ready with a known-to-be good interface	923	whether a file descriptor is really ready with a known-to-be good interface
796	such as poll (fortunately in our Xlib example, Xlib already does this on	924	such as poll (fortunately in our Xlib example, Xlib already does this on
797	its own, so its quite safe to use).	925	its own, so its quite safe to use).
		926
		927	=head3 The special problem of disappearing file descriptors
		928
		929	Some backends (e.g kqueue, epoll) need to be told about closing a file
		930	descriptor (either by calling C<close> explicitly or by any other means,
		931	such as C<dup>). The reason is that you register interest in some file
		932	descriptor, but when it goes away, the operating system will silently drop
		933	this interest. If another file descriptor with the same number then is
		934	registered with libev, there is no efficient way to see that this is, in
		935	fact, a different file descriptor.
		936
		937	To avoid having to explicitly tell libev about such cases, libev follows
		938	the following policy: Each time C<ev_io_set> is being called, libev
		939	will assume that this is potentially a new file descriptor, otherwise
		940	it is assumed that the file descriptor stays the same. That means that
		941	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		942	descriptor even if the file descriptor number itself did not change.
		943
		944	This is how one would do it normally anyway, the important point is that
		945	the libev application should not optimise around libev but should leave
		946	optimisations to libev.
		947
		948
		949	=head3 Watcher-Specific Functions
798		950
799	=over 4	951	=over 4
800		952
801	=item ev_io_init (ev_io *, callback, int fd, int events)	953	=item ev_io_init (ev_io *, callback, int fd, int events)
802		954
…		…
814		966
815	The events being watched.	967	The events being watched.
816		968
817	=back	969	=back
818		970
819	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	971	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
820	readable, but only once. Since it is likely line-buffered, you could	972	readable, but only once. Since it is likely line-buffered, you could
821	attempt to read a whole line in the callback:	973	attempt to read a whole line in the callback.
822		974
823	static void	975	static void
824	stdin_readable_cb (struct ev_loop loop, struct ev_io w, int revents)	976	stdin_readable_cb (struct ev_loop loop, struct ev_io w, int revents)
825	{	977	{
826	ev_io_stop (loop, w);	978	ev_io_stop (loop, w);
…		…
856		1008
857	The callback is guarenteed to be invoked only when its timeout has passed,	1009	The callback is guarenteed to be invoked only when its timeout has passed,
858	but if multiple timers become ready during the same loop iteration then	1010	but if multiple timers become ready during the same loop iteration then
859	order of execution is undefined.	1011	order of execution is undefined.
860		1012
		1013	=head3 Watcher-Specific Functions and Data Members
		1014
861	=over 4	1015	=over 4
862		1016
863	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1017	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
864		1018
865	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1019	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
…		…
878	=item ev_timer_again (loop)	1032	=item ev_timer_again (loop)
879		1033
880	This will act as if the timer timed out and restart it again if it is	1034	This will act as if the timer timed out and restart it again if it is
881	repeating. The exact semantics are:	1035	repeating. The exact semantics are:
882		1036
		1037	If the timer is pending, its pending status is cleared.
		1038
883	If the timer is started but nonrepeating, stop it.	1039	If the timer is started but nonrepeating, stop it (as if it timed out).
884		1040
885	If the timer is repeating, either start it if necessary (with the repeat	1041	If the timer is repeating, either start it if necessary (with the
886	value), or reset the running timer to the repeat value.	1042	C<repeat> value), or reset the running timer to the C<repeat> value.
887		1043
888	This sounds a bit complicated, but here is a useful and typical	1044	This sounds a bit complicated, but here is a useful and typical
889	example: Imagine you have a tcp connection and you want a so-called	1045	example: Imagine you have a tcp connection and you want a so-called idle
890	idle timeout, that is, you want to be called when there have been,	1046	timeout, that is, you want to be called when there have been, say, 60
891	say, 60 seconds of inactivity on the socket. The easiest way to do	1047	seconds of inactivity on the socket. The easiest way to do this is to
892	this is to configure an C<ev_timer> with C<after>=C<repeat>=C<60> and calling	1048	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
893	C<ev_timer_again> each time you successfully read or write some data. If	1049	C<ev_timer_again> each time you successfully read or write some data. If
894	you go into an idle state where you do not expect data to travel on the	1050	you go into an idle state where you do not expect data to travel on the
895	socket, you can stop the timer, and again will automatically restart it if	1051	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
896	need be.	1052	automatically restart it if need be.
897		1053
898	You can also ignore the C<after> value and C<ev_timer_start> altogether	1054	That means you can ignore the C<after> value and C<ev_timer_start>
899	and only ever use the C<repeat> value:	1055	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
900		1056
901	ev_timer_init (timer, callback, 0., 5.);	1057	ev_timer_init (timer, callback, 0., 5.);
902	ev_timer_again (loop, timer);	1058	ev_timer_again (loop, timer);
903	...	1059	...
904	timer->again = 17.;	1060	timer->again = 17.;
905	ev_timer_again (loop, timer);	1061	ev_timer_again (loop, timer);
906	...	1062	...
907	timer->again = 10.;	1063	timer->again = 10.;
908	ev_timer_again (loop, timer);	1064	ev_timer_again (loop, timer);
909		1065
910	This is more efficient then stopping/starting the timer eahc time you want	1066	This is more slightly efficient then stopping/starting the timer each time
911	to modify its timeout value.	1067	you want to modify its timeout value.
912		1068
913	=item ev_tstamp repeat [read-write]	1069	=item ev_tstamp repeat [read-write]
914		1070
915	The current C<repeat> value. Will be used each time the watcher times out	1071	The current C<repeat> value. Will be used each time the watcher times out
916	or C<ev_timer_again> is called and determines the next timeout (if any),	1072	or C<ev_timer_again> is called and determines the next timeout (if any),
917	which is also when any modifications are taken into account.	1073	which is also when any modifications are taken into account.
918		1074
919	=back	1075	=back
920		1076
921	Example: create a timer that fires after 60 seconds.	1077	Example: Create a timer that fires after 60 seconds.
922		1078
923	static void	1079	static void
924	one_minute_cb (struct ev_loop loop, struct ev_timer w, int revents)	1080	one_minute_cb (struct ev_loop loop, struct ev_timer w, int revents)
925	{	1081	{
926	.. one minute over, w is actually stopped right here	1082	.. one minute over, w is actually stopped right here
…		…
928		1084
929	struct ev_timer mytimer;	1085	struct ev_timer mytimer;
930	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1086	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
931	ev_timer_start (loop, &mytimer);	1087	ev_timer_start (loop, &mytimer);
932		1088
933	Example: create a timeout timer that times out after 10 seconds of	1089	Example: Create a timeout timer that times out after 10 seconds of
934	inactivity.	1090	inactivity.
