ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.188 by root, Tue Sep 30 18:35:47 2008 UTC vs.
Revision 1.227 by root, Wed Mar 4 14:33:10 2009 UTC

…		…
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	// a single header file is required	11	// a single header file is required
12	#include <ev.h>	12	#include <ev.h>
13		13
		14	#include <stdio.h> // for puts
		15
14	// every watcher type has its own typedef'd struct	16	// every watcher type has its own typedef'd struct
15	// with the name ev_<type>	17	// with the name ev_TYPE
16	ev_io stdin_watcher;	18	ev_io stdin_watcher;
17	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
18		20
19	// all watcher callbacks have a similar signature	21	// all watcher callbacks have a similar signature
20	// this callback is called when data is readable on stdin	22	// this callback is called when data is readable on stdin
21	static void	23	static void
22	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
23	{	25	{
24	puts ("stdin ready");	26	puts ("stdin ready");
25	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
26	// with its corresponding stop function.	28	// with its corresponding stop function.
27	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
…		…
30	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_unloop (EV_A_ EVUNLOOP_ALL);
31	}	33	}
32		34
33	// another callback, this time for a time-out	35	// another callback, this time for a time-out
34	static void	36	static void
35	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
36	{	38	{
37	puts ("timeout");	39	puts ("timeout");
38	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_loop to stop iterating
39	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_unloop (EV_A_ EVUNLOOP_ONE);
40	}	42	}
…		…
103	Libev is very configurable. In this manual the default (and most common)	105	Libev is very configurable. In this manual the default (and most common)
104	configuration will be described, which supports multiple event loops. For	106	configuration will be described, which supports multiple event loops. For
105	more info about various configuration options please have a look at	107	more info about various configuration options please have a look at
106	B<EMBED> section in this manual. If libev was configured without support	108	B<EMBED> section in this manual. If libev was configured without support
107	for multiple event loops, then all functions taking an initial argument of	109	for multiple event loops, then all functions taking an initial argument of
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	110	name C<loop> (which is always of type C<ev_loop *>) will not have
109	this argument.	111	this argument.
110		112
111	=head2 TIME REPRESENTATION	113	=head2 TIME REPRESENTATION
112		114
113	Libev represents time as a single floating point number, representing the	115	Libev represents time as a single floating point number, representing the
…		…
276		278
277	=back	279	=back
278		280
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	281	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
280		282
281	An event loop is described by a C<struct ev_loop *>. The library knows two	283	An event loop is described by a C<struct ev_loop *> (the C<struct>
282	types of such loops, the I<default> loop, which supports signals and child	284	is I<not> optional in this case, as there is also an C<ev_loop>
283	events, and dynamically created loops which do not.	285	I<function>).
		286
		287	The library knows two types of such loops, the I<default> loop, which
		288	supports signals and child events, and dynamically created loops which do
		289	not.
284		290
285	=over 4	291	=over 4
286		292
287	=item struct ev_loop *ev_default_loop (unsigned int flags)	293	=item struct ev_loop *ev_default_loop (unsigned int flags)
288		294
…		…
294	If you don't know what event loop to use, use the one returned from this	300	If you don't know what event loop to use, use the one returned from this
295	function.	301	function.
296		302
297	Note that this function is I<not> thread-safe, so if you want to use it	303	Note that this function is I<not> thread-safe, so if you want to use it
298	from multiple threads, you have to lock (note also that this is unlikely,	304	from multiple threads, you have to lock (note also that this is unlikely,
299	as loops cannot bes hared easily between threads anyway).	305	as loops cannot be shared easily between threads anyway).
300		306
301	The default loop is the only loop that can handle C<ev_signal> and	307	The default loop is the only loop that can handle C<ev_signal> and
302	C<ev_child> watchers, and to do this, it always registers a handler	308	C<ev_child> watchers, and to do this, it always registers a handler
303	for C<SIGCHLD>. If this is a problem for your application you can either	309	for C<SIGCHLD>. If this is a problem for your application you can either
304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	310	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
380	=item C<EVBACKEND_EPOLL> (value 4, Linux)	386	=item C<EVBACKEND_EPOLL> (value 4, Linux)
381		387
382	For few fds, this backend is a bit little slower than poll and select,	388	For few fds, this backend is a bit little slower than poll and select,
383	but it scales phenomenally better. While poll and select usually scale	389	but it scales phenomenally better. While poll and select usually scale
384	like O(total_fds) where n is the total number of fds (or the highest fd),	390	like O(total_fds) where n is the total number of fds (or the highest fd),
385	epoll scales either O(1) or O(active_fds). The epoll design has a number	391	epoll scales either O(1) or O(active_fds).
386	of shortcomings, such as silently dropping events in some hard-to-detect	392
387	cases and requiring a system call per fd change, no fork support and bad	393	The epoll mechanism deserves honorable mention as the most misdesigned
388	support for dup.	394	of the more advanced event mechanisms: mere annoyances include silently
		395	dropping file descriptors, requiring a system call per change per file
		396	descriptor (and unnecessary guessing of parameters), problems with dup and
		397	so on. The biggest issue is fork races, however - if a program forks then
		398	I<both> parent and child process have to recreate the epoll set, which can
		399	take considerable time (one syscall per file descriptor) and is of course
		400	hard to detect.
		401
		402	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		403	of course I<doesn't>, and epoll just loves to report events for totally
		404	I<different> file descriptors (even already closed ones, so one cannot
		405	even remove them from the set) than registered in the set (especially
		406	on SMP systems). Libev tries to counter these spurious notifications by
		407	employing an additional generation counter and comparing that against the
		408	events to filter out spurious ones, recreating the set when required.
389		409
390	While stopping, setting and starting an I/O watcher in the same iteration	410	While stopping, setting and starting an I/O watcher in the same iteration
391	will result in some caching, there is still a system call per such incident	411	will result in some caching, there is still a system call per such
392	(because the fd could point to a different file description now), so its	412	incident (because the same I<file descriptor> could point to a different
393	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	413	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
394	very well if you register events for both fds.	414	file descriptors might not work very well if you register events for both
395		415	file descriptors.
396	Please note that epoll sometimes generates spurious notifications, so you
397	need to use non-blocking I/O or other means to avoid blocking when no data
398	(or space) is available.
399		416
400	Best performance from this backend is achieved by not unregistering all	417	Best performance from this backend is achieved by not unregistering all
401	watchers for a file descriptor until it has been closed, if possible,	418	watchers for a file descriptor until it has been closed, if possible,
402	i.e. keep at least one watcher active per fd at all times. Stopping and	419	i.e. keep at least one watcher active per fd at all times. Stopping and
403	starting a watcher (without re-setting it) also usually doesn't cause	420	starting a watcher (without re-setting it) also usually doesn't cause
404	extra overhead.	421	extra overhead. A fork can both result in spurious notifications as well
		422	as in libev having to destroy and recreate the epoll object, which can
		423	take considerable time and thus should be avoided.
		424
		425	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		426	faster than epoll for maybe up to a hundred file descriptors, depending on
		427	the usage. So sad.
405		428
406	While nominally embeddable in other event loops, this feature is broken in	429	While nominally embeddable in other event loops, this feature is broken in
407	all kernel versions tested so far.	430	all kernel versions tested so far.
408		431
409	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	432	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
410	C<EVBACKEND_POLL>.	433	C<EVBACKEND_POLL>.
411		434
412	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	435	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
413		436
414	Kqueue deserves special mention, as at the time of this writing, it was	437	Kqueue deserves special mention, as at the time of this writing, it
415	broken on all BSDs except NetBSD (usually it doesn't work reliably with	438	was broken on all BSDs except NetBSD (usually it doesn't work reliably
416	anything but sockets and pipes, except on Darwin, where of course it's	439	with anything but sockets and pipes, except on Darwin, where of course
417	completely useless). For this reason it's not being "auto-detected" unless	440	it's completely useless). Unlike epoll, however, whose brokenness
418	you explicitly specify it in the flags (i.e. using C<EVBACKEND_KQUEUE>) or	441	is by design, these kqueue bugs can (and eventually will) be fixed
419	libev was compiled on a known-to-be-good (-enough) system like NetBSD.	442	without API changes to existing programs. For this reason it's not being
		443	"auto-detected" unless you explicitly specify it in the flags (i.e. using
		444	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		445	system like NetBSD.
420		446
421	You still can embed kqueue into a normal poll or select backend and use it	447	You still can embed kqueue into a normal poll or select backend and use it
422	only for sockets (after having made sure that sockets work with kqueue on	448	only for sockets (after having made sure that sockets work with kqueue on
423	the target platform). See C<ev_embed> watchers for more info.	449	the target platform). See C<ev_embed> watchers for more info.
424		450
425	It scales in the same way as the epoll backend, but the interface to the	451	It scales in the same way as the epoll backend, but the interface to the
426	kernel is more efficient (which says nothing about its actual speed, of	452	kernel is more efficient (which says nothing about its actual speed, of
427	course). While stopping, setting and starting an I/O watcher does never	453	course). While stopping, setting and starting an I/O watcher does never
428	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	454	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
429	two event changes per incident. Support for C<fork ()> is very bad and it	455	two event changes per incident. Support for C<fork ()> is very bad (but
430	drops fds silently in similarly hard-to-detect cases.	456	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		457	cases
431		458
432	This backend usually performs well under most conditions.	459	This backend usually performs well under most conditions.
433		460
434	While nominally embeddable in other event loops, this doesn't work	461	While nominally embeddable in other event loops, this doesn't work
435	everywhere, so you might need to test for this. And since it is broken	462	everywhere, so you might need to test for this. And since it is broken
436	almost everywhere, you should only use it when you have a lot of sockets	463	almost everywhere, you should only use it when you have a lot of sockets
437	(for which it usually works), by embedding it into another event loop	464	(for which it usually works), by embedding it into another event loop
438	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,	465	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
439	using it only for sockets.	466	also broken on OS X)) and, did I mention it, using it only for sockets.
440		467
441	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with	468	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
442	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with	469	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
443	C<NOTE_EOF>.	470	C<NOTE_EOF>.
444		471
…		…
464	might perform better.	491	might perform better.
465		492
466	On the positive side, with the exception of the spurious readiness	493	On the positive side, with the exception of the spurious readiness
467	notifications, this backend actually performed fully to specification	494	notifications, this backend actually performed fully to specification
468	in all tests and is fully embeddable, which is a rare feat among the	495	in all tests and is fully embeddable, which is a rare feat among the
469	OS-specific backends.	496	OS-specific backends (I vastly prefer correctness over speed hacks).
470		497
471	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	498	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
472	C<EVBACKEND_POLL>.	499	C<EVBACKEND_POLL>.
473		500
474	=item C<EVBACKEND_ALL>	501	=item C<EVBACKEND_ALL>
…		…
527	responsibility to either stop all watchers cleanly yourself I<before>	554	responsibility to either stop all watchers cleanly yourself I<before>
528	calling this function, or cope with the fact afterwards (which is usually	555	calling this function, or cope with the fact afterwards (which is usually
529	the easiest thing, you can just ignore the watchers and/or C<free ()> them	556	the easiest thing, you can just ignore the watchers and/or C<free ()> them
530	for example).	557	for example).
531		558
532	Note that certain global state, such as signal state, will not be freed by	559	Note that certain global state, such as signal state (and installed signal
533	this function, and related watchers (such as signal and child watchers)	560	handlers), will not be freed by this function, and related watchers (such
534	would need to be stopped manually.	561	as signal and child watchers) would need to be stopped manually.
