ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.36 by root, Sat Nov 24 07:14:26 2007 UTC vs.
Revision 1.158 by root, Wed May 21 12:51:38 2008 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
		9	=head2 EXAMPLE PROGRAM
		10
		11	// a single header file is required
		12	#include <ev.h>
		13
		14	// every watcher type has its own typedef'd struct
		15	// with the name ev_<type>
		16	ev_io stdin_watcher;
		17	ev_timer timeout_watcher;
		18
		19	// all watcher callbacks have a similar signature
		20	// this callback is called when data is readable on stdin
		21	static void
		22	stdin_cb (EV_P_ struct ev_io *w, int revents)
		23	{
		24	puts ("stdin ready");
		25	// for one-shot events, one must manually stop the watcher
		26	// with its corresponding stop function.
		27	ev_io_stop (EV_A_ w);
		28
		29	// this causes all nested ev_loop's to stop iterating
		30	ev_unloop (EV_A_ EVUNLOOP_ALL);
		31	}
		32
		33	// another callback, this time for a time-out
		34	static void
		35	timeout_cb (EV_P_ struct ev_timer *w, int revents)
		36	{
		37	puts ("timeout");
		38	// this causes the innermost ev_loop to stop iterating
		39	ev_unloop (EV_A_ EVUNLOOP_ONE);
		40	}
		41
		42	int
		43	main (void)
		44	{
		45	// use the default event loop unless you have special needs
		46	struct ev_loop *loop = ev_default_loop (0);
		47
		48	// initialise an io watcher, then start it
		49	// this one will watch for stdin to become readable
		50	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		51	ev_io_start (loop, &stdin_watcher);
		52
		53	// initialise a timer watcher, then start it
		54	// simple non-repeating 5.5 second timeout
		55	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		56	ev_timer_start (loop, &timeout_watcher);
		57
		58	// now wait for events to arrive
		59	ev_loop (loop, 0);
		60
		61	// unloop was called, so exit
		62	return 0;
		63	}
		64
9	=head1 DESCRIPTION	65	=head1 DESCRIPTION
10		66
		67	The newest version of this document is also available as an html-formatted
		68	web page you might find easier to navigate when reading it for the first
		69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		70
11	Libev is an event loop: you register interest in certain events (such as a	71	Libev is an event loop: you register interest in certain events (such as a
12	file descriptor being readable or a timeout occuring), and it will manage	72	file descriptor being readable or a timeout occurring), and it will manage
13	these event sources and provide your program with events.	73	these event sources and provide your program with events.
14		74
15	To do this, it must take more or less complete control over your process	75	To do this, it must take more or less complete control over your process
16	(or thread) by executing the I<event loop> handler, and will then	76	(or thread) by executing the I<event loop> handler, and will then
17	communicate events via a callback mechanism.	77	communicate events via a callback mechanism.
…		…
19	You register interest in certain events by registering so-called I<event	79	You register interest in certain events by registering so-called I<event
20	watchers>, which are relatively small C structures you initialise with the	80	watchers>, which are relatively small C structures you initialise with the
21	details of the event, and then hand it over to libev by I<starting> the	81	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	82	watcher.
23		83
24	=head1 FEATURES	84	=head2 FEATURES
25		85
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
29	events (related to SIGCHLD), and event watchers dealing with the event	89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	90	with customised rescheduling (C<ev_periodic>), synchronous signals
		91	(C<ev_signal>), process status change events (C<ev_child>), and event
		92	watchers dealing with the event loop mechanism itself (C<ev_idle>,
		93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as
		94	file watchers (C<ev_stat>) and even limited support for fork events
		95	(C<ev_fork>).
		96
		97	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	99	for example).
33		100
34	=head1 CONVENTIONS	101	=head2 CONVENTIONS
35		102
36	Libev is very configurable. In this manual the default configuration	103	Libev is very configurable. In this manual the default (and most common)
37	will be described, which supports multiple event loops. For more info	104	configuration will be described, which supports multiple event loops. For
38	about various configuration options please have a look at the file	105	more info about various configuration options please have a look at
39	F<README.embed> in the libev distribution. If libev was configured without	106	B<EMBED> section in this manual. If libev was configured without support
40	support for multiple event loops, then all functions taking an initial	107	for multiple event loops, then all functions taking an initial argument of
41	argument of name C<loop> (which is always of type C<struct ev_loop *>)	108	name C<loop> (which is always of type C<struct ev_loop *>) will not have
42	will not have this argument.	109	this argument.
43		110
44	=head1 TIME REPRESENTATION	111	=head2 TIME REPRESENTATION
45		112
46	Libev represents time as a single floating point number, representing the	113	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	114	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	115	the beginning of 1970, details are complicated, don't ask). This type is
49	called C<ev_tstamp>, which is what you should use too. It usually aliases	116	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the C<double> type in C, and when you need to do any calculations on	117	to the C<double> type in C, and when you need to do any calculations on
51	it, you should treat it as such.	118	it, you should treat it as some floatingpoint value. Unlike the name
52		119	component C<stamp> might indicate, it is also used for time differences
		120	throughout libev.
53		121
54	=head1 GLOBAL FUNCTIONS	122	=head1 GLOBAL FUNCTIONS
55		123
56	These functions can be called anytime, even before initialising the	124	These functions can be called anytime, even before initialising the
57	library in any way.	125	library in any way.
…		…
62		130
63	Returns the current time as libev would use it. Please note that the	131	Returns the current time as libev would use it. Please note that the
64	C<ev_now> function is usually faster and also often returns the timestamp	132	C<ev_now> function is usually faster and also often returns the timestamp
65	you actually want to know.	133	you actually want to know.
66		134
		135	=item ev_sleep (ev_tstamp interval)
		136
		137	Sleep for the given interval: The current thread will be blocked until
		138	either it is interrupted or the given time interval has passed. Basically
		139	this is a subsecond-resolution C<sleep ()>.
		140
67	=item int ev_version_major ()	141	=item int ev_version_major ()
68		142
69	=item int ev_version_minor ()	143	=item int ev_version_minor ()
70		144
71	You can find out the major and minor version numbers of the library	145	You can find out the major and minor ABI version numbers of the library
72	you linked against by calling the functions C<ev_version_major> and	146	you linked against by calling the functions C<ev_version_major> and
73	C<ev_version_minor>. If you want, you can compare against the global	147	C<ev_version_minor>. If you want, you can compare against the global
74	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	148	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
75	version of the library your program was compiled against.	149	version of the library your program was compiled against.
76		150
		151	These version numbers refer to the ABI version of the library, not the
		152	release version.
		153
77	Usually, it's a good idea to terminate if the major versions mismatch,	154	Usually, it's a good idea to terminate if the major versions mismatch,
78	as this indicates an incompatible change. Minor versions are usually	155	as this indicates an incompatible change. Minor versions are usually
79	compatible to older versions, so a larger minor version alone is usually	156	compatible to older versions, so a larger minor version alone is usually
80	not a problem.	157	not a problem.
81		158
82	Example: make sure we haven't accidentally been linked against the wrong	159	Example: Make sure we haven't accidentally been linked against the wrong
83	version:	160	version.
