ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.349 by root, Mon Jan 10 01:58:55 2011 UTC vs.
Revision 1.444 by root, Mon Oct 29 00:00:22 2018 UTC

		1	=encoding utf-8
		2
1	=head1 NAME	3	=head1 NAME
2		4
3	libev - a high performance full-featured event loop written in C	5	libev - a high performance full-featured event loop written in C
4		6
5	=head1 SYNOPSIS	7	=head1 SYNOPSIS
…		…
58	ev_timer_start (loop, &timeout_watcher);	60	ev_timer_start (loop, &timeout_watcher);
59		61
60	// now wait for events to arrive	62	// now wait for events to arrive
61	ev_run (loop, 0);	63	ev_run (loop, 0);
62		64
63	// unloop was called, so exit	65	// break was called, so exit
64	return 0;	66	return 0;
65	}	67	}
66		68
67	=head1 ABOUT THIS DOCUMENT	69	=head1 ABOUT THIS DOCUMENT
68		70
…		…
82		84
83	=head1 WHAT TO READ WHEN IN A HURRY	85	=head1 WHAT TO READ WHEN IN A HURRY
84		86
85	This manual tries to be very detailed, but unfortunately, this also makes	87	This manual tries to be very detailed, but unfortunately, this also makes
86	it very long. If you just want to know the basics of libev, I suggest	88	it very long. If you just want to know the basics of libev, I suggest
87	reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and	89	reading L</ANATOMY OF A WATCHER>, then the L</EXAMPLE PROGRAM> above and
88	look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and	90	look up the missing functions in L</GLOBAL FUNCTIONS> and the C<ev_io> and
89	C<ev_timer> sections in L<WATCHER TYPES>.	91	C<ev_timer> sections in L</WATCHER TYPES>.
90		92
91	=head1 ABOUT LIBEV	93	=head1 ABOUT LIBEV
92		94
93	Libev is an event loop: you register interest in certain events (such as a	95	Libev is an event loop: you register interest in certain events (such as a
94	file descriptor being readable or a timeout occurring), and it will manage	96	file descriptor being readable or a timeout occurring), and it will manage
…		…
174	=item ev_tstamp ev_time ()	176	=item ev_tstamp ev_time ()
175		177
176	Returns the current time as libev would use it. Please note that the	178	Returns the current time as libev would use it. Please note that the
177	C<ev_now> function is usually faster and also often returns the timestamp	179	C<ev_now> function is usually faster and also often returns the timestamp
178	you actually want to know. Also interesting is the combination of	180	you actually want to know. Also interesting is the combination of
179	C<ev_update_now> and C<ev_now>.	181	C<ev_now_update> and C<ev_now>.
180		182
181	=item ev_sleep (ev_tstamp interval)	183	=item ev_sleep (ev_tstamp interval)
182		184
183	Sleep for the given interval: The current thread will be blocked until	185	Sleep for the given interval: The current thread will be blocked
184	either it is interrupted or the given time interval has passed. Basically	186	until either it is interrupted or the given time interval has
		187	passed (approximately - it might return a bit earlier even if not
		188	interrupted). Returns immediately if C<< interval <= 0 >>.
		189
185	this is a sub-second-resolution C<sleep ()>.	190	Basically this is a sub-second-resolution C<sleep ()>.
		191
		192	The range of the C<interval> is limited - libev only guarantees to work
		193	with sleep times of up to one day (C<< interval <= 86400 >>).
186		194
187	=item int ev_version_major ()	195	=item int ev_version_major ()
188		196
189	=item int ev_version_minor ()	197	=item int ev_version_minor ()
190		198
…		…
241	the current system, you would need to look at C<ev_embeddable_backends ()	249	the current system, you would need to look at C<ev_embeddable_backends ()
242	& ev_supported_backends ()>, likewise for recommended ones.	250	& ev_supported_backends ()>, likewise for recommended ones.
243		251
244	See the description of C<ev_embed> watchers for more info.	252	See the description of C<ev_embed> watchers for more info.
245		253
246	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	254	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
247		255
248	Sets the allocation function to use (the prototype is similar - the	256	Sets the allocation function to use (the prototype is similar - the
249	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	257	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
250	used to allocate and free memory (no surprises here). If it returns zero	258	used to allocate and free memory (no surprises here). If it returns zero
251	when memory needs to be allocated (C<size != 0>), the library might abort	259	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
277	}	285	}
278		286
279	...	287	...
280	ev_set_allocator (persistent_realloc);	288	ev_set_allocator (persistent_realloc);
281		289
282	=item ev_set_syserr_cb (void (*cb)(const char *msg))	290	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
283		291
284	Set the callback function to call on a retryable system call error (such	292	Set the callback function to call on a retryable system call error (such
285	as failed select, poll, epoll_wait). The message is a printable string	293	as failed select, poll, epoll_wait). The message is a printable string
286	indicating the system call or subsystem causing the problem. If this	294	indicating the system call or subsystem causing the problem. If this
287	callback is set, then libev will expect it to remedy the situation, no	295	callback is set, then libev will expect it to remedy the situation, no
…		…
305		313
306	This function can be used to "simulate" a signal receive. It is completely	314	This function can be used to "simulate" a signal receive. It is completely
307	safe to call this function at any time, from any context, including signal	315	safe to call this function at any time, from any context, including signal
308	handlers or random threads.	316	handlers or random threads.
309		317
310	It's main use is to customise signal handling in your process, especially	318	Its main use is to customise signal handling in your process, especially
311	in the presence of threads. For example, you could block signals	319	in the presence of threads. For example, you could block signals
312	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when	320	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
313	creating any loops), and in one thread, use C<sigwait> or any other	321	creating any loops), and in one thread, use C<sigwait> or any other
314	mechanism to wait for signals, then "deliver" them to libev by calling	322	mechanism to wait for signals, then "deliver" them to libev by calling
315	C<ev_feed_signal>.	323	C<ev_feed_signal>.
…		…
390		398
391	If this flag bit is or'ed into the flag value (or the program runs setuid	399	If this flag bit is or'ed into the flag value (or the program runs setuid
392	or setgid) then libev will I<not> look at the environment variable	400	or setgid) then libev will I<not> look at the environment variable
393	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	401	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
394	override the flags completely if it is found in the environment. This is	402	override the flags completely if it is found in the environment. This is
395	useful to try out specific backends to test their performance, or to work	403	useful to try out specific backends to test their performance, to work
396	around bugs.	404	around bugs, or to make libev threadsafe (accessing environment variables
		405	cannot be done in a threadsafe way, but usually it works if no other
		406	thread modifies them).
397		407
398	=item C<EVFLAG_FORKCHECK>	408	=item C<EVFLAG_FORKCHECK>
399		409
400	Instead of calling C<ev_loop_fork> manually after a fork, you can also	410	Instead of calling C<ev_loop_fork> manually after a fork, you can also
401	make libev check for a fork in each iteration by enabling this flag.	411	make libev check for a fork in each iteration by enabling this flag.
402		412
403	This works by calling C<getpid ()> on every iteration of the loop,	413	This works by calling C<getpid ()> on every iteration of the loop,
404	and thus this might slow down your event loop if you do a lot of loop	414	and thus this might slow down your event loop if you do a lot of loop
405	iterations and little real work, but is usually not noticeable (on my	415	iterations and little real work, but is usually not noticeable (on my
406	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	416	GNU/Linux system for example, C<getpid> is actually a simple 5-insn
407	without a system call and thus I<very> fast, but my GNU/Linux system also has	417	sequence without a system call and thus I<very> fast, but my GNU/Linux
408	C<pthread_atfork> which is even faster).	418	system also has C<pthread_atfork> which is even faster). (Update: glibc
		419	versions 2.25 apparently removed the C<getpid> optimisation again).
409		420
410	The big advantage of this flag is that you can forget about fork (and	421	The big advantage of this flag is that you can forget about fork (and
411	forget about forgetting to tell libev about forking) when you use this	422	forget about forgetting to tell libev about forking, although you still
412	flag.	423	have to ignore C<SIGPIPE>) when you use this flag.
413		424
414	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	425	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
415	environment variable.	426	environment variable.
416		427
417	=item C<EVFLAG_NOINOTIFY>	428	=item C<EVFLAG_NOINOTIFY>
…		…
435	example) that can't properly initialise their signal masks.	446	example) that can't properly initialise their signal masks.
436		447
437	=item C<EVFLAG_NOSIGMASK>	448	=item C<EVFLAG_NOSIGMASK>
438		449
439	When this flag is specified, then libev will avoid to modify the signal	450	When this flag is specified, then libev will avoid to modify the signal
440	mask. Specifically, this means you ahve to make sure signals are unblocked	451	mask. Specifically, this means you have to make sure signals are unblocked
441	when you want to receive them.	452	when you want to receive them.
442		453
443	This behaviour is useful when you want to do your own signal handling, or	454	This behaviour is useful when you want to do your own signal handling, or
444	want to handle signals only in specific threads and want to avoid libev	455	want to handle signals only in specific threads and want to avoid libev
445	unblocking the signals.	456	unblocking the signals.
		457
		458	It's also required by POSIX in a threaded program, as libev calls
		459	C<sigprocmask>, whose behaviour is officially unspecified.
446		460
447	This flag's behaviour will become the default in future versions of libev.	461	This flag's behaviour will become the default in future versions of libev.
448		462
449	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	463	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
450		464
…		…
480	=item C<EVBACKEND_EPOLL> (value 4, Linux)	494	=item C<EVBACKEND_EPOLL> (value 4, Linux)
481		495
482	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9	496	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
483	kernels).	497	kernels).
484		498
485	For few fds, this backend is a bit little slower than poll and select,	499	For few fds, this backend is a bit little slower than poll and select, but
486	but it scales phenomenally better. While poll and select usually scale	500	it scales phenomenally better. While poll and select usually scale like
487	like O(total_fds) where n is the total number of fds (or the highest fd),	501	O(total_fds) where total_fds is the total number of fds (or the highest
488	epoll scales either O(1) or O(active_fds).	502	fd), epoll scales either O(1) or O(active_fds).
489		503
490	The epoll mechanism deserves honorable mention as the most misdesigned	504	The epoll mechanism deserves honorable mention as the most misdesigned
491	of the more advanced event mechanisms: mere annoyances include silently	505	of the more advanced event mechanisms: mere annoyances include silently
492	dropping file descriptors, requiring a system call per change per file	506	dropping file descriptors, requiring a system call per change per file
493	descriptor (and unnecessary guessing of parameters), problems with dup,	507	descriptor (and unnecessary guessing of parameters), problems with dup,
…		…
496	0.1ms) and so on. The biggest issue is fork races, however - if a program	510	0.1ms) and so on. The biggest issue is fork races, however - if a program
497	forks then I<both> parent and child process have to recreate the epoll	511	forks then I<both> parent and child process have to recreate the epoll
498	set, which can take considerable time (one syscall per file descriptor)	512	set, which can take considerable time (one syscall per file descriptor)
499	and is of course hard to detect.	513	and is of course hard to detect.
500		514
501	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but	515	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
502	of course I<doesn't>, and epoll just loves to report events for totally	516	but of course I<doesn't>, and epoll just loves to report events for
503	I<different> file descriptors (even already closed ones, so one cannot	517	totally I<different> file descriptors (even already closed ones, so
504	even remove them from the set) than registered in the set (especially	518	one cannot even remove them from the set) than registered in the set
505	on SMP systems). Libev tries to counter these spurious notifications by	519	(especially on SMP systems). Libev tries to counter these spurious
506	employing an additional generation counter and comparing that against the	520	notifications by employing an additional generation counter and comparing
507	events to filter out spurious ones, recreating the set when required. Last	521	that against the events to filter out spurious ones, recreating the set
		522	when required. Epoll also erroneously rounds down timeouts, but gives you
		523	no way to know when and by how much, so sometimes you have to busy-wait
		524	because epoll returns immediately despite a nonzero timeout. And last
508	not least, it also refuses to work with some file descriptors which work	525	not least, it also refuses to work with some file descriptors which work
509	perfectly fine with C<select> (files, many character devices...).	526	perfectly fine with C<select> (files, many character devices...).
510		527
511	Epoll is truly the train wreck analog among event poll mechanisms.	528	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		529	cobbled together in a hurry, no thought to design or interaction with
		530	others. Oh, the pain, will it ever stop...
512		531
513	While stopping, setting and starting an I/O watcher in the same iteration	532	While stopping, setting and starting an I/O watcher in the same iteration
514	will result in some caching, there is still a system call per such	533	will result in some caching, there is still a system call per such
515	incident (because the same I<file descriptor> could point to a different	534	incident (because the same I<file descriptor> could point to a different
516	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	535	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
553		572
554	It scales in the same way as the epoll backend, but the interface to the	573	It scales in the same way as the epoll backend, but the interface to the
555	kernel is more efficient (which says nothing about its actual speed, of	574	kernel is more efficient (which says nothing about its actual speed, of
556	course). While stopping, setting and starting an I/O watcher does never	575	course). While stopping, setting and starting an I/O watcher does never
557	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	576	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
558	two event changes per incident. Support for C<fork ()> is very bad (but	577	two event changes per incident. Support for C<fork ()> is very bad (you
559	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect	578	might have to leak fd's on fork, but it's more sane than epoll) and it
560	cases	579	drops fds silently in similarly hard-to-detect cases.
561		580
562	This backend usually performs well under most conditions.	581	This backend usually performs well under most conditions.
563		582
564	While nominally embeddable in other event loops, this doesn't work	583	While nominally embeddable in other event loops, this doesn't work
565	everywhere, so you might need to test for this. And since it is broken	584	everywhere, so you might need to test for this. And since it is broken
…		…
582	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	601	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
583		602
584	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	603	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
585	it's really slow, but it still scales very well (O(active_fds)).	604	it's really slow, but it still scales very well (O(active_fds)).
586		605
587	Please note that Solaris event ports can deliver a lot of spurious
588	notifications, so you need to use non-blocking I/O or other means to avoid
589	blocking when no data (or space) is available.
590
591	While this backend scales well, it requires one system call per active	606	While this backend scales well, it requires one system call per active
592	file descriptor per loop iteration. For small and medium numbers of file	607	file descriptor per loop iteration. For small and medium numbers of file
593	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	608	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
594	might perform better.	609	might perform better.
595		610
596	On the positive side, with the exception of the spurious readiness	611	On the positive side, this backend actually performed fully to
597	notifications, this backend actually performed fully to specification
598	in all tests and is fully embeddable, which is a rare feat among the	612	specification in all tests and is fully embeddable, which is a rare feat
599	OS-specific backends (I vastly prefer correctness over speed hacks).	613	among the OS-specific backends (I vastly prefer correctness over speed
		614	hacks).
		615
		616	On the negative side, the interface is I<bizarre> - so bizarre that
		617	even sun itself gets it wrong in their code examples: The event polling
		618	function sometimes returns events to the caller even though an error
		619	occurred, but with no indication whether it has done so or not (yes, it's
		620	even documented that way) - deadly for edge-triggered interfaces where you
		621	absolutely have to know whether an event occurred or not because you have
		622	to re-arm the watcher.