535		562
536	In general it is not advisable to call this function except in the	563	In general it is not advisable to call this function except in the
537	rare occasion where you really need to free e.g. the signal handling	564	rare occasion where you really need to free e.g. the signal handling
538	pipe fds. If you need dynamically allocated loops it is better to use	565	pipe fds. If you need dynamically allocated loops it is better to use
539	C<ev_loop_new> and C<ev_loop_destroy>).	566	C<ev_loop_new> and C<ev_loop_destroy>).
…		…
631	the loop.	658	the loop.
632		659
633	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	660	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
634	necessary) and will handle those and any already outstanding ones. It	661	necessary) and will handle those and any already outstanding ones. It
635	will block your process until at least one new event arrives (which could	662	will block your process until at least one new event arrives (which could
636	be an event internal to libev itself, so there is no guarentee that a	663	be an event internal to libev itself, so there is no guarantee that a
637	user-registered callback will be called), and will return after one	664	user-registered callback will be called), and will return after one
638	iteration of the loop.	665	iteration of the loop.
639		666
640	This is useful if you are waiting for some external event in conjunction	667	This is useful if you are waiting for some external event in conjunction
641	with something not expressible using other libev watchers (i.e. "roll your	668	with something not expressible using other libev watchers (i.e. "roll your
…		…
685	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	712	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
686	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	713	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
687		714
688	This "unloop state" will be cleared when entering C<ev_loop> again.	715	This "unloop state" will be cleared when entering C<ev_loop> again.
689		716
		717	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		718
690	=item ev_ref (loop)	719	=item ev_ref (loop)
691		720
692	=item ev_unref (loop)	721	=item ev_unref (loop)
693		722
694	Ref/unref can be used to add or remove a reference count on the event	723	Ref/unref can be used to add or remove a reference count on the event
…		…
708	respectively).	737	respectively).
709		738
710	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	739	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
711	running when nothing else is active.	740	running when nothing else is active.
712		741
713	struct ev_signal exitsig;	742	ev_signal exitsig;
714	ev_signal_init (&exitsig, sig_cb, SIGINT);	743	ev_signal_init (&exitsig, sig_cb, SIGINT);
715	ev_signal_start (loop, &exitsig);	744	ev_signal_start (loop, &exitsig);
716	evf_unref (loop);	745	evf_unref (loop);
717		746
718	Example: For some weird reason, unregister the above signal handler again.	747	Example: For some weird reason, unregister the above signal handler again.
…		…
766	they fire on, say, one-second boundaries only.	795	they fire on, say, one-second boundaries only.
767		796
768	=item ev_loop_verify (loop)	797	=item ev_loop_verify (loop)
769		798
770	This function only does something when C<EV_VERIFY> support has been	799	This function only does something when C<EV_VERIFY> support has been
771	compiled in. which is the default for non-minimal builds. It tries to go	800	compiled in, which is the default for non-minimal builds. It tries to go
772	through all internal structures and checks them for validity. If anything	801	through all internal structures and checks them for validity. If anything
773	is found to be inconsistent, it will print an error message to standard	802	is found to be inconsistent, it will print an error message to standard
774	error and call C<abort ()>.	803	error and call C<abort ()>.
775		804
776	This can be used to catch bugs inside libev itself: under normal	805	This can be used to catch bugs inside libev itself: under normal
…		…
780	=back	809	=back
781		810
782		811
783	=head1 ANATOMY OF A WATCHER	812	=head1 ANATOMY OF A WATCHER
784		813
		814	In the following description, uppercase C<TYPE> in names stands for the
		815	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		816	watchers and C<ev_io_start> for I/O watchers.
		817
785	A watcher is a structure that you create and register to record your	818	A watcher is a structure that you create and register to record your
786	interest in some event. For instance, if you want to wait for STDIN to	819	interest in some event. For instance, if you want to wait for STDIN to
787	become readable, you would create an C<ev_io> watcher for that:	820	become readable, you would create an C<ev_io> watcher for that:
788		821
789	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	822	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
790	{	823	{
791	ev_io_stop (w);	824	ev_io_stop (w);
792	ev_unloop (loop, EVUNLOOP_ALL);	825	ev_unloop (loop, EVUNLOOP_ALL);
793	}	826	}
794		827
795	struct ev_loop *loop = ev_default_loop (0);	828	struct ev_loop *loop = ev_default_loop (0);
		829
796	struct ev_io stdin_watcher;	830	ev_io stdin_watcher;
		831
797	ev_init (&stdin_watcher, my_cb);	832	ev_init (&stdin_watcher, my_cb);
798	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	833	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
799	ev_io_start (loop, &stdin_watcher);	834	ev_io_start (loop, &stdin_watcher);
		835
800	ev_loop (loop, 0);	836	ev_loop (loop, 0);
801		837
802	As you can see, you are responsible for allocating the memory for your	838	As you can see, you are responsible for allocating the memory for your
803	watcher structures (and it is usually a bad idea to do this on the stack,	839	watcher structures (and it is I<usually> a bad idea to do this on the
804	although this can sometimes be quite valid).	840	stack).
		841
		842	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		843	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
805		844
806	Each watcher structure must be initialised by a call to C<ev_init	845	Each watcher structure must be initialised by a call to C<ev_init
807	(watcher *, callback)>, which expects a callback to be provided. This	846	(watcher *, callback)>, which expects a callback to be provided. This
808	callback gets invoked each time the event occurs (or, in the case of I/O	847	callback gets invoked each time the event occurs (or, in the case of I/O
809	watchers, each time the event loop detects that the file descriptor given	848	watchers, each time the event loop detects that the file descriptor given
810	is readable and/or writable).	849	is readable and/or writable).
811		850
812	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	851	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
813	with arguments specific to this watcher type. There is also a macro	852	macro to configure it, with arguments specific to the watcher type. There
814	to combine initialisation and setting in one call: C<< ev_<type>_init	853	is also a macro to combine initialisation and setting in one call: C<<
815	(watcher *, callback, ...) >>.	854	ev_TYPE_init (watcher *, callback, ...) >>.
816		855
817	To make the watcher actually watch out for events, you have to start it	856	To make the watcher actually watch out for events, you have to start it
818	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	857	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
819	*) >>), and you can stop watching for events at any time by calling the	858	*) >>), and you can stop watching for events at any time by calling the
820	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	859	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
821		860
822	As long as your watcher is active (has been started but not stopped) you	861	As long as your watcher is active (has been started but not stopped) you
823	must not touch the values stored in it. Most specifically you must never	862	must not touch the values stored in it. Most specifically you must never
824	reinitialise it or call its C<set> macro.	863	reinitialise it or call its C<ev_TYPE_set> macro.
825		864
826	Each and every callback receives the event loop pointer as first, the	865	Each and every callback receives the event loop pointer as first, the
827	registered watcher structure as second, and a bitset of received events as	866	registered watcher structure as second, and a bitset of received events as
828	third argument.	867	third argument.
829		868
…		…
892	=item C<EV_ERROR>	931	=item C<EV_ERROR>
893		932
894	An unspecified error has occurred, the watcher has been stopped. This might	933	An unspecified error has occurred, the watcher has been stopped. This might
895	happen because the watcher could not be properly started because libev	934	happen because the watcher could not be properly started because libev
896	ran out of memory, a file descriptor was found to be closed or any other	935	ran out of memory, a file descriptor was found to be closed or any other
		936	problem. Libev considers these application bugs.
		937
897	problem. You best act on it by reporting the problem and somehow coping	938	You best act on it by reporting the problem and somehow coping with the
898	with the watcher being stopped.	939	watcher being stopped. Note that well-written programs should not receive
		940	an error ever, so when your watcher receives it, this usually indicates a
		941	bug in your program.
899		942
900	Libev will usually signal a few "dummy" events together with an error, for	943	Libev will usually signal a few "dummy" events together with an error, for
901	example it might indicate that a fd is readable or writable, and if your	944	example it might indicate that a fd is readable or writable, and if your
902	callbacks is well-written it can just attempt the operation and cope with	945	callbacks is well-written it can just attempt the operation and cope with
903	the error from read() or write(). This will not work in multi-threaded	946	the error from read() or write(). This will not work in multi-threaded
…		…
906		949
907	=back	950	=back
908		951
909	=head2 GENERIC WATCHER FUNCTIONS	952	=head2 GENERIC WATCHER FUNCTIONS
910		953
911	In the following description, C<TYPE> stands for the watcher type,
912	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
913
914	=over 4	954	=over 4
915		955
916	=item C<ev_init> (ev_TYPE *watcher, callback)	956	=item C<ev_init> (ev_TYPE *watcher, callback)
917		957
918	This macro initialises the generic portion of a watcher. The contents	958	This macro initialises the generic portion of a watcher. The contents
…		…
923	which rolls both calls into one.	963	which rolls both calls into one.
924		964
925	You can reinitialise a watcher at any time as long as it has been stopped	965	You can reinitialise a watcher at any time as long as it has been stopped
926	(or never started) and there are no pending events outstanding.	966	(or never started) and there are no pending events outstanding.
927		967
928	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	968	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
929	int revents)>.	969	int revents)>.
930		970
931	Example: Initialise an C<ev_io> watcher in two steps.	971	Example: Initialise an C<ev_io> watcher in two steps.
932		972
933	ev_io w;	973	ev_io w;
…		…
967		1007
968	ev_io_start (EV_DEFAULT_UC, &w);	1008	ev_io_start (EV_DEFAULT_UC, &w);
969		1009
970	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1010	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
971		1011
972	Stops the given watcher again (if active) and clears the pending	1012	Stops the given watcher if active, and clears the pending status (whether
		1013	the watcher was active or not).
		1014
973	status. It is possible that stopped watchers are pending (for example,	1015	It is possible that stopped watchers are pending - for example,
974	non-repeating timers are being stopped when they become pending), but	1016	non-repeating timers are being stopped when they become pending - but
975	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1017	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
976	you want to free or reuse the memory used by the watcher it is therefore a	1018	pending. If you want to free or reuse the memory used by the watcher it is
977	good idea to always call its C<ev_TYPE_stop> function.	1019	therefore a good idea to always call its C<ev_TYPE_stop> function.
978		1020
979	=item bool ev_is_active (ev_TYPE *watcher)	1021	=item bool ev_is_active (ev_TYPE *watcher)
980		1022
981	Returns a true value iff the watcher is active (i.e. it has been started	1023	Returns a true value iff the watcher is active (i.e. it has been started
982	and not yet been stopped). As long as a watcher is active you must not modify	1024	and not yet been stopped). As long as a watcher is active you must not modify
…		…
1024	The default priority used by watchers when no priority has been set is	1066	The default priority used by watchers when no priority has been set is
1025	always C<0>, which is supposed to not be too high and not be too low :).	1067	always C<0>, which is supposed to not be too high and not be too low :).
1026		1068
1027	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1069	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
1028	fine, as long as you do not mind that the priority value you query might	1070	fine, as long as you do not mind that the priority value you query might
1029	or might not have been adjusted to be within valid range.	1071	or might not have been clamped to the valid range.
1030		1072
1031	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1073	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1032		1074
1033	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1075	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1034	C<loop> nor C<revents> need to be valid as long as the watcher callback	1076	C<loop> nor C<revents> need to be valid as long as the watcher callback
…		…
1056	member, you can also "subclass" the watcher type and provide your own	1098	member, you can also "subclass" the watcher type and provide your own
1057	data:	1099	data:
1058		1100
1059	struct my_io	1101	struct my_io
1060	{	1102	{
1061	struct ev_io io;	1103	ev_io io;
1062	int otherfd;	1104	int otherfd;
1063	void *somedata;	1105	void *somedata;
1064	struct whatever *mostinteresting;	1106	struct whatever *mostinteresting;
1065	};	1107	};
1066		1108
…		…
1069	ev_io_init (&w.io, my_cb, fd, EV_READ);	1111	ev_io_init (&w.io, my_cb, fd, EV_READ);
1070		1112
1071	And since your callback will be called with a pointer to the watcher, you	1113	And since your callback will be called with a pointer to the watcher, you
1072	can cast it back to your own type:	1114	can cast it back to your own type:
1073		1115
1074	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1116	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1075	{	1117	{
1076	struct my_io *w = (struct my_io *)w_;	1118	struct my_io *w = (struct my_io *)w_;
1077	...	1119	...