84		161
85	assert (("libev version mismatch",	162	assert (("libev version mismatch",
86	ev_version_major () == EV_VERSION_MAJOR	163	ev_version_major () == EV_VERSION_MAJOR
87	&& ev_version_minor () >= EV_VERSION_MINOR));	164	&& ev_version_minor () >= EV_VERSION_MINOR));
88		165
…		…
118		195
119	See the description of C<ev_embed> watchers for more info.	196	See the description of C<ev_embed> watchers for more info.
120		197
121	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	198	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
122		199
123	Sets the allocation function to use (the prototype is similar to the	200	Sets the allocation function to use (the prototype is similar - the
124	realloc C function, the semantics are identical). It is used to allocate	201	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
125	and free memory (no surprises here). If it returns zero when memory	202	used to allocate and free memory (no surprises here). If it returns zero
126	needs to be allocated, the library might abort or take some potentially	203	when memory needs to be allocated (C<size != 0>), the library might abort
127	destructive action. The default is your system realloc function.	204	or take some potentially destructive action.
		205
		206	Since some systems (at least OpenBSD and Darwin) fail to implement
		207	correct C<realloc> semantics, libev will use a wrapper around the system
		208	C<realloc> and C<free> functions by default.
128		209
129	You could override this function in high-availability programs to, say,	210	You could override this function in high-availability programs to, say,
130	free some memory if it cannot allocate memory, to use a special allocator,	211	free some memory if it cannot allocate memory, to use a special allocator,
131	or even to sleep a while and retry until some memory is available.	212	or even to sleep a while and retry until some memory is available.
132		213
133	Example: replace the libev allocator with one that waits a bit and then	214	Example: Replace the libev allocator with one that waits a bit and then
134	retries: better than mine).	215	retries (example requires a standards-compliant C<realloc>).
135		216
136	static void *	217	static void *
137	persistent_realloc (void *ptr, long size)	218	persistent_realloc (void *ptr, size_t size)
138	{	219	{
139	for (;;)	220	for (;;)
140	{	221	{
141	void *newptr = realloc (ptr, size);	222	void *newptr = realloc (ptr, size);
142		223
…		…
158	callback is set, then libev will expect it to remedy the sitution, no	239	callback is set, then libev will expect it to remedy the sitution, no
159	matter what, when it returns. That is, libev will generally retry the	240	matter what, when it returns. That is, libev will generally retry the
160	requested operation, or, if the condition doesn't go away, do bad stuff	241	requested operation, or, if the condition doesn't go away, do bad stuff
161	(such as abort).	242	(such as abort).
162		243
163	Example: do the same thing as libev does internally:	244	Example: This is basically the same thing that libev does internally, too.
164		245
165	static void	246	static void
166	fatal_error (const char *msg)	247	fatal_error (const char *msg)
167	{	248	{
168	perror (msg);	249	perror (msg);
…		…
178		259
179	An event loop is described by a C<struct ev_loop *>. The library knows two	260	An event loop is described by a C<struct ev_loop *>. The library knows two
180	types of such loops, the I<default> loop, which supports signals and child	261	types of such loops, the I<default> loop, which supports signals and child
181	events, and dynamically created loops which do not.	262	events, and dynamically created loops which do not.
182		263
183	If you use threads, a common model is to run the default event loop
184	in your main thread (or in a separate thread) and for each thread you
185	create, you also create another event loop. Libev itself does no locking
186	whatsoever, so if you mix calls to the same event loop in different
187	threads, make sure you lock (this is usually a bad idea, though, even if
188	done correctly, because it's hideous and inefficient).
189
190	=over 4	264	=over 4
191		265
192	=item struct ev_loop *ev_default_loop (unsigned int flags)	266	=item struct ev_loop *ev_default_loop (unsigned int flags)
193		267
194	This will initialise the default event loop if it hasn't been initialised	268	This will initialise the default event loop if it hasn't been initialised
…		…
196	false. If it already was initialised it simply returns it (and ignores the	270	false. If it already was initialised it simply returns it (and ignores the
197	flags. If that is troubling you, check C<ev_backend ()> afterwards).	271	flags. If that is troubling you, check C<ev_backend ()> afterwards).
198		272
199	If you don't know what event loop to use, use the one returned from this	273	If you don't know what event loop to use, use the one returned from this
200	function.	274	function.
		275
		276	Note that this function is I<not> thread-safe, so if you want to use it
		277	from multiple threads, you have to lock (note also that this is unlikely,
		278	as loops cannot bes hared easily between threads anyway).
		279
		280	The default loop is the only loop that can handle C<ev_signal> and
		281	C<ev_child> watchers, and to do this, it always registers a handler
		282	for C<SIGCHLD>. If this is a problem for your app you can either
		283	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		284	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		285	C<ev_default_init>.
201		286
202	The flags argument can be used to specify special behaviour or specific	287	The flags argument can be used to specify special behaviour or specific
203	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	288	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
204		289
205	The following flags are supported:	290	The following flags are supported:
…		…
218	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	303	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
219	override the flags completely if it is found in the environment. This is	304	override the flags completely if it is found in the environment. This is
220	useful to try out specific backends to test their performance, or to work	305	useful to try out specific backends to test their performance, or to work
221	around bugs.	306	around bugs.
222		307
		308	=item C<EVFLAG_FORKCHECK>
		309
		310	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		311	a fork, you can also make libev check for a fork in each iteration by
		312	enabling this flag.
		313
		314	This works by calling C<getpid ()> on every iteration of the loop,
		315	and thus this might slow down your event loop if you do a lot of loop
		316	iterations and little real work, but is usually not noticeable (on my
		317	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
		318	without a syscall and thus I<very> fast, but my GNU/Linux system also has
		319	C<pthread_atfork> which is even faster).
		320
		321	The big advantage of this flag is that you can forget about fork (and
		322	forget about forgetting to tell libev about forking) when you use this
		323	flag.
		324
		325	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		326	environment variable.
		327
223	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	328	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
224		329
225	This is your standard select(2) backend. Not I<completely> standard, as	330	This is your standard select(2) backend. Not I<completely> standard, as
226	libev tries to roll its own fd_set with no limits on the number of fds,	331	libev tries to roll its own fd_set with no limits on the number of fds,
227	but if that fails, expect a fairly low limit on the number of fds when	332	but if that fails, expect a fairly low limit on the number of fds when
228	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	333	using this backend. It doesn't scale too well (O(highest_fd)), but its
229	the fastest backend for a low number of fds.	334	usually the fastest backend for a low number of (low-numbered :) fds.
		335
		336	To get good performance out of this backend you need a high amount of
		337	parallelity (most of the file descriptors should be busy). If you are
		338	writing a server, you should C<accept ()> in a loop to accept as many
		339	connections as possible during one iteration. You might also want to have
		340	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		341	readiness notifications you get per iteration.
230		342
231	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	343	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
232		344
233	And this is your standard poll(2) backend. It's more complicated than	345	And this is your standard poll(2) backend. It's more complicated
234	select, but handles sparse fds better and has no artificial limit on the	346	than select, but handles sparse fds better and has no artificial
235	number of fds you can use (except it will slow down considerably with a	347	limit on the number of fds you can use (except it will slow down
236	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	348	considerably with a lot of inactive fds). It scales similarly to select,
		349	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		350	performance tips.