935		1091
936	static void	1092	static void
937	timeout_cb (struct ev_loop loop, struct ev_timer w, int revents)	1093	timeout_cb (struct ev_loop loop, struct ev_timer w, int revents)
938	{	1094	{
…		…
958	but on wallclock time (absolute time). You can tell a periodic watcher	1114	but on wallclock time (absolute time). You can tell a periodic watcher
959	to trigger "at" some specific point in time. For example, if you tell a	1115	to trigger "at" some specific point in time. For example, if you tell a
960	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1116	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
961	+ 10.>) and then reset your system clock to the last year, then it will	1117	+ 10.>) and then reset your system clock to the last year, then it will
962	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1118	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
963	roughly 10 seconds later and of course not if you reset your system time	1119	roughly 10 seconds later).
964	again).
965		1120
966	They can also be used to implement vastly more complex timers, such as	1121	They can also be used to implement vastly more complex timers, such as
967	triggering an event on eahc midnight, local time.	1122	triggering an event on each midnight, local time or other, complicated,
		1123	rules.
968		1124
969	As with timers, the callback is guarenteed to be invoked only when the	1125	As with timers, the callback is guarenteed to be invoked only when the
970	time (C<at>) has been passed, but if multiple periodic timers become ready	1126	time (C<at>) has been passed, but if multiple periodic timers become ready
971	during the same loop iteration then order of execution is undefined.	1127	during the same loop iteration then order of execution is undefined.
972		1128
		1129	=head3 Watcher-Specific Functions and Data Members
		1130
973	=over 4	1131	=over 4
974		1132
975	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1133	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
976		1134
977	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1135	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
…		…
979	Lots of arguments, lets sort it out... There are basically three modes of	1137	Lots of arguments, lets sort it out... There are basically three modes of
980	operation, and we will explain them from simplest to complex:	1138	operation, and we will explain them from simplest to complex:
981		1139
982	=over 4	1140	=over 4
983		1141
984	=item * absolute timer (interval = reschedule_cb = 0)	1142	=item * absolute timer (at = time, interval = reschedule_cb = 0)
985		1143
986	In this configuration the watcher triggers an event at the wallclock time	1144	In this configuration the watcher triggers an event at the wallclock time
987	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1145	C<at> and doesn't repeat. It will not adjust when a time jump occurs,
988	that is, if it is to be run at January 1st 2011 then it will run when the	1146	that is, if it is to be run at January 1st 2011 then it will run when the
989	system time reaches or surpasses this time.	1147	system time reaches or surpasses this time.
990		1148
991	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1149	=item * non-repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
992		1150
993	In this mode the watcher will always be scheduled to time out at the next	1151	In this mode the watcher will always be scheduled to time out at the next
994	C<at + N * interval> time (for some integer N) and then repeat, regardless	1152	C<at + N * interval> time (for some integer N, which can also be negative)
995	of any time jumps.	1153	and then repeat, regardless of any time jumps.
996		1154
997	This can be used to create timers that do not drift with respect to system	1155	This can be used to create timers that do not drift with respect to system
998	time:	1156	time:
999		1157
1000	ev_periodic_set (&periodic, 0., 3600., 0);	1158	ev_periodic_set (&periodic, 0., 3600., 0);
…		…
1006		1164
1007	Another way to think about it (for the mathematically inclined) is that	1165	Another way to think about it (for the mathematically inclined) is that
1008	C<ev_periodic> will try to run the callback in this mode at the next possible	1166	C<ev_periodic> will try to run the callback in this mode at the next possible
1009	time where C<time = at (mod interval)>, regardless of any time jumps.	1167	time where C<time = at (mod interval)>, regardless of any time jumps.
1010		1168
		1169	For numerical stability it is preferable that the C<at> value is near
		1170	C<ev_now ()> (the current time), but there is no range requirement for
		1171	this value.
		1172
1011	=item * manual reschedule mode (reschedule_cb = callback)	1173	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1012		1174
1013	In this mode the values for C<interval> and C<at> are both being	1175	In this mode the values for C<interval> and C<at> are both being
1014	ignored. Instead, each time the periodic watcher gets scheduled, the	1176	ignored. Instead, each time the periodic watcher gets scheduled, the
1015	reschedule callback will be called with the watcher as first, and the	1177	reschedule callback will be called with the watcher as first, and the
1016	current time as second argument.	1178	current time as second argument.
1017		1179
1018	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1180	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
1019	ever, or make any event loop modifications>. If you need to stop it,	1181	ever, or make any event loop modifications>. If you need to stop it,
1020	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	1182	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
1021	starting a prepare watcher).	1183	starting an C<ev_prepare> watcher, which is legal).
1022		1184
1023	Its prototype is C<ev_tstamp (reschedule_cb)(struct ev_periodic w,	1185	Its prototype is C<ev_tstamp (reschedule_cb)(struct ev_periodic w,
1024	ev_tstamp now)>, e.g.:	1186	ev_tstamp now)>, e.g.:
1025		1187
1026	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1188	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
…		…
1049	Simply stops and restarts the periodic watcher again. This is only useful	1211	Simply stops and restarts the periodic watcher again. This is only useful
1050	when you changed some parameters or the reschedule callback would return	1212	when you changed some parameters or the reschedule callback would return
1051	a different time than the last time it was called (e.g. in a crond like	1213	a different time than the last time it was called (e.g. in a crond like
1052	program when the crontabs have changed).	1214	program when the crontabs have changed).
1053		1215
		1216	=item ev_tstamp offset [read-write]
		1217
		1218	When repeating, this contains the offset value, otherwise this is the
		1219	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1220
		1221	Can be modified any time, but changes only take effect when the periodic
		1222	timer fires or C<ev_periodic_again> is being called.
		1223
1054	=item ev_tstamp interval [read-write]	1224	=item ev_tstamp interval [read-write]
1055		1225
1056	The current interval value. Can be modified any time, but changes only	1226	The current interval value. Can be modified any time, but changes only
1057	take effect when the periodic timer fires or C<ev_periodic_again> is being	1227	take effect when the periodic timer fires or C<ev_periodic_again> is being
1058	called.	1228	called.