1078	}	1120	}
1079		1121
…		…
1097	programmers):	1139	programmers):
1098		1140
1099	#include <stddef.h>	1141	#include <stddef.h>
1100		1142
1101	static void	1143	static void
1102	t1_cb (EV_P_ struct ev_timer *w, int revents)	1144	t1_cb (EV_P_ ev_timer *w, int revents)
1103	{	1145	{
1104	struct my_biggy big = (struct my_biggy *	1146	struct my_biggy big = (struct my_biggy *
1105	(((char *)w) - offsetof (struct my_biggy, t1));	1147	(((char *)w) - offsetof (struct my_biggy, t1));
1106	}	1148	}
1107		1149
1108	static void	1150	static void
1109	t2_cb (EV_P_ struct ev_timer *w, int revents)	1151	t2_cb (EV_P_ ev_timer *w, int revents)
1110	{	1152	{
1111	struct my_biggy big = (struct my_biggy *	1153	struct my_biggy big = (struct my_biggy *
1112	(((char *)w) - offsetof (struct my_biggy, t2));	1154	(((char *)w) - offsetof (struct my_biggy, t2));
1113	}	1155	}
1114		1156
…		…
1249	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1291	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1250	readable, but only once. Since it is likely line-buffered, you could	1292	readable, but only once. Since it is likely line-buffered, you could
1251	attempt to read a whole line in the callback.	1293	attempt to read a whole line in the callback.
1252		1294
1253	static void	1295	static void
1254	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1296	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1255	{	1297	{
1256	ev_io_stop (loop, w);	1298	ev_io_stop (loop, w);
1257	.. read from stdin here (or from w->fd) and handle any I/O errors	1299	.. read from stdin here (or from w->fd) and handle any I/O errors
1258	}	1300	}
1259		1301
1260	...	1302	...
1261	struct ev_loop *loop = ev_default_init (0);	1303	struct ev_loop *loop = ev_default_init (0);
1262	struct ev_io stdin_readable;	1304	ev_io stdin_readable;
1263	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1305	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1264	ev_io_start (loop, &stdin_readable);	1306	ev_io_start (loop, &stdin_readable);
1265	ev_loop (loop, 0);	1307	ev_loop (loop, 0);
1266		1308
1267		1309
…		…
1278		1320
1279	The callback is guaranteed to be invoked only I<after> its timeout has	1321	The callback is guaranteed to be invoked only I<after> its timeout has
1280	passed, but if multiple timers become ready during the same loop iteration	1322	passed, but if multiple timers become ready during the same loop iteration
1281	then order of execution is undefined.	1323	then order of execution is undefined.
1282		1324
		1325	=head3 Be smart about timeouts
		1326
		1327	Many real-world problems involve some kind of timeout, usually for error
		1328	recovery. A typical example is an HTTP request - if the other side hangs,
		1329	you want to raise some error after a while.
		1330
		1331	What follows are some ways to handle this problem, from obvious and
		1332	inefficient to smart and efficient.
		1333
		1334	In the following, a 60 second activity timeout is assumed - a timeout that
		1335	gets reset to 60 seconds each time there is activity (e.g. each time some
		1336	data or other life sign was received).
		1337
		1338	=over 4
		1339
		1340	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1341
		1342	This is the most obvious, but not the most simple way: In the beginning,
		1343	start the watcher:
		1344
		1345	ev_timer_init (timer, callback, 60., 0.);
		1346	ev_timer_start (loop, timer);
		1347
		1348	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1349	and start it again:
		1350
		1351	ev_timer_stop (loop, timer);
		1352	ev_timer_set (timer, 60., 0.);
		1353	ev_timer_start (loop, timer);
		1354
		1355	This is relatively simple to implement, but means that each time there is
		1356	some activity, libev will first have to remove the timer from its internal
		1357	data structure and then add it again. Libev tries to be fast, but it's
		1358	still not a constant-time operation.
		1359
		1360	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1361
		1362	This is the easiest way, and involves using C<ev_timer_again> instead of
		1363	C<ev_timer_start>.
		1364
		1365	To implement this, configure an C<ev_timer> with a C<repeat> value
		1366	of C<60> and then call C<ev_timer_again> at start and each time you
		1367	successfully read or write some data. If you go into an idle state where
		1368	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1369	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1370
		1371	That means you can ignore both the C<ev_timer_start> function and the
		1372	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1373	member and C<ev_timer_again>.
		1374
		1375	At start:
		1376
		1377	ev_timer_init (timer, callback);
		1378	timer->repeat = 60.;
		1379	ev_timer_again (loop, timer);
		1380
		1381	Each time there is some activity:
		1382
		1383	ev_timer_again (loop, timer);
		1384
		1385	It is even possible to change the time-out on the fly, regardless of
		1386	whether the watcher is active or not:
		1387
		1388	timer->repeat = 30.;
		1389	ev_timer_again (loop, timer);
		1390
		1391	This is slightly more efficient then stopping/starting the timer each time
		1392	you want to modify its timeout value, as libev does not have to completely
		1393	remove and re-insert the timer from/into its internal data structure.
		1394
		1395	It is, however, even simpler than the "obvious" way to do it.
		1396
		1397	=item 3. Let the timer time out, but then re-arm it as required.
		1398
		1399	This method is more tricky, but usually most efficient: Most timeouts are
		1400	relatively long compared to the intervals between other activity - in
		1401	our example, within 60 seconds, there are usually many I/O events with
		1402	associated activity resets.
		1403
		1404	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1405	but remember the time of last activity, and check for a real timeout only
		1406	within the callback:
		1407
		1408	ev_tstamp last_activity; // time of last activity
		1409
		1410	static void
		1411	callback (EV_P_ ev_timer *w, int revents)
		1412	{
		1413	ev_tstamp now = ev_now (EV_A);
		1414	ev_tstamp timeout = last_activity + 60.;
		1415
		1416	// if last_activity + 60. is older than now, we did time out
		1417	if (timeout < now)
		1418	{
		1419	// timeout occured, take action
		1420	}
		1421	else
		1422	{
		1423	// callback was invoked, but there was some activity, re-arm
		1424	// the watcher to fire in last_activity + 60, which is
		1425	// guaranteed to be in the future, so "again" is positive:
		1426	w->repeat = timeout - now;
		1427	ev_timer_again (EV_A_ w);
		1428	}
		1429	}
		1430
		1431	To summarise the callback: first calculate the real timeout (defined
		1432	as "60 seconds after the last activity"), then check if that time has
		1433	been reached, which means something I<did>, in fact, time out. Otherwise
		1434	the callback was invoked too early (C<timeout> is in the future), so
		1435	re-schedule the timer to fire at that future time, to see if maybe we have
		1436	a timeout then.
		1437
		1438	Note how C<ev_timer_again> is used, taking advantage of the
		1439	C<ev_timer_again> optimisation when the timer is already running.
		1440
		1441	This scheme causes more callback invocations (about one every 60 seconds
		1442	minus half the average time between activity), but virtually no calls to
		1443	libev to change the timeout.
		1444
		1445	To start the timer, simply initialise the watcher and set C<last_activity>
		1446	to the current time (meaning we just have some activity :), then call the
		1447	callback, which will "do the right thing" and start the timer:
		1448
		1449	ev_timer_init (timer, callback);
		1450	last_activity = ev_now (loop);
		1451	callback (loop, timer, EV_TIMEOUT);
		1452
		1453	And when there is some activity, simply store the current time in
		1454	C<last_activity>, no libev calls at all:
		1455
		1456	last_actiivty = ev_now (loop);
		1457
		1458	This technique is slightly more complex, but in most cases where the
		1459	time-out is unlikely to be triggered, much more efficient.
		1460
		1461	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1462	callback :) - just change the timeout and invoke the callback, which will
		1463	fix things for you.
		1464
		1465	=item 4. Wee, just use a double-linked list for your timeouts.
		1466
		1467	If there is not one request, but many thousands (millions...), all
		1468	employing some kind of timeout with the same timeout value, then one can
		1469	do even better:
		1470
		1471	When starting the timeout, calculate the timeout value and put the timeout
		1472	at the I<end> of the list.
		1473
		1474	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1475	the list is expected to fire (for example, using the technique #3).
		1476
		1477	When there is some activity, remove the timer from the list, recalculate
		1478	the timeout, append it to the end of the list again, and make sure to
		1479	update the C<ev_timer> if it was taken from the beginning of the list.
		1480
		1481	This way, one can manage an unlimited number of timeouts in O(1) time for
		1482	starting, stopping and updating the timers, at the expense of a major
		1483	complication, and having to use a constant timeout. The constant timeout
		1484	ensures that the list stays sorted.
		1485
		1486	=back
		1487
		1488	So which method the best?
		1489
		1490	Method #2 is a simple no-brain-required solution that is adequate in most
		1491	situations. Method #3 requires a bit more thinking, but handles many cases
		1492	better, and isn't very complicated either. In most case, choosing either
		1493	one is fine, with #3 being better in typical situations.
		1494
		1495	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1496	rather complicated, but extremely efficient, something that really pays
		1497	off after the first million or so of active timers, i.e. it's usually
		1498	overkill :)
		1499
1283	=head3 The special problem of time updates	1500	=head3 The special problem of time updates
1284		1501
1285	Establishing the current time is a costly operation (it usually takes at	1502	Establishing the current time is a costly operation (it usually takes at
1286	least two system calls): EV therefore updates its idea of the current	1503	least two system calls): EV therefore updates its idea of the current
1287	time only before and after C<ev_loop> collects new events, which causes a	1504	time only before and after C<ev_loop> collects new events, which causes a
…		…
1330	If the timer is started but non-repeating, stop it (as if it timed out).	1547	If the timer is started but non-repeating, stop it (as if it timed out).
1331		1548
1332	If the timer is repeating, either start it if necessary (with the	1549	If the timer is repeating, either start it if necessary (with the
1333	C<repeat> value), or reset the running timer to the C<repeat> value.	1550	C<repeat> value), or reset the running timer to the C<repeat> value.
1334		1551
1335	This sounds a bit complicated, but here is a useful and typical	1552	This sounds a bit complicated, see "Be smart about timeouts", above, for a
1336	example: Imagine you have a TCP connection and you want a so-called idle	1553	usage example.
1337	timeout, that is, you want to be called when there have been, say, 60
1338	seconds of inactivity on the socket. The easiest way to do this is to
1339	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1340	C<ev_timer_again> each time you successfully read or write some data. If
1341	you go into an idle state where you do not expect data to travel on the
1342	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1343	automatically restart it if need be.
1344
1345	That means you can ignore the C<after> value and C<ev_timer_start>
1346	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1347
1348	ev_timer_init (timer, callback, 0., 5.);
1349	ev_timer_again (loop, timer);
1350	...
1351	timer->again = 17.;
1352	ev_timer_again (loop, timer);
1353	...
1354	timer->again = 10.;
1355	ev_timer_again (loop, timer);
1356
1357	This is more slightly efficient then stopping/starting the timer each time
1358	you want to modify its timeout value.
1359
1360	Note, however, that it is often even more efficient to remember the
1361	time of the last activity and let the timer time-out naturally. In the
1362	callback, you then check whether the time-out is real, or, if there was
1363	some activity, you reschedule the watcher to time-out in "last_activity +
1364	timeout - ev_now ()" seconds.