237		351
238	=item C<EVBACKEND_EPOLL> (value 4, Linux)	352	=item C<EVBACKEND_EPOLL> (value 4, Linux)
239		353
240	For few fds, this backend is a bit little slower than poll and select,	354	For few fds, this backend is a bit little slower than poll and select,
241	but it scales phenomenally better. While poll and select usually scale like	355	but it scales phenomenally better. While poll and select usually scale
242	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	356	like O(total_fds) where n is the total number of fds (or the highest fd),
243	either O(1) or O(active_fds).	357	epoll scales either O(1) or O(active_fds). The epoll design has a number
		358	of shortcomings, such as silently dropping events in some hard-to-detect
		359	cases and requiring a syscall per fd change, no fork support and bad
		360	support for dup.
244		361
245	While stopping and starting an I/O watcher in the same iteration will	362	While stopping, setting and starting an I/O watcher in the same iteration
246	result in some caching, there is still a syscall per such incident	363	will result in some caching, there is still a syscall per such incident
247	(because the fd could point to a different file description now), so its	364	(because the fd could point to a different file description now), so its
248	best to avoid that. Also, dup()ed file descriptors might not work very	365	best to avoid that. Also, C<dup ()>'ed file descriptors might not work
249	well if you register events for both fds.	366	very well if you register events for both fds.
250		367
251	Please note that epoll sometimes generates spurious notifications, so you	368	Please note that epoll sometimes generates spurious notifications, so you
252	need to use non-blocking I/O or other means to avoid blocking when no data	369	need to use non-blocking I/O or other means to avoid blocking when no data
253	(or space) is available.	370	(or space) is available.
254		371
		372	Best performance from this backend is achieved by not unregistering all
		373	watchers for a file descriptor until it has been closed, if possible, i.e.
		374	keep at least one watcher active per fd at all times.
		375
		376	While nominally embeddeble in other event loops, this feature is broken in
		377	all kernel versions tested so far.
		378
255	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	379	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
256		380
257	Kqueue deserves special mention, as at the time of this writing, it	381	Kqueue deserves special mention, as at the time of this writing, it
258	was broken on all BSDs except NetBSD (usually it doesn't work with	382	was broken on all BSDs except NetBSD (usually it doesn't work reliably
259	anything but sockets and pipes, except on Darwin, where of course its	383	with anything but sockets and pipes, except on Darwin, where of course
260	completely useless). For this reason its not being "autodetected"	384	it's completely useless). For this reason it's not being "autodetected"
261	unless you explicitly specify it explicitly in the flags (i.e. using	385	unless you explicitly specify it explicitly in the flags (i.e. using
262	C<EVBACKEND_KQUEUE>).	386	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		387	system like NetBSD.
		388
		389	You still can embed kqueue into a normal poll or select backend and use it
		390	only for sockets (after having made sure that sockets work with kqueue on
		391	the target platform). See C<ev_embed> watchers for more info.
263		392
264	It scales in the same way as the epoll backend, but the interface to the	393	It scales in the same way as the epoll backend, but the interface to the
265	kernel is more efficient (which says nothing about its actual speed, of	394	kernel is more efficient (which says nothing about its actual speed, of
266	course). While starting and stopping an I/O watcher does not cause an	395	course). While stopping, setting and starting an I/O watcher does never
267	extra syscall as with epoll, it still adds up to four event changes per	396	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to
268	incident, so its best to avoid that.	397	two event changes per incident, support for C<fork ()> is very bad and it
		398	drops fds silently in similarly hard-to-detect cases.
		399
		400	This backend usually performs well under most conditions.
		401
		402	While nominally embeddable in other event loops, this doesn't work
		403	everywhere, so you might need to test for this. And since it is broken
		404	almost everywhere, you should only use it when you have a lot of sockets
		405	(for which it usually works), by embedding it into another event loop
		406	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for
		407	sockets.
269		408
270	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	409	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
271		410
272	This is not implemented yet (and might never be).	411	This is not implemented yet (and might never be, unless you send me an
		412	implementation). According to reports, C</dev/poll> only supports sockets
		413	and is not embeddable, which would limit the usefulness of this backend
		414	immensely.
273		415
274	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	416	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
275		417
276	This uses the Solaris 10 port mechanism. As with everything on Solaris,	418	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
277	it's really slow, but it still scales very well (O(active_fds)).	419	it's really slow, but it still scales very well (O(active_fds)).
278		420
279	Please note that solaris ports can result in a lot of spurious	421	Please note that solaris event ports can deliver a lot of spurious
280	notifications, so you need to use non-blocking I/O or other means to avoid	422	notifications, so you need to use non-blocking I/O or other means to avoid
281	blocking when no data (or space) is available.	423	blocking when no data (or space) is available.
		424
		425	While this backend scales well, it requires one system call per active
		426	file descriptor per loop iteration. For small and medium numbers of file
		427	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		428	might perform better.
		429
		430	On the positive side, ignoring the spurious readiness notifications, this
		431	backend actually performed to specification in all tests and is fully
		432	embeddable, which is a rare feat among the OS-specific backends.
282		433
283	=item C<EVBACKEND_ALL>	434	=item C<EVBACKEND_ALL>
284		435
285	Try all backends (even potentially broken ones that wouldn't be tried	436	Try all backends (even potentially broken ones that wouldn't be tried
286	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	437	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
287	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	438	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
288		439
		440	It is definitely not recommended to use this flag.
		441
289	=back	442	=back
290		443
291	If one or more of these are ored into the flags value, then only these	444	If one or more of these are ored into the flags value, then only these
292	backends will be tried (in the reverse order as given here). If none are	445	backends will be tried (in the reverse order as listed here). If none are
293	specified, most compiled-in backend will be tried, usually in reverse	446	specified, all backends in C<ev_recommended_backends ()> will be tried.
294	order of their flag values :)
295		447
296	The most typical usage is like this:	448	The most typical usage is like this:
297		449
298	if (!ev_default_loop (0))	450	if (!ev_default_loop (0))
299	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	451	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
…		…
314	Similar to C<ev_default_loop>, but always creates a new event loop that is	466	Similar to C<ev_default_loop>, but always creates a new event loop that is
315	always distinct from the default loop. Unlike the default loop, it cannot	467	always distinct from the default loop. Unlike the default loop, it cannot
316	handle signal and child watchers, and attempts to do so will be greeted by	468	handle signal and child watchers, and attempts to do so will be greeted by
317	undefined behaviour (or a failed assertion if assertions are enabled).	469	undefined behaviour (or a failed assertion if assertions are enabled).
318		470
		471	Note that this function I<is> thread-safe, and the recommended way to use
		472	libev with threads is indeed to create one loop per thread, and using the
		473	default loop in the "main" or "initial" thread.
		474
319	Example: try to create a event loop that uses epoll and nothing else.	475	Example: Try to create a event loop that uses epoll and nothing else.