		623
		624	Fortunately libev seems to be able to work around these idiocies.
600		625
601	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	626	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
602	C<EVBACKEND_POLL>.	627	C<EVBACKEND_POLL>.
603		628
604	=item C<EVBACKEND_ALL>	629	=item C<EVBACKEND_ALL>
…		…
658	If you need dynamically allocated loops it is better to use C<ev_loop_new>	683	If you need dynamically allocated loops it is better to use C<ev_loop_new>
659	and C<ev_loop_destroy>.	684	and C<ev_loop_destroy>.
660		685
661	=item ev_loop_fork (loop)	686	=item ev_loop_fork (loop)
662		687
663	This function sets a flag that causes subsequent C<ev_run> iterations to	688	This function sets a flag that causes subsequent C<ev_run> iterations
664	reinitialise the kernel state for backends that have one. Despite the	689	to reinitialise the kernel state for backends that have one. Despite
665	name, you can call it anytime, but it makes most sense after forking, in	690	the name, you can call it anytime you are allowed to start or stop
666	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the	691	watchers (except inside an C<ev_prepare> callback), but it makes most
		692	sense after forking, in the child process. You I<must> call it (or use
667	child before resuming or calling C<ev_run>.	693	C<EVFLAG_FORKCHECK>) in the child before resuming or calling C<ev_run>.
668		694
		695	In addition, if you want to reuse a loop (via this function or
		696	C<EVFLAG_FORKCHECK>), you I<also> have to ignore C<SIGPIPE>.
		697
669	Again, you I<have> to call it on I<any> loop that you want to re-use after	698	Again, you I<have> to call it on I<any> loop that you want to re-use after
670	a fork, I<even if you do not plan to use the loop in the parent>. This is	699	a fork, I<even if you do not plan to use the loop in the parent>. This is
671	because some kernel interfaces *cough* I<kqueue> *cough* do funny things	700	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
672	during fork.	701	during fork.
673		702
674	On the other hand, you only need to call this function in the child	703	On the other hand, you only need to call this function in the child
…		…
744		773
745	This function is rarely useful, but when some event callback runs for a	774	This function is rarely useful, but when some event callback runs for a
746	very long time without entering the event loop, updating libev's idea of	775	very long time without entering the event loop, updating libev's idea of
747	the current time is a good idea.	776	the current time is a good idea.
748		777
749	See also L<The special problem of time updates> in the C<ev_timer> section.	778	See also L</The special problem of time updates> in the C<ev_timer> section.
750		779
751	=item ev_suspend (loop)	780	=item ev_suspend (loop)
752		781
753	=item ev_resume (loop)	782	=item ev_resume (loop)
754		783
…		…
772	without a previous call to C<ev_suspend>.	801	without a previous call to C<ev_suspend>.
773		802
774	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the	803	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
775	event loop time (see C<ev_now_update>).	804	event loop time (see C<ev_now_update>).
776		805
777	=item ev_run (loop, int flags)	806	=item bool ev_run (loop, int flags)
778		807
779	Finally, this is it, the event handler. This function usually is called	808	Finally, this is it, the event handler. This function usually is called
780	after you have initialised all your watchers and you want to start	809	after you have initialised all your watchers and you want to start
781	handling events. It will ask the operating system for any new events, call	810	handling events. It will ask the operating system for any new events, call
782	the watcher callbacks, an then repeat the whole process indefinitely: This	811	the watcher callbacks, and then repeat the whole process indefinitely: This
783	is why event loops are called I<loops>.	812	is why event loops are called I<loops>.
784		813
785	If the flags argument is specified as C<0>, it will keep handling events	814	If the flags argument is specified as C<0>, it will keep handling events
786	until either no event watchers are active anymore or C<ev_break> was	815	until either no event watchers are active anymore or C<ev_break> was
787	called.	816	called.
		817
		818	The return value is false if there are no more active watchers (which
		819	usually means "all jobs done" or "deadlock"), and true in all other cases
		820	(which usually means " you should call C<ev_run> again").
788		821
789	Please note that an explicit C<ev_break> is usually better than	822	Please note that an explicit C<ev_break> is usually better than
790	relying on all watchers to be stopped when deciding when a program has	823	relying on all watchers to be stopped when deciding when a program has
791	finished (especially in interactive programs), but having a program	824	finished (especially in interactive programs), but having a program
792	that automatically loops as long as it has to and no longer by virtue	825	that automatically loops as long as it has to and no longer by virtue
793	of relying on its watchers stopping correctly, that is truly a thing of	826	of relying on its watchers stopping correctly, that is truly a thing of
794	beauty.	827	beauty.
795		828
796	This function is also I<mostly> exception-safe - you can break out of	829	This function is I<mostly> exception-safe - you can break out of a
797	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++	830	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
798	exception and so on. This does not decrement the C<ev_depth> value, nor	831	exception and so on. This does not decrement the C<ev_depth> value, nor
799	will it clear any outstanding C<EVBREAK_ONE> breaks.	832	will it clear any outstanding C<EVBREAK_ONE> breaks.
800		833
801	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle	834	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
802	those events and any already outstanding ones, but will not wait and	835	those events and any already outstanding ones, but will not wait and
…		…
814	This is useful if you are waiting for some external event in conjunction	847	This is useful if you are waiting for some external event in conjunction
815	with something not expressible using other libev watchers (i.e. "roll your	848	with something not expressible using other libev watchers (i.e. "roll your
816	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	849	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
817	usually a better approach for this kind of thing.	850	usually a better approach for this kind of thing.
818		851
819	Here are the gory details of what C<ev_run> does:	852	Here are the gory details of what C<ev_run> does (this is for your
		853	understanding, not a guarantee that things will work exactly like this in
		854	future versions):
820		855
821	- Increment loop depth.	856	- Increment loop depth.
822	- Reset the ev_break status.	857	- Reset the ev_break status.
823	- Before the first iteration, call any pending watchers.	858	- Before the first iteration, call any pending watchers.
824	LOOP:	859	LOOP:
…		…
857	anymore.	892	anymore.
858		893
859	... queue jobs here, make sure they register event watchers as long	894	... queue jobs here, make sure they register event watchers as long
860	... as they still have work to do (even an idle watcher will do..)	895	... as they still have work to do (even an idle watcher will do..)
861	ev_run (my_loop, 0);	896	ev_run (my_loop, 0);
862	... jobs done or somebody called unloop. yeah!	897	... jobs done or somebody called break. yeah!
863		898
864	=item ev_break (loop, how)	899	=item ev_break (loop, how)
865		900
866	Can be used to make a call to C<ev_run> return early (but only after it	901	Can be used to make a call to C<ev_run> return early (but only after it
867	has processed all outstanding events). The C<how> argument must be either	902	has processed all outstanding events). The C<how> argument must be either
…		…
930	overhead for the actual polling but can deliver many events at once.	965	overhead for the actual polling but can deliver many events at once.
931		966
932	By setting a higher I<io collect interval> you allow libev to spend more	967	By setting a higher I<io collect interval> you allow libev to spend more
933	time collecting I/O events, so you can handle more events per iteration,	968	time collecting I/O events, so you can handle more events per iteration,
934	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	969	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
935	C<ev_timer>) will be not affected. Setting this to a non-null value will	970	C<ev_timer>) will not be affected. Setting this to a non-null value will
936	introduce an additional C<ev_sleep ()> call into most loop iterations. The	971	introduce an additional C<ev_sleep ()> call into most loop iterations. The
937	sleep time ensures that libev will not poll for I/O events more often then	972	sleep time ensures that libev will not poll for I/O events more often then
938	once per this interval, on average.	973	once per this interval, on average (as long as the host time resolution is
		974	good enough).
939		975
940	Likewise, by setting a higher I<timeout collect interval> you allow libev	976	Likewise, by setting a higher I<timeout collect interval> you allow libev
941	to spend more time collecting timeouts, at the expense of increased	977	to spend more time collecting timeouts, at the expense of increased
942	latency/jitter/inexactness (the watcher callback will be called	978	latency/jitter/inexactness (the watcher callback will be called
943	later). C<ev_io> watchers will not be affected. Setting this to a non-null	979	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
989	invoke the actual watchers inside another context (another thread etc.).	1025	invoke the actual watchers inside another context (another thread etc.).
990		1026
991	If you want to reset the callback, use C<ev_invoke_pending> as new	1027	If you want to reset the callback, use C<ev_invoke_pending> as new
992	callback.	1028	callback.
993		1029
994	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))	1030	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
995		1031
996	Sometimes you want to share the same loop between multiple threads. This	1032	Sometimes you want to share the same loop between multiple threads. This
997	can be done relatively simply by putting mutex_lock/unlock calls around	1033	can be done relatively simply by putting mutex_lock/unlock calls around
998	each call to a libev function.	1034	each call to a libev function.
999		1035
1000	However, C<ev_run> can run an indefinite time, so it is not feasible	1036	However, C<ev_run> can run an indefinite time, so it is not feasible
1001	to wait for it to return. One way around this is to wake up the event	1037	to wait for it to return. One way around this is to wake up the event
1002	loop via C<ev_break> and C<av_async_send>, another way is to set these	1038	loop via C<ev_break> and C<ev_async_send>, another way is to set these
1003	I<release> and I<acquire> callbacks on the loop.	1039	I<release> and I<acquire> callbacks on the loop.
1004		1040
1005	When set, then C<release> will be called just before the thread is	1041	When set, then C<release> will be called just before the thread is
1006	suspended waiting for new events, and C<acquire> is called just	1042	suspended waiting for new events, and C<acquire> is called just
1007	afterwards.	1043	afterwards.
…		…
1147		1183
1148	=item C<EV_PREPARE>	1184	=item C<EV_PREPARE>
1149		1185
1150	=item C<EV_CHECK>	1186	=item C<EV_CHECK>
1151		1187
1152	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts	1188	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1153	to gather new events, and all C<ev_check> watchers are invoked just after	1189	gather new events, and all C<ev_check> watchers are queued (not invoked)
1154	C<ev_run> has gathered them, but before it invokes any callbacks for any	1190	just after C<ev_run> has gathered them, but before it queues any callbacks
		1191	for any received events. That means C<ev_prepare> watchers are the last
		1192	watchers invoked before the event loop sleeps or polls for new events, and
		1193	C<ev_check> watchers will be invoked before any other watchers of the same
		1194	or lower priority within an event loop iteration.
		1195
1155	received events. Callbacks of both watcher types can start and stop as	1196	Callbacks of both watcher types can start and stop as many watchers as
1156	many watchers as they want, and all of them will be taken into account	1197	they want, and all of them will be taken into account (for example, a
1157	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1198	C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1158	C<ev_run> from blocking).	1199	blocking).
1159		1200
1160	=item C<EV_EMBED>	1201	=item C<EV_EMBED>
1161		1202
1162	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1203	The embedded event loop specified in the C<ev_embed> watcher needs attention.
1163		1204
…		…
1286		1327
1287	=item callback ev_cb (ev_TYPE *watcher)	1328	=item callback ev_cb (ev_TYPE *watcher)
1288		1329
1289	Returns the callback currently set on the watcher.	1330	Returns the callback currently set on the watcher.
1290		1331
1291	=item ev_cb_set (ev_TYPE *watcher, callback)	1332	=item ev_set_cb (ev_TYPE *watcher, callback)
1292		1333
1293	Change the callback. You can change the callback at virtually any time	1334	Change the callback. You can change the callback at virtually any time
1294	(modulo threads).	1335	(modulo threads).
1295		1336
1296	=item ev_set_priority (ev_TYPE *watcher, int priority)	1337	=item ev_set_priority (ev_TYPE *watcher, int priority)
…		…
1314	or might not have been clamped to the valid range.	1355	or might not have been clamped to the valid range.
1315		1356
1316	The default priority used by watchers when no priority has been set is	1357	The default priority used by watchers when no priority has been set is
1317	always C<0>, which is supposed to not be too high and not be too low :).	1358	always C<0>, which is supposed to not be too high and not be too low :).
1318		1359
1319	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of	1360	See L</WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1320	priorities.	1361	priorities.
1321		1362
1322	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1363	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1323		1364
1324	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1365	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
…		…
1349	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related	1390	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1350	functions that do not need a watcher.	1391	functions that do not need a watcher.
1351		1392
1352	=back	1393	=back
1353		1394
1354	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1395	See also the L</ASSOCIATING CUSTOM DATA WITH A WATCHER> and L</BUILDING YOUR
1355		1396	OWN COMPOSITE WATCHERS> idioms.
1356	Each watcher has, by default, a member C<void *data> that you can change
1357	and read at any time: libev will completely ignore it. This can be used
1358	to associate arbitrary data with your watcher. If you need more data and
1359	don't want to allocate memory and store a pointer to it in that data
1360	member, you can also "subclass" the watcher type and provide your own
1361	data:
1362
1363	struct my_io
1364	{
1365	ev_io io;
1366	int otherfd;
1367	void *somedata;
1368	struct whatever *mostinteresting;
1369	};
1370
1371	...
1372	struct my_io w;
1373	ev_io_init (&w.io, my_cb, fd, EV_READ);
1374
1375	And since your callback will be called with a pointer to the watcher, you
1376	can cast it back to your own type:
1377
1378	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1379	{
1380	struct my_io *w = (struct my_io *)w_;
1381	...
1382	}
1383
1384	More interesting and less C-conformant ways of casting your callback type
1385	instead have been omitted.
1386
1387	Another common scenario is to use some data structure with multiple
1388	embedded watchers:
1389
1390	struct my_biggy
1391	{
1392	int some_data;
1393	ev_timer t1;
1394	ev_timer t2;
1395	}
1396
1397	In this case getting the pointer to C<my_biggy> is a bit more
1398	complicated: Either you store the address of your C<my_biggy> struct
1399	in the C<data> member of the watcher (for woozies), or you need to use
1400	some pointer arithmetic using C<offsetof> inside your watchers (for real
1401	programmers):
1402
1403	#include <stddef.h>
1404
1405	static void
1406	t1_cb (EV_P_ ev_timer *w, int revents)
1407	{
1408	struct my_biggy big = (struct my_biggy *)
1409	(((char *)w) - offsetof (struct my_biggy, t1));
1410	}
1411
1412	static void
1413	t2_cb (EV_P_ ev_timer *w, int revents)
1414	{
1415	struct my_biggy big = (struct my_biggy *)
1416	(((char *)w) - offsetof (struct my_biggy, t2));
1417	}
1418		1397
1419	=head2 WATCHER STATES	1398	=head2 WATCHER STATES
1420		1399
1421	There are various watcher states mentioned throughout this manual -	1400	There are various watcher states mentioned throughout this manual -
1422	active, pending and so on. In this section these states and the rules to	1401	active, pending and so on. In this section these states and the rules to
1423	transition between them will be described in more detail - and while these	1402	transition between them will be described in more detail - and while these
1424	rules might look complicated, they usually do "the right thing".	1403	rules might look complicated, they usually do "the right thing".