1365		1554
1366	=item ev_tstamp repeat [read-write]	1555	=item ev_tstamp repeat [read-write]
1367		1556
1368	The current C<repeat> value. Will be used each time the watcher times out	1557	The current C<repeat> value. Will be used each time the watcher times out
1369	or C<ev_timer_again> is called, and determines the next timeout (if any),	1558	or C<ev_timer_again> is called, and determines the next timeout (if any),
…		…
1374	=head3 Examples	1563	=head3 Examples
1375		1564
1376	Example: Create a timer that fires after 60 seconds.	1565	Example: Create a timer that fires after 60 seconds.
1377		1566
1378	static void	1567	static void
1379	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1568	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1380	{	1569	{
1381	.. one minute over, w is actually stopped right here	1570	.. one minute over, w is actually stopped right here
1382	}	1571	}
1383		1572
1384	struct ev_timer mytimer;	1573	ev_timer mytimer;
1385	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1574	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1386	ev_timer_start (loop, &mytimer);	1575	ev_timer_start (loop, &mytimer);
1387		1576
1388	Example: Create a timeout timer that times out after 10 seconds of	1577	Example: Create a timeout timer that times out after 10 seconds of
1389	inactivity.	1578	inactivity.
1390		1579
1391	static void	1580	static void
1392	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1581	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1393	{	1582	{
1394	.. ten seconds without any activity	1583	.. ten seconds without any activity
1395	}	1584	}
1396		1585
1397	struct ev_timer mytimer;	1586	ev_timer mytimer;
1398	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1587	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1399	ev_timer_again (&mytimer); /* start timer */	1588	ev_timer_again (&mytimer); /* start timer */
1400	ev_loop (loop, 0);	1589	ev_loop (loop, 0);
1401		1590
1402	// and in some piece of code that gets executed on any "activity":	1591	// and in some piece of code that gets executed on any "activity":
…		…
1407	=head2 C<ev_periodic> - to cron or not to cron?	1596	=head2 C<ev_periodic> - to cron or not to cron?
1408		1597
1409	Periodic watchers are also timers of a kind, but they are very versatile	1598	Periodic watchers are also timers of a kind, but they are very versatile
1410	(and unfortunately a bit complex).	1599	(and unfortunately a bit complex).
1411		1600
1412	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1601	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1413	but on wall clock time (absolute time). You can tell a periodic watcher	1602	relative time, the physical time that passes) but on wall clock time
1414	to trigger after some specific point in time. For example, if you tell a	1603	(absolute time, the thing you can read on your calender or clock). The
1415	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1604	difference is that wall clock time can run faster or slower than real
1416	+ 10.>, that is, an absolute time not a delay) and then reset your system	1605	time, and time jumps are not uncommon (e.g. when you adjust your
1417	clock to January of the previous year, then it will take more than year	1606	wrist-watch).
1418	to trigger the event (unlike an C<ev_timer>, which would still trigger
1419	roughly 10 seconds later as it uses a relative timeout).
1420		1607
		1608	You can tell a periodic watcher to trigger after some specific point
		1609	in time: for example, if you tell a periodic watcher to trigger "in 10
		1610	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1611	not a delay) and then reset your system clock to January of the previous
		1612	year, then it will take a year or more to trigger the event (unlike an
		1613	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1614	it, as it uses a relative timeout).
		1615
1421	C<ev_periodic>s can also be used to implement vastly more complex timers,	1616	C<ev_periodic> watchers can also be used to implement vastly more complex
1422	such as triggering an event on each "midnight, local time", or other	1617	timers, such as triggering an event on each "midnight, local time", or
1423	complicated rules.	1618	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1619	those cannot react to time jumps.
1424		1620
1425	As with timers, the callback is guaranteed to be invoked only when the	1621	As with timers, the callback is guaranteed to be invoked only when the
1426	time (C<at>) has passed, but if multiple periodic timers become ready	1622	point in time where it is supposed to trigger has passed, but if multiple
1427	during the same loop iteration, then order of execution is undefined.	1623	periodic timers become ready during the same loop iteration, then order of
		1624	execution is undefined.
1428		1625
1429	=head3 Watcher-Specific Functions and Data Members	1626	=head3 Watcher-Specific Functions and Data Members
1430		1627
1431	=over 4	1628	=over 4
1432		1629
1433	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1630	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1434		1631
1435	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1632	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1436		1633
1437	Lots of arguments, lets sort it out... There are basically three modes of	1634	Lots of arguments, let's sort it out... There are basically three modes of
1438	operation, and we will explain them from simplest to most complex:	1635	operation, and we will explain them from simplest to most complex:
1439		1636
1440	=over 4	1637	=over 4
1441		1638
1442	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1639	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1443		1640
1444	In this configuration the watcher triggers an event after the wall clock	1641	In this configuration the watcher triggers an event after the wall clock
1445	time C<at> has passed. It will not repeat and will not adjust when a time	1642	time C<offset> has passed. It will not repeat and will not adjust when a
1446	jump occurs, that is, if it is to be run at January 1st 2011 then it will	1643	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1447	only run when the system clock reaches or surpasses this time.	1644	will be stopped and invoked when the system clock reaches or surpasses
		1645	this point in time.
1448		1646
1449	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1647	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1450		1648
1451	In this mode the watcher will always be scheduled to time out at the next	1649	In this mode the watcher will always be scheduled to time out at the next
1452	C<at + N * interval> time (for some integer N, which can also be negative)	1650	C<offset + N * interval> time (for some integer N, which can also be
1453	and then repeat, regardless of any time jumps.	1651	negative) and then repeat, regardless of any time jumps. The C<offset>
		1652	argument is merely an offset into the C<interval> periods.
1454		1653
1455	This can be used to create timers that do not drift with respect to the	1654	This can be used to create timers that do not drift with respect to the
1456	system clock, for example, here is a C<ev_periodic> that triggers each	1655	system clock, for example, here is an C<ev_periodic> that triggers each
1457	hour, on the hour:	1656	hour, on the hour (with respect to UTC):
1458		1657
1459	ev_periodic_set (&periodic, 0., 3600., 0);	1658	ev_periodic_set (&periodic, 0., 3600., 0);
1460		1659
1461	This doesn't mean there will always be 3600 seconds in between triggers,	1660	This doesn't mean there will always be 3600 seconds in between triggers,
1462	but only that the callback will be called when the system time shows a	1661	but only that the callback will be called when the system time shows a
1463	full hour (UTC), or more correctly, when the system time is evenly divisible	1662	full hour (UTC), or more correctly, when the system time is evenly divisible
1464	by 3600.	1663	by 3600.
1465		1664
1466	Another way to think about it (for the mathematically inclined) is that	1665	Another way to think about it (for the mathematically inclined) is that
1467	C<ev_periodic> will try to run the callback in this mode at the next possible	1666	C<ev_periodic> will try to run the callback in this mode at the next possible
1468	time where C<time = at (mod interval)>, regardless of any time jumps.	1667	time where C<time = offset (mod interval)>, regardless of any time jumps.
1469		1668
1470	For numerical stability it is preferable that the C<at> value is near	1669	For numerical stability it is preferable that the C<offset> value is near
1471	C<ev_now ()> (the current time), but there is no range requirement for	1670	C<ev_now ()> (the current time), but there is no range requirement for
1472	this value, and in fact is often specified as zero.	1671	this value, and in fact is often specified as zero.
1473		1672
1474	Note also that there is an upper limit to how often a timer can fire (CPU	1673	Note also that there is an upper limit to how often a timer can fire (CPU
1475	speed for example), so if C<interval> is very small then timing stability	1674	speed for example), so if C<interval> is very small then timing stability
1476	will of course deteriorate. Libev itself tries to be exact to be about one	1675	will of course deteriorate. Libev itself tries to be exact to be about one
1477	millisecond (if the OS supports it and the machine is fast enough).	1676	millisecond (if the OS supports it and the machine is fast enough).
1478		1677
1479	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1678	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1480		1679
1481	In this mode the values for C<interval> and C<at> are both being	1680	In this mode the values for C<interval> and C<offset> are both being
1482	ignored. Instead, each time the periodic watcher gets scheduled, the	1681	ignored. Instead, each time the periodic watcher gets scheduled, the
1483	reschedule callback will be called with the watcher as first, and the	1682	reschedule callback will be called with the watcher as first, and the
1484	current time as second argument.	1683	current time as second argument.
1485		1684
1486	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1685	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1487	ever, or make ANY event loop modifications whatsoever>.	1686	or make ANY other event loop modifications whatsoever, unless explicitly
		1687	allowed by documentation here>.
1488		1688
1489	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	1689	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1490	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	1690	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1491	only event loop modification you are allowed to do).	1691	only event loop modification you are allowed to do).
1492		1692
1493	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	1693	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1494	*w, ev_tstamp now)>, e.g.:	1694	*w, ev_tstamp now)>, e.g.:
1495		1695
		1696	static ev_tstamp
1496	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1697	my_rescheduler (ev_periodic *w, ev_tstamp now)
1497	{	1698	{
1498	return now + 60.;	1699	return now + 60.;
1499	}	1700	}
1500		1701
1501	It must return the next time to trigger, based on the passed time value	1702	It must return the next time to trigger, based on the passed time value
…		…
1521	a different time than the last time it was called (e.g. in a crond like	1722	a different time than the last time it was called (e.g. in a crond like
1522	program when the crontabs have changed).	1723	program when the crontabs have changed).
1523		1724
1524	=item ev_tstamp ev_periodic_at (ev_periodic *)	1725	=item ev_tstamp ev_periodic_at (ev_periodic *)
1525		1726
1526	When active, returns the absolute time that the watcher is supposed to	1727	When active, returns the absolute time that the watcher is supposed
1527	trigger next.	1728	to trigger next. This is not the same as the C<offset> argument to
		1729	C<ev_periodic_set>, but indeed works even in interval and manual
		1730	rescheduling modes.
1528		1731
1529	=item ev_tstamp offset [read-write]	1732	=item ev_tstamp offset [read-write]
1530		1733
1531	When repeating, this contains the offset value, otherwise this is the	1734	When repeating, this contains the offset value, otherwise this is the
1532	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	1735	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		1736	although libev might modify this value for better numerical stability).
1533		1737
1534	Can be modified any time, but changes only take effect when the periodic	1738	Can be modified any time, but changes only take effect when the periodic
1535	timer fires or C<ev_periodic_again> is being called.	1739	timer fires or C<ev_periodic_again> is being called.
1536		1740
1537	=item ev_tstamp interval [read-write]	1741	=item ev_tstamp interval [read-write]
1538		1742
1539	The current interval value. Can be modified any time, but changes only	1743	The current interval value. Can be modified any time, but changes only
1540	take effect when the periodic timer fires or C<ev_periodic_again> is being	1744	take effect when the periodic timer fires or C<ev_periodic_again> is being
1541	called.	1745	called.
1542		1746
1543	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	1747	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1544		1748
1545	The current reschedule callback, or C<0>, if this functionality is	1749	The current reschedule callback, or C<0>, if this functionality is
1546	switched off. Can be changed any time, but changes only take effect when	1750	switched off. Can be changed any time, but changes only take effect when
1547	the periodic timer fires or C<ev_periodic_again> is being called.	1751	the periodic timer fires or C<ev_periodic_again> is being called.
1548		1752
…		…
1553	Example: Call a callback every hour, or, more precisely, whenever the	1757	Example: Call a callback every hour, or, more precisely, whenever the
1554	system time is divisible by 3600. The callback invocation times have	1758	system time is divisible by 3600. The callback invocation times have
1555	potentially a lot of jitter, but good long-term stability.	1759	potentially a lot of jitter, but good long-term stability.