…		…
1061		1231
1062	The current reschedule callback, or C<0>, if this functionality is	1232	The current reschedule callback, or C<0>, if this functionality is
1063	switched off. Can be changed any time, but changes only take effect when	1233	switched off. Can be changed any time, but changes only take effect when
1064	the periodic timer fires or C<ev_periodic_again> is being called.	1234	the periodic timer fires or C<ev_periodic_again> is being called.
1065		1235
		1236	=item ev_tstamp at [read-only]
		1237
		1238	When active, contains the absolute time that the watcher is supposed to
		1239	trigger next.
		1240
1066	=back	1241	=back
1067		1242
1068	Example: call a callback every hour, or, more precisely, whenever the	1243	Example: Call a callback every hour, or, more precisely, whenever the
1069	system clock is divisible by 3600. The callback invocation times have	1244	system clock is divisible by 3600. The callback invocation times have
1070	potentially a lot of jittering, but good long-term stability.	1245	potentially a lot of jittering, but good long-term stability.
1071		1246
1072	static void	1247	static void
1073	clock_cb (struct ev_loop loop, struct ev_io w, int revents)	1248	clock_cb (struct ev_loop loop, struct ev_io w, int revents)
…		…
1077		1252
1078	struct ev_periodic hourly_tick;	1253	struct ev_periodic hourly_tick;
1079	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1254	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1080	ev_periodic_start (loop, &hourly_tick);	1255	ev_periodic_start (loop, &hourly_tick);
1081		1256
1082	Example: the same as above, but use a reschedule callback to do it:	1257	Example: The same as above, but use a reschedule callback to do it:
1083		1258
1084	#include <math.h>	1259	#include <math.h>
1085		1260
1086	static ev_tstamp	1261	static ev_tstamp
1087	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1262	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
…		…
1089	return fmod (now, 3600.) + 3600.;	1264	return fmod (now, 3600.) + 3600.;
1090	}	1265	}
1091		1266
1092	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1267	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1093		1268
1094	Example: call a callback every hour, starting now:	1269	Example: Call a callback every hour, starting now:
1095		1270
1096	struct ev_periodic hourly_tick;	1271	struct ev_periodic hourly_tick;
1097	ev_periodic_init (&hourly_tick, clock_cb,	1272	ev_periodic_init (&hourly_tick, clock_cb,
1098	fmod (ev_now (loop), 3600.), 3600., 0);	1273	fmod (ev_now (loop), 3600.), 3600., 0);
1099	ev_periodic_start (loop, &hourly_tick);	1274	ev_periodic_start (loop, &hourly_tick);
…		…
1111	with the kernel (thus it coexists with your own signal handlers as long	1286	with the kernel (thus it coexists with your own signal handlers as long
1112	as you don't register any with libev). Similarly, when the last signal	1287	as you don't register any with libev). Similarly, when the last signal
1113	watcher for a signal is stopped libev will reset the signal handler to	1288	watcher for a signal is stopped libev will reset the signal handler to
1114	SIG_DFL (regardless of what it was set to before).	1289	SIG_DFL (regardless of what it was set to before).
1115		1290
		1291	=head3 Watcher-Specific Functions and Data Members
		1292
1116	=over 4	1293	=over 4
1117		1294
1118	=item ev_signal_init (ev_signal *, callback, int signum)	1295	=item ev_signal_init (ev_signal *, callback, int signum)
1119		1296
1120	=item ev_signal_set (ev_signal *, int signum)	1297	=item ev_signal_set (ev_signal *, int signum)
…		…
1131		1308
1132	=head2 C<ev_child> - watch out for process status changes	1309	=head2 C<ev_child> - watch out for process status changes
1133		1310
1134	Child watchers trigger when your process receives a SIGCHLD in response to	1311	Child watchers trigger when your process receives a SIGCHLD in response to
1135	some child status changes (most typically when a child of yours dies).	1312	some child status changes (most typically when a child of yours dies).
		1313
		1314	=head3 Watcher-Specific Functions and Data Members
1136		1315
1137	=over 4	1316	=over 4
1138		1317
1139	=item ev_child_init (ev_child *, callback, int pid)	1318	=item ev_child_init (ev_child *, callback, int pid)
1140		1319
…		…
1160	The process exit/trace status caused by C<rpid> (see your systems	1339	The process exit/trace status caused by C<rpid> (see your systems
1161	C<waitpid> and C<sys/wait.h> documentation for details).	1340	C<waitpid> and C<sys/wait.h> documentation for details).
1162		1341
1163	=back	1342	=back
1164		1343
1165	Example: try to exit cleanly on SIGINT and SIGTERM.	1344	Example: Try to exit cleanly on SIGINT and SIGTERM.
1166		1345
1167	static void	1346	static void
1168	sigint_cb (struct ev_loop loop, struct ev_signal w, int revents)	1347	sigint_cb (struct ev_loop loop, struct ev_signal w, int revents)
1169	{	1348	{
1170	ev_unloop (loop, EVUNLOOP_ALL);	1349	ev_unloop (loop, EVUNLOOP_ALL);
…		…
1185	not exist" is a status change like any other. The condition "path does	1364	not exist" is a status change like any other. The condition "path does
1186	not exist" is signified by the C<st_nlink> field being zero (which is	1365	not exist" is signified by the C<st_nlink> field being zero (which is
1187	otherwise always forced to be at least one) and all the other fields of	1366	otherwise always forced to be at least one) and all the other fields of
1188	the stat buffer having unspecified contents.	1367	the stat buffer having unspecified contents.
1189		1368
		1369	The path I<should> be absolute and I<must not> end in a slash. If it is
		1370	relative and your working directory changes, the behaviour is undefined.
		1371
1190	Since there is no standard to do this, the portable implementation simply	1372	Since there is no standard to do this, the portable implementation simply
1191	calls C<stat (2)> regulalry on the path to see if it changed somehow. You	1373	calls C<stat (2)> regularly on the path to see if it changed somehow. You
1192	can specify a recommended polling interval for this case. If you specify	1374	can specify a recommended polling interval for this case. If you specify
1193	a polling interval of C<0> (highly recommended!) then a I<suitable,	1375	a polling interval of C<0> (highly recommended!) then a I<suitable,
1194	unspecified default> value will be used (which you can expect to be around	1376	unspecified default> value will be used (which you can expect to be around
1195	five seconds, although this might change dynamically). Libev will also	1377	five seconds, although this might change dynamically). Libev will also
1196	impose a minimum interval which is currently around C<0.1>, but thats	1378	impose a minimum interval which is currently around C<0.1>, but thats
…		…
1198		1380
1199	This watcher type is not meant for massive numbers of stat watchers,	1381	This watcher type is not meant for massive numbers of stat watchers,
1200	as even with OS-supported change notifications, this can be	1382	as even with OS-supported change notifications, this can be
1201	resource-intensive.	1383	resource-intensive.