320		476
321	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	477	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
322	if (!epoller)	478	if (!epoller)
323	fatal ("no epoll found here, maybe it hides under your chair");	479	fatal ("no epoll found here, maybe it hides under your chair");
324		480
325	=item ev_default_destroy ()	481	=item ev_default_destroy ()
326		482
327	Destroys the default loop again (frees all memory and kernel state	483	Destroys the default loop again (frees all memory and kernel state
328	etc.). This stops all registered event watchers (by not touching them in	484	etc.). None of the active event watchers will be stopped in the normal
329	any way whatsoever, although you cannot rely on this :).	485	sense, so e.g. C<ev_is_active> might still return true. It is your
		486	responsibility to either stop all watchers cleanly yoursef I<before>
		487	calling this function, or cope with the fact afterwards (which is usually
		488	the easiest thing, you can just ignore the watchers and/or C<free ()> them
		489	for example).
		490
		491	Note that certain global state, such as signal state, will not be freed by
		492	this function, and related watchers (such as signal and child watchers)
		493	would need to be stopped manually.
		494
		495	In general it is not advisable to call this function except in the
		496	rare occasion where you really need to free e.g. the signal handling
		497	pipe fds. If you need dynamically allocated loops it is better to use
		498	C<ev_loop_new> and C<ev_loop_destroy>).
330		499
331	=item ev_loop_destroy (loop)	500	=item ev_loop_destroy (loop)
332		501
333	Like C<ev_default_destroy>, but destroys an event loop created by an	502	Like C<ev_default_destroy>, but destroys an event loop created by an
334	earlier call to C<ev_loop_new>.	503	earlier call to C<ev_loop_new>.
335		504
336	=item ev_default_fork ()	505	=item ev_default_fork ()
337		506
		507	This function sets a flag that causes subsequent C<ev_loop> iterations
338	This function reinitialises the kernel state for backends that have	508	to reinitialise the kernel state for backends that have one. Despite the
339	one. Despite the name, you can call it anytime, but it makes most sense	509	name, you can call it anytime, but it makes most sense after forking, in
340	after forking, in either the parent or child process (or both, but that	510	the child process (or both child and parent, but that again makes little
341	again makes little sense).	511	sense). You I<must> call it in the child before using any of the libev
		512	functions, and it will only take effect at the next C<ev_loop> iteration.
342		513
343	You I<must> call this function in the child process after forking if and	514	On the other hand, you only need to call this function in the child
344	only if you want to use the event library in both processes. If you just	515	process if and only if you want to use the event library in the child. If
345	fork+exec, you don't have to call it.	516	you just fork+exec, you don't have to call it at all.
346		517
347	The function itself is quite fast and it's usually not a problem to call	518	The function itself is quite fast and it's usually not a problem to call
348	it just in case after a fork. To make this easy, the function will fit in	519	it just in case after a fork. To make this easy, the function will fit in
349	quite nicely into a call to C<pthread_atfork>:	520	quite nicely into a call to C<pthread_atfork>:
350		521
351	pthread_atfork (0, 0, ev_default_fork);	522	pthread_atfork (0, 0, ev_default_fork);
352		523
353	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
354	without calling this function, so if you force one of those backends you
355	do not need to care.
356
357	=item ev_loop_fork (loop)	524	=item ev_loop_fork (loop)
358		525
359	Like C<ev_default_fork>, but acts on an event loop created by	526	Like C<ev_default_fork>, but acts on an event loop created by
360	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	527	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
361	after fork, and how you do this is entirely your own problem.	528	after fork, and how you do this is entirely your own problem.
		529
		530	=item int ev_is_default_loop (loop)
		531
		532	Returns true when the given loop actually is the default loop, false otherwise.
		533
		534	=item unsigned int ev_loop_count (loop)
		535
		536	Returns the count of loop iterations for the loop, which is identical to
		537	the number of times libev did poll for new events. It starts at C<0> and
		538	happily wraps around with enough iterations.
		539
		540	This value can sometimes be useful as a generation counter of sorts (it
		541	"ticks" the number of loop iterations), as it roughly corresponds with
		542	C<ev_prepare> and C<ev_check> calls.
362		543
363	=item unsigned int ev_backend (loop)	544	=item unsigned int ev_backend (loop)
364		545
365	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	546	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
366	use.	547	use.
…		…
369		550
370	Returns the current "event loop time", which is the time the event loop	551	Returns the current "event loop time", which is the time the event loop
371	received events and started processing them. This timestamp does not	552	received events and started processing them. This timestamp does not
372	change as long as callbacks are being processed, and this is also the base	553	change as long as callbacks are being processed, and this is also the base
373	time used for relative timers. You can treat it as the timestamp of the	554	time used for relative timers. You can treat it as the timestamp of the
374	event occuring (or more correctly, libev finding out about it).	555	event occurring (or more correctly, libev finding out about it).
375		556
376	=item ev_loop (loop, int flags)	557	=item ev_loop (loop, int flags)
377		558
378	Finally, this is it, the event handler. This function usually is called	559	Finally, this is it, the event handler. This function usually is called
379	after you initialised all your watchers and you want to start handling	560	after you initialised all your watchers and you want to start handling
…		…
400	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	581	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
401	usually a better approach for this kind of thing.	582	usually a better approach for this kind of thing.
402		583
403	Here are the gory details of what C<ev_loop> does:	584	Here are the gory details of what C<ev_loop> does:
404		585
405	* If there are no active watchers (reference count is zero), return.	586	- Before the first iteration, call any pending watchers.
406	- Queue prepare watchers and then call all outstanding watchers.	587	* If EVFLAG_FORKCHECK was used, check for a fork.
		588	- If a fork was detected, queue and call all fork watchers.
		589	- Queue and call all prepare watchers.
407	- If we have been forked, recreate the kernel state.	590	- If we have been forked, recreate the kernel state.
408	- Update the kernel state with all outstanding changes.	591	- Update the kernel state with all outstanding changes.
409	- Update the "event loop time".	592	- Update the "event loop time".
410	- Calculate for how long to block.	593	- Calculate for how long to sleep or block, if at all
		594	(active idle watchers, EVLOOP_NONBLOCK or not having
		595	any active watchers at all will result in not sleeping).
		596	- Sleep if the I/O and timer collect interval say so.
411	- Block the process, waiting for any events.	597	- Block the process, waiting for any events.
412	- Queue all outstanding I/O (fd) events.	598	- Queue all outstanding I/O (fd) events.
413	- Update the "event loop time" and do time jump handling.	599	- Update the "event loop time" and do time jump handling.
414	- Queue all outstanding timers.	600	- Queue all outstanding timers.
415	- Queue all outstanding periodics.	601	- Queue all outstanding periodics.
416	- If no events are pending now, queue all idle watchers.	602	- If no events are pending now, queue all idle watchers.
417	- Queue all check watchers.	603	- Queue all check watchers.
418	- Call all queued watchers in reverse order (i.e. check watchers first).	604	- Call all queued watchers in reverse order (i.e. check watchers first).
419	Signals and child watchers are implemented as I/O watchers, and will	605	Signals and child watchers are implemented as I/O watchers, and will
420	be handled here by queueing them when their watcher gets executed.	606	be handled here by queueing them when their watcher gets executed.
421	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	607	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
422	were used, return, otherwise continue with step *.	608	were used, or there are no active watchers, return, otherwise
		609	continue with step *.