1556		1760
1557	static void	1761	static void
1558	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1762	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1559	{	1763	{
1560	... its now a full hour (UTC, or TAI or whatever your clock follows)	1764	... its now a full hour (UTC, or TAI or whatever your clock follows)
1561	}	1765	}
1562		1766
1563	struct ev_periodic hourly_tick;	1767	ev_periodic hourly_tick;
1564	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1768	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1565	ev_periodic_start (loop, &hourly_tick);	1769	ev_periodic_start (loop, &hourly_tick);
1566		1770
1567	Example: The same as above, but use a reschedule callback to do it:	1771	Example: The same as above, but use a reschedule callback to do it:
1568		1772
1569	#include <math.h>	1773	#include <math.h>
1570		1774
1571	static ev_tstamp	1775	static ev_tstamp
1572	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1776	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1573	{	1777	{
1574	return now + (3600. - fmod (now, 3600.));	1778	return now + (3600. - fmod (now, 3600.));
1575	}	1779	}
1576		1780
1577	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1781	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1578		1782
1579	Example: Call a callback every hour, starting now:	1783	Example: Call a callback every hour, starting now:
1580		1784
1581	struct ev_periodic hourly_tick;	1785	ev_periodic hourly_tick;
1582	ev_periodic_init (&hourly_tick, clock_cb,	1786	ev_periodic_init (&hourly_tick, clock_cb,
1583	fmod (ev_now (loop), 3600.), 3600., 0);	1787	fmod (ev_now (loop), 3600.), 3600., 0);
1584	ev_periodic_start (loop, &hourly_tick);	1788	ev_periodic_start (loop, &hourly_tick);
1585		1789
1586		1790
…		…
1628	=head3 Examples	1832	=head3 Examples
1629		1833
1630	Example: Try to exit cleanly on SIGINT.	1834	Example: Try to exit cleanly on SIGINT.
1631		1835
1632	static void	1836	static void
1633	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	1837	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1634	{	1838	{
1635	ev_unloop (loop, EVUNLOOP_ALL);	1839	ev_unloop (loop, EVUNLOOP_ALL);
1636	}	1840	}
1637		1841
1638	struct ev_signal signal_watcher;	1842	ev_signal signal_watcher;
1639	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	1843	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1640	ev_signal_start (loop, &signal_watcher);	1844	ev_signal_start (loop, &signal_watcher);
1641		1845
1642		1846
1643	=head2 C<ev_child> - watch out for process status changes	1847	=head2 C<ev_child> - watch out for process status changes
…		…
1718	its completion.	1922	its completion.
1719		1923
1720	ev_child cw;	1924	ev_child cw;
1721		1925
1722	static void	1926	static void
1723	child_cb (EV_P_ struct ev_child *w, int revents)	1927	child_cb (EV_P_ ev_child *w, int revents)
1724	{	1928	{
1725	ev_child_stop (EV_A_ w);	1929	ev_child_stop (EV_A_ w);
1726	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	1930	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1727	}	1931	}
1728		1932
…		…
1743		1947
1744		1948
1745	=head2 C<ev_stat> - did the file attributes just change?	1949	=head2 C<ev_stat> - did the file attributes just change?
1746		1950
1747	This watches a file system path for attribute changes. That is, it calls	1951	This watches a file system path for attribute changes. That is, it calls
1748	C<stat> regularly (or when the OS says it changed) and sees if it changed	1952	C<stat> on that path in regular intervals (or when the OS says it changed)
1749	compared to the last time, invoking the callback if it did.	1953	and sees if it changed compared to the last time, invoking the callback if
		1954	it did.
1750		1955
1751	The path does not need to exist: changing from "path exists" to "path does	1956	The path does not need to exist: changing from "path exists" to "path does
1752	not exist" is a status change like any other. The condition "path does	1957	not exist" is a status change like any other. The condition "path does not
1753	not exist" is signified by the C<st_nlink> field being zero (which is	1958	exist" (or more correctly "path cannot be stat'ed") is signified by the
1754	otherwise always forced to be at least one) and all the other fields of	1959	C<st_nlink> field being zero (which is otherwise always forced to be at
1755	the stat buffer having unspecified contents.	1960	least one) and all the other fields of the stat buffer having unspecified
		1961	contents.
1756		1962
1757	The path I<should> be absolute and I<must not> end in a slash. If it is	1963	The path I<must not> end in a slash or contain special components such as
		1964	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1758	relative and your working directory changes, the behaviour is undefined.	1965	your working directory changes, then the behaviour is undefined.
1759		1966
1760	Since there is no standard kernel interface to do this, the portable	1967	Since there is no portable change notification interface available, the
1761	implementation simply calls C<stat (2)> regularly on the path to see if	1968	portable implementation simply calls C<stat(2)> regularly on the path
1762	it changed somehow. You can specify a recommended polling interval for	1969	to see if it changed somehow. You can specify a recommended polling
1763	this case. If you specify a polling interval of C<0> (highly recommended!)	1970	interval for this case. If you specify a polling interval of C<0> (highly
1764	then a I<suitable, unspecified default> value will be used (which	1971	recommended!) then a I<suitable, unspecified default> value will be used
1765	you can expect to be around five seconds, although this might change	1972	(which you can expect to be around five seconds, although this might
1766	dynamically). Libev will also impose a minimum interval which is currently	1973	change dynamically). Libev will also impose a minimum interval which is
1767	around C<0.1>, but thats usually overkill.	1974	currently around C<0.1>, but that's usually overkill.
1768		1975
1769	This watcher type is not meant for massive numbers of stat watchers,	1976	This watcher type is not meant for massive numbers of stat watchers,
1770	as even with OS-supported change notifications, this can be	1977	as even with OS-supported change notifications, this can be
1771	resource-intensive.	1978	resource-intensive.
1772		1979
1773	At the time of this writing, the only OS-specific interface implemented	1980	At the time of this writing, the only OS-specific interface implemented
1774	is the Linux inotify interface (implementing kqueue support is left as	1981	is the Linux inotify interface (implementing kqueue support is left as an
1775	an exercise for the reader. Note, however, that the author sees no way	1982	exercise for the reader. Note, however, that the author sees no way of
1776	of implementing C<ev_stat> semantics with kqueue).	1983	implementing C<ev_stat> semantics with kqueue, except as a hint).
1777		1984
1778	=head3 ABI Issues (Largefile Support)	1985	=head3 ABI Issues (Largefile Support)
1779		1986
1780	Libev by default (unless the user overrides this) uses the default	1987	Libev by default (unless the user overrides this) uses the default
1781	compilation environment, which means that on systems with large file	1988	compilation environment, which means that on systems with large file
1782	support disabled by default, you get the 32 bit version of the stat	1989	support disabled by default, you get the 32 bit version of the stat
1783	structure. When using the library from programs that change the ABI to	1990	structure. When using the library from programs that change the ABI to
1784	use 64 bit file offsets the programs will fail. In that case you have to	1991	use 64 bit file offsets the programs will fail. In that case you have to
1785	compile libev with the same flags to get binary compatibility. This is	1992	compile libev with the same flags to get binary compatibility. This is
1786	obviously the case with any flags that change the ABI, but the problem is	1993	obviously the case with any flags that change the ABI, but the problem is
1787	most noticeably disabled with ev_stat and large file support.	1994	most noticeably displayed with ev_stat and large file support.
1788		1995
1789	The solution for this is to lobby your distribution maker to make large	1996	The solution for this is to lobby your distribution maker to make large
1790	file interfaces available by default (as e.g. FreeBSD does) and not	1997	file interfaces available by default (as e.g. FreeBSD does) and not
1791	optional. Libev cannot simply switch on large file support because it has	1998	optional. Libev cannot simply switch on large file support because it has
1792	to exchange stat structures with application programs compiled using the	1999	to exchange stat structures with application programs compiled using the
1793	default compilation environment.	2000	default compilation environment.
1794		2001
1795	=head3 Inotify and Kqueue	2002	=head3 Inotify and Kqueue
1796		2003
1797	When C<inotify (7)> support has been compiled into libev (generally only	2004	When C<inotify (7)> support has been compiled into libev and present at
1798	available with Linux) and present at runtime, it will be used to speed up	2005	runtime, it will be used to speed up change detection where possible. The
1799	change detection where possible. The inotify descriptor will be created lazily	2006	inotify descriptor will be created lazily when the first C<ev_stat>
1800	when the first C<ev_stat> watcher is being started.	2007	watcher is being started.
1801		2008
1802	Inotify presence does not change the semantics of C<ev_stat> watchers	2009	Inotify presence does not change the semantics of C<ev_stat> watchers
1803	except that changes might be detected earlier, and in some cases, to avoid	2010	except that changes might be detected earlier, and in some cases, to avoid
1804	making regular C<stat> calls. Even in the presence of inotify support	2011	making regular C<stat> calls. Even in the presence of inotify support
1805	there are many cases where libev has to resort to regular C<stat> polling,	2012	there are many cases where libev has to resort to regular C<stat> polling,
1806	but as long as the path exists, libev usually gets away without polling.	2013	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2014	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2015	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2016	xfs are fully working) libev usually gets away without polling.
1807		2017
1808	There is no support for kqueue, as apparently it cannot be used to	2018	There is no support for kqueue, as apparently it cannot be used to
1809	implement this functionality, due to the requirement of having a file	2019	implement this functionality, due to the requirement of having a file
1810	descriptor open on the object at all times, and detecting renames, unlinks	2020	descriptor open on the object at all times, and detecting renames, unlinks
1811	etc. is difficult.	2021	etc. is difficult.
1812		2022
		2023	=head3 C<stat ()> is a synchronous operation
		2024
		2025	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2026	the process. The exception are C<ev_stat> watchers - those call C<stat
		2027	()>, which is a synchronous operation.
		2028
		2029	For local paths, this usually doesn't matter: unless the system is very
		2030	busy or the intervals between stat's are large, a stat call will be fast,
		2031	as the path data is usually in memory already (except when starting the
		2032	watcher).
		2033
		2034	For networked file systems, calling C<stat ()> can block an indefinite
		2035	time due to network issues, and even under good conditions, a stat call
		2036	often takes multiple milliseconds.
		2037
		2038	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2039	paths, although this is fully supported by libev.
		2040
1813	=head3 The special problem of stat time resolution	2041	=head3 The special problem of stat time resolution
1814		2042
1815	The C<stat ()> system call only supports full-second resolution portably, and	2043	The C<stat ()> system call only supports full-second resolution portably,
1816	even on systems where the resolution is higher, most file systems still	2044	and even on systems where the resolution is higher, most file systems
1817	only support whole seconds.	2045	still only support whole seconds.
1818		2046
1819	That means that, if the time is the only thing that changes, you can	2047	That means that, if the time is the only thing that changes, you can
1820	easily miss updates: on the first update, C<ev_stat> detects a change and	2048	easily miss updates: on the first update, C<ev_stat> detects a change and
1821	calls your callback, which does something. When there is another update	2049	calls your callback, which does something. When there is another update
1822	within the same second, C<ev_stat> will be unable to detect unless the	2050	within the same second, C<ev_stat> will be unable to detect unless the
…		…
1965		2193
1966	=head3 Watcher-Specific Functions and Data Members	2194	=head3 Watcher-Specific Functions and Data Members
1967		2195
1968	=over 4	2196	=over 4
1969		2197
1970	=item ev_idle_init (ev_signal *, callback)	2198	=item ev_idle_init (ev_idle *, callback)
1971		2199
1972	Initialises and configures the idle watcher - it has no parameters of any	2200	Initialises and configures the idle watcher - it has no parameters of any
1973	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2201	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1974	believe me.	2202	believe me.