1202		1384
1203	At the time of this writing, no specific OS backends are implemented, but	1385	At the time of this writing, only the Linux inotify interface is
1204	if demand increases, at least a kqueue and inotify backend will be added.	1386	implemented (implementing kqueue support is left as an exercise for the
		1387	reader). Inotify will be used to give hints only and should not change the
		1388	semantics of C<ev_stat> watchers, which means that libev sometimes needs
		1389	to fall back to regular polling again even with inotify, but changes are
		1390	usually detected immediately, and if the file exists there will be no
		1391	polling.
		1392
		1393	=head3 Watcher-Specific Functions and Data Members
1205		1394
1206	=over 4	1395	=over 4
1207		1396
1208	=item ev_stat_init (ev_stat , callback, const char path, ev_tstamp interval)	1397	=item ev_stat_init (ev_stat , callback, const char path, ev_tstamp interval)
1209		1398
…		…
1273	ev_stat_start (loop, &passwd);	1462	ev_stat_start (loop, &passwd);
1274		1463
1275		1464
1276	=head2 C<ev_idle> - when you've got nothing better to do...	1465	=head2 C<ev_idle> - when you've got nothing better to do...
1277		1466
1278	Idle watchers trigger events when there are no other events are pending	1467	Idle watchers trigger events when no other events of the same or higher
1279	(prepare, check and other idle watchers do not count). That is, as long	1468	priority are pending (prepare, check and other idle watchers do not
1280	as your process is busy handling sockets or timeouts (or even signals,	1469	count).
1281	imagine) it will not be triggered. But when your process is idle all idle	1470
1282	watchers are being called again and again, once per event loop iteration -	1471	That is, as long as your process is busy handling sockets or timeouts
		1472	(or even signals, imagine) of the same or higher priority it will not be
		1473	triggered. But when your process is idle (or only lower-priority watchers
		1474	are pending), the idle watchers are being called once per event loop
1283	until stopped, that is, or your process receives more events and becomes	1475	iteration - until stopped, that is, or your process receives more events
1284	busy.	1476	and becomes busy again with higher priority stuff.
1285		1477
1286	The most noteworthy effect is that as long as any idle watchers are	1478	The most noteworthy effect is that as long as any idle watchers are
1287	active, the process will not block when waiting for new events.	1479	active, the process will not block when waiting for new events.
1288		1480
1289	Apart from keeping your process non-blocking (which is a useful	1481	Apart from keeping your process non-blocking (which is a useful
1290	effect on its own sometimes), idle watchers are a good place to do	1482	effect on its own sometimes), idle watchers are a good place to do
1291	"pseudo-background processing", or delay processing stuff to after the	1483	"pseudo-background processing", or delay processing stuff to after the
1292	event loop has handled all outstanding events.	1484	event loop has handled all outstanding events.
1293		1485
		1486	=head3 Watcher-Specific Functions and Data Members
		1487
1294	=over 4	1488	=over 4
1295		1489
1296	=item ev_idle_init (ev_signal *, callback)	1490	=item ev_idle_init (ev_signal *, callback)
1297		1491
1298	Initialises and configures the idle watcher - it has no parameters of any	1492	Initialises and configures the idle watcher - it has no parameters of any
1299	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1493	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1300	believe me.	1494	believe me.
1301		1495
1302	=back	1496	=back
1303		1497
1304	Example: dynamically allocate an C<ev_idle>, start it, and in the	1498	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1305	callback, free it. Alos, use no error checking, as usual.	1499	callback, free it. Also, use no error checking, as usual.
1306		1500
1307	static void	1501	static void
1308	idle_cb (struct ev_loop loop, struct ev_idle w, int revents)	1502	idle_cb (struct ev_loop loop, struct ev_idle w, int revents)
1309	{	1503	{
1310	free (w);	1504	free (w);
…		…
1355	with priority higher than or equal to the event loop and one coroutine	1549	with priority higher than or equal to the event loop and one coroutine
1356	of lower priority, but only once, using idle watchers to keep the event	1550	of lower priority, but only once, using idle watchers to keep the event
1357	loop from blocking if lower-priority coroutines are active, thus mapping	1551	loop from blocking if lower-priority coroutines are active, thus mapping
1358	low-priority coroutines to idle/background tasks).	1552	low-priority coroutines to idle/background tasks).
1359		1553
		1554	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1555	priority, to ensure that they are being run before any other watchers
		1556	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1557	too) should not activate ("feed") events into libev. While libev fully
		1558	supports this, they will be called before other C<ev_check> watchers did
		1559	their job. As C<ev_check> watchers are often used to embed other event
		1560	loops those other event loops might be in an unusable state until their
		1561	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		1562	others).
		1563
		1564	=head3 Watcher-Specific Functions and Data Members
		1565
1360	=over 4	1566	=over 4
1361		1567
1362	=item ev_prepare_init (ev_prepare *, callback)	1568	=item ev_prepare_init (ev_prepare *, callback)
1363		1569
1364	=item ev_check_init (ev_check *, callback)	1570	=item ev_check_init (ev_check *, callback)
…		…
1367	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1573	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1368	macros, but using them is utterly, utterly and completely pointless.	1574	macros, but using them is utterly, utterly and completely pointless.
1369		1575
1370	=back	1576	=back
1371		1577
1372	Example: To include a library such as adns, you would add IO watchers	1578	There are a number of principal ways to embed other event loops or modules
1373	and a timeout watcher in a prepare handler, as required by libadns, and	1579	into libev. Here are some ideas on how to include libadns into libev
		1580	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1581	use for an actually working example. Another Perl module named C<EV::Glib>
		1582	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV
		1583	into the Glib event loop).
		1584
		1585	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1374	in a check watcher, destroy them and call into libadns. What follows is	1586	and in a check watcher, destroy them and call into libadns. What follows
1375	pseudo-code only of course:	1587	is pseudo-code only of course. This requires you to either use a low
		1588	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1589	the callbacks for the IO/timeout watchers might not have been called yet.