1425		1404
1426	=over 4	1405	=over 4
1427		1406
1428	=item initialiased	1407	=item initialised
1429		1408
1430	Before a watcher can be registered with the event looop it has to be	1409	Before a watcher can be registered with the event loop it has to be
1431	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to	1410	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
1432	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.	1411	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1433		1412
1434	In this state it is simply some block of memory that is suitable for use	1413	In this state it is simply some block of memory that is suitable for
1435	in an event loop. It can be moved around, freed, reused etc. at will.	1414	use in an event loop. It can be moved around, freed, reused etc. at
		1415	will - as long as you either keep the memory contents intact, or call
		1416	C<ev_TYPE_init> again.
1436		1417
1437	=item started/running/active	1418	=item started/running/active
1438		1419
1439	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes	1420	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1440	property of the event loop, and is actively waiting for events. While in	1421	property of the event loop, and is actively waiting for events. While in
…		…
1468	latter will clear any pending state the watcher might be in, regardless	1449	latter will clear any pending state the watcher might be in, regardless
1469	of whether it was active or not, so stopping a watcher explicitly before	1450	of whether it was active or not, so stopping a watcher explicitly before
1470	freeing it is often a good idea.	1451	freeing it is often a good idea.
1471		1452
1472	While stopped (and not pending) the watcher is essentially in the	1453	While stopped (and not pending) the watcher is essentially in the
1473	initialised state, that is it can be reused, moved, modified in any way	1454	initialised state, that is, it can be reused, moved, modified in any way
1474	you wish.	1455	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1456	it again).
1475		1457
1476	=back	1458	=back
1477		1459
1478	=head2 WATCHER PRIORITY MODELS	1460	=head2 WATCHER PRIORITY MODELS
1479		1461
…		…
1608	In general you can register as many read and/or write event watchers per	1590	In general you can register as many read and/or write event watchers per
1609	fd as you want (as long as you don't confuse yourself). Setting all file	1591	fd as you want (as long as you don't confuse yourself). Setting all file
1610	descriptors to non-blocking mode is also usually a good idea (but not	1592	descriptors to non-blocking mode is also usually a good idea (but not
1611	required if you know what you are doing).	1593	required if you know what you are doing).
1612		1594
1613	If you cannot use non-blocking mode, then force the use of a
1614	known-to-be-good backend (at the time of this writing, this includes only
1615	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1616	descriptors for which non-blocking operation makes no sense (such as
1617	files) - libev doesn't guarantee any specific behaviour in that case.
1618
1619	Another thing you have to watch out for is that it is quite easy to	1595	Another thing you have to watch out for is that it is quite easy to
1620	receive "spurious" readiness notifications, that is your callback might	1596	receive "spurious" readiness notifications, that is, your callback might
1621	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1597	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1622	because there is no data. Not only are some backends known to create a	1598	because there is no data. It is very easy to get into this situation even
1623	lot of those (for example Solaris ports), it is very easy to get into	1599	with a relatively standard program structure. Thus it is best to always
1624	this situation even with a relatively standard program structure. Thus	1600	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1625	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1626	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1601	preferable to a program hanging until some data arrives.
1627		1602
1628	If you cannot run the fd in non-blocking mode (for example you should	1603	If you cannot run the fd in non-blocking mode (for example you should
1629	not play around with an Xlib connection), then you have to separately	1604	not play around with an Xlib connection), then you have to separately
1630	re-test whether a file descriptor is really ready with a known-to-be good	1605	re-test whether a file descriptor is really ready with a known-to-be good
1631	interface such as poll (fortunately in our Xlib example, Xlib already	1606	interface such as poll (fortunately in the case of Xlib, it already does
1632	does this on its own, so its quite safe to use). Some people additionally	1607	this on its own, so its quite safe to use). Some people additionally
1633	use C<SIGALRM> and an interval timer, just to be sure you won't block	1608	use C<SIGALRM> and an interval timer, just to be sure you won't block
1634	indefinitely.	1609	indefinitely.
1635		1610
1636	But really, best use non-blocking mode.	1611	But really, best use non-blocking mode.
1637		1612
…		…
1665		1640
1666	There is no workaround possible except not registering events	1641	There is no workaround possible except not registering events
1667	for potentially C<dup ()>'ed file descriptors, or to resort to	1642	for potentially C<dup ()>'ed file descriptors, or to resort to
1668	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1643	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1669		1644
		1645	=head3 The special problem of files
		1646
		1647	Many people try to use C<select> (or libev) on file descriptors
		1648	representing files, and expect it to become ready when their program
		1649	doesn't block on disk accesses (which can take a long time on their own).
		1650
		1651	However, this cannot ever work in the "expected" way - you get a readiness
		1652	notification as soon as the kernel knows whether and how much data is
		1653	there, and in the case of open files, that's always the case, so you
		1654	always get a readiness notification instantly, and your read (or possibly
		1655	write) will still block on the disk I/O.
		1656
		1657	Another way to view it is that in the case of sockets, pipes, character
		1658	devices and so on, there is another party (the sender) that delivers data
		1659	on its own, but in the case of files, there is no such thing: the disk
		1660	will not send data on its own, simply because it doesn't know what you
		1661	wish to read - you would first have to request some data.
		1662
		1663	Since files are typically not-so-well supported by advanced notification
		1664	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1665	to files, even though you should not use it. The reason for this is
		1666	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1667	usually a tty, often a pipe, but also sometimes files or special devices
		1668	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1669	F</dev/urandom>), and even though the file might better be served with
		1670	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1671	it "just works" instead of freezing.
		1672
		1673	So avoid file descriptors pointing to files when you know it (e.g. use
		1674	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1675	when you rarely read from a file instead of from a socket, and want to
		1676	reuse the same code path.
		1677
1670	=head3 The special problem of fork	1678	=head3 The special problem of fork
1671		1679
1672	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1680	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1673	useless behaviour. Libev fully supports fork, but needs to be told about	1681	useless behaviour. Libev fully supports fork, but needs to be told about
1674	it in the child.	1682	it in the child if you want to continue to use it in the child.
1675		1683
1676	To support fork in your programs, you either have to call	1684	To support fork in your child processes, you have to call C<ev_loop_fork
1677	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1685	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1678	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1686	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1679	C<EVBACKEND_POLL>.
1680		1687
1681	=head3 The special problem of SIGPIPE	1688	=head3 The special problem of SIGPIPE
1682		1689
1683	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1690	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1684	when writing to a pipe whose other end has been closed, your program gets	1691	when writing to a pipe whose other end has been closed, your program gets
…		…
1782	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1789	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1783	monotonic clock option helps a lot here).	1790	monotonic clock option helps a lot here).
1784		1791
1785	The callback is guaranteed to be invoked only I<after> its timeout has	1792	The callback is guaranteed to be invoked only I<after> its timeout has
1786	passed (not I<at>, so on systems with very low-resolution clocks this	1793	passed (not I<at>, so on systems with very low-resolution clocks this
1787	might introduce a small delay). If multiple timers become ready during the	1794	might introduce a small delay, see "the special problem of being too
		1795	early", below). If multiple timers become ready during the same loop
1788	same loop iteration then the ones with earlier time-out values are invoked	1796	iteration then the ones with earlier time-out values are invoked before
1789	before ones of the same priority with later time-out values (but this is	1797	ones of the same priority with later time-out values (but this is no
1790	no longer true when a callback calls C<ev_run> recursively).	1798	longer true when a callback calls C<ev_run> recursively).
1791		1799
1792	=head3 Be smart about timeouts	1800	=head3 Be smart about timeouts
1793		1801
1794	Many real-world problems involve some kind of timeout, usually for error	1802	Many real-world problems involve some kind of timeout, usually for error
1795	recovery. A typical example is an HTTP request - if the other side hangs,	1803	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1870		1878
1871	In this case, it would be more efficient to leave the C<ev_timer> alone,	1879	In this case, it would be more efficient to leave the C<ev_timer> alone,
1872	but remember the time of last activity, and check for a real timeout only	1880	but remember the time of last activity, and check for a real timeout only
1873	within the callback:	1881	within the callback:
1874		1882
		1883	ev_tstamp timeout = 60.;
1875	ev_tstamp last_activity; // time of last activity	1884	ev_tstamp last_activity; // time of last activity
		1885	ev_timer timer;
1876		1886
1877	static void	1887	static void
1878	callback (EV_P_ ev_timer *w, int revents)	1888	callback (EV_P_ ev_timer *w, int revents)
1879	{	1889	{
1880	ev_tstamp now = ev_now (EV_A);	1890	// calculate when the timeout would happen
1881	ev_tstamp timeout = last_activity + 60.;	1891	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1882		1892
1883	// if last_activity + 60. is older than now, we did time out	1893	// if negative, it means we the timeout already occurred
1884	if (timeout < now)	1894	if (after < 0.)
1885	{	1895	{
1886	// timeout occurred, take action	1896	// timeout occurred, take action
1887	}	1897	}
1888	else	1898	else
1889	{	1899	{
1890	// callback was invoked, but there was some activity, re-arm	1900	// callback was invoked, but there was some recent
1891	// the watcher to fire in last_activity + 60, which is	1901	// activity. simply restart the timer to time out
1892	// guaranteed to be in the future, so "again" is positive:	1902	// after "after" seconds, which is the earliest time
1893	w->repeat = timeout - now;	1903	// the timeout can occur.
		1904	ev_timer_set (w, after, 0.);
1894	ev_timer_again (EV_A_ w);	1905	ev_timer_start (EV_A_ w);
1895	}	1906	}
1896	}	1907	}
1897		1908
1898	To summarise the callback: first calculate the real timeout (defined	1909	To summarise the callback: first calculate in how many seconds the
1899	as "60 seconds after the last activity"), then check if that time has	1910	timeout will occur (by calculating the absolute time when it would occur,
1900	been reached, which means something I<did>, in fact, time out. Otherwise	1911	C<last_activity + timeout>, and subtracting the current time, C<ev_now
1901	the callback was invoked too early (C<timeout> is in the future), so	1912	(EV_A)> from that).
1902	re-schedule the timer to fire at that future time, to see if maybe we have
1903	a timeout then.
1904		1913
1905	Note how C<ev_timer_again> is used, taking advantage of the	1914	If this value is negative, then we are already past the timeout, i.e. we
1906	C<ev_timer_again> optimisation when the timer is already running.	1915	timed out, and need to do whatever is needed in this case.
		1916
		1917	Otherwise, we now the earliest time at which the timeout would trigger,
		1918	and simply start the timer with this timeout value.
		1919
		1920	In other words, each time the callback is invoked it will check whether
		1921	the timeout occurred. If not, it will simply reschedule itself to check
		1922	again at the earliest time it could time out. Rinse. Repeat.
1907		1923
1908	This scheme causes more callback invocations (about one every 60 seconds	1924	This scheme causes more callback invocations (about one every 60 seconds
1909	minus half the average time between activity), but virtually no calls to	1925	minus half the average time between activity), but virtually no calls to
1910	libev to change the timeout.	1926	libev to change the timeout.
1911		1927
1912	To start the timer, simply initialise the watcher and set C<last_activity>	1928	To start the machinery, simply initialise the watcher and set
1913	to the current time (meaning we just have some activity :), then call the	1929	C<last_activity> to the current time (meaning there was some activity just
1914	callback, which will "do the right thing" and start the timer:	1930	now), then call the callback, which will "do the right thing" and start
		1931	the timer:
1915		1932
		1933	last_activity = ev_now (EV_A);
1916	ev_init (timer, callback);	1934	ev_init (&timer, callback);
1917	last_activity = ev_now (loop);	1935	callback (EV_A_ &timer, 0);
1918	callback (loop, timer, EV_TIMER);
1919		1936
1920	And when there is some activity, simply store the current time in	1937	When there is some activity, simply store the current time in
1921	C<last_activity>, no libev calls at all:	1938	C<last_activity>, no libev calls at all:
1922		1939
		1940	if (activity detected)
1923	last_activity = ev_now (loop);	1941	last_activity = ev_now (EV_A);
		1942
		1943	When your timeout value changes, then the timeout can be changed by simply
		1944	providing a new value, stopping the timer and calling the callback, which
		1945	will again do the right thing (for example, time out immediately :).
		1946
		1947	timeout = new_value;
		1948	ev_timer_stop (EV_A_ &timer);
		1949	callback (EV_A_ &timer, 0);
1924		1950
1925	This technique is slightly more complex, but in most cases where the	1951	This technique is slightly more complex, but in most cases where the
1926	time-out is unlikely to be triggered, much more efficient.	1952	time-out is unlikely to be triggered, much more efficient.
1927
1928	Changing the timeout is trivial as well (if it isn't hard-coded in the
1929	callback :) - just change the timeout and invoke the callback, which will
1930	fix things for you.
1931		1953
1932	=item 4. Wee, just use a double-linked list for your timeouts.	1954	=item 4. Wee, just use a double-linked list for your timeouts.
1933		1955
1934	If there is not one request, but many thousands (millions...), all	1956	If there is not one request, but many thousands (millions...), all
1935	employing some kind of timeout with the same timeout value, then one can	1957	employing some kind of timeout with the same timeout value, then one can
…		…
1962	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is	1984	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1963	rather complicated, but extremely efficient, something that really pays	1985	rather complicated, but extremely efficient, something that really pays
1964	off after the first million or so of active timers, i.e. it's usually	1986	off after the first million or so of active timers, i.e. it's usually
1965	overkill :)	1987	overkill :)
1966		1988
		1989	=head3 The special problem of being too early
		1990
		1991	If you ask a timer to call your callback after three seconds, then
		1992	you expect it to be invoked after three seconds - but of course, this
		1993	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		1994	guaranteed to any precision by libev - imagine somebody suspending the
		1995	process with a STOP signal for a few hours for example.
		1996
		1997	So, libev tries to invoke your callback as soon as possible I<after> the
		1998	delay has occurred, but cannot guarantee this.
		1999
		2000	A less obvious failure mode is calling your callback too early: many event
		2001	loops compare timestamps with a "elapsed delay >= requested delay", but
		2002	this can cause your callback to be invoked much earlier than you would
		2003	expect.
		2004
		2005	To see why, imagine a system with a clock that only offers full second
		2006	resolution (think windows if you can't come up with a broken enough OS
		2007	yourself). If you schedule a one-second timer at the time 500.9, then the
		2008	event loop will schedule your timeout to elapse at a system time of 500
		2009	(500.9 truncated to the resolution) + 1, or 501.
		2010
		2011	If an event library looks at the timeout 0.1s later, it will see "501 >=
		2012	501" and invoke the callback 0.1s after it was started, even though a
		2013	one-second delay was requested - this is being "too early", despite best
		2014	intentions.
		2015
		2016	This is the reason why libev will never invoke the callback if the elapsed
		2017	delay equals the requested delay, but only when the elapsed delay is
		2018	larger than the requested delay. In the example above, libev would only invoke
		2019	the callback at system time 502, or 1.1s after the timer was started.
		2020
		2021	So, while libev cannot guarantee that your callback will be invoked
		2022	exactly when requested, it I<can> and I<does> guarantee that the requested
		2023	delay has actually elapsed, or in other words, it always errs on the "too
		2024	late" side of things.