1975		2203
…		…
1979		2207
1980	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2208	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1981	callback, free it. Also, use no error checking, as usual.	2209	callback, free it. Also, use no error checking, as usual.
1982		2210
1983	static void	2211	static void
1984	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2212	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1985	{	2213	{
1986	free (w);	2214	free (w);
1987	// now do something you wanted to do when the program has	2215	// now do something you wanted to do when the program has
1988	// no longer anything immediate to do.	2216	// no longer anything immediate to do.
1989	}	2217	}
1990		2218
1991	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2219	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1992	ev_idle_init (idle_watcher, idle_cb);	2220	ev_idle_init (idle_watcher, idle_cb);
1993	ev_idle_start (loop, idle_cb);	2221	ev_idle_start (loop, idle_cb);
1994		2222
1995		2223
1996	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2224	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
…		…
2077		2305
2078	static ev_io iow [nfd];	2306	static ev_io iow [nfd];
2079	static ev_timer tw;	2307	static ev_timer tw;
2080		2308
2081	static void	2309	static void
2082	io_cb (ev_loop *loop, ev_io *w, int revents)	2310	io_cb (struct ev_loop *loop, ev_io *w, int revents)
2083	{	2311	{
2084	}	2312	}
2085		2313
2086	// create io watchers for each fd and a timer before blocking	2314	// create io watchers for each fd and a timer before blocking
2087	static void	2315	static void
2088	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2316	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
2089	{	2317	{
2090	int timeout = 3600000;	2318	int timeout = 3600000;
2091	struct pollfd fds [nfd];	2319	struct pollfd fds [nfd];
2092	// actual code will need to loop here and realloc etc.	2320	// actual code will need to loop here and realloc etc.
2093	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2321	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
…		…
2108	}	2336	}
2109	}	2337	}
2110		2338
2111	// stop all watchers after blocking	2339	// stop all watchers after blocking
2112	static void	2340	static void
2113	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2341	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
2114	{	2342	{
2115	ev_timer_stop (loop, &tw);	2343	ev_timer_stop (loop, &tw);
2116		2344
2117	for (int i = 0; i < nfd; ++i)	2345	for (int i = 0; i < nfd; ++i)
2118	{	2346	{
…		…
2214	some fds have to be watched and handled very quickly (with low latency),	2442	some fds have to be watched and handled very quickly (with low latency),
2215	and even priorities and idle watchers might have too much overhead. In	2443	and even priorities and idle watchers might have too much overhead. In
2216	this case you would put all the high priority stuff in one loop and all	2444	this case you would put all the high priority stuff in one loop and all
2217	the rest in a second one, and embed the second one in the first.	2445	the rest in a second one, and embed the second one in the first.
2218		2446
2219	As long as the watcher is active, the callback will be invoked every time	2447	As long as the watcher is active, the callback will be invoked every
2220	there might be events pending in the embedded loop. The callback must then	2448	time there might be events pending in the embedded loop. The callback
2221	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2449	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2222	their callbacks (you could also start an idle watcher to give the embedded	2450	sweep and invoke their callbacks (the callback doesn't need to invoke the
2223	loop strictly lower priority for example). You can also set the callback	2451	C<ev_embed_sweep> function directly, it could also start an idle watcher
2224	to C<0>, in which case the embed watcher will automatically execute the	2452	to give the embedded loop strictly lower priority for example).
2225	embedded loop sweep.
2226		2453
2227	As long as the watcher is started it will automatically handle events. The	2454	You can also set the callback to C<0>, in which case the embed watcher
2228	callback will be invoked whenever some events have been handled. You can	2455	will automatically execute the embedded loop sweep whenever necessary.
2229	set the callback to C<0> to avoid having to specify one if you are not
2230	interested in that.
2231		2456
2232	Also, there have not currently been made special provisions for forking:	2457	Fork detection will be handled transparently while the C<ev_embed> watcher
2233	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2458	is active, i.e., the embedded loop will automatically be forked when the
2234	but you will also have to stop and restart any C<ev_embed> watchers	2459	embedding loop forks. In other cases, the user is responsible for calling
2235	yourself - but you can use a fork watcher to handle this automatically,	2460	C<ev_loop_fork> on the embedded loop.
2236	and future versions of libev might do just that.
2237		2461
2238	Unfortunately, not all backends are embeddable: only the ones returned by	2462	Unfortunately, not all backends are embeddable: only the ones returned by
2239	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2463	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2240	portable one.	2464	portable one.
2241		2465
…		…
2286	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2510	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2287	used).	2511	used).
2288		2512
2289	struct ev_loop *loop_hi = ev_default_init (0);	2513	struct ev_loop *loop_hi = ev_default_init (0);
2290	struct ev_loop *loop_lo = 0;	2514	struct ev_loop *loop_lo = 0;
2291	struct ev_embed embed;	2515	ev_embed embed;
2292		2516
2293	// see if there is a chance of getting one that works	2517	// see if there is a chance of getting one that works
2294	// (remember that a flags value of 0 means autodetection)	2518	// (remember that a flags value of 0 means autodetection)
2295	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2519	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2296	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2520	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2310	kqueue implementation). Store the kqueue/socket-only event loop in	2534	kqueue implementation). Store the kqueue/socket-only event loop in
2311	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2535	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2312		2536
2313	struct ev_loop *loop = ev_default_init (0);	2537	struct ev_loop *loop = ev_default_init (0);
2314	struct ev_loop *loop_socket = 0;	2538	struct ev_loop *loop_socket = 0;
2315	struct ev_embed embed;	2539	ev_embed embed;
2316		2540
2317	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2541	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2318	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2542	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2319	{	2543	{
2320	ev_embed_init (&embed, 0, loop_socket);	2544	ev_embed_init (&embed, 0, loop_socket);
…		…
2384	=over 4	2608	=over 4
2385		2609
2386	=item queueing from a signal handler context	2610	=item queueing from a signal handler context
2387		2611
2388	To implement race-free queueing, you simply add to the queue in the signal	2612	To implement race-free queueing, you simply add to the queue in the signal
2389	handler but you block the signal handler in the watcher callback. Here is an example that does that for	2613	handler but you block the signal handler in the watcher callback. Here is
2390	some fictitious SIGUSR1 handler:	2614	an example that does that for some fictitious SIGUSR1 handler:
2391		2615
2392	static ev_async mysig;	2616	static ev_async mysig;
2393		2617
2394	static void	2618	static void
2395	sigusr1_handler (void)	2619	sigusr1_handler (void)
…		…
2461	=over 4	2685	=over 4
2462		2686
2463	=item ev_async_init (ev_async *, callback)	2687	=item ev_async_init (ev_async *, callback)
2464		2688
2465	Initialises and configures the async watcher - it has no parameters of any	2689	Initialises and configures the async watcher - it has no parameters of any
2466	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	2690	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2467	trust me.	2691	trust me.
2468		2692
2469	=item ev_async_send (loop, ev_async *)	2693	=item ev_async_send (loop, ev_async *)
2470		2694
2471	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	2695	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2472	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	2696	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2473	C<ev_feed_event>, this call is safe to do from other threads, signal or	2697	C<ev_feed_event>, this call is safe to do from other threads, signal or
2474	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	2698	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2475	section below on what exactly this means).	2699	section below on what exactly this means).
2476		2700
		2701	Note that, as with other watchers in libev, multiple events might get
		2702	compressed into a single callback invocation (another way to look at this
		2703	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		2704	reset when the event loop detects that).
		2705
2477	This call incurs the overhead of a system call only once per loop iteration,	2706	This call incurs the overhead of a system call only once per event loop
2478	so while the overhead might be noticeable, it doesn't apply to repeated	2707	iteration, so while the overhead might be noticeable, it doesn't apply to
2479	calls to C<ev_async_send>.	2708	repeated calls to C<ev_async_send> for the same event loop.
2480		2709
2481	=item bool = ev_async_pending (ev_async *)	2710	=item bool = ev_async_pending (ev_async *)
2482		2711
2483	Returns a non-zero value when C<ev_async_send> has been called on the	2712	Returns a non-zero value when C<ev_async_send> has been called on the
2484	watcher but the event has not yet been processed (or even noted) by the	2713	watcher but the event has not yet been processed (or even noted) by the
…		…
2487	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	2716	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2488	the loop iterates next and checks for the watcher to have become active,	2717	the loop iterates next and checks for the watcher to have become active,
2489	it will reset the flag again. C<ev_async_pending> can be used to very	2718	it will reset the flag again. C<ev_async_pending> can be used to very
2490	quickly check whether invoking the loop might be a good idea.	2719	quickly check whether invoking the loop might be a good idea.
2491		2720
2492	Not that this does I<not> check whether the watcher itself is pending, only	2721	Not that this does I<not> check whether the watcher itself is pending,
2493	whether it has been requested to make this watcher pending.	2722	only whether it has been requested to make this watcher pending: there
		2723	is a time window between the event loop checking and resetting the async
		2724	notification, and the callback being invoked.
2494		2725
2495	=back	2726	=back
2496		2727
2497		2728
2498	=head1 OTHER FUNCTIONS	2729	=head1 OTHER FUNCTIONS
…		…
2502	=over 4	2733	=over 4
2503		2734
2504	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	2735	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2505		2736
2506	This function combines a simple timer and an I/O watcher, calls your	2737	This function combines a simple timer and an I/O watcher, calls your
2507	callback on whichever event happens first and automatically stop both	2738	callback on whichever event happens first and automatically stops both
2508	watchers. This is useful if you want to wait for a single event on an fd	2739	watchers. This is useful if you want to wait for a single event on an fd
2509	or timeout without having to allocate/configure/start/stop/free one or	2740	or timeout without having to allocate/configure/start/stop/free one or
2510	more watchers yourself.	2741	more watchers yourself.
2511		2742
2512	If C<fd> is less than 0, then no I/O watcher will be started and events	2743	If C<fd> is less than 0, then no I/O watcher will be started and the
2513	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	2744	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2514	C<events> set will be created and started.	2745	the given C<fd> and C<events> set will be created and started.
2515		2746
2516	If C<timeout> is less than 0, then no timeout watcher will be	2747	If C<timeout> is less than 0, then no timeout watcher will be
2517	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	2748	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2518	repeat = 0) will be started. While C<0> is a valid timeout, it is of	2749	repeat = 0) will be started. C<0> is a valid timeout.
2519	dubious value.
2520		2750
2521	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	2751	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2522	passed an C<revents> set like normal event callbacks (a combination of	2752	passed an C<revents> set like normal event callbacks (a combination of
2523	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	2753	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2524	value passed to C<ev_once>:	2754	value passed to C<ev_once>. Note that it is possible to receive I<both>
		2755	a timeout and an io event at the same time - you probably should give io
		2756	events precedence.
		2757
		2758	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2525		2759
2526	static void stdin_ready (int revents, void *arg)	2760	static void stdin_ready (int revents, void *arg)
2527	{	2761	{
		2762	if (revents & EV_READ)
		2763	/* stdin might have data for us, joy! */;
2528	if (revents & EV_TIMEOUT)	2764	else if (revents & EV_TIMEOUT)
2529	/* doh, nothing entered */;	2765	/* doh, nothing entered */;
2530	else if (revents & EV_READ)
2531	/* stdin might have data for us, joy! */;
2532	}	2766	}
2533		2767
2534	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	2768	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2535		2769
2536	=item ev_feed_event (ev_loop *, watcher *, int revents)	2770	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
2537		2771
2538	Feeds the given event set into the event loop, as if the specified event	2772	Feeds the given event set into the event loop, as if the specified event
2539	had happened for the specified watcher (which must be a pointer to an	2773	had happened for the specified watcher (which must be a pointer to an
2540	initialised but not necessarily started event watcher).	2774	initialised but not necessarily started event watcher).