423		610
424	Example: queue some jobs and then loop until no events are outsanding	611	Example: Queue some jobs and then loop until no events are outstanding
425	anymore.	612	anymore.
426		613
427	... queue jobs here, make sure they register event watchers as long	614	... queue jobs here, make sure they register event watchers as long
428	... as they still have work to do (even an idle watcher will do..)	615	... as they still have work to do (even an idle watcher will do..)
429	ev_loop (my_loop, 0);	616	ev_loop (my_loop, 0);
…		…
433		620
434	Can be used to make a call to C<ev_loop> return early (but only after it	621	Can be used to make a call to C<ev_loop> return early (but only after it
435	has processed all outstanding events). The C<how> argument must be either	622	has processed all outstanding events). The C<how> argument must be either
436	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	623	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
437	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	624	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
		625
		626	This "unloop state" will be cleared when entering C<ev_loop> again.
438		627
439	=item ev_ref (loop)	628	=item ev_ref (loop)
440		629
441	=item ev_unref (loop)	630	=item ev_unref (loop)
442		631
…		…
447	returning, ev_unref() after starting, and ev_ref() before stopping it. For	636	returning, ev_unref() after starting, and ev_ref() before stopping it. For
448	example, libev itself uses this for its internal signal pipe: It is not	637	example, libev itself uses this for its internal signal pipe: It is not
449	visible to the libev user and should not keep C<ev_loop> from exiting if	638	visible to the libev user and should not keep C<ev_loop> from exiting if
450	no event watchers registered by it are active. It is also an excellent	639	no event watchers registered by it are active. It is also an excellent
451	way to do this for generic recurring timers or from within third-party	640	way to do this for generic recurring timers or from within third-party
452	libraries. Just remember to I<unref after start> and I<ref before stop>.	641	libraries. Just remember to I<unref after start> and I<ref before stop>
		642	(but only if the watcher wasn't active before, or was active before,
		643	respectively).
453		644
454	Example: create a signal watcher, but keep it from keeping C<ev_loop>	645	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
455	running when nothing else is active.	646	running when nothing else is active.
456		647
457	struct dv_signal exitsig;	648	struct ev_signal exitsig;
458	ev_signal_init (&exitsig, sig_cb, SIGINT);	649	ev_signal_init (&exitsig, sig_cb, SIGINT);
459	ev_signal_start (myloop, &exitsig);	650	ev_signal_start (loop, &exitsig);
460	evf_unref (myloop);	651	evf_unref (loop);
461		652
462	Example: for some weird reason, unregister the above signal handler again.	653	Example: For some weird reason, unregister the above signal handler again.
463		654
464	ev_ref (myloop);	655	ev_ref (loop);
465	ev_signal_stop (myloop, &exitsig);	656	ev_signal_stop (loop, &exitsig);
		657
		658	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		659
		660	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		661
		662	These advanced functions influence the time that libev will spend waiting
		663	for events. Both are by default C<0>, meaning that libev will try to
		664	invoke timer/periodic callbacks and I/O callbacks with minimum latency.
		665
		666	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		667	allows libev to delay invocation of I/O and timer/periodic callbacks to
		668	increase efficiency of loop iterations.
		669
		670	The background is that sometimes your program runs just fast enough to
		671	handle one (or very few) event(s) per loop iteration. While this makes
		672	the program responsive, it also wastes a lot of CPU time to poll for new
		673	events, especially with backends like C<select ()> which have a high
		674	overhead for the actual polling but can deliver many events at once.
		675
		676	By setting a higher I<io collect interval> you allow libev to spend more
		677	time collecting I/O events, so you can handle more events per iteration,
		678	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		679	C<ev_timer>) will be not affected. Setting this to a non-null value will
		680	introduce an additional C<ev_sleep ()> call into most loop iterations.
		681
		682	Likewise, by setting a higher I<timeout collect interval> you allow libev
		683	to spend more time collecting timeouts, at the expense of increased
		684	latency (the watcher callback will be called later). C<ev_io> watchers
		685	will not be affected. Setting this to a non-null value will not introduce
		686	any overhead in libev.
		687
		688	Many (busy) programs can usually benefit by setting the io collect
		689	interval to a value near C<0.1> or so, which is often enough for
		690	interactive servers (of course not for games), likewise for timeouts. It
		691	usually doesn't make much sense to set it to a lower value than C<0.01>,
		692	as this approsaches the timing granularity of most systems.
466		693
467	=back	694	=back
		695
468		696
469	=head1 ANATOMY OF A WATCHER	697	=head1 ANATOMY OF A WATCHER
470		698
471	A watcher is a structure that you create and register to record your	699	A watcher is a structure that you create and register to record your
472	interest in some event. For instance, if you want to wait for STDIN to	700	interest in some event. For instance, if you want to wait for STDIN to
…		…
539	The signal specified in the C<ev_signal> watcher has been received by a thread.	767	The signal specified in the C<ev_signal> watcher has been received by a thread.
540		768
541	=item C<EV_CHILD>	769	=item C<EV_CHILD>
542		770
543	The pid specified in the C<ev_child> watcher has received a status change.	771	The pid specified in the C<ev_child> watcher has received a status change.
		772
		773	=item C<EV_STAT>
		774
		775	The path specified in the C<ev_stat> watcher changed its attributes somehow.
544		776
545	=item C<EV_IDLE>	777	=item C<EV_IDLE>
546		778
547	The C<ev_idle> watcher has determined that you have nothing better to do.	779	The C<ev_idle> watcher has determined that you have nothing better to do.
548		780
…		…
556	received events. Callbacks of both watcher types can start and stop as	788	received events. Callbacks of both watcher types can start and stop as
557	many watchers as they want, and all of them will be taken into account	789	many watchers as they want, and all of them will be taken into account
558	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	790	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
559	C<ev_loop> from blocking).	791	C<ev_loop> from blocking).
560		792
		793	=item C<EV_EMBED>
		794
		795	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		796
		797	=item C<EV_FORK>
		798
		799	The event loop has been resumed in the child process after fork (see
		800	C<ev_fork>).
		801
		802	=item C<EV_ASYNC>
		803
		804	The given async watcher has been asynchronously notified (see C<ev_async>).
		805
561	=item C<EV_ERROR>	806	=item C<EV_ERROR>
562		807
563	An unspecified error has occured, the watcher has been stopped. This might	808	An unspecified error has occured, the watcher has been stopped. This might
564	happen because the watcher could not be properly started because libev	809	happen because the watcher could not be properly started because libev
565	ran out of memory, a file descriptor was found to be closed or any other	810	ran out of memory, a file descriptor was found to be closed or any other
…		…
572	with the error from read() or write(). This will not work in multithreaded	817	with the error from read() or write(). This will not work in multithreaded
573	programs, though, so beware.	818	programs, though, so beware.
574		819
575	=back	820	=back
576		821
577	=head2 SUMMARY OF GENERIC WATCHER FUNCTIONS	822	=head2 GENERIC WATCHER FUNCTIONS
578		823
579	In the following description, C<TYPE> stands for the watcher type,	824	In the following description, C<TYPE> stands for the watcher type,
580	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.	825	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
581		826
582	=over 4	827	=over 4
…		…
591	which rolls both calls into one.	836	which rolls both calls into one.