		2025
1967	=head3 The special problem of time updates	2026	=head3 The special problem of time updates
1968		2027
1969	Establishing the current time is a costly operation (it usually takes at	2028	Establishing the current time is a costly operation (it usually takes
1970	least two system calls): EV therefore updates its idea of the current	2029	at least one system call): EV therefore updates its idea of the current
1971	time only before and after C<ev_run> collects new events, which causes a	2030	time only before and after C<ev_run> collects new events, which causes a
1972	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2031	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1973	lots of events in one iteration.	2032	lots of events in one iteration.
1974		2033
1975	The relative timeouts are calculated relative to the C<ev_now ()>	2034	The relative timeouts are calculated relative to the C<ev_now ()>
1976	time. This is usually the right thing as this timestamp refers to the time	2035	time. This is usually the right thing as this timestamp refers to the time
1977	of the event triggering whatever timeout you are modifying/starting. If	2036	of the event triggering whatever timeout you are modifying/starting. If
1978	you suspect event processing to be delayed and you I<need> to base the	2037	you suspect event processing to be delayed and you I<need> to base the
1979	timeout on the current time, use something like this to adjust for this:	2038	timeout on the current time, use something like the following to adjust
		2039	for it:
1980		2040
1981	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2041	ev_timer_set (&timer, after + (ev_time () - ev_now ()), 0.);
1982		2042
1983	If the event loop is suspended for a long time, you can also force an	2043	If the event loop is suspended for a long time, you can also force an
1984	update of the time returned by C<ev_now ()> by calling C<ev_now_update	2044	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1985	()>.	2045	()>, although that will push the event time of all outstanding events
		2046	further into the future.
		2047
		2048	=head3 The special problem of unsynchronised clocks
		2049
		2050	Modern systems have a variety of clocks - libev itself uses the normal
		2051	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2052	jumps).
		2053
		2054	Neither of these clocks is synchronised with each other or any other clock
		2055	on the system, so C<ev_time ()> might return a considerably different time
		2056	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2057	a call to C<gettimeofday> might return a second count that is one higher
		2058	than a directly following call to C<time>.
		2059
		2060	The moral of this is to only compare libev-related timestamps with
		2061	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2062	a second or so.
		2063
		2064	One more problem arises due to this lack of synchronisation: if libev uses
		2065	the system monotonic clock and you compare timestamps from C<ev_time>
		2066	or C<ev_now> from when you started your timer and when your callback is
		2067	invoked, you will find that sometimes the callback is a bit "early".
		2068
		2069	This is because C<ev_timer>s work in real time, not wall clock time, so
		2070	libev makes sure your callback is not invoked before the delay happened,
		2071	I<measured according to the real time>, not the system clock.
		2072
		2073	If your timeouts are based on a physical timescale (e.g. "time out this
		2074	connection after 100 seconds") then this shouldn't bother you as it is
		2075	exactly the right behaviour.
		2076
		2077	If you want to compare wall clock/system timestamps to your timers, then
		2078	you need to use C<ev_periodic>s, as these are based on the wall clock
		2079	time, where your comparisons will always generate correct results.
1986		2080
1987	=head3 The special problems of suspended animation	2081	=head3 The special problems of suspended animation
1988		2082
1989	When you leave the server world it is quite customary to hit machines that	2083	When you leave the server world it is quite customary to hit machines that
1990	can suspend/hibernate - what happens to the clocks during such a suspend?	2084	can suspend/hibernate - what happens to the clocks during such a suspend?
…		…
2020		2114
2021	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2115	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
2022		2116
2023	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2117	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
2024		2118
2025	Configure the timer to trigger after C<after> seconds. If C<repeat>	2119	Configure the timer to trigger after C<after> seconds (fractional and
2026	is C<0.>, then it will automatically be stopped once the timeout is	2120	negative values are supported). If C<repeat> is C<0.>, then it will
2027	reached. If it is positive, then the timer will automatically be	2121	automatically be stopped once the timeout is reached. If it is positive,
2028	configured to trigger again C<repeat> seconds later, again, and again,	2122	then the timer will automatically be configured to trigger again C<repeat>
2029	until stopped manually.	2123	seconds later, again, and again, until stopped manually.
2030		2124
2031	The timer itself will do a best-effort at avoiding drift, that is, if	2125	The timer itself will do a best-effort at avoiding drift, that is, if
2032	you configure a timer to trigger every 10 seconds, then it will normally	2126	you configure a timer to trigger every 10 seconds, then it will normally
2033	trigger at exactly 10 second intervals. If, however, your program cannot	2127	trigger at exactly 10 second intervals. If, however, your program cannot
2034	keep up with the timer (because it takes longer than those 10 seconds to	2128	keep up with the timer (because it takes longer than those 10 seconds to
2035	do stuff) the timer will not fire more than once per event loop iteration.	2129	do stuff) the timer will not fire more than once per event loop iteration.
2036		2130
2037	=item ev_timer_again (loop, ev_timer *)	2131	=item ev_timer_again (loop, ev_timer *)
2038		2132
2039	This will act as if the timer timed out and restart it again if it is	2133	This will act as if the timer timed out, and restarts it again if it is
2040	repeating. The exact semantics are:	2134	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2135	timeout to the C<repeat> value and calling C<ev_timer_start>.
2041		2136
		2137	The exact semantics are as in the following rules, all of which will be
		2138	applied to the watcher:
		2139
		2140	=over 4
		2141
2042	If the timer is pending, its pending status is cleared.	2142	=item If the timer is pending, the pending status is always cleared.
2043		2143
2044	If the timer is started but non-repeating, stop it (as if it timed out).	2144	=item If the timer is started but non-repeating, stop it (as if it timed
		2145	out, without invoking it).
2045		2146
2046	If the timer is repeating, either start it if necessary (with the	2147	=item If the timer is repeating, make the C<repeat> value the new timeout
2047	C<repeat> value), or reset the running timer to the C<repeat> value.	2148	and start the timer, if necessary.
2048		2149
		2150	=back
		2151
2049	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a	2152	This sounds a bit complicated, see L</Be smart about timeouts>, above, for a
2050	usage example.	2153	usage example.
2051		2154
2052	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)	2155	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
2053		2156
2054	Returns the remaining time until a timer fires. If the timer is active,	2157	Returns the remaining time until a timer fires. If the timer is active,
…		…
2107	Periodic watchers are also timers of a kind, but they are very versatile	2210	Periodic watchers are also timers of a kind, but they are very versatile
2108	(and unfortunately a bit complex).	2211	(and unfortunately a bit complex).
2109		2212
2110	Unlike C<ev_timer>, periodic watchers are not based on real time (or	2213	Unlike C<ev_timer>, periodic watchers are not based on real time (or
2111	relative time, the physical time that passes) but on wall clock time	2214	relative time, the physical time that passes) but on wall clock time
2112	(absolute time, the thing you can read on your calender or clock). The	2215	(absolute time, the thing you can read on your calendar or clock). The
2113	difference is that wall clock time can run faster or slower than real	2216	difference is that wall clock time can run faster or slower than real
2114	time, and time jumps are not uncommon (e.g. when you adjust your	2217	time, and time jumps are not uncommon (e.g. when you adjust your
2115	wrist-watch).	2218	wrist-watch).
2116		2219
2117	You can tell a periodic watcher to trigger after some specific point	2220	You can tell a periodic watcher to trigger after some specific point
…		…
2122	C<ev_timer>, which would still trigger roughly 10 seconds after starting	2225	C<ev_timer>, which would still trigger roughly 10 seconds after starting
2123	it, as it uses a relative timeout).	2226	it, as it uses a relative timeout).
2124		2227
2125	C<ev_periodic> watchers can also be used to implement vastly more complex	2228	C<ev_periodic> watchers can also be used to implement vastly more complex
2126	timers, such as triggering an event on each "midnight, local time", or	2229	timers, such as triggering an event on each "midnight, local time", or
2127	other complicated rules. This cannot be done with C<ev_timer> watchers, as	2230	other complicated rules. This cannot easily be done with C<ev_timer>
2128	those cannot react to time jumps.	2231	watchers, as those cannot react to time jumps.
2129		2232
2130	As with timers, the callback is guaranteed to be invoked only when the	2233	As with timers, the callback is guaranteed to be invoked only when the
2131	point in time where it is supposed to trigger has passed. If multiple	2234	point in time where it is supposed to trigger has passed. If multiple
2132	timers become ready during the same loop iteration then the ones with	2235	timers become ready during the same loop iteration then the ones with
2133	earlier time-out values are invoked before ones with later time-out values	2236	earlier time-out values are invoked before ones with later time-out values
…		…
2174		2277
2175	Another way to think about it (for the mathematically inclined) is that	2278	Another way to think about it (for the mathematically inclined) is that
2176	C<ev_periodic> will try to run the callback in this mode at the next possible	2279	C<ev_periodic> will try to run the callback in this mode at the next possible
2177	time where C<time = offset (mod interval)>, regardless of any time jumps.	2280	time where C<time = offset (mod interval)>, regardless of any time jumps.
2178		2281
2179	For numerical stability it is preferable that the C<offset> value is near	2282	The C<interval> I<MUST> be positive, and for numerical stability, the
2180	C<ev_now ()> (the current time), but there is no range requirement for	2283	interval value should be higher than C<1/8192> (which is around 100
2181	this value, and in fact is often specified as zero.	2284	microseconds) and C<offset> should be higher than C<0> and should have
		2285	at most a similar magnitude as the current time (say, within a factor of
		2286	ten). Typical values for offset are, in fact, C<0> or something between
		2287	C<0> and C<interval>, which is also the recommended range.
2182		2288
2183	Note also that there is an upper limit to how often a timer can fire (CPU	2289	Note also that there is an upper limit to how often a timer can fire (CPU
2184	speed for example), so if C<interval> is very small then timing stability	2290	speed for example), so if C<interval> is very small then timing stability
2185	will of course deteriorate. Libev itself tries to be exact to be about one	2291	will of course deteriorate. Libev itself tries to be exact to be about one
2186	millisecond (if the OS supports it and the machine is fast enough).	2292	millisecond (if the OS supports it and the machine is fast enough).
…		…
2216		2322
2217	NOTE: I<< This callback must always return a time that is higher than or	2323	NOTE: I<< This callback must always return a time that is higher than or
2218	equal to the passed C<now> value >>.	2324	equal to the passed C<now> value >>.
2219		2325
2220	This can be used to create very complex timers, such as a timer that	2326	This can be used to create very complex timers, such as a timer that
2221	triggers on "next midnight, local time". To do this, you would calculate the	2327	triggers on "next midnight, local time". To do this, you would calculate
2222	next midnight after C<now> and return the timestamp value for this. How	2328	the next midnight after C<now> and return the timestamp value for
2223	you do this is, again, up to you (but it is not trivial, which is the main	2329	this. Here is a (completely untested, no error checking) example on how to
2224	reason I omitted it as an example).	2330	do this:
		2331
		2332	#include <time.h>
		2333
		2334	static ev_tstamp
		2335	my_rescheduler (ev_periodic *w, ev_tstamp now)
		2336	{
		2337	time_t tnow = (time_t)now;
		2338	struct tm tm;
		2339	localtime_r (&tnow, &tm);
		2340
		2341	tm.tm_sec = tm.tm_min = tm.tm_hour = 0; // midnight current day
		2342	++tm.tm_mday; // midnight next day
		2343
		2344	return mktime (&tm);
		2345	}
		2346
		2347	Note: this code might run into trouble on days that have more then two
		2348	midnights (beginning and end).
2225		2349
2226	=back	2350	=back
2227		2351
2228	=item ev_periodic_again (loop, ev_periodic *)	2352	=item ev_periodic_again (loop, ev_periodic *)
2229		2353
…		…
2294		2418
2295	ev_periodic hourly_tick;	2419	ev_periodic hourly_tick;
2296	ev_periodic_init (&hourly_tick, clock_cb,	2420	ev_periodic_init (&hourly_tick, clock_cb,
2297	fmod (ev_now (loop), 3600.), 3600., 0);	2421	fmod (ev_now (loop), 3600.), 3600., 0);
2298	ev_periodic_start (loop, &hourly_tick);	2422	ev_periodic_start (loop, &hourly_tick);
2299		2423
2300		2424
2301	=head2 C<ev_signal> - signal me when a signal gets signalled!	2425	=head2 C<ev_signal> - signal me when a signal gets signalled!
2302		2426
2303	Signal watchers will trigger an event when the process receives a specific	2427	Signal watchers will trigger an event when the process receives a specific
2304	signal one or more times. Even though signals are very asynchronous, libev	2428	signal one or more times. Even though signals are very asynchronous, libev
…		…
2314	only within the same loop, i.e. you can watch for C<SIGINT> in your	2438	only within the same loop, i.e. you can watch for C<SIGINT> in your
2315	default loop and for C<SIGIO> in another loop, but you cannot watch for	2439	default loop and for C<SIGIO> in another loop, but you cannot watch for
2316	C<SIGINT> in both the default loop and another loop at the same time. At	2440	C<SIGINT> in both the default loop and another loop at the same time. At
2317	the moment, C<SIGCHLD> is permanently tied to the default loop.	2441	the moment, C<SIGCHLD> is permanently tied to the default loop.
2318		2442
2319	When the first watcher gets started will libev actually register something	2443	Only after the first watcher for a signal is started will libev actually
2320	with the kernel (thus it coexists with your own signal handlers as long as	2444	register something with the kernel. It thus coexists with your own signal
2321	you don't register any with libev for the same signal).	2445	handlers as long as you don't register any with libev for the same signal.
2322		2446
2323	If possible and supported, libev will install its handlers with	2447	If possible and supported, libev will install its handlers with
2324	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should	2448	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2325	not be unduly interrupted. If you have a problem with system calls getting	2449	not be unduly interrupted. If you have a problem with system calls getting
2326	interrupted by signals you can block all signals in an C<ev_check> watcher	2450	interrupted by signals you can block all signals in an C<ev_check> watcher
…		…
2329	=head3 The special problem of inheritance over fork/execve/pthread_create	2453	=head3 The special problem of inheritance over fork/execve/pthread_create
2330		2454
2331	Both the signal mask (C<sigprocmask>) and the signal disposition	2455	Both the signal mask (C<sigprocmask>) and the signal disposition
2332	(C<sigaction>) are unspecified after starting a signal watcher (and after	2456	(C<sigaction>) are unspecified after starting a signal watcher (and after
2333	stopping it again), that is, libev might or might not block the signal,	2457	stopping it again), that is, libev might or might not block the signal,
2334	and might or might not set or restore the installed signal handler.	2458	and might or might not set or restore the installed signal handler (but
		2459	see C<EVFLAG_NOSIGMASK>).
2335		2460
2336	While this does not matter for the signal disposition (libev never	2461	While this does not matter for the signal disposition (libev never
2337	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on	2462	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2338	C<execve>), this matters for the signal mask: many programs do not expect	2463	C<execve>), this matters for the signal mask: many programs do not expect
2339	certain signals to be blocked.	2464	certain signals to be blocked.
…		…
2510		2635
2511	=head2 C<ev_stat> - did the file attributes just change?	2636	=head2 C<ev_stat> - did the file attributes just change?