2541		2775
2542	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	2776	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)
2543		2777
2544	Feed an event on the given fd, as if a file descriptor backend detected	2778	Feed an event on the given fd, as if a file descriptor backend detected
2545	the given events it.	2779	the given events it.
2546		2780
2547	=item ev_feed_signal_event (ev_loop *loop, int signum)	2781	=item ev_feed_signal_event (struct ev_loop *loop, int signum)
2548		2782
2549	Feed an event as if the given signal occurred (C<loop> must be the default	2783	Feed an event as if the given signal occurred (C<loop> must be the default
2550	loop!).	2784	loop!).
2551		2785
2552	=back	2786	=back
…		…
2673	}	2907	}
2674		2908
2675	myclass obj;	2909	myclass obj;
2676	ev::io iow;	2910	ev::io iow;
2677	iow.set <myclass, &myclass::io_cb> (&obj);	2911	iow.set <myclass, &myclass::io_cb> (&obj);
		2912
		2913	=item w->set (object *)
		2914
		2915	This is an B<experimental> feature that might go away in a future version.
		2916
		2917	This is a variation of a method callback - leaving out the method to call
		2918	will default the method to C<operator ()>, which makes it possible to use
		2919	functor objects without having to manually specify the C<operator ()> all
		2920	the time. Incidentally, you can then also leave out the template argument
		2921	list.
		2922
		2923	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		2924	int revents)>.
		2925
		2926	See the method-C<set> above for more details.
		2927
		2928	Example: use a functor object as callback.
		2929
		2930	struct myfunctor
		2931	{
		2932	void operator() (ev::io &w, int revents)
		2933	{
		2934	...
		2935	}
		2936	}
		2937
		2938	myfunctor f;
		2939
		2940	ev::io w;
		2941	w.set (&f);
2678		2942
2679	=item w->set<function> (void *data = 0)	2943	=item w->set<function> (void *data = 0)
2680		2944
2681	Also sets a callback, but uses a static method or plain function as	2945	Also sets a callback, but uses a static method or plain function as
2682	callback. The optional C<data> argument will be stored in the watcher's	2946	callback. The optional C<data> argument will be stored in the watcher's
…		…
2782	Tony Arcieri has written a ruby extension that offers access to a subset	3046	Tony Arcieri has written a ruby extension that offers access to a subset
2783	of the libev API and adds file handle abstractions, asynchronous DNS and	3047	of the libev API and adds file handle abstractions, asynchronous DNS and
2784	more on top of it. It can be found via gem servers. Its homepage is at	3048	more on top of it. It can be found via gem servers. Its homepage is at
2785	L<http://rev.rubyforge.org/>.	3049	L<http://rev.rubyforge.org/>.
2786		3050
		3051	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3052	makes rev work even on mingw.
		3053
2787	=item D	3054	=item D
2788		3055
2789	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3056	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2790	be found at L<http://proj.llucax.com.ar/wiki/evd>.	3057	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3058
		3059	=item Ocaml
		3060
		3061	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3062	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
2791		3063
2792	=back	3064	=back
2793		3065
2794		3066
2795	=head1 MACRO MAGIC	3067	=head1 MACRO MAGIC
…		…
2896		3168
2897	#define EV_STANDALONE 1	3169	#define EV_STANDALONE 1
2898	#include "ev.h"	3170	#include "ev.h"
2899		3171
2900	Both header files and implementation files can be compiled with a C++	3172	Both header files and implementation files can be compiled with a C++
2901	compiler (at least, thats a stated goal, and breakage will be treated	3173	compiler (at least, that's a stated goal, and breakage will be treated
2902	as a bug).	3174	as a bug).
2903		3175
2904	You need the following files in your source tree, or in a directory	3176	You need the following files in your source tree, or in a directory
2905	in your include path (e.g. in libev/ when using -Ilibev):	3177	in your include path (e.g. in libev/ when using -Ilibev):
2906		3178
…		…
2962	keeps libev from including F<config.h>, and it also defines dummy	3234	keeps libev from including F<config.h>, and it also defines dummy
2963	implementations for some libevent functions (such as logging, which is not	3235	implementations for some libevent functions (such as logging, which is not
2964	supported). It will also not define any of the structs usually found in	3236	supported). It will also not define any of the structs usually found in
2965	F<event.h> that are not directly supported by the libev core alone.	3237	F<event.h> that are not directly supported by the libev core alone.
2966		3238
		3239	In stanbdalone mode, libev will still try to automatically deduce the
		3240	configuration, but has to be more conservative.
		3241
2967	=item EV_USE_MONOTONIC	3242	=item EV_USE_MONOTONIC
2968		3243
2969	If defined to be C<1>, libev will try to detect the availability of the	3244	If defined to be C<1>, libev will try to detect the availability of the
2970	monotonic clock option at both compile time and runtime. Otherwise no use	3245	monotonic clock option at both compile time and runtime. Otherwise no
2971	of the monotonic clock option will be attempted. If you enable this, you	3246	use of the monotonic clock option will be attempted. If you enable this,
2972	usually have to link against librt or something similar. Enabling it when	3247	you usually have to link against librt or something similar. Enabling it
2973	the functionality isn't available is safe, though, although you have	3248	when the functionality isn't available is safe, though, although you have
2974	to make sure you link against any libraries where the C<clock_gettime>	3249	to make sure you link against any libraries where the C<clock_gettime>
2975	function is hiding in (often F<-lrt>).	3250	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2976		3251
2977	=item EV_USE_REALTIME	3252	=item EV_USE_REALTIME
2978		3253
2979	If defined to be C<1>, libev will try to detect the availability of the	3254	If defined to be C<1>, libev will try to detect the availability of the
2980	real-time clock option at compile time (and assume its availability at	3255	real-time clock option at compile time (and assume its availability
2981	runtime if successful). Otherwise no use of the real-time clock option will	3256	at runtime if successful). Otherwise no use of the real-time clock
2982	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3257	option will be attempted. This effectively replaces C<gettimeofday>
2983	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3258	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2984	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3259	correctness. See the note about libraries in the description of
		3260	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3261	C<EV_USE_CLOCK_SYSCALL>.
		3262
		3263	=item EV_USE_CLOCK_SYSCALL
		3264
		3265	If defined to be C<1>, libev will try to use a direct syscall instead
		3266	of calling the system-provided C<clock_gettime> function. This option
		3267	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3268	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3269	programs needlessly. Using a direct syscall is slightly slower (in
		3270	theory), because no optimised vdso implementation can be used, but avoids
		3271	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3272	higher, as it simplifies linking (no need for C<-lrt>).
2985		3273
2986	=item EV_USE_NANOSLEEP	3274	=item EV_USE_NANOSLEEP
2987		3275
2988	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3276	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2989	and will use it for delays. Otherwise it will use C<select ()>.	3277	and will use it for delays. Otherwise it will use C<select ()>.
…		…
3005		3293
3006	=item EV_SELECT_USE_FD_SET	3294	=item EV_SELECT_USE_FD_SET
3007		3295
3008	If defined to C<1>, then the select backend will use the system C<fd_set>	3296	If defined to C<1>, then the select backend will use the system C<fd_set>
3009	structure. This is useful if libev doesn't compile due to a missing	3297	structure. This is useful if libev doesn't compile due to a missing
3010	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3298	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
3011	exotic systems. This usually limits the range of file descriptors to some	3299	on exotic systems. This usually limits the range of file descriptors to
3012	low limit such as 1024 or might have other limitations (winsocket only	3300	some low limit such as 1024 or might have other limitations (winsocket
3013	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3301	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
3014	influence the size of the C<fd_set> used.	3302	configures the maximum size of the C<fd_set>.
3015		3303
3016	=item EV_SELECT_IS_WINSOCKET	3304	=item EV_SELECT_IS_WINSOCKET
3017		3305
3018	When defined to C<1>, the select backend will assume that	3306	When defined to C<1>, the select backend will assume that
3019	select/socket/connect etc. don't understand file descriptors but	3307	select/socket/connect etc. don't understand file descriptors but
…		…
3306	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	3594	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3307		3595
3308	#include "ev_cpp.h"	3596	#include "ev_cpp.h"
3309	#include "ev.c"	3597	#include "ev.c"
3310		3598
		3599	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3311		3600
3312	=head1 THREADS AND COROUTINES	3601	=head2 THREADS AND COROUTINES
3313		3602
3314	=head2 THREADS	3603	=head3 THREADS
3315		3604
3316	All libev functions are reentrant and thread-safe unless explicitly	3605	All libev functions are reentrant and thread-safe unless explicitly
3317	documented otherwise, but it uses no locking itself. This means that you	3606	documented otherwise, but libev implements no locking itself. This means
3318	can use as many loops as you want in parallel, as long as there are no	3607	that you can use as many loops as you want in parallel, as long as there
3319	concurrent calls into any libev function with the same loop parameter	3608	are no concurrent calls into any libev function with the same loop
3320	(C<ev_default_*> calls have an implicit default loop parameter, of	3609	parameter (C<ev_default_*> calls have an implicit default loop parameter,
3321	course): libev guarantees that different event loops share no data	3610	of course): libev guarantees that different event loops share no data
3322	structures that need any locking.	3611	structures that need any locking.
3323		3612
3324	Or to put it differently: calls with different loop parameters can be done	3613	Or to put it differently: calls with different loop parameters can be done
3325	concurrently from multiple threads, calls with the same loop parameter	3614	concurrently from multiple threads, calls with the same loop parameter
3326	must be done serially (but can be done from different threads, as long as	3615	must be done serially (but can be done from different threads, as long as
…		…
3366	default loop and triggering an C<ev_async> watcher from the default loop	3655	default loop and triggering an C<ev_async> watcher from the default loop
3367	watcher callback into the event loop interested in the signal.	3656	watcher callback into the event loop interested in the signal.
3368		3657
3369	=back	3658	=back
3370		3659
3371	=head2 COROUTINES	3660	=head3 COROUTINES
3372		3661
3373	Libev is much more accommodating to coroutines ("cooperative threads"):	3662	Libev is very accommodating to coroutines ("cooperative threads"):
3374	libev fully supports nesting calls to it's functions from different	3663	libev fully supports nesting calls to its functions from different
3375	coroutines (e.g. you can call C<ev_loop> on the same loop from two	3664	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3376	different coroutines and switch freely between both coroutines running the	3665	different coroutines, and switch freely between both coroutines running the
3377	loop, as long as you don't confuse yourself). The only exception is that	3666	loop, as long as you don't confuse yourself). The only exception is that
3378	you must not do this from C<ev_periodic> reschedule callbacks.	3667	you must not do this from C<ev_periodic> reschedule callbacks.
3379		3668
3380	Care has been taken to ensure that libev does not keep local state inside	3669	Care has been taken to ensure that libev does not keep local state inside
3381	C<ev_loop>, and other calls do not usually allow coroutine switches.	3670	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		3671	they do not call any callbacks.
3382		3672
		3673	=head2 COMPILER WARNINGS
3383		3674
3384	=head1 COMPLEXITIES	3675	Depending on your compiler and compiler settings, you might get no or a
		3676	lot of warnings when compiling libev code. Some people are apparently
		3677	scared by this.
3385		3678
3386	In this section the complexities of (many of) the algorithms used inside	3679	However, these are unavoidable for many reasons. For one, each compiler
3387	libev will be explained. For complexity discussions about backends see the	3680	has different warnings, and each user has different tastes regarding
3388	documentation for C<ev_default_init>.	3681	warning options. "Warn-free" code therefore cannot be a goal except when
		3682	targeting a specific compiler and compiler-version.