1376		1590
1377	static ev_io iow [nfd];	1591	static ev_io iow [nfd];
1378	static ev_timer tw;	1592	static ev_timer tw;
1379		1593
1380	static void	1594	static void
1381	io_cb (ev_loop loop, ev_io w, int revents)	1595	io_cb (ev_loop loop, ev_io w, int revents)
1382	{	1596	{
1383	// set the relevant poll flags
1384	// could also call adns_processreadable etc. here
1385	struct pollfd fd = (struct pollfd )w->data;
1386	if (revents & EV_READ ) fd->revents \|= fd->events & POLLIN;
1387	if (revents & EV_WRITE) fd->revents \|= fd->events & POLLOUT;
1388	}	1597	}
1389		1598
1390	// create io watchers for each fd and a timer before blocking	1599	// create io watchers for each fd and a timer before blocking
1391	static void	1600	static void
1392	adns_prepare_cb (ev_loop loop, ev_prepare w, int revents)	1601	adns_prepare_cb (ev_loop loop, ev_prepare w, int revents)
1393	{	1602	{
1394	int timeout = 3600000;truct pollfd fds [nfd];	1603	int timeout = 3600000;
		1604	struct pollfd fds [nfd];
1395	// actual code will need to loop here and realloc etc.	1605	// actual code will need to loop here and realloc etc.
1396	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	1606	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1397		1607
1398	/* the callback is illegal, but won't be called as we stop during check */	1608	/* the callback is illegal, but won't be called as we stop during check */
1399	ev_timer_init (&tw, 0, timeout * 1e-3);	1609	ev_timer_init (&tw, 0, timeout * 1e-3);
1400	ev_timer_start (loop, &tw);	1610	ev_timer_start (loop, &tw);
1401		1611
1402	// create on ev_io per pollfd	1612	// create one ev_io per pollfd
1403	for (int i = 0; i < nfd; ++i)	1613	for (int i = 0; i < nfd; ++i)
1404	{	1614	{
1405	ev_io_init (iow + i, io_cb, fds [i].fd,	1615	ev_io_init (iow + i, io_cb, fds [i].fd,
1406	((fds [i].events & POLLIN ? EV_READ : 0)	1616	((fds [i].events & POLLIN ? EV_READ : 0)
1407	\| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	1617	\| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1408		1618
1409	fds [i].revents = 0;	1619	fds [i].revents = 0;
1410	iow [i].data = fds + i;
1411	ev_io_start (loop, iow + i);	1620	ev_io_start (loop, iow + i);
1412	}	1621	}
1413	}	1622	}
1414		1623
1415	// stop all watchers after blocking	1624	// stop all watchers after blocking
…		…
1417	adns_check_cb (ev_loop loop, ev_check w, int revents)	1626	adns_check_cb (ev_loop loop, ev_check w, int revents)
1418	{	1627	{
1419	ev_timer_stop (loop, &tw);	1628	ev_timer_stop (loop, &tw);
1420		1629
1421	for (int i = 0; i < nfd; ++i)	1630	for (int i = 0; i < nfd; ++i)
		1631	{
		1632	// set the relevant poll flags
		1633	// could also call adns_processreadable etc. here
		1634	struct pollfd *fd = fds + i;
		1635	int revents = ev_clear_pending (iow + i);
		1636	if (revents & EV_READ ) fd->revents \|= fd->events & POLLIN;
		1637	if (revents & EV_WRITE) fd->revents \|= fd->events & POLLOUT;
		1638
		1639	// now stop the watcher
1422	ev_io_stop (loop, iow + i);	1640	ev_io_stop (loop, iow + i);
		1641	}
1423		1642
1424	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	1643	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1644	}
		1645
		1646	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		1647	in the prepare watcher and would dispose of the check watcher.
		1648
		1649	Method 3: If the module to be embedded supports explicit event
		1650	notification (adns does), you can also make use of the actual watcher
		1651	callbacks, and only destroy/create the watchers in the prepare watcher.
		1652
		1653	static void
		1654	timer_cb (EV_P_ ev_timer *w, int revents)
		1655	{
		1656	adns_state ads = (adns_state)w->data;
		1657	update_now (EV_A);
		1658
		1659	adns_processtimeouts (ads, &tv_now);
		1660	}
		1661
		1662	static void
		1663	io_cb (EV_P_ ev_io *w, int revents)
		1664	{
		1665	adns_state ads = (adns_state)w->data;
		1666	update_now (EV_A);
		1667
		1668	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		1669	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		1670	}
		1671
		1672	// do not ever call adns_afterpoll
		1673
		1674	Method 4: Do not use a prepare or check watcher because the module you
		1675	want to embed is too inflexible to support it. Instead, youc na override
		1676	their poll function. The drawback with this solution is that the main
		1677	loop is now no longer controllable by EV. The C<Glib::EV> module does
		1678	this.
		1679
		1680	static gint
		1681	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		1682	{
		1683	int got_events = 0;
		1684
		1685	for (n = 0; n < nfds; ++n)
		1686	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		1687
		1688	if (timeout >= 0)
		1689	// create/start timer
		1690
		1691	// poll
		1692	ev_loop (EV_A_ 0);
		1693
		1694	// stop timer again
		1695	if (timeout >= 0)
		1696	ev_timer_stop (EV_A_ &to);
		1697
		1698	// stop io watchers again - their callbacks should have set
		1699	for (n = 0; n < nfds; ++n)
		1700	ev_io_stop (EV_A_ iow [n]);
		1701
		1702	return got_events;
1425	}	1703	}
1426		1704
1427		1705
1428	=head2 C<ev_embed> - when one backend isn't enough...	1706	=head2 C<ev_embed> - when one backend isn't enough...
1429		1707
…		…
1493	ev_embed_start (loop_hi, &embed);	1771	ev_embed_start (loop_hi, &embed);
1494	}	1772	}
1495	else	1773	else
1496	loop_lo = loop_hi;	1774	loop_lo = loop_hi;
1497		1775
		1776	=head3 Watcher-Specific Functions and Data Members
		1777
1498	=over 4	1778	=over 4
1499		1779
1500	=item ev_embed_init (ev_embed , callback, struct ev_loop embedded_loop)	1780	=item ev_embed_init (ev_embed , callback, struct ev_loop embedded_loop)
1501		1781
1502	=item ev_embed_set (ev_embed , callback, struct ev_loop embedded_loop)	1782	=item ev_embed_set (ev_embed , callback, struct ev_loop embedded_loop)
…		…
1528	event loop blocks next and before C<ev_check> watchers are being called,	1808	event loop blocks next and before C<ev_check> watchers are being called,
1529	and only in the child after the fork. If whoever good citizen calling	1809	and only in the child after the fork. If whoever good citizen calling
1530	C<ev_default_fork> cheats and calls it in the wrong process, the fork	1810	C<ev_default_fork> cheats and calls it in the wrong process, the fork
1531	handlers will be invoked, too, of course.	1811	handlers will be invoked, too, of course.