2512		2637
2513	This watches a file system path for attribute changes. That is, it calls	2638	This watches a file system path for attribute changes. That is, it calls
2514	C<stat> on that path in regular intervals (or when the OS says it changed)	2639	C<stat> on that path in regular intervals (or when the OS says it changed)
2515	and sees if it changed compared to the last time, invoking the callback if	2640	and sees if it changed compared to the last time, invoking the callback
2516	it did.	2641	if it did. Starting the watcher C<stat>'s the file, so only changes that
		2642	happen after the watcher has been started will be reported.
2517		2643
2518	The path does not need to exist: changing from "path exists" to "path does	2644	The path does not need to exist: changing from "path exists" to "path does
2519	not exist" is a status change like any other. The condition "path does not	2645	not exist" is a status change like any other. The condition "path does not
2520	exist" (or more correctly "path cannot be stat'ed") is signified by the	2646	exist" (or more correctly "path cannot be stat'ed") is signified by the
2521	C<st_nlink> field being zero (which is otherwise always forced to be at	2647	C<st_nlink> field being zero (which is otherwise always forced to be at
…		…
2751	Apart from keeping your process non-blocking (which is a useful	2877	Apart from keeping your process non-blocking (which is a useful
2752	effect on its own sometimes), idle watchers are a good place to do	2878	effect on its own sometimes), idle watchers are a good place to do
2753	"pseudo-background processing", or delay processing stuff to after the	2879	"pseudo-background processing", or delay processing stuff to after the
2754	event loop has handled all outstanding events.	2880	event loop has handled all outstanding events.
2755		2881
		2882	=head3 Abusing an C<ev_idle> watcher for its side-effect
		2883
		2884	As long as there is at least one active idle watcher, libev will never
		2885	sleep unnecessarily. Or in other words, it will loop as fast as possible.
		2886	For this to work, the idle watcher doesn't need to be invoked at all - the
		2887	lowest priority will do.
		2888
		2889	This mode of operation can be useful together with an C<ev_check> watcher,
		2890	to do something on each event loop iteration - for example to balance load
		2891	between different connections.
		2892
		2893	See L</Abusing an ev_check watcher for its side-effect> for a longer
		2894	example.
		2895
2756	=head3 Watcher-Specific Functions and Data Members	2896	=head3 Watcher-Specific Functions and Data Members
2757		2897
2758	=over 4	2898	=over 4
2759		2899
2760	=item ev_idle_init (ev_idle *, callback)	2900	=item ev_idle_init (ev_idle *, callback)
…		…
2771	callback, free it. Also, use no error checking, as usual.	2911	callback, free it. Also, use no error checking, as usual.
2772		2912
2773	static void	2913	static void
2774	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)	2914	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
2775	{	2915	{
		2916	// stop the watcher
		2917	ev_idle_stop (loop, w);
		2918
		2919	// now we can free it
2776	free (w);	2920	free (w);
		2921
2777	// now do something you wanted to do when the program has	2922	// now do something you wanted to do when the program has
2778	// no longer anything immediate to do.	2923	// no longer anything immediate to do.
2779	}	2924	}
2780		2925
2781	ev_idle *idle_watcher = malloc (sizeof (ev_idle));	2926	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
…		…
2783	ev_idle_start (loop, idle_watcher);	2928	ev_idle_start (loop, idle_watcher);
2784		2929
2785		2930
2786	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2931	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
2787		2932
2788	Prepare and check watchers are usually (but not always) used in pairs:	2933	Prepare and check watchers are often (but not always) used in pairs:
2789	prepare watchers get invoked before the process blocks and check watchers	2934	prepare watchers get invoked before the process blocks and check watchers
2790	afterwards.	2935	afterwards.
2791		2936
2792	You I<must not> call C<ev_run> or similar functions that enter	2937	You I<must not> call C<ev_run> (or similar functions that enter the
2793	the current event loop from either C<ev_prepare> or C<ev_check>	2938	current event loop) or C<ev_loop_fork> from either C<ev_prepare> or
2794	watchers. Other loops than the current one are fine, however. The	2939	C<ev_check> watchers. Other loops than the current one are fine,
2795	rationale behind this is that you do not need to check for recursion in	2940	however. The rationale behind this is that you do not need to check
2796	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2941	for recursion in those watchers, i.e. the sequence will always be
2797	C<ev_check> so if you have one watcher of each kind they will always be	2942	C<ev_prepare>, blocking, C<ev_check> so if you have one watcher of each
2798	called in pairs bracketing the blocking call.	2943	kind they will always be called in pairs bracketing the blocking call.
2799		2944
2800	Their main purpose is to integrate other event mechanisms into libev and	2945	Their main purpose is to integrate other event mechanisms into libev and
2801	their use is somewhat advanced. They could be used, for example, to track	2946	their use is somewhat advanced. They could be used, for example, to track
2802	variable changes, implement your own watchers, integrate net-snmp or a	2947	variable changes, implement your own watchers, integrate net-snmp or a
2803	coroutine library and lots more. They are also occasionally useful if	2948	coroutine library and lots more. They are also occasionally useful if
…		…
2821	with priority higher than or equal to the event loop and one coroutine	2966	with priority higher than or equal to the event loop and one coroutine
2822	of lower priority, but only once, using idle watchers to keep the event	2967	of lower priority, but only once, using idle watchers to keep the event
2823	loop from blocking if lower-priority coroutines are active, thus mapping	2968	loop from blocking if lower-priority coroutines are active, thus mapping
2824	low-priority coroutines to idle/background tasks).	2969	low-priority coroutines to idle/background tasks).
2825		2970
2826	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2971	When used for this purpose, it is recommended to give C<ev_check> watchers
2827	priority, to ensure that they are being run before any other watchers	2972	highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
2828	after the poll (this doesn't matter for C<ev_prepare> watchers).	2973	any other watchers after the poll (this doesn't matter for C<ev_prepare>
		2974	watchers).
2829		2975
2830	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not	2976	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
2831	activate ("feed") events into libev. While libev fully supports this, they	2977	activate ("feed") events into libev. While libev fully supports this, they
2832	might get executed before other C<ev_check> watchers did their job. As	2978	might get executed before other C<ev_check> watchers did their job. As
2833	C<ev_check> watchers are often used to embed other (non-libev) event	2979	C<ev_check> watchers are often used to embed other (non-libev) event
2834	loops those other event loops might be in an unusable state until their	2980	loops those other event loops might be in an unusable state until their
2835	C<ev_check> watcher ran (always remind yourself to coexist peacefully with	2981	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
2836	others).	2982	others).
		2983
		2984	=head3 Abusing an C<ev_check> watcher for its side-effect
		2985
		2986	C<ev_check> (and less often also C<ev_prepare>) watchers can also be
		2987	useful because they are called once per event loop iteration. For
		2988	example, if you want to handle a large number of connections fairly, you
		2989	normally only do a bit of work for each active connection, and if there
		2990	is more work to do, you wait for the next event loop iteration, so other
		2991	connections have a chance of making progress.
		2992
		2993	Using an C<ev_check> watcher is almost enough: it will be called on the
		2994	next event loop iteration. However, that isn't as soon as possible -
		2995	without external events, your C<ev_check> watcher will not be invoked.
		2996
		2997	This is where C<ev_idle> watchers come in handy - all you need is a
		2998	single global idle watcher that is active as long as you have one active
		2999	C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
		3000	will not sleep, and the C<ev_check> watcher makes sure a callback gets
		3001	invoked. Neither watcher alone can do that.
2837		3002
2838	=head3 Watcher-Specific Functions and Data Members	3003	=head3 Watcher-Specific Functions and Data Members
2839		3004
2840	=over 4	3005	=over 4
2841		3006
…		…
3042		3207
3043	=over 4	3208	=over 4
3044		3209
3045	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	3210	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
3046		3211
3047	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	3212	=item ev_embed_set (ev_embed *, struct ev_loop *embedded_loop)
3048		3213
3049	Configures the watcher to embed the given loop, which must be	3214	Configures the watcher to embed the given loop, which must be
3050	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3215	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
3051	invoked automatically, otherwise it is the responsibility of the callback	3216	invoked automatically, otherwise it is the responsibility of the callback
3052	to invoke it (it will continue to be called until the sweep has been done,	3217	to invoke it (it will continue to be called until the sweep has been done,
…		…
3073	used).	3238	used).
3074		3239
3075	struct ev_loop *loop_hi = ev_default_init (0);	3240	struct ev_loop *loop_hi = ev_default_init (0);
3076	struct ev_loop *loop_lo = 0;	3241	struct ev_loop *loop_lo = 0;
3077	ev_embed embed;	3242	ev_embed embed;
3078		3243
3079	// see if there is a chance of getting one that works	3244	// see if there is a chance of getting one that works
3080	// (remember that a flags value of 0 means autodetection)	3245	// (remember that a flags value of 0 means autodetection)
3081	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	3246	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
3082	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	3247	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
3083	: 0;	3248	: 0;
…		…
3097	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	3262	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
3098		3263
3099	struct ev_loop *loop = ev_default_init (0);	3264	struct ev_loop *loop = ev_default_init (0);
3100	struct ev_loop *loop_socket = 0;	3265	struct ev_loop *loop_socket = 0;
3101	ev_embed embed;	3266	ev_embed embed;
3102		3267
3103	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	3268	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
3104	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	3269	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
3105	{	3270	{
3106	ev_embed_init (&embed, 0, loop_socket);	3271	ev_embed_init (&embed, 0, loop_socket);
3107	ev_embed_start (loop, &embed);	3272	ev_embed_start (loop, &embed);
…		…
3115		3280
3116	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	3281	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
3117		3282
3118	Fork watchers are called when a C<fork ()> was detected (usually because	3283	Fork watchers are called when a C<fork ()> was detected (usually because
3119	whoever is a good citizen cared to tell libev about it by calling	3284	whoever is a good citizen cared to tell libev about it by calling
3120	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the	3285	C<ev_loop_fork>). The invocation is done before the event loop blocks next
3121	event loop blocks next and before C<ev_check> watchers are being called,	3286	and before C<ev_check> watchers are being called, and only in the child
3122	and only in the child after the fork. If whoever good citizen calling	3287	after the fork. If whoever good citizen calling C<ev_default_fork> cheats
3123	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3288	and calls it in the wrong process, the fork handlers will be invoked, too,
3124	handlers will be invoked, too, of course.	3289	of course.
3125		3290
3126	=head3 The special problem of life after fork - how is it possible?	3291	=head3 The special problem of life after fork - how is it possible?
3127		3292
3128	Most uses of C<fork()> consist of forking, then some simple calls to set	3293	Most uses of C<fork ()> consist of forking, then some simple calls to set
3129	up/change the process environment, followed by a call to C<exec()>. This	3294	up/change the process environment, followed by a call to C<exec()>. This
3130	sequence should be handled by libev without any problems.	3295	sequence should be handled by libev without any problems.
3131		3296
3132	This changes when the application actually wants to do event handling	3297	This changes when the application actually wants to do event handling
3133	in the child, or both parent in child, in effect "continuing" after the	3298	in the child, or both parent in child, in effect "continuing" after the
…		…
3210	atexit (program_exits);	3375	atexit (program_exits);
3211		3376
3212		3377
3213	=head2 C<ev_async> - how to wake up an event loop	3378	=head2 C<ev_async> - how to wake up an event loop
3214		3379
3215	In general, you cannot use an C<ev_run> from multiple threads or other	3380	In general, you cannot use an C<ev_loop> from multiple threads or other
3216	asynchronous sources such as signal handlers (as opposed to multiple event	3381	asynchronous sources such as signal handlers (as opposed to multiple event
3217	loops - those are of course safe to use in different threads).	3382	loops - those are of course safe to use in different threads).
3218		3383
3219	Sometimes, however, you need to wake up an event loop you do not control,	3384	Sometimes, however, you need to wake up an event loop you do not control,
3220	for example because it belongs to another thread. This is what C<ev_async>	3385	for example because it belongs to another thread. This is what C<ev_async>
…		…
3222	it by calling C<ev_async_send>, which is thread- and signal safe.	3387	it by calling C<ev_async_send>, which is thread- and signal safe.
3223		3388
3224	This functionality is very similar to C<ev_signal> watchers, as signals,	3389	This functionality is very similar to C<ev_signal> watchers, as signals,
3225	too, are asynchronous in nature, and signals, too, will be compressed	3390	too, are asynchronous in nature, and signals, too, will be compressed
3226	(i.e. the number of callback invocations may be less than the number of	3391	(i.e. the number of callback invocations may be less than the number of
3227	C<ev_async_sent> calls). In fact, you could use signal watchers as a kind	3392	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
3228	of "global async watchers" by using a watcher on an otherwise unused	3393	of "global async watchers" by using a watcher on an otherwise unused
3229	signal, and C<ev_feed_signal> to signal this watcher from another thread,	3394	signal, and C<ev_feed_signal> to signal this watcher from another thread,
3230	even without knowing which loop owns the signal.	3395	even without knowing which loop owns the signal.
3231
3232	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
3233	just the default loop.
3234		3396
3235	=head3 Queueing	3397	=head3 Queueing
3236		3398
3237	C<ev_async> does not support queueing of data in any way. The reason	3399	C<ev_async> does not support queueing of data in any way. The reason
3238	is that the author does not know of a simple (or any) algorithm for a	3400	is that the author does not know of a simple (or any) algorithm for a
…		…
3330	trust me.	3492	trust me.
3331		3493
3332	=item ev_async_send (loop, ev_async *)	3494	=item ev_async_send (loop, ev_async *)
3333		3495
3334	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3496	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3335	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3497	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3498	returns.
		3499
3336	C<ev_feed_event>, this call is safe to do from other threads, signal or	3500	Unlike C<ev_feed_event>, this call is safe to do from other threads,
3337	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3501	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3338	section below on what exactly this means).	3502	embedding section below on what exactly this means).
3339		3503
3340	Note that, as with other watchers in libev, multiple events might get	3504	Note that, as with other watchers in libev, multiple events might get
3341	compressed into a single callback invocation (another way to look at this	3505	compressed into a single callback invocation (another way to look at
3342	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,	3506	this is that C<ev_async> watchers are level-triggered: they are set on
3343	reset when the event loop detects that).	3507	C<ev_async_send>, reset when the event loop detects that).
3344		3508
3345	This call incurs the overhead of a system call only once per event loop	3509	This call incurs the overhead of at most one extra system call per event
3346	iteration, so while the overhead might be noticeable, it doesn't apply to	3510	loop iteration, if the event loop is blocked, and no syscall at all if
3347	repeated calls to C<ev_async_send> for the same event loop.	3511	the event loop (or your program) is processing events. That means that
		3512	repeated calls are basically free (there is no need to avoid calls for
		3513	performance reasons) and that the overhead becomes smaller (typically
		3514	zero) under load.
3348		3515
3349	=item bool = ev_async_pending (ev_async *)	3516	=item bool = ev_async_pending (ev_async *)
3350		3517
3351	Returns a non-zero value when C<ev_async_send> has been called on the	3518	Returns a non-zero value when C<ev_async_send> has been called on the
3352	watcher but the event has not yet been processed (or even noted) by the	3519	watcher but the event has not yet been processed (or even noted) by the
…		…
3369		3536
3370	There are some other functions of possible interest. Described. Here. Now.	3537	There are some other functions of possible interest. Described. Here. Now.