3389		3683
3390	All of the following are about amortised time: If an array needs to be	3684	Another reason is that some compiler warnings require elaborate
3391	extended, libev needs to realloc and move the whole array, but this	3685	workarounds, or other changes to the code that make it less clear and less
3392	happens asymptotically never with higher number of elements, so O(1) might	3686	maintainable.
3393	mean it might do a lengthy realloc operation in rare cases, but on average
3394	it is much faster and asymptotically approaches constant time.
3395		3687
3396	=over 4	3688	And of course, some compiler warnings are just plain stupid, or simply
		3689	wrong (because they don't actually warn about the condition their message
		3690	seems to warn about). For example, certain older gcc versions had some
		3691	warnings that resulted an extreme number of false positives. These have
		3692	been fixed, but some people still insist on making code warn-free with
		3693	such buggy versions.
3397		3694
3398	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	3695	While libev is written to generate as few warnings as possible,
		3696	"warn-free" code is not a goal, and it is recommended not to build libev
		3697	with any compiler warnings enabled unless you are prepared to cope with
		3698	them (e.g. by ignoring them). Remember that warnings are just that:
		3699	warnings, not errors, or proof of bugs.
3399		3700
3400	This means that, when you have a watcher that triggers in one hour and
3401	there are 100 watchers that would trigger before that then inserting will
3402	have to skip roughly seven (C<ld 100>) of these watchers.
3403		3701
3404	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	3702	=head2 VALGRIND
3405		3703
3406	That means that changing a timer costs less than removing/adding them	3704	Valgrind has a special section here because it is a popular tool that is
3407	as only the relative motion in the event queue has to be paid for.	3705	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3408		3706
3409	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	3707	If you think you found a bug (memory leak, uninitialised data access etc.)
		3708	in libev, then check twice: If valgrind reports something like:
3410		3709
3411	These just add the watcher into an array or at the head of a list.	3710	==2274== definitely lost: 0 bytes in 0 blocks.
		3711	==2274== possibly lost: 0 bytes in 0 blocks.
		3712	==2274== still reachable: 256 bytes in 1 blocks.
3412		3713
3413	=item Stopping check/prepare/idle/fork/async watchers: O(1)	3714	Then there is no memory leak, just as memory accounted to global variables
		3715	is not a memleak - the memory is still being referenced, and didn't leak.
3414		3716
3415	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	3717	Similarly, under some circumstances, valgrind might report kernel bugs
		3718	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		3719	although an acceptable workaround has been found here), or it might be
		3720	confused.
3416		3721
3417	These watchers are stored in lists then need to be walked to find the	3722	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3418	correct watcher to remove. The lists are usually short (you don't usually	3723	make it into some kind of religion.
3419	have many watchers waiting for the same fd or signal).
3420		3724
3421	=item Finding the next timer in each loop iteration: O(1)	3725	If you are unsure about something, feel free to contact the mailing list
		3726	with the full valgrind report and an explanation on why you think this
		3727	is a bug in libev (best check the archives, too :). However, don't be
		3728	annoyed when you get a brisk "this is no bug" answer and take the chance
		3729	of learning how to interpret valgrind properly.
3422		3730
3423	By virtue of using a binary or 4-heap, the next timer is always found at a	3731	If you need, for some reason, empty reports from valgrind for your project
3424	fixed position in the storage array.	3732	I suggest using suppression lists.
3425		3733
3426	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3427		3734
3428	A change means an I/O watcher gets started or stopped, which requires	3735	=head1 PORTABILITY NOTES
3429	libev to recalculate its status (and possibly tell the kernel, depending
3430	on backend and whether C<ev_io_set> was used).
3431		3736
3432	=item Activating one watcher (putting it into the pending state): O(1)
3433
3434	=item Priority handling: O(number_of_priorities)
3435
3436	Priorities are implemented by allocating some space for each
3437	priority. When doing priority-based operations, libev usually has to
3438	linearly search all the priorities, but starting/stopping and activating
3439	watchers becomes O(1) with respect to priority handling.
3440
3441	=item Sending an ev_async: O(1)
3442
3443	=item Processing ev_async_send: O(number_of_async_watchers)
3444
3445	=item Processing signals: O(max_signal_number)
3446
3447	Sending involves a system call I<iff> there were no other C<ev_async_send>
3448	calls in the current loop iteration. Checking for async and signal events
3449	involves iterating over all running async watchers or all signal numbers.
3450
3451	=back
3452
3453
3454	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	3737	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
3455		3738
3456	Win32 doesn't support any of the standards (e.g. POSIX) that libev	3739	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3457	requires, and its I/O model is fundamentally incompatible with the POSIX	3740	requires, and its I/O model is fundamentally incompatible with the POSIX
3458	model. Libev still offers limited functionality on this platform in	3741	model. Libev still offers limited functionality on this platform in
3459	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	3742	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
…		…
3546	wrap all I/O functions and provide your own fd management, but the cost of	3829	wrap all I/O functions and provide your own fd management, but the cost of
3547	calling select (O(n²)) will likely make this unworkable.	3830	calling select (O(n²)) will likely make this unworkable.
3548		3831
3549	=back	3832	=back
3550		3833
3551
3552	=head1 PORTABILITY REQUIREMENTS	3834	=head2 PORTABILITY REQUIREMENTS
3553		3835
3554	In addition to a working ISO-C implementation, libev relies on a few	3836	In addition to a working ISO-C implementation and of course the
3555	additional extensions:	3837	backend-specific APIs, libev relies on a few additional extensions:
3556		3838
3557	=over 4	3839	=over 4
3558		3840
3559	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	3841	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3560	calling conventions regardless of C<ev_watcher_type *>.	3842	calling conventions regardless of C<ev_watcher_type *>.
…		…
3585	except the initial one, and run the default loop in the initial thread as	3867	except the initial one, and run the default loop in the initial thread as
3586	well.	3868	well.
3587		3869
3588	=item C<long> must be large enough for common memory allocation sizes	3870	=item C<long> must be large enough for common memory allocation sizes
3589		3871
3590	To improve portability and simplify using libev, libev uses C<long>	3872	To improve portability and simplify its API, libev uses C<long> internally
3591	internally instead of C<size_t> when allocating its data structures. On	3873	instead of C<size_t> when allocating its data structures. On non-POSIX
3592	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	3874	systems (Microsoft...) this might be unexpectedly low, but is still at
3593	is still at least 31 bits everywhere, which is enough for hundreds of	3875	least 31 bits everywhere, which is enough for hundreds of millions of
3594	millions of watchers.	3876	watchers.
3595		3877
3596	=item C<double> must hold a time value in seconds with enough accuracy	3878	=item C<double> must hold a time value in seconds with enough accuracy
3597		3879
3598	The type C<double> is used to represent timestamps. It is required to	3880	The type C<double> is used to represent timestamps. It is required to
3599	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	3881	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
…		…
3603	=back	3885	=back
3604		3886
3605	If you know of other additional requirements drop me a note.	3887	If you know of other additional requirements drop me a note.
3606		3888
3607		3889
3608	=head1 COMPILER WARNINGS	3890	=head1 ALGORITHMIC COMPLEXITIES
3609		3891
3610	Depending on your compiler and compiler settings, you might get no or a	3892	In this section the complexities of (many of) the algorithms used inside
3611	lot of warnings when compiling libev code. Some people are apparently	3893	libev will be documented. For complexity discussions about backends see
3612	scared by this.	3894	the documentation for C<ev_default_init>.
3613		3895
3614	However, these are unavoidable for many reasons. For one, each compiler	3896	All of the following are about amortised time: If an array needs to be
3615	has different warnings, and each user has different tastes regarding	3897	extended, libev needs to realloc and move the whole array, but this
3616	warning options. "Warn-free" code therefore cannot be a goal except when	3898	happens asymptotically rarer with higher number of elements, so O(1) might
3617	targeting a specific compiler and compiler-version.	3899	mean that libev does a lengthy realloc operation in rare cases, but on
		3900	average it is much faster and asymptotically approaches constant time.
3618		3901
3619	Another reason is that some compiler warnings require elaborate	3902	=over 4
3620	workarounds, or other changes to the code that make it less clear and less
3621	maintainable.
3622		3903
3623	And of course, some compiler warnings are just plain stupid, or simply	3904	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3624	wrong (because they don't actually warn about the condition their message
3625	seems to warn about).
3626		3905
3627	While libev is written to generate as few warnings as possible,	3906	This means that, when you have a watcher that triggers in one hour and
3628	"warn-free" code is not a goal, and it is recommended not to build libev	3907	there are 100 watchers that would trigger before that, then inserting will
3629	with any compiler warnings enabled unless you are prepared to cope with	3908	have to skip roughly seven (C<ld 100>) of these watchers.
3630	them (e.g. by ignoring them). Remember that warnings are just that:
3631	warnings, not errors, or proof of bugs.
3632		3909
		3910	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3633		3911
3634	=head1 VALGRIND	3912	That means that changing a timer costs less than removing/adding them,
		3913	as only the relative motion in the event queue has to be paid for.
3635		3914
3636	Valgrind has a special section here because it is a popular tool that is	3915	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3637	highly useful, but valgrind reports are very hard to interpret.
3638		3916
3639	If you think you found a bug (memory leak, uninitialised data access etc.)	3917	These just add the watcher into an array or at the head of a list.
3640	in libev, then check twice: If valgrind reports something like:
3641		3918
3642	==2274== definitely lost: 0 bytes in 0 blocks.	3919	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3643	==2274== possibly lost: 0 bytes in 0 blocks.
3644	==2274== still reachable: 256 bytes in 1 blocks.
3645		3920
3646	Then there is no memory leak. Similarly, under some circumstances,	3921	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3647	valgrind might report kernel bugs as if it were a bug in libev, or it
3648	might be confused (it is a very good tool, but only a tool).
3649		3922
3650	If you are unsure about something, feel free to contact the mailing list	3923	These watchers are stored in lists, so they need to be walked to find the
3651	with the full valgrind report and an explanation on why you think this is	3924	correct watcher to remove. The lists are usually short (you don't usually
3652	a bug in libev. However, don't be annoyed when you get a brisk "this is	3925	have many watchers waiting for the same fd or signal: one is typical, two
3653	no bug" answer and take the chance of learning how to interpret valgrind	3926	is rare).
3654	properly.
3655		3927
3656	If you need, for some reason, empty reports from valgrind for your project	3928	=item Finding the next timer in each loop iteration: O(1)
3657	I suggest using suppression lists.	3929
		3930	By virtue of using a binary or 4-heap, the next timer is always found at a
		3931	fixed position in the storage array.
		3932
		3933	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		3934
		3935	A change means an I/O watcher gets started or stopped, which requires
		3936	libev to recalculate its status (and possibly tell the kernel, depending
		3937	on backend and whether C<ev_io_set> was used).
		3938
		3939	=item Activating one watcher (putting it into the pending state): O(1)
		3940
		3941	=item Priority handling: O(number_of_priorities)
		3942
		3943	Priorities are implemented by allocating some space for each
		3944	priority. When doing priority-based operations, libev usually has to
		3945	linearly search all the priorities, but starting/stopping and activating
		3946	watchers becomes O(1) with respect to priority handling.
		3947
		3948	=item Sending an ev_async: O(1)
		3949
		3950	=item Processing ev_async_send: O(number_of_async_watchers)
		3951
		3952	=item Processing signals: O(max_signal_number)
		3953
		3954	Sending involves a system call I<iff> there were no other C<ev_async_send>
		3955	calls in the current loop iteration. Checking for async and signal events
		3956	involves iterating over all running async watchers or all signal numbers.
		3957
		3958	=back
3658		3959
3659		3960
3660	=head1 AUTHOR	3961	=head1 AUTHOR
3661		3962
3662	Marc Lehmann <libev@schmorp.de>.	3963	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3663		3964

Diff Legend

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