1532		1812
		1813	=head3 Watcher-Specific Functions and Data Members
		1814
1533	=over 4	1815	=over 4
1534		1816
1535	=item ev_fork_init (ev_signal *, callback)	1817	=item ev_fork_init (ev_signal *, callback)
1536		1818
1537	Initialises and configures the fork watcher - it has no parameters of any	1819	Initialises and configures the fork watcher - it has no parameters of any
…		…
1633		1915
1634	To use it,	1916	To use it,
1635		1917
1636	#include <ev++.h>	1918	#include <ev++.h>
1637		1919
1638	(it is not installed by default). This automatically includes F<ev.h>	1920	This automatically includes F<ev.h> and puts all of its definitions (many
1639	and puts all of its definitions (many of them macros) into the global	1921	of them macros) into the global namespace. All C++ specific things are
1640	namespace. All C++ specific things are put into the C<ev> namespace.	1922	put into the C<ev> namespace. It should support all the same embedding
		1923	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1641		1924
1642	It should support all the same embedding options as F<ev.h>, most notably	1925	Care has been taken to keep the overhead low. The only data member the C++
1643	C<EV_MULTIPLICITY>.	1926	classes add (compared to plain C-style watchers) is the event loop pointer
		1927	that the watcher is associated with (or no additional members at all if
		1928	you disable C<EV_MULTIPLICITY> when embedding libev).
		1929
		1930	Currently, functions, and static and non-static member functions can be
		1931	used as callbacks. Other types should be easy to add as long as they only
		1932	need one additional pointer for context. If you need support for other
		1933	types of functors please contact the author (preferably after implementing
		1934	it).
1644		1935
1645	Here is a list of things available in the C<ev> namespace:	1936	Here is a list of things available in the C<ev> namespace:
1646		1937
1647	=over 4	1938	=over 4
1648		1939
…		…
1664		1955
1665	All of those classes have these methods:	1956	All of those classes have these methods:
1666		1957
1667	=over 4	1958	=over 4
1668		1959
1669	=item ev::TYPE::TYPE (object , object::method )	1960	=item ev::TYPE::TYPE ()
1670		1961
1671	=item ev::TYPE::TYPE (object , object::method , struct ev_loop *)	1962	=item ev::TYPE::TYPE (struct ev_loop *)
1672		1963
1673	=item ev::TYPE::~TYPE	1964	=item ev::TYPE::~TYPE
1674		1965
1675	The constructor takes a pointer to an object and a method pointer to	1966	The constructor (optionally) takes an event loop to associate the watcher
1676	the event handler callback to call in this class. The constructor calls	1967	with. If it is omitted, it will use C<EV_DEFAULT>.
1677	C<ev_init> for you, which means you have to call the C<set> method	1968
1678	before starting it. If you do not specify a loop then the constructor	1969	The constructor calls C<ev_init> for you, which means you have to call the
1679	automatically associates the default loop with this watcher.	1970	C<set> method before starting it.
		1971
		1972	It will not set a callback, however: You have to call the templated C<set>
		1973	method to set a callback before you can start the watcher.
		1974
		1975	(The reason why you have to use a method is a limitation in C++ which does
		1976	not allow explicit template arguments for constructors).
1680		1977
1681	The destructor automatically stops the watcher if it is active.	1978	The destructor automatically stops the watcher if it is active.
		1979
		1980	=item w->set<class, &class::method> (object *)
		1981
		1982	This method sets the callback method to call. The method has to have a
		1983	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		1984	first argument and the C<revents> as second. The object must be given as
		1985	parameter and is stored in the C<data> member of the watcher.
		1986
		1987	This method synthesizes efficient thunking code to call your method from
		1988	the C callback that libev requires. If your compiler can inline your
		1989	callback (i.e. it is visible to it at the place of the C<set> call and
		1990	your compiler is good :), then the method will be fully inlined into the
		1991	thunking function, making it as fast as a direct C callback.
		1992
		1993	Example: simple class declaration and watcher initialisation
		1994
		1995	struct myclass
		1996	{
		1997	void io_cb (ev::io &w, int revents) { }
		1998	}
		1999
		2000	myclass obj;
		2001	ev::io iow;
		2002	iow.set <myclass, &myclass::io_cb> (&obj);
		2003
		2004	=item w->set<function> (void *data = 0)
		2005
		2006	Also sets a callback, but uses a static method or plain function as
		2007	callback. The optional C<data> argument will be stored in the watcher's
		2008	C<data> member and is free for you to use.
		2009
		2010	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2011
		2012	See the method-C<set> above for more details.
		2013
		2014	Example:
		2015
		2016	static void io_cb (ev::io &w, int revents) { }
		2017	iow.set <io_cb> ();
1682		2018
1683	=item w->set (struct ev_loop *)	2019	=item w->set (struct ev_loop *)
1684		2020
1685	Associates a different C<struct ev_loop> with this watcher. You can only	2021	Associates a different C<struct ev_loop> with this watcher. You can only
1686	do this when the watcher is inactive (and not pending either).	2022	do this when the watcher is inactive (and not pending either).
1687		2023
1688	=item w->set ([args])	2024	=item w->set ([args])
1689		2025
1690	Basically the same as C<ev_TYPE_set>, with the same args. Must be	2026	Basically the same as C<ev_TYPE_set>, with the same args. Must be
1691	called at least once. Unlike the C counterpart, an active watcher gets	2027	called at least once. Unlike the C counterpart, an active watcher gets
1692	automatically stopped and restarted.	2028	automatically stopped and restarted when reconfiguring it with this
		2029	method.
1693		2030
1694	=item w->start ()	2031	=item w->start ()
1695		2032
1696	Starts the watcher. Note that there is no C<loop> argument as the	2033	Starts the watcher. Note that there is no C<loop> argument, as the
1697	constructor already takes the loop.	2034	constructor already stores the event loop.