3371		3538
3372	=over 4	3539	=over 4
3373		3540
3374	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3541	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback, arg)
3375		3542
3376	This function combines a simple timer and an I/O watcher, calls your	3543	This function combines a simple timer and an I/O watcher, calls your
3377	callback on whichever event happens first and automatically stops both	3544	callback on whichever event happens first and automatically stops both
3378	watchers. This is useful if you want to wait for a single event on an fd	3545	watchers. This is useful if you want to wait for a single event on an fd
3379	or timeout without having to allocate/configure/start/stop/free one or	3546	or timeout without having to allocate/configure/start/stop/free one or
…		…
3407	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3574	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3408		3575
3409	=item ev_feed_fd_event (loop, int fd, int revents)	3576	=item ev_feed_fd_event (loop, int fd, int revents)
3410		3577
3411	Feed an event on the given fd, as if a file descriptor backend detected	3578	Feed an event on the given fd, as if a file descriptor backend detected
3412	the given events it.	3579	the given events.
3413		3580
3414	=item ev_feed_signal_event (loop, int signum)	3581	=item ev_feed_signal_event (loop, int signum)
3415		3582
3416	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,	3583	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3417	which is async-safe.	3584	which is async-safe.
…		…
3423		3590
3424	This section explains some common idioms that are not immediately	3591	This section explains some common idioms that are not immediately
3425	obvious. Note that examples are sprinkled over the whole manual, and this	3592	obvious. Note that examples are sprinkled over the whole manual, and this
3426	section only contains stuff that wouldn't fit anywhere else.	3593	section only contains stuff that wouldn't fit anywhere else.
3427		3594
3428	=over 4	3595	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
3429		3596
3430	=item Model/nested event loop invocations and exit conditions.	3597	Each watcher has, by default, a C<void *data> member that you can read
		3598	or modify at any time: libev will completely ignore it. This can be used
		3599	to associate arbitrary data with your watcher. If you need more data and
		3600	don't want to allocate memory separately and store a pointer to it in that
		3601	data member, you can also "subclass" the watcher type and provide your own
		3602	data:
		3603
		3604	struct my_io
		3605	{
		3606	ev_io io;
		3607	int otherfd;
		3608	void *somedata;
		3609	struct whatever *mostinteresting;
		3610	};
		3611
		3612	...
		3613	struct my_io w;
		3614	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3615
		3616	And since your callback will be called with a pointer to the watcher, you
		3617	can cast it back to your own type:
		3618
		3619	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3620	{
		3621	struct my_io *w = (struct my_io *)w_;
		3622	...
		3623	}
		3624
		3625	More interesting and less C-conformant ways of casting your callback
		3626	function type instead have been omitted.
		3627
		3628	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3629
		3630	Another common scenario is to use some data structure with multiple
		3631	embedded watchers, in effect creating your own watcher that combines
		3632	multiple libev event sources into one "super-watcher":
		3633
		3634	struct my_biggy
		3635	{
		3636	int some_data;
		3637	ev_timer t1;
		3638	ev_timer t2;
		3639	}
		3640
		3641	In this case getting the pointer to C<my_biggy> is a bit more
		3642	complicated: Either you store the address of your C<my_biggy> struct in
		3643	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3644	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3645	real programmers):
		3646
		3647	#include <stddef.h>
		3648
		3649	static void
		3650	t1_cb (EV_P_ ev_timer *w, int revents)
		3651	{
		3652	struct my_biggy big = (struct my_biggy *)
		3653	(((char *)w) - offsetof (struct my_biggy, t1));
		3654	}
		3655
		3656	static void
		3657	t2_cb (EV_P_ ev_timer *w, int revents)
		3658	{
		3659	struct my_biggy big = (struct my_biggy *)
		3660	(((char *)w) - offsetof (struct my_biggy, t2));
		3661	}
		3662
		3663	=head2 AVOIDING FINISHING BEFORE RETURNING
		3664
		3665	Often you have structures like this in event-based programs:
		3666
		3667	callback ()
		3668	{
		3669	free (request);
		3670	}
		3671
		3672	request = start_new_request (..., callback);
		3673
		3674	The intent is to start some "lengthy" operation. The C<request> could be
		3675	used to cancel the operation, or do other things with it.
		3676
		3677	It's not uncommon to have code paths in C<start_new_request> that
		3678	immediately invoke the callback, for example, to report errors. Or you add
		3679	some caching layer that finds that it can skip the lengthy aspects of the
		3680	operation and simply invoke the callback with the result.
		3681
		3682	The problem here is that this will happen I<before> C<start_new_request>
		3683	has returned, so C<request> is not set.
		3684
		3685	Even if you pass the request by some safer means to the callback, you
		3686	might want to do something to the request after starting it, such as
		3687	canceling it, which probably isn't working so well when the callback has
		3688	already been invoked.
		3689
		3690	A common way around all these issues is to make sure that
		3691	C<start_new_request> I<always> returns before the callback is invoked. If
		3692	C<start_new_request> immediately knows the result, it can artificially
		3693	delay invoking the callback by using a C<prepare> or C<idle> watcher for
		3694	example, or more sneakily, by reusing an existing (stopped) watcher and
		3695	pushing it into the pending queue:
		3696
		3697	ev_set_cb (watcher, callback);
		3698	ev_feed_event (EV_A_ watcher, 0);
		3699
		3700	This way, C<start_new_request> can safely return before the callback is
		3701	invoked, while not delaying callback invocation too much.
		3702
		3703	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
3431		3704
3432	Often (especially in GUI toolkits) there are places where you have	3705	Often (especially in GUI toolkits) there are places where you have
3433	I<modal> interaction, which is most easily implemented by recursively	3706	I<modal> interaction, which is most easily implemented by recursively
3434	invoking C<ev_run>.	3707	invoking C<ev_run>.
3435		3708
3436	This brings the problem of exiting - a callback might want to finish the	3709	This brings the problem of exiting - a callback might want to finish the
3437	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but	3710	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
3438	a modal "Are you sure?" dialog is still waiting), or just the nested one	3711	a modal "Are you sure?" dialog is still waiting), or just the nested one
3439	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some	3712	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
3440	other combination: In these cases, C<ev_break> will not work alone.	3713	other combination: In these cases, a simple C<ev_break> will not work.
3441		3714
3442	The solution is to maintain "break this loop" variable for each C<ev_run>	3715	The solution is to maintain "break this loop" variable for each C<ev_run>
3443	invocation, and use a loop around C<ev_run> until the condition is	3716	invocation, and use a loop around C<ev_run> until the condition is
3444	triggered, using C<EVRUN_ONCE>:	3717	triggered, using C<EVRUN_ONCE>:
3445		3718
…		…
3447	int exit_main_loop = 0;	3720	int exit_main_loop = 0;
3448		3721
3449	while (!exit_main_loop)	3722	while (!exit_main_loop)
3450	ev_run (EV_DEFAULT_ EVRUN_ONCE);	3723	ev_run (EV_DEFAULT_ EVRUN_ONCE);
3451		3724
3452	// in a model watcher	3725	// in a modal watcher
3453	int exit_nested_loop = 0;	3726	int exit_nested_loop = 0;
3454		3727
3455	while (!exit_nested_loop)	3728	while (!exit_nested_loop)
3456	ev_run (EV_A_ EVRUN_ONCE);	3729	ev_run (EV_A_ EVRUN_ONCE);
3457		3730
…		…
3464	exit_main_loop = 1;	3737	exit_main_loop = 1;
3465		3738
3466	// exit both	3739	// exit both
3467	exit_main_loop = exit_nested_loop = 1;	3740	exit_main_loop = exit_nested_loop = 1;
3468		3741
3469	=back	3742	=head2 THREAD LOCKING EXAMPLE
		3743
		3744	Here is a fictitious example of how to run an event loop in a different
		3745	thread from where callbacks are being invoked and watchers are
		3746	created/added/removed.
		3747
		3748	For a real-world example, see the C<EV::Loop::Async> perl module,
		3749	which uses exactly this technique (which is suited for many high-level
		3750	languages).
		3751
		3752	The example uses a pthread mutex to protect the loop data, a condition
		3753	variable to wait for callback invocations, an async watcher to notify the
		3754	event loop thread and an unspecified mechanism to wake up the main thread.
		3755
		3756	First, you need to associate some data with the event loop:
		3757
		3758	typedef struct {
		3759	mutex_t lock; /* global loop lock */
		3760	ev_async async_w;
		3761	thread_t tid;
		3762	cond_t invoke_cv;
		3763	} userdata;
		3764
		3765	void prepare_loop (EV_P)
		3766	{
		3767	// for simplicity, we use a static userdata struct.
		3768	static userdata u;
		3769
		3770	ev_async_init (&u->async_w, async_cb);
		3771	ev_async_start (EV_A_ &u->async_w);
		3772
		3773	pthread_mutex_init (&u->lock, 0);
		3774	pthread_cond_init (&u->invoke_cv, 0);
		3775
		3776	// now associate this with the loop
		3777	ev_set_userdata (EV_A_ u);
		3778	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3779	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3780
		3781	// then create the thread running ev_run
		3782	pthread_create (&u->tid, 0, l_run, EV_A);
		3783	}
		3784
		3785	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3786	solely to wake up the event loop so it takes notice of any new watchers
		3787	that might have been added:
		3788
		3789	static void
		3790	async_cb (EV_P_ ev_async *w, int revents)
		3791	{
		3792	// just used for the side effects
		3793	}
		3794
		3795	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3796	protecting the loop data, respectively.
		3797
		3798	static void
		3799	l_release (EV_P)
		3800	{
		3801	userdata *u = ev_userdata (EV_A);
		3802	pthread_mutex_unlock (&u->lock);
		3803	}
		3804
		3805	static void
		3806	l_acquire (EV_P)
		3807	{
		3808	userdata *u = ev_userdata (EV_A);
		3809	pthread_mutex_lock (&u->lock);
		3810	}
		3811
		3812	The event loop thread first acquires the mutex, and then jumps straight
		3813	into C<ev_run>:
		3814
		3815	void *
		3816	l_run (void *thr_arg)
		3817	{
		3818	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3819
		3820	l_acquire (EV_A);
		3821	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3822	ev_run (EV_A_ 0);
		3823	l_release (EV_A);
		3824
		3825	return 0;
		3826	}
		3827
		3828	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3829	signal the main thread via some unspecified mechanism (signals? pipe
		3830	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3831	have been called (in a while loop because a) spurious wakeups are possible
		3832	and b) skipping inter-thread-communication when there are no pending
		3833	watchers is very beneficial):
		3834
		3835	static void
		3836	l_invoke (EV_P)
		3837	{
		3838	userdata *u = ev_userdata (EV_A);
		3839
		3840	while (ev_pending_count (EV_A))
		3841	{
		3842	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3843	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3844	}
		3845	}
		3846
		3847	Now, whenever the main thread gets told to invoke pending watchers, it
		3848	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3849	thread to continue:
		3850
		3851	static void
		3852	real_invoke_pending (EV_P)
		3853	{
		3854	userdata *u = ev_userdata (EV_A);
		3855
		3856	pthread_mutex_lock (&u->lock);
		3857	ev_invoke_pending (EV_A);
		3858	pthread_cond_signal (&u->invoke_cv);
		3859	pthread_mutex_unlock (&u->lock);
		3860	}
		3861
		3862	Whenever you want to start/stop a watcher or do other modifications to an
		3863	event loop, you will now have to lock:
		3864
		3865	ev_timer timeout_watcher;
		3866	userdata *u = ev_userdata (EV_A);
		3867
		3868	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3869
		3870	pthread_mutex_lock (&u->lock);
		3871	ev_timer_start (EV_A_ &timeout_watcher);
		3872	ev_async_send (EV_A_ &u->async_w);
		3873	pthread_mutex_unlock (&u->lock);
		3874
		3875	Note that sending the C<ev_async> watcher is required because otherwise
		3876	an event loop currently blocking in the kernel will have no knowledge
		3877	about the newly added timer. By waking up the loop it will pick up any new
		3878	watchers in the next event loop iteration.
		3879
		3880	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3881
		3882	While the overhead of a callback that e.g. schedules a thread is small, it
		3883	is still an overhead. If you embed libev, and your main usage is with some
		3884	kind of threads or coroutines, you might want to customise libev so that
		3885	doesn't need callbacks anymore.
		3886
		3887	Imagine you have coroutines that you can switch to using a function
		3888	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3889	and that due to some magic, the currently active coroutine is stored in a
		3890	global called C<current_coro>. Then you can build your own "wait for libev
		3891	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3892	the differing C<;> conventions):
		3893
		3894	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3895	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3896
		3897	That means instead of having a C callback function, you store the
		3898	coroutine to switch to in each watcher, and instead of having libev call
		3899	your callback, you instead have it switch to that coroutine.
		3900
		3901	A coroutine might now wait for an event with a function called
		3902	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3903	matter when, or whether the watcher is active or not when this function is
		3904	called):
		3905
		3906	void
		3907	wait_for_event (ev_watcher *w)
		3908	{
		3909	ev_set_cb (w, current_coro);
		3910	switch_to (libev_coro);
		3911	}
		3912
		3913	That basically suspends the coroutine inside C<wait_for_event> and
		3914	continues the libev coroutine, which, when appropriate, switches back to
		3915	this or any other coroutine.
		3916
		3917	You can do similar tricks if you have, say, threads with an event queue -
		3918	instead of storing a coroutine, you store the queue object and instead of
		3919	switching to a coroutine, you push the watcher onto the queue and notify
		3920	any waiters.
		3921
		3922	To embed libev, see L</EMBEDDING>, but in short, it's easiest to create two
		3923	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3924
		3925	// my_ev.h
		3926	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3927	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3928	#include "../libev/ev.h"
		3929
		3930	// my_ev.c
		3931	#define EV_H "my_ev.h"
		3932	#include "../libev/ev.c"
		3933
		3934	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3935	F<my_ev.c> into your project. When properly specifying include paths, you
		3936	can even use F<ev.h> as header file name directly.
3470		3937
3471		3938
3472	=head1 LIBEVENT EMULATION	3939	=head1 LIBEVENT EMULATION
3473		3940
3474	Libev offers a compatibility emulation layer for libevent. It cannot	3941	Libev offers a compatibility emulation layer for libevent. It cannot
…		…
3504		3971
3505	=back	3972	=back
3506		3973
3507	=head1 C++ SUPPORT	3974	=head1 C++ SUPPORT
3508		3975
		3976	=head2 C API
		3977
		3978	The normal C API should work fine when used from C++: both ev.h and the
		3979	libev sources can be compiled as C++. Therefore, code that uses the C API
		3980	will work fine.
		3981
		3982	Proper exception specifications might have to be added to callbacks passed
		3983	to libev: exceptions may be thrown only from watcher callbacks, all
		3984	other callbacks (allocator, syserr, loop acquire/release and periodic
		3985	reschedule callbacks) must not throw exceptions, and might need a C<throw
		3986	()> specification. If you have code that needs to be compiled as both C
		3987	and C++ you can use the C<EV_THROW> macro for this:
		3988
		3989	static void
		3990	fatal_error (const char *msg) EV_THROW
		3991	{
		3992	perror (msg);
		3993	abort ();
		3994	}
		3995
		3996	...