592		837
593	You can reinitialise a watcher at any time as long as it has been stopped	838	You can reinitialise a watcher at any time as long as it has been stopped
594	(or never started) and there are no pending events outstanding.	839	(or never started) and there are no pending events outstanding.
595		840
596	The callbakc is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	841	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,
597	int revents)>.	842	int revents)>.
598		843
599	=item C<ev_TYPE_set> (ev_TYPE *, [args])	844	=item C<ev_TYPE_set> (ev_TYPE *, [args])
600		845
601	This macro initialises the type-specific parts of a watcher. You need to	846	This macro initialises the type-specific parts of a watcher. You need to
…		…
636	=item bool ev_is_pending (ev_TYPE *watcher)	881	=item bool ev_is_pending (ev_TYPE *watcher)
637		882
638	Returns a true value iff the watcher is pending, (i.e. it has outstanding	883	Returns a true value iff the watcher is pending, (i.e. it has outstanding
639	events but its callback has not yet been invoked). As long as a watcher	884	events but its callback has not yet been invoked). As long as a watcher
640	is pending (but not active) you must not call an init function on it (but	885	is pending (but not active) you must not call an init function on it (but
641	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	886	C<ev_TYPE_set> is safe), you must not change its priority, and you must
642	libev (e.g. you cnanot C<free ()> it).	887	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		888	it).
643		889
644	=item callback = ev_cb (ev_TYPE *watcher)	890	=item callback ev_cb (ev_TYPE *watcher)
645		891
646	Returns the callback currently set on the watcher.	892	Returns the callback currently set on the watcher.
647		893
648	=item ev_cb_set (ev_TYPE *watcher, callback)	894	=item ev_cb_set (ev_TYPE *watcher, callback)
649		895
650	Change the callback. You can change the callback at virtually any time	896	Change the callback. You can change the callback at virtually any time
651	(modulo threads).	897	(modulo threads).
		898
		899	=item ev_set_priority (ev_TYPE *watcher, priority)
		900
		901	=item int ev_priority (ev_TYPE *watcher)
		902
		903	Set and query the priority of the watcher. The priority is a small
		904	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		905	(default: C<-2>). Pending watchers with higher priority will be invoked
		906	before watchers with lower priority, but priority will not keep watchers
		907	from being executed (except for C<ev_idle> watchers).
		908
		909	This means that priorities are I<only> used for ordering callback
		910	invocation after new events have been received. This is useful, for
		911	example, to reduce latency after idling, or more often, to bind two
		912	watchers on the same event and make sure one is called first.
		913
		914	If you need to suppress invocation when higher priority events are pending
		915	you need to look at C<ev_idle> watchers, which provide this functionality.
		916
		917	You I<must not> change the priority of a watcher as long as it is active or
		918	pending.
		919
		920	The default priority used by watchers when no priority has been set is
		921	always C<0>, which is supposed to not be too high and not be too low :).
		922
		923	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		924	fine, as long as you do not mind that the priority value you query might
		925	or might not have been adjusted to be within valid range.
		926
		927	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		928
		929	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		930	C<loop> nor C<revents> need to be valid as long as the watcher callback
		931	can deal with that fact.
		932
		933	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		934
		935	If the watcher is pending, this function returns clears its pending status
		936	and returns its C<revents> bitset (as if its callback was invoked). If the
		937	watcher isn't pending it does nothing and returns C<0>.
652		938
653	=back	939	=back
654		940
655		941
656	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	942	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
…		…
677	{	963	{
678	struct my_io *w = (struct my_io *)w_;	964	struct my_io *w = (struct my_io *)w_;
679	...	965	...
680	}	966	}
681		967
682	More interesting and less C-conformant ways of catsing your callback type	968	More interesting and less C-conformant ways of casting your callback type
683	have been omitted....	969	instead have been omitted.
		970
		971	Another common scenario is having some data structure with multiple
		972	watchers:
		973
		974	struct my_biggy
		975	{
		976	int some_data;
		977	ev_timer t1;
		978	ev_timer t2;
		979	}
		980
		981	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		982	you need to use C<offsetof>:
		983
		984	#include <stddef.h>
		985
		986	static void
		987	t1_cb (EV_P_ struct ev_timer *w, int revents)
		988	{
		989	struct my_biggy big = (struct my_biggy *
		990	(((char *)w) - offsetof (struct my_biggy, t1));
		991	}
		992
		993	static void
		994	t2_cb (EV_P_ struct ev_timer *w, int revents)
		995	{
		996	struct my_biggy big = (struct my_biggy *
		997	(((char *)w) - offsetof (struct my_biggy, t2));
		998	}
684		999
685		1000
686	=head1 WATCHER TYPES	1001	=head1 WATCHER TYPES
687		1002
688	This section describes each watcher in detail, but will not repeat	1003	This section describes each watcher in detail, but will not repeat
689	information given in the last section.	1004	information given in the last section. Any initialisation/set macros,
		1005	functions and members specific to the watcher type are explained.
690		1006
		1007	Members are additionally marked with either I<[read-only]>, meaning that,
		1008	while the watcher is active, you can look at the member and expect some
		1009	sensible content, but you must not modify it (you can modify it while the
		1010	watcher is stopped to your hearts content), or I<[read-write]>, which
		1011	means you can expect it to have some sensible content while the watcher
		1012	is active, but you can also modify it. Modifying it may not do something
		1013	sensible or take immediate effect (or do anything at all), but libev will
		1014	not crash or malfunction in any way.
691		1015
		1016
692	=head2 C<ev_io> - is this file descriptor readable or writable	1017	=head2 C<ev_io> - is this file descriptor readable or writable?
693		1018
694	I/O watchers check whether a file descriptor is readable or writable	1019	I/O watchers check whether a file descriptor is readable or writable
695	in each iteration of the event loop (This behaviour is called	1020	in each iteration of the event loop, or, more precisely, when reading
696	level-triggering because you keep receiving events as long as the	1021	would not block the process and writing would at least be able to write
697	condition persists. Remember you can stop the watcher if you don't want to	1022	some data. This behaviour is called level-triggering because you keep
698	act on the event and neither want to receive future events).	1023	receiving events as long as the condition persists. Remember you can stop
		1024	the watcher if you don't want to act on the event and neither want to
		1025	receive future events.
699		1026
700	In general you can register as many read and/or write event watchers per	1027	In general you can register as many read and/or write event watchers per
701	fd as you want (as long as you don't confuse yourself). Setting all file	1028	fd as you want (as long as you don't confuse yourself). Setting all file
702	descriptors to non-blocking mode is also usually a good idea (but not	1029	descriptors to non-blocking mode is also usually a good idea (but not
703	required if you know what you are doing).	1030	required if you know what you are doing).
704		1031
705	You have to be careful with dup'ed file descriptors, though. Some backends
706	(the linux epoll backend is a notable example) cannot handle dup'ed file
707	descriptors correctly if you register interest in two or more fds pointing
708	to the same underlying file/socket etc. description (that is, they share
709	the same underlying "file open").