1698		2035
1699	=item w->stop ()	2036	=item w->stop ()
1700		2037
1701	Stops the watcher if it is active. Again, no C<loop> argument.	2038	Stops the watcher if it is active. Again, no C<loop> argument.
1702		2039
1703	=item w->again () C<ev::timer>, C<ev::periodic> only	2040	=item w->again () (C<ev::timer>, C<ev::periodic> only)
1704		2041
1705	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding	2042	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
1706	C<ev_TYPE_again> function.	2043	C<ev_TYPE_again> function.
1707		2044
1708	=item w->sweep () C<ev::embed> only	2045	=item w->sweep () (C<ev::embed> only)
1709		2046
1710	Invokes C<ev_embed_sweep>.	2047	Invokes C<ev_embed_sweep>.
1711		2048
1712	=item w->update () C<ev::stat> only	2049	=item w->update () (C<ev::stat> only)
1713		2050
1714	Invokes C<ev_stat_stat>.	2051	Invokes C<ev_stat_stat>.
1715		2052
1716	=back	2053	=back
1717		2054
…		…
1727		2064
1728	myclass ();	2065	myclass ();
1729	}	2066	}
1730		2067
1731	myclass::myclass (int fd)	2068	myclass::myclass (int fd)
1732	: io (this, &myclass::io_cb),
1733	idle (this, &myclass::idle_cb)
1734	{	2069	{
		2070	io .set <myclass, &myclass::io_cb > (this);
		2071	idle.set <myclass, &myclass::idle_cb> (this);
		2072
1735	io.start (fd, ev::READ);	2073	io.start (fd, ev::READ);
1736	}	2074	}
1737		2075
1738		2076
1739	=head1 MACRO MAGIC	2077	=head1 MACRO MAGIC
1740		2078
1741	Libev can be compiled with a variety of options, the most fundemantal is	2079	Libev can be compiled with a variety of options, the most fundamantal
1742	C<EV_MULTIPLICITY>. This option determines wether (most) functions and	2080	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
1743	callbacks have an initial C<struct ev_loop *> argument.	2081	functions and callbacks have an initial C<struct ev_loop *> argument.
1744		2082
1745	To make it easier to write programs that cope with either variant, the	2083	To make it easier to write programs that cope with either variant, the
1746	following macros are defined:	2084	following macros are defined:
1747		2085
1748	=over 4	2086	=over 4
…		…
1780	Similar to the other two macros, this gives you the value of the default	2118	Similar to the other two macros, this gives you the value of the default
1781	loop, if multiple loops are supported ("ev loop default").	2119	loop, if multiple loops are supported ("ev loop default").
1782		2120
1783	=back	2121	=back
1784		2122
1785	Example: Declare and initialise a check watcher, working regardless of	2123	Example: Declare and initialise a check watcher, utilising the above
1786	wether multiple loops are supported or not.	2124	macros so it will work regardless of whether multiple loops are supported
		2125	or not.
1787		2126
1788	static void	2127	static void
1789	check_cb (EV_P_ ev_timer *w, int revents)	2128	check_cb (EV_P_ ev_timer *w, int revents)
1790	{	2129	{
1791	ev_check_stop (EV_A_ w);	2130	ev_check_stop (EV_A_ w);
…		…
1793		2132
1794	ev_check check;	2133	ev_check check;
1795	ev_check_init (&check, check_cb);	2134	ev_check_init (&check, check_cb);
1796	ev_check_start (EV_DEFAULT_ &check);	2135	ev_check_start (EV_DEFAULT_ &check);
1797	ev_loop (EV_DEFAULT_ 0);	2136	ev_loop (EV_DEFAULT_ 0);
1798
1799		2137
1800	=head1 EMBEDDING	2138	=head1 EMBEDDING
1801		2139
1802	Libev can (and often is) directly embedded into host	2140	Libev can (and often is) directly embedded into host
1803	applications. Examples of applications that embed it include the Deliantra	2141	applications. Examples of applications that embed it include the Deliantra
…		…
1843	ev_vars.h	2181	ev_vars.h
1844	ev_wrap.h	2182	ev_wrap.h
1845		2183
1846	ev_win32.c required on win32 platforms only	2184	ev_win32.c required on win32 platforms only
1847		2185
1848	ev_select.c only when select backend is enabled (which is by default)	2186	ev_select.c only when select backend is enabled (which is enabled by default)
1849	ev_poll.c only when poll backend is enabled (disabled by default)	2187	ev_poll.c only when poll backend is enabled (disabled by default)
1850	ev_epoll.c only when the epoll backend is enabled (disabled by default)	2188	ev_epoll.c only when the epoll backend is enabled (disabled by default)
1851	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	2189	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1852	ev_port.c only when the solaris port backend is enabled (disabled by default)	2190	ev_port.c only when the solaris port backend is enabled (disabled by default)
1853		2191
…		…
1978		2316
1979	=item EV_USE_DEVPOLL	2317	=item EV_USE_DEVPOLL
1980		2318
1981	reserved for future expansion, works like the USE symbols above.	2319	reserved for future expansion, works like the USE symbols above.
1982		2320
		2321	=item EV_USE_INOTIFY
		2322
		2323	If defined to be C<1>, libev will compile in support for the Linux inotify
		2324	interface to speed up C<ev_stat> watchers. Its actual availability will
		2325	be detected at runtime.
		2326
1983	=item EV_H	2327	=item EV_H
1984		2328
1985	The name of the F<ev.h> header file used to include it. The default if	2329	The name of the F<ev.h> header file used to include it. The default if
1986	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This	2330	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This
1987	can be used to virtually rename the F<ev.h> header file in case of conflicts.	2331	can be used to virtually rename the F<ev.h> header file in case of conflicts.
…		…
2010	will have the C<struct ev_loop *> as first argument, and you can create	2354	will have the C<struct ev_loop *> as first argument, and you can create
2011	additional independent event loops. Otherwise there will be no support	2355	additional independent event loops. Otherwise there will be no support
2012	for multiple event loops and there is no first event loop pointer	2356	for multiple event loops and there is no first event loop pointer
2013	argument. Instead, all functions act on the single default loop.	2357	argument. Instead, all functions act on the single default loop.
2014		2358
		2359	=item EV_MINPRI
		2360
		2361	=item EV_MAXPRI
		2362
		2363	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2364	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2365	provide for more priorities by overriding those symbols (usually defined
		2366	to be C<-2> and C<2>, respectively).