		3997	ev_set_syserr_cb (fatal_error);
		3998
		3999	The only API functions that can currently throw exceptions are C<ev_run>,
		4000	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		4001	because it runs cleanup watchers).
		4002
		4003	Throwing exceptions in watcher callbacks is only supported if libev itself
		4004	is compiled with a C++ compiler or your C and C++ environments allow
		4005	throwing exceptions through C libraries (most do).
		4006
		4007	=head2 C++ API
		4008
3509	Libev comes with some simplistic wrapper classes for C++ that mainly allow	4009	Libev comes with some simplistic wrapper classes for C++ that mainly allow
3510	you to use some convenience methods to start/stop watchers and also change	4010	you to use some convenience methods to start/stop watchers and also change
3511	the callback model to a model using method callbacks on objects.	4011	the callback model to a model using method callbacks on objects.
3512		4012
3513	To use it,	4013	To use it,
3514		4014
3515	#include <ev++.h>	4015	#include <ev++.h>
3516		4016
3517	This automatically includes F<ev.h> and puts all of its definitions (many	4017	This automatically includes F<ev.h> and puts all of its definitions (many
3518	of them macros) into the global namespace. All C++ specific things are	4018	of them macros) into the global namespace. All C++ specific things are
3519	put into the C<ev> namespace. It should support all the same embedding	4019	put into the C<ev> namespace. It should support all the same embedding
…		…
3528	with C<operator ()> can be used as callbacks. Other types should be easy	4028	with C<operator ()> can be used as callbacks. Other types should be easy
3529	to add as long as they only need one additional pointer for context. If	4029	to add as long as they only need one additional pointer for context. If
3530	you need support for other types of functors please contact the author	4030	you need support for other types of functors please contact the author
3531	(preferably after implementing it).	4031	(preferably after implementing it).
3532		4032
		4033	For all this to work, your C++ compiler either has to use the same calling
		4034	conventions as your C compiler (for static member functions), or you have
		4035	to embed libev and compile libev itself as C++.
		4036
3533	Here is a list of things available in the C<ev> namespace:	4037	Here is a list of things available in the C<ev> namespace:
3534		4038
3535	=over 4	4039	=over 4
3536		4040
3537	=item C<ev::READ>, C<ev::WRITE> etc.	4041	=item C<ev::READ>, C<ev::WRITE> etc.
…		…
3546	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.	4050	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
3547		4051
3548	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of	4052	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
3549	the same name in the C<ev> namespace, with the exception of C<ev_signal>	4053	the same name in the C<ev> namespace, with the exception of C<ev_signal>
3550	which is called C<ev::sig> to avoid clashes with the C<signal> macro	4054	which is called C<ev::sig> to avoid clashes with the C<signal> macro
3551	defines by many implementations.	4055	defined by many implementations.
3552		4056
3553	All of those classes have these methods:	4057	All of those classes have these methods:
3554		4058
3555	=over 4	4059	=over 4
3556		4060
…		…
3618	void operator() (ev::io &w, int revents)	4122	void operator() (ev::io &w, int revents)
3619	{	4123	{
3620	...	4124	...
3621	}	4125	}
3622	}	4126	}
3623		4127
3624	myfunctor f;	4128	myfunctor f;
3625		4129
3626	ev::io w;	4130	ev::io w;
3627	w.set (&f);	4131	w.set (&f);
3628		4132
…		…
3646	Associates a different C<struct ev_loop> with this watcher. You can only	4150	Associates a different C<struct ev_loop> with this watcher. You can only
3647	do this when the watcher is inactive (and not pending either).	4151	do this when the watcher is inactive (and not pending either).
3648		4152
3649	=item w->set ([arguments])	4153	=item w->set ([arguments])
3650		4154
3651	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this	4155	Basically the same as C<ev_TYPE_set> (except for C<ev::embed> watchers>),
3652	method or a suitable start method must be called at least once. Unlike the	4156	with the same arguments. Either this method or a suitable start method
3653	C counterpart, an active watcher gets automatically stopped and restarted	4157	must be called at least once. Unlike the C counterpart, an active watcher
3654	when reconfiguring it with this method.	4158	gets automatically stopped and restarted when reconfiguring it with this
		4159	method.
		4160
		4161	For C<ev::embed> watchers this method is called C<set_embed>, to avoid
		4162	clashing with the C<set (loop)> method.
3655		4163
3656	=item w->start ()	4164	=item w->start ()
3657		4165
3658	Starts the watcher. Note that there is no C<loop> argument, as the	4166	Starts the watcher. Note that there is no C<loop> argument, as the
3659	constructor already stores the event loop.	4167	constructor already stores the event loop.
…		…
3689	watchers in the constructor.	4197	watchers in the constructor.
3690		4198
3691	class myclass	4199	class myclass
3692	{	4200	{
3693	ev::io io ; void io_cb (ev::io &w, int revents);	4201	ev::io io ; void io_cb (ev::io &w, int revents);
3694	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);	4202	ev::io io2 ; void io2_cb (ev::io &w, int revents);
3695	ev::idle idle; void idle_cb (ev::idle &w, int revents);	4203	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3696		4204
3697	myclass (int fd)	4205	myclass (int fd)
3698	{	4206	{
3699	io .set <myclass, &myclass::io_cb > (this);	4207	io .set <myclass, &myclass::io_cb > (this);
…		…
3750	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.	4258	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
3751		4259
3752	=item D	4260	=item D
3753		4261
3754	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	4262	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
3755	be found at L<http://proj.llucax.com.ar/wiki/evd>.	4263	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
3756		4264
3757	=item Ocaml	4265	=item Ocaml
3758		4266
3759	Erkki Seppala has written Ocaml bindings for libev, to be found at	4267	Erkki Seppala has written Ocaml bindings for libev, to be found at
3760	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	4268	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
…		…
3763		4271
3764	Brian Maher has written a partial interface to libev for lua (at the	4272	Brian Maher has written a partial interface to libev for lua (at the
3765	time of this writing, only C<ev_io> and C<ev_timer>), to be found at	4273	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
3766	L<http://github.com/brimworks/lua-ev>.	4274	L<http://github.com/brimworks/lua-ev>.
3767		4275
		4276	=item Javascript
		4277
		4278	Node.js (L<http://nodejs.org>) uses libev as the underlying event library.
		4279
		4280	=item Others
		4281
		4282	There are others, and I stopped counting.
		4283
3768	=back	4284	=back
3769		4285
3770		4286
3771	=head1 MACRO MAGIC	4287	=head1 MACRO MAGIC
3772		4288
…		…
3808	suitable for use with C<EV_A>.	4324	suitable for use with C<EV_A>.
3809		4325
3810	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4326	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3811		4327
3812	Similar to the other two macros, this gives you the value of the default	4328	Similar to the other two macros, this gives you the value of the default
3813	loop, if multiple loops are supported ("ev loop default").	4329	loop, if multiple loops are supported ("ev loop default"). The default loop
		4330	will be initialised if it isn't already initialised.
		4331
		4332	For non-multiplicity builds, these macros do nothing, so you always have
		4333	to initialise the loop somewhere.
3814		4334
3815	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>	4335	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3816		4336
3817	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the	4337	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3818	default loop has been initialised (C<UC> == unchecked). Their behaviour	4338	default loop has been initialised (C<UC> == unchecked). Their behaviour
…		…
3885	ev_vars.h	4405	ev_vars.h
3886	ev_wrap.h	4406	ev_wrap.h
3887		4407
3888	ev_win32.c required on win32 platforms only	4408	ev_win32.c required on win32 platforms only
3889		4409
3890	ev_select.c only when select backend is enabled (which is enabled by default)	4410	ev_select.c only when select backend is enabled
3891	ev_poll.c only when poll backend is enabled (disabled by default)	4411	ev_poll.c only when poll backend is enabled
3892	ev_epoll.c only when the epoll backend is enabled (disabled by default)	4412	ev_epoll.c only when the epoll backend is enabled
3893	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	4413	ev_kqueue.c only when the kqueue backend is enabled
3894	ev_port.c only when the solaris port backend is enabled (disabled by default)	4414	ev_port.c only when the solaris port backend is enabled
3895		4415
3896	F<ev.c> includes the backend files directly when enabled, so you only need	4416	F<ev.c> includes the backend files directly when enabled, so you only need
3897	to compile this single file.	4417	to compile this single file.
3898		4418
3899	=head3 LIBEVENT COMPATIBILITY API	4419	=head3 LIBEVENT COMPATIBILITY API
…		…
3963	supported). It will also not define any of the structs usually found in	4483	supported). It will also not define any of the structs usually found in
3964	F<event.h> that are not directly supported by the libev core alone.	4484	F<event.h> that are not directly supported by the libev core alone.
3965		4485
3966	In standalone mode, libev will still try to automatically deduce the	4486	In standalone mode, libev will still try to automatically deduce the
3967	configuration, but has to be more conservative.	4487	configuration, but has to be more conservative.
		4488
		4489	=item EV_USE_FLOOR
		4490
		4491	If defined to be C<1>, libev will use the C<floor ()> function for its
		4492	periodic reschedule calculations, otherwise libev will fall back on a
		4493	portable (slower) implementation. If you enable this, you usually have to
		4494	link against libm or something equivalent. Enabling this when the C<floor>
		4495	function is not available will fail, so the safe default is to not enable
		4496	this.
3968		4497
3969	=item EV_USE_MONOTONIC	4498	=item EV_USE_MONOTONIC
3970		4499
3971	If defined to be C<1>, libev will try to detect the availability of the	4500	If defined to be C<1>, libev will try to detect the availability of the
3972	monotonic clock option at both compile time and runtime. Otherwise no	4501	monotonic clock option at both compile time and runtime. Otherwise no
…		…
4057		4586
4058	If programs implement their own fd to handle mapping on win32, then this	4587	If programs implement their own fd to handle mapping on win32, then this
4059	macro can be used to override the C<close> function, useful to unregister	4588	macro can be used to override the C<close> function, useful to unregister
4060	file descriptors again. Note that the replacement function has to close	4589	file descriptors again. Note that the replacement function has to close
4061	the underlying OS handle.	4590	the underlying OS handle.
		4591
		4592	=item EV_USE_WSASOCKET
		4593
		4594	If defined to be C<1>, libev will use C<WSASocket> to create its internal
		4595	communication socket, which works better in some environments. Otherwise,
		4596	the normal C<socket> function will be used, which works better in other
		4597	environments.
4062		4598
4063	=item EV_USE_POLL	4599	=item EV_USE_POLL
4064		4600
4065	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4601	If defined to be C<1>, libev will compile in support for the C<poll>(2)
4066	backend. Otherwise it will be enabled on non-win32 platforms. It	4602	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
4102	If defined to be C<1>, libev will compile in support for the Linux inotify	4638	If defined to be C<1>, libev will compile in support for the Linux inotify
4103	interface to speed up C<ev_stat> watchers. Its actual availability will	4639	interface to speed up C<ev_stat> watchers. Its actual availability will
4104	be detected at runtime. If undefined, it will be enabled if the headers	4640	be detected at runtime. If undefined, it will be enabled if the headers
4105	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4641	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
4106		4642
		4643	=item EV_NO_SMP
		4644
		4645	If defined to be C<1>, libev will assume that memory is always coherent
		4646	between threads, that is, threads can be used, but threads never run on
		4647	different cpus (or different cpu cores). This reduces dependencies
		4648	and makes libev faster.
		4649
		4650	=item EV_NO_THREADS
		4651
		4652	If defined to be C<1>, libev will assume that it will never be called from
		4653	different threads (that includes signal handlers), which is a stronger
		4654	assumption than C<EV_NO_SMP>, above. This reduces dependencies and makes
		4655	libev faster.
		4656
4107	=item EV_ATOMIC_T	4657	=item EV_ATOMIC_T
4108		4658
4109	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	4659	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
4110	access is atomic with respect to other threads or signal contexts. No such	4660	access is atomic with respect to other threads or signal contexts. No
4111	type is easily found in the C language, so you can provide your own type	4661	such type is easily found in the C language, so you can provide your own
4112	that you know is safe for your purposes. It is used both for signal handler "locking"	4662	type that you know is safe for your purposes. It is used both for signal
4113	as well as for signal and thread safety in C<ev_async> watchers.	4663	handler "locking" as well as for signal and thread safety in C<ev_async>
		4664	watchers.
4114		4665
4115	In the absence of this define, libev will use C<sig_atomic_t volatile>	4666	In the absence of this define, libev will use C<sig_atomic_t volatile>
4116	(from F<signal.h>), which is usually good enough on most platforms.	4667	(from F<signal.h>), which is usually good enough on most platforms.
4117		4668
4118	=item EV_H (h)	4669	=item EV_H (h)
…		…
4145	will have the C<struct ev_loop *> as first argument, and you can create	4696	will have the C<struct ev_loop *> as first argument, and you can create
4146	additional independent event loops. Otherwise there will be no support	4697	additional independent event loops. Otherwise there will be no support
4147	for multiple event loops and there is no first event loop pointer	4698	for multiple event loops and there is no first event loop pointer
4148	argument. Instead, all functions act on the single default loop.	4699	argument. Instead, all functions act on the single default loop.
4149		4700
		4701	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4702	default loop when multiplicity is switched off - you always have to
		4703	initialise the loop manually in this case.
		4704
4150	=item EV_MINPRI	4705	=item EV_MINPRI
4151		4706
4152	=item EV_MAXPRI	4707	=item EV_MAXPRI
4153		4708
4154	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4709	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
4190	#define EV_USE_POLL 1	4745	#define EV_USE_POLL 1
4191	#define EV_CHILD_ENABLE 1	4746	#define EV_CHILD_ENABLE 1
4192	#define EV_ASYNC_ENABLE 1	4747	#define EV_ASYNC_ENABLE 1
4193		4748
4194	The actual value is a bitset, it can be a combination of the following	4749	The actual value is a bitset, it can be a combination of the following
4195	values:	4750	values (by default, all of these are enabled):
4196		4751
4197	=over 4	4752	=over 4
4198		4753
4199	=item C<1> - faster/larger code	4754	=item C<1> - faster/larger code
4200		4755
…		…
4204	code size by roughly 30% on amd64).	4759	code size by roughly 30% on amd64).
4205		4760
4206	When optimising for size, use of compiler flags such as C<-Os> with	4761	When optimising for size, use of compiler flags such as C<-Os> with
4207	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of	4762	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
4208	assertions.	4763	assertions.
		4764
		4765	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4766	(e.g. gcc with C<-Os>).
4209		4767
4210	=item C<2> - faster/larger data structures	4768	=item C<2> - faster/larger data structures
4211		4769
4212	Replaces the small 2-heap for timer management by a faster 4-heap, larger	4770	Replaces the small 2-heap for timer management by a faster 4-heap, larger
4213	hash table sizes and so on. This will usually further increase code size	4771	hash table sizes and so on. This will usually further increase code size
4214	and can additionally have an effect on the size of data structures at	4772	and can additionally have an effect on the size of data structures at
4215	runtime.	4773	runtime.