710
711	If you must do this, then force the use of a known-to-be-good backend	1032	If you must do this, then force the use of a known-to-be-good backend
712	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1033	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
713	C<EVBACKEND_POLL>).	1034	C<EVBACKEND_POLL>).
714		1035
		1036	Another thing you have to watch out for is that it is quite easy to
		1037	receive "spurious" readiness notifications, that is your callback might
		1038	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		1039	because there is no data. Not only are some backends known to create a
		1040	lot of those (for example solaris ports), it is very easy to get into
		1041	this situation even with a relatively standard program structure. Thus
		1042	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		1043	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		1044
		1045	If you cannot run the fd in non-blocking mode (for example you should not
		1046	play around with an Xlib connection), then you have to seperately re-test
		1047	whether a file descriptor is really ready with a known-to-be good interface
		1048	such as poll (fortunately in our Xlib example, Xlib already does this on
		1049	its own, so its quite safe to use).
		1050
		1051	=head3 The special problem of disappearing file descriptors
		1052
		1053	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1054	descriptor (either by calling C<close> explicitly or by any other means,
		1055	such as C<dup>). The reason is that you register interest in some file
		1056	descriptor, but when it goes away, the operating system will silently drop
		1057	this interest. If another file descriptor with the same number then is
		1058	registered with libev, there is no efficient way to see that this is, in
		1059	fact, a different file descriptor.
		1060
		1061	To avoid having to explicitly tell libev about such cases, libev follows
		1062	the following policy: Each time C<ev_io_set> is being called, libev
		1063	will assume that this is potentially a new file descriptor, otherwise
		1064	it is assumed that the file descriptor stays the same. That means that
		1065	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1066	descriptor even if the file descriptor number itself did not change.
		1067
		1068	This is how one would do it normally anyway, the important point is that
		1069	the libev application should not optimise around libev but should leave
		1070	optimisations to libev.
		1071
		1072	=head3 The special problem of dup'ed file descriptors
		1073
		1074	Some backends (e.g. epoll), cannot register events for file descriptors,
		1075	but only events for the underlying file descriptions. That means when you
		1076	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1077	events for them, only one file descriptor might actually receive events.
		1078
		1079	There is no workaround possible except not registering events
		1080	for potentially C<dup ()>'ed file descriptors, or to resort to
		1081	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1082
		1083	=head3 The special problem of fork
		1084
		1085	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1086	useless behaviour. Libev fully supports fork, but needs to be told about
		1087	it in the child.
		1088
		1089	To support fork in your programs, you either have to call
		1090	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
		1091	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
		1092	C<EVBACKEND_POLL>.
		1093
		1094	=head3 The special problem of SIGPIPE
		1095
		1096	While not really specific to libev, it is easy to forget about SIGPIPE:
		1097	when reading from a pipe whose other end has been closed, your program
		1098	gets send a SIGPIPE, which, by default, aborts your program. For most
		1099	programs this is sensible behaviour, for daemons, this is usually
		1100	undesirable.
		1101
		1102	So when you encounter spurious, unexplained daemon exits, make sure you
		1103	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1104	somewhere, as that would have given you a big clue).
		1105
		1106
		1107	=head3 Watcher-Specific Functions
		1108
715	=over 4	1109	=over 4
716		1110
717	=item ev_io_init (ev_io *, callback, int fd, int events)	1111	=item ev_io_init (ev_io *, callback, int fd, int events)
718		1112
719	=item ev_io_set (ev_io *, int fd, int events)	1113	=item ev_io_set (ev_io *, int fd, int events)
720		1114
721	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	1115	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
722	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	1116	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or
723	EV_WRITE> to receive the given events.	1117	C<EV_READ | EV_WRITE> to receive the given events.
724		1118
725	Please note that most of the more scalable backend mechanisms (for example	1119	=item int fd [read-only]
726	epoll and solaris ports) can result in spurious readyness notifications	1120
727	for file descriptors, so you practically need to use non-blocking I/O (and	1121	The file descriptor being watched.
728	treat callback invocation as hint only), or retest separately with a safe	1122
729	interface before doing I/O (XLib can do this), or force the use of either	1123	=item int events [read-only]
730	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>, which don't suffer from this	1124
731	problem. Also note that it is quite easy to have your callback invoked	1125	The events being watched.
732	when the readyness condition is no longer valid even when employing
733	typical ways of handling events, so its a good idea to use non-blocking
734	I/O unconditionally.
735		1126
736	=back	1127	=back
737		1128
		1129	=head3 Examples
		1130
738	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	1131	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
739	readable, but only once. Since it is likely line-buffered, you could	1132	readable, but only once. Since it is likely line-buffered, you could
740	attempt to read a whole line in the callback:	1133	attempt to read a whole line in the callback.
741		1134
742	static void	1135	static void
743	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1136	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
744	{	1137	{
745	ev_io_stop (loop, w);	1138	ev_io_stop (loop, w);
…		…
752	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1145	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
753	ev_io_start (loop, &stdin_readable);	1146	ev_io_start (loop, &stdin_readable);
754	ev_loop (loop, 0);	1147	ev_loop (loop, 0);
755		1148
756		1149
757	=head2 C<ev_timer> - relative and optionally recurring timeouts	1150	=head2 C<ev_timer> - relative and optionally repeating timeouts
758		1151
759	Timer watchers are simple relative timers that generate an event after a	1152	Timer watchers are simple relative timers that generate an event after a
760	given time, and optionally repeating in regular intervals after that.	1153	given time, and optionally repeating in regular intervals after that.
761		1154
762	The timers are based on real time, that is, if you register an event that	1155	The timers are based on real time, that is, if you register an event that
763	times out after an hour and you reset your system clock to last years	1156	times out after an hour and you reset your system clock to january last
764	time, it will still time out after (roughly) and hour. "Roughly" because	1157	year, it will still time out after (roughly) and hour. "Roughly" because
765	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1158	detecting time jumps is hard, and some inaccuracies are unavoidable (the
766	monotonic clock option helps a lot here).	1159	monotonic clock option helps a lot here).
767		1160
768	The relative timeouts are calculated relative to the C<ev_now ()>	1161	The relative timeouts are calculated relative to the C<ev_now ()>
769	time. This is usually the right thing as this timestamp refers to the time	1162	time. This is usually the right thing as this timestamp refers to the time
…		…
771	you suspect event processing to be delayed and you I<need> to base the timeout	1164	you suspect event processing to be delayed and you I<need> to base the timeout
772	on the current time, use something like this to adjust for this:	1165	on the current time, use something like this to adjust for this:
773		1166
774	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1167	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
775		1168
776	The callback is guarenteed to be invoked only when its timeout has passed,	1169	The callback is guarenteed to be invoked only after its timeout has passed,
777	but if multiple timers become ready during the same loop iteration then	1170	but if multiple timers become ready during the same loop iteration then
778	order of execution is undefined.	1171	order of execution is undefined.
779		1172
		1173	=head3 Watcher-Specific Functions and Data Members
		1174
780	=over 4	1175	=over 4
781		1176
782	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1177	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
783		1178
784	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1179	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
785		1180
786	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1181	Configure the timer to trigger after C<after> seconds. If C<repeat>
787	C<0.>, then it will automatically be stopped. If it is positive, then the	1182	is C<0.>, then it will automatically be stopped once the timeout is
788	timer will automatically be configured to trigger again C<repeat> seconds	1183	reached. If it is positive, then the timer will automatically be
789	later, again, and again, until stopped manually.	1184	configured to trigger again C<repeat> seconds later, again, and again,
		1185	until stopped manually.