		2367
		2368	When doing priority-based operations, libev usually has to linearly search
		2369	all the priorities, so having many of them (hundreds) uses a lot of space
		2370	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2371	fine.
		2372
		2373	If your embedding app does not need any priorities, defining these both to
		2374	C<0> will save some memory and cpu.
		2375
2015	=item EV_PERIODIC_ENABLE	2376	=item EV_PERIODIC_ENABLE
2016		2377
2017	If undefined or defined to be C<1>, then periodic timers are supported. If	2378	If undefined or defined to be C<1>, then periodic timers are supported. If
2018	defined to be C<0>, then they are not. Disabling them saves a few kB of	2379	defined to be C<0>, then they are not. Disabling them saves a few kB of
2019	code.	2380	code.
2020		2381
		2382	=item EV_IDLE_ENABLE
		2383
		2384	If undefined or defined to be C<1>, then idle watchers are supported. If
		2385	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2386	code.
		2387
2021	=item EV_EMBED_ENABLE	2388	=item EV_EMBED_ENABLE
2022		2389
2023	If undefined or defined to be C<1>, then embed watchers are supported. If	2390	If undefined or defined to be C<1>, then embed watchers are supported. If
2024	defined to be C<0>, then they are not.	2391	defined to be C<0>, then they are not.
2025		2392
…		…
2042	=item EV_PID_HASHSIZE	2409	=item EV_PID_HASHSIZE
2043		2410
2044	C<ev_child> watchers use a small hash table to distribute workload by	2411	C<ev_child> watchers use a small hash table to distribute workload by
2045	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	2412	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
2046	than enough. If you need to manage thousands of children you might want to	2413	than enough. If you need to manage thousands of children you might want to
2047	increase this value.	2414	increase this value (I<must> be a power of two).
		2415
		2416	=item EV_INOTIFY_HASHSIZE
		2417
		2418	C<ev_staz> watchers use a small hash table to distribute workload by
		2419	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		2420	usually more than enough. If you need to manage thousands of C<ev_stat>
		2421	watchers you might want to increase this value (I<must> be a power of
		2422	two).
2048		2423
2049	=item EV_COMMON	2424	=item EV_COMMON
2050		2425
2051	By default, all watchers have a C<void *data> member. By redefining	2426	By default, all watchers have a C<void *data> member. By redefining
2052	this macro to a something else you can include more and other types of	2427	this macro to a something else you can include more and other types of
…		…
2081	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file	2456	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
2082	will be compiled. It is pretty complex because it provides its own header	2457	will be compiled. It is pretty complex because it provides its own header
2083	file.	2458	file.
2084		2459
2085	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	2460	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2086	that everybody includes and which overrides some autoconf choices:	2461	that everybody includes and which overrides some configure choices:
2087		2462
		2463	#define EV_MINIMAL 1
2088	#define EV_USE_POLL 0	2464	#define EV_USE_POLL 0
2089	#define EV_MULTIPLICITY 0	2465	#define EV_MULTIPLICITY 0
2090	#define EV_PERIODICS 0	2466	#define EV_PERIODIC_ENABLE 0
		2467	#define EV_STAT_ENABLE 0
		2468	#define EV_FORK_ENABLE 0
2091	#define EV_CONFIG_H <config.h>	2469	#define EV_CONFIG_H <config.h>
		2470	#define EV_MINPRI 0
		2471	#define EV_MAXPRI 0
2092		2472
2093	#include "ev++.h"	2473	#include "ev++.h"
2094		2474
2095	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	2475	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2096		2476
…		…
2102		2482
2103	In this section the complexities of (many of) the algorithms used inside	2483	In this section the complexities of (many of) the algorithms used inside
2104	libev will be explained. For complexity discussions about backends see the	2484	libev will be explained. For complexity discussions about backends see the
2105	documentation for C<ev_default_init>.	2485	documentation for C<ev_default_init>.
2106		2486
		2487	All of the following are about amortised time: If an array needs to be
		2488	extended, libev needs to realloc and move the whole array, but this
		2489	happens asymptotically never with higher number of elements, so O(1) might
		2490	mean it might do a lengthy realloc operation in rare cases, but on average
		2491	it is much faster and asymptotically approaches constant time.
		2492
2107	=over 4	2493	=over 4
2108		2494
2109	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	2495	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2110		2496
		2497	This means that, when you have a watcher that triggers in one hour and
		2498	there are 100 watchers that would trigger before that then inserting will
		2499	have to skip those 100 watchers.
		2500
2111	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)	2501	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
2112		2502
		2503	That means that for changing a timer costs less than removing/adding them
		2504	as only the relative motion in the event queue has to be paid for.
		2505
2113	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	2506	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
2114		2507
		2508	These just add the watcher into an array or at the head of a list.
2115	=item Stopping check/prepare/idle watchers: O(1)	2509	=item Stopping check/prepare/idle watchers: O(1)
2116		2510
2117	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))	2511	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		2512
		2513	These watchers are stored in lists then need to be walked to find the
		2514	correct watcher to remove. The lists are usually short (you don't usually
		2515	have many watchers waiting for the same fd or signal).
2118		2516
2119	=item Finding the next timer per loop iteration: O(1)	2517	=item Finding the next timer per loop iteration: O(1)
2120		2518
2121	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	2519	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2122		2520
		2521	A change means an I/O watcher gets started or stopped, which requires
		2522	libev to recalculate its status (and possibly tell the kernel).
		2523
2123	=item Activating one watcher: O(1)	2524	=item Activating one watcher: O(1)
2124		2525
		2526	=item Priority handling: O(number_of_priorities)
		2527
		2528	Priorities are implemented by allocating some space for each
		2529	priority. When doing priority-based operations, libev usually has to
		2530	linearly search all the priorities.
		2531
2125	=back	2532	=back
2126		2533
2127		2534
2128	=head1 AUTHOR	2535	=head1 AUTHOR
2129		2536

Diff Legend

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

Comparing libev/ev.pod (file contents): Revision 1.53 by root, Tue Nov 27 20:15:02 2007 UTC vs. Revision 1.88 by ayin, Tue Dec 18 13:06:18 2007 UTC

Diff Legend

Comparing libev/ev.pod (file contents):
Revision 1.53 by root, Tue Nov 27 20:15:02 2007 UTC vs.
Revision 1.88 by ayin, Tue Dec 18 13:06:18 2007 UTC