4216		4774
		4775	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4776	(e.g. gcc with C<-Os>).
		4777
4217	=item C<4> - full API configuration	4778	=item C<4> - full API configuration
4218		4779
4219	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and	4780	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
4220	enables multiplicity (C<EV_MULTIPLICITY>=1).	4781	enables multiplicity (C<EV_MULTIPLICITY>=1).
4221		4782
…		…
4251		4812
4252	With an intelligent-enough linker (gcc+binutils are intelligent enough	4813	With an intelligent-enough linker (gcc+binutils are intelligent enough
4253	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by	4814	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
4254	your program might be left out as well - a binary starting a timer and an	4815	your program might be left out as well - a binary starting a timer and an
4255	I/O watcher then might come out at only 5Kb.	4816	I/O watcher then might come out at only 5Kb.
		4817
		4818	=item EV_API_STATIC
		4819
		4820	If this symbol is defined (by default it is not), then all identifiers
		4821	will have static linkage. This means that libev will not export any
		4822	identifiers, and you cannot link against libev anymore. This can be useful
		4823	when you embed libev, only want to use libev functions in a single file,
		4824	and do not want its identifiers to be visible.
		4825
		4826	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4827	wants to use libev.
		4828
		4829	This option only works when libev is compiled with a C compiler, as C++
		4830	doesn't support the required declaration syntax.
4256		4831
4257	=item EV_AVOID_STDIO	4832	=item EV_AVOID_STDIO
4258		4833
4259	If this is set to C<1> at compiletime, then libev will avoid using stdio	4834	If this is set to C<1> at compiletime, then libev will avoid using stdio
4260	functions (printf, scanf, perror etc.). This will increase the code size	4835	functions (printf, scanf, perror etc.). This will increase the code size
…		…
4404	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4979	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4405		4980
4406	#include "ev_cpp.h"	4981	#include "ev_cpp.h"
4407	#include "ev.c"	4982	#include "ev.c"
4408		4983
4409	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES	4984	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
4410		4985
4411	=head2 THREADS AND COROUTINES	4986	=head2 THREADS AND COROUTINES
4412		4987
4413	=head3 THREADS	4988	=head3 THREADS
4414		4989
…		…
4465	default loop and triggering an C<ev_async> watcher from the default loop	5040	default loop and triggering an C<ev_async> watcher from the default loop
4466	watcher callback into the event loop interested in the signal.	5041	watcher callback into the event loop interested in the signal.
4467		5042
4468	=back	5043	=back
4469		5044
4470	=head4 THREAD LOCKING EXAMPLE	5045	See also L</THREAD LOCKING EXAMPLE>.
4471
4472	Here is a fictitious example of how to run an event loop in a different
4473	thread than where callbacks are being invoked and watchers are
4474	created/added/removed.
4475
4476	For a real-world example, see the C<EV::Loop::Async> perl module,
4477	which uses exactly this technique (which is suited for many high-level
4478	languages).
4479
4480	The example uses a pthread mutex to protect the loop data, a condition
4481	variable to wait for callback invocations, an async watcher to notify the
4482	event loop thread and an unspecified mechanism to wake up the main thread.
4483
4484	First, you need to associate some data with the event loop:
4485
4486	typedef struct {
4487	mutex_t lock; /* global loop lock */
4488	ev_async async_w;
4489	thread_t tid;
4490	cond_t invoke_cv;
4491	} userdata;
4492
4493	void prepare_loop (EV_P)
4494	{
4495	// for simplicity, we use a static userdata struct.
4496	static userdata u;
4497
4498	ev_async_init (&u->async_w, async_cb);
4499	ev_async_start (EV_A_ &u->async_w);
4500
4501	pthread_mutex_init (&u->lock, 0);
4502	pthread_cond_init (&u->invoke_cv, 0);
4503
4504	// now associate this with the loop
4505	ev_set_userdata (EV_A_ u);
4506	ev_set_invoke_pending_cb (EV_A_ l_invoke);
4507	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4508
4509	// then create the thread running ev_loop
4510	pthread_create (&u->tid, 0, l_run, EV_A);
4511	}
4512
4513	The callback for the C<ev_async> watcher does nothing: the watcher is used
4514	solely to wake up the event loop so it takes notice of any new watchers
4515	that might have been added:
4516
4517	static void
4518	async_cb (EV_P_ ev_async *w, int revents)
4519	{
4520	// just used for the side effects
4521	}
4522
4523	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4524	protecting the loop data, respectively.
4525
4526	static void
4527	l_release (EV_P)
4528	{
4529	userdata *u = ev_userdata (EV_A);
4530	pthread_mutex_unlock (&u->lock);
4531	}
4532
4533	static void
4534	l_acquire (EV_P)
4535	{
4536	userdata *u = ev_userdata (EV_A);
4537	pthread_mutex_lock (&u->lock);
4538	}
4539
4540	The event loop thread first acquires the mutex, and then jumps straight
4541	into C<ev_run>:
4542
4543	void *
4544	l_run (void *thr_arg)
4545	{
4546	struct ev_loop *loop = (struct ev_loop *)thr_arg;
4547
4548	l_acquire (EV_A);
4549	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4550	ev_run (EV_A_ 0);
4551	l_release (EV_A);
4552
4553	return 0;
4554	}
4555
4556	Instead of invoking all pending watchers, the C<l_invoke> callback will
4557	signal the main thread via some unspecified mechanism (signals? pipe
4558	writes? C<Async::Interrupt>?) and then waits until all pending watchers
4559	have been called (in a while loop because a) spurious wakeups are possible
4560	and b) skipping inter-thread-communication when there are no pending
4561	watchers is very beneficial):
4562
4563	static void
4564	l_invoke (EV_P)
4565	{
4566	userdata *u = ev_userdata (EV_A);
4567
4568	while (ev_pending_count (EV_A))
4569	{
4570	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4571	pthread_cond_wait (&u->invoke_cv, &u->lock);
4572	}
4573	}
4574
4575	Now, whenever the main thread gets told to invoke pending watchers, it
4576	will grab the lock, call C<ev_invoke_pending> and then signal the loop
4577	thread to continue:
4578
4579	static void
4580	real_invoke_pending (EV_P)
4581	{
4582	userdata *u = ev_userdata (EV_A);
4583
4584	pthread_mutex_lock (&u->lock);
4585	ev_invoke_pending (EV_A);
4586	pthread_cond_signal (&u->invoke_cv);
4587	pthread_mutex_unlock (&u->lock);
4588	}
4589
4590	Whenever you want to start/stop a watcher or do other modifications to an
4591	event loop, you will now have to lock:
4592
4593	ev_timer timeout_watcher;
4594	userdata *u = ev_userdata (EV_A);
4595
4596	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4597
4598	pthread_mutex_lock (&u->lock);
4599	ev_timer_start (EV_A_ &timeout_watcher);
4600	ev_async_send (EV_A_ &u->async_w);
4601	pthread_mutex_unlock (&u->lock);
4602
4603	Note that sending the C<ev_async> watcher is required because otherwise
4604	an event loop currently blocking in the kernel will have no knowledge
4605	about the newly added timer. By waking up the loop it will pick up any new
4606	watchers in the next event loop iteration.
4607		5046
4608	=head3 COROUTINES	5047	=head3 COROUTINES
4609		5048
4610	Libev is very accommodating to coroutines ("cooperative threads"):	5049	Libev is very accommodating to coroutines ("cooperative threads"):
4611	libev fully supports nesting calls to its functions from different	5050	libev fully supports nesting calls to its functions from different
…		…
4776	requires, and its I/O model is fundamentally incompatible with the POSIX	5215	requires, and its I/O model is fundamentally incompatible with the POSIX
4777	model. Libev still offers limited functionality on this platform in	5216	model. Libev still offers limited functionality on this platform in
4778	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5217	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4779	descriptors. This only applies when using Win32 natively, not when using	5218	descriptors. This only applies when using Win32 natively, not when using
4780	e.g. cygwin. Actually, it only applies to the microsofts own compilers,	5219	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
4781	as every compielr comes with a slightly differently broken/incompatible	5220	as every compiler comes with a slightly differently broken/incompatible
4782	environment.	5221	environment.
4783		5222
4784	Lifting these limitations would basically require the full	5223	Lifting these limitations would basically require the full
4785	re-implementation of the I/O system. If you are into this kind of thing,	5224	re-implementation of the I/O system. If you are into this kind of thing,
4786	then note that glib does exactly that for you in a very portable way (note	5225	then note that glib does exactly that for you in a very portable way (note
…		…
4880	structure (guaranteed by POSIX but not by ISO C for example), but it also	5319	structure (guaranteed by POSIX but not by ISO C for example), but it also
4881	assumes that the same (machine) code can be used to call any watcher	5320	assumes that the same (machine) code can be used to call any watcher
4882	callback: The watcher callbacks have different type signatures, but libev	5321	callback: The watcher callbacks have different type signatures, but libev
4883	calls them using an C<ev_watcher *> internally.	5322	calls them using an C<ev_watcher *> internally.
4884		5323
		5324	=item null pointers and integer zero are represented by 0 bytes
		5325
		5326	Libev uses C<memset> to initialise structs and arrays to C<0> bytes, and
		5327	relies on this setting pointers and integers to null.
		5328
4885	=item pointer accesses must be thread-atomic	5329	=item pointer accesses must be thread-atomic
4886		5330
4887	Accessing a pointer value must be atomic, it must both be readable and	5331	Accessing a pointer value must be atomic, it must both be readable and
4888	writable in one piece - this is the case on all current architectures.	5332	writable in one piece - this is the case on all current architectures.
4889		5333
…		…
4902	thread" or will block signals process-wide, both behaviours would	5346	thread" or will block signals process-wide, both behaviours would
4903	be compatible with libev. Interaction between C<sigprocmask> and	5347	be compatible with libev. Interaction between C<sigprocmask> and
4904	C<pthread_sigmask> could complicate things, however.	5348	C<pthread_sigmask> could complicate things, however.
4905		5349
4906	The most portable way to handle signals is to block signals in all threads	5350	The most portable way to handle signals is to block signals in all threads
4907	except the initial one, and run the default loop in the initial thread as	5351	except the initial one, and run the signal handling loop in the initial
4908	well.	5352	thread as well.
4909		5353
4910	=item C<long> must be large enough for common memory allocation sizes	5354	=item C<long> must be large enough for common memory allocation sizes
4911		5355
4912	To improve portability and simplify its API, libev uses C<long> internally	5356	To improve portability and simplify its API, libev uses C<long> internally
4913	instead of C<size_t> when allocating its data structures. On non-POSIX	5357	instead of C<size_t> when allocating its data structures. On non-POSIX
…		…
4919		5363
4920	The type C<double> is used to represent timestamps. It is required to	5364	The type C<double> is used to represent timestamps. It is required to
4921	have at least 51 bits of mantissa (and 9 bits of exponent), which is	5365	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4922	good enough for at least into the year 4000 with millisecond accuracy	5366	good enough for at least into the year 4000 with millisecond accuracy
4923	(the design goal for libev). This requirement is overfulfilled by	5367	(the design goal for libev). This requirement is overfulfilled by
4924	implementations using IEEE 754, which is basically all existing ones. With	5368	implementations using IEEE 754, which is basically all existing ones.
		5369
4925	IEEE 754 doubles, you get microsecond accuracy until at least 2200.	5370	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5371	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5372	is either obsolete or somebody patched it to use C<long double> or
		5373	something like that, just kidding).
4926		5374
4927	=back	5375	=back
4928		5376
4929	If you know of other additional requirements drop me a note.	5377	If you know of other additional requirements drop me a note.
4930		5378
…		…
4992	=item Processing ev_async_send: O(number_of_async_watchers)	5440	=item Processing ev_async_send: O(number_of_async_watchers)
4993		5441
4994	=item Processing signals: O(max_signal_number)	5442	=item Processing signals: O(max_signal_number)
4995		5443
4996	Sending involves a system call I<iff> there were no other C<ev_async_send>	5444	Sending involves a system call I<iff> there were no other C<ev_async_send>
4997	calls in the current loop iteration. Checking for async and signal events	5445	calls in the current loop iteration and the loop is currently
		5446	blocked. Checking for async and signal events involves iterating over all
4998	involves iterating over all running async watchers or all signal numbers.	5447	running async watchers or all signal numbers.
4999		5448
5000	=back	5449	=back
5001		5450
5002		5451
5003	=head1 PORTING FROM LIBEV 3.X TO 4.X	5452	=head1 PORTING FROM LIBEV 3.X TO 4.X
…		…
5012	=over 4	5461	=over 4
5013		5462
5014	=item C<EV_COMPAT3> backwards compatibility mechanism	5463	=item C<EV_COMPAT3> backwards compatibility mechanism
5015		5464
5016	The backward compatibility mechanism can be controlled by	5465	The backward compatibility mechanism can be controlled by
5017	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>	5466	C<EV_COMPAT3>. See L</"PREPROCESSOR SYMBOLS/MACROS"> in the L</EMBEDDING>
5018	section.	5467	section.
5019		5468
5020	=item C<ev_default_destroy> and C<ev_default_fork> have been removed	5469	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
5021		5470
5022	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:	5471	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
…		…
5065	=over 4	5514	=over 4
5066		5515
5067	=item active	5516	=item active
5068		5517
5069	A watcher is active as long as it has been started and not yet stopped.	5518	A watcher is active as long as it has been started and not yet stopped.
5070	See L<WATCHER STATES> for details.	5519	See L</WATCHER STATES> for details.
5071		5520
5072	=item application	5521	=item application
5073		5522
5074	In this document, an application is whatever is using libev.	5523	In this document, an application is whatever is using libev.
5075		5524
…		…
5111	watchers and events.	5560	watchers and events.
5112		5561
5113	=item pending	5562	=item pending
5114		5563
5115	A watcher is pending as soon as the corresponding event has been	5564	A watcher is pending as soon as the corresponding event has been
5116	detected. See L<WATCHER STATES> for details.	5565	detected. See L</WATCHER STATES> for details.
5117		5566
5118	=item real time	5567	=item real time
5119		5568
5120	The physical time that is observed. It is apparently strictly monotonic :)	5569	The physical time that is observed. It is apparently strictly monotonic :)
5121		5570
5122	=item wall-clock time	5571	=item wall-clock time
5123		5572
5124	The time and date as shown on clocks. Unlike real time, it can actually	5573	The time and date as shown on clocks. Unlike real time, it can actually
5125	be wrong and jump forwards and backwards, e.g. when the you adjust your	5574	be wrong and jump forwards and backwards, e.g. when you adjust your
5126	clock.	5575	clock.
5127		5576
5128	=item watcher	5577	=item watcher
5129		5578
5130	A data structure that describes interest in certain events. Watchers need	5579	A data structure that describes interest in certain events. Watchers need
…		…
5133	=back	5582	=back
5134		5583
5135	=head1 AUTHOR	5584	=head1 AUTHOR
5136		5585
5137	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael	5586	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
5138	Magnusson and Emanuele Giaquinta.	5587	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
5139		5588

Diff Legend

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