790		1186
791	The timer itself will do a best-effort at avoiding drift, that is, if you	1187	The timer itself will do a best-effort at avoiding drift, that is, if
792	configure a timer to trigger every 10 seconds, then it will trigger at	1188	you configure a timer to trigger every 10 seconds, then it will normally
793	exactly 10 second intervals. If, however, your program cannot keep up with	1189	trigger at exactly 10 second intervals. If, however, your program cannot
794	the timer (because it takes longer than those 10 seconds to do stuff) the	1190	keep up with the timer (because it takes longer than those 10 seconds to
795	timer will not fire more than once per event loop iteration.	1191	do stuff) the timer will not fire more than once per event loop iteration.
796		1192
797	=item ev_timer_again (loop)	1193	=item ev_timer_again (loop, ev_timer *)
798		1194
799	This will act as if the timer timed out and restart it again if it is	1195	This will act as if the timer timed out and restart it again if it is
800	repeating. The exact semantics are:	1196	repeating. The exact semantics are:
801		1197
		1198	If the timer is pending, its pending status is cleared.
		1199
802	If the timer is started but nonrepeating, stop it.	1200	If the timer is started but nonrepeating, stop it (as if it timed out).
803		1201
804	If the timer is repeating, either start it if necessary (with the repeat	1202	If the timer is repeating, either start it if necessary (with the
805	value), or reset the running timer to the repeat value.	1203	C<repeat> value), or reset the running timer to the C<repeat> value.
806		1204
807	This sounds a bit complicated, but here is a useful and typical	1205	This sounds a bit complicated, but here is a useful and typical
808	example: Imagine you have a tcp connection and you want a so-called idle	1206	example: Imagine you have a tcp connection and you want a so-called idle
809	timeout, that is, you want to be called when there have been, say, 60	1207	timeout, that is, you want to be called when there have been, say, 60
810	seconds of inactivity on the socket. The easiest way to do this is to	1208	seconds of inactivity on the socket. The easiest way to do this is to
811	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	1209	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
812	time you successfully read or write some data. If you go into an idle	1210	C<ev_timer_again> each time you successfully read or write some data. If
813	state where you do not expect data to travel on the socket, you can stop	1211	you go into an idle state where you do not expect data to travel on the
814	the timer, and again will automatically restart it if need be.	1212	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
		1213	automatically restart it if need be.
		1214
		1215	That means you can ignore the C<after> value and C<ev_timer_start>
		1216	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
		1217
		1218	ev_timer_init (timer, callback, 0., 5.);
		1219	ev_timer_again (loop, timer);
		1220	...
		1221	timer->again = 17.;
		1222	ev_timer_again (loop, timer);
		1223	...
		1224	timer->again = 10.;
		1225	ev_timer_again (loop, timer);
		1226
		1227	This is more slightly efficient then stopping/starting the timer each time
		1228	you want to modify its timeout value.
		1229
		1230	=item ev_tstamp repeat [read-write]
		1231
		1232	The current C<repeat> value. Will be used each time the watcher times out
		1233	or C<ev_timer_again> is called and determines the next timeout (if any),
		1234	which is also when any modifications are taken into account.
815		1235
816	=back	1236	=back
817		1237
		1238	=head3 Examples
		1239
818	Example: create a timer that fires after 60 seconds.	1240	Example: Create a timer that fires after 60 seconds.
819		1241
820	static void	1242	static void
821	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1243	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
822	{	1244	{
823	.. one minute over, w is actually stopped right here	1245	.. one minute over, w is actually stopped right here
…		…
825		1247
826	struct ev_timer mytimer;	1248	struct ev_timer mytimer;
827	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1249	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
828	ev_timer_start (loop, &mytimer);	1250	ev_timer_start (loop, &mytimer);
829		1251
830	Example: create a timeout timer that times out after 10 seconds of	1252	Example: Create a timeout timer that times out after 10 seconds of
831	inactivity.	1253	inactivity.
832		1254
833	static void	1255	static void
834	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1256	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
835	{	1257	{
…		…
844	// and in some piece of code that gets executed on any "activity":	1266	// and in some piece of code that gets executed on any "activity":
845	// reset the timeout to start ticking again at 10 seconds	1267	// reset the timeout to start ticking again at 10 seconds
846	ev_timer_again (&mytimer);	1268	ev_timer_again (&mytimer);
847		1269
848		1270
849	=head2 C<ev_periodic> - to cron or not to cron	1271	=head2 C<ev_periodic> - to cron or not to cron?
850		1272
851	Periodic watchers are also timers of a kind, but they are very versatile	1273	Periodic watchers are also timers of a kind, but they are very versatile
852	(and unfortunately a bit complex).	1274	(and unfortunately a bit complex).
853		1275
854	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1276	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
855	but on wallclock time (absolute time). You can tell a periodic watcher	1277	but on wallclock time (absolute time). You can tell a periodic watcher
856	to trigger "at" some specific point in time. For example, if you tell a	1278	to trigger after some specific point in time. For example, if you tell a
857	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	1279	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
858	+ 10.>) and then reset your system clock to the last year, then it will	1280	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1281	clock to january of the previous year, then it will take more than year
859	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1282	to trigger the event (unlike an C<ev_timer>, which would still trigger
860	roughly 10 seconds later and of course not if you reset your system time	1283	roughly 10 seconds later as it uses a relative timeout).
861	again).
862		1284
863	They can also be used to implement vastly more complex timers, such as	1285	C<ev_periodic>s can also be used to implement vastly more complex timers,
864	triggering an event on eahc midnight, local time.	1286	such as triggering an event on each "midnight, local time", or other
		1287	complicated, rules.
865		1288
866	As with timers, the callback is guarenteed to be invoked only when the	1289	As with timers, the callback is guarenteed to be invoked only when the
867	time (C<at>) has been passed, but if multiple periodic timers become ready	1290	time (C<at>) has passed, but if multiple periodic timers become ready
868	during the same loop iteration then order of execution is undefined.	1291	during the same loop iteration then order of execution is undefined.
		1292
		1293	=head3 Watcher-Specific Functions and Data Members
869		1294
870	=over 4	1295	=over 4
871		1296
872	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1297	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
873		1298
…		…
876	Lots of arguments, lets sort it out... There are basically three modes of	1301	Lots of arguments, lets sort it out... There are basically three modes of
877	operation, and we will explain them from simplest to complex:	1302	operation, and we will explain them from simplest to complex:
878		1303
879	=over 4	1304	=over 4
880		1305
881	=item * absolute timer (interval = reschedule_cb = 0)	1306	=item * absolute timer (at = time, interval = reschedule_cb = 0)
882		1307
883	In this configuration the watcher triggers an event at the wallclock time	1308	In this configuration the watcher triggers an event after the wallclock
884	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1309	time C<at> has passed and doesn't repeat. It will not adjust when a time
885	that is, if it is to be run at January 1st 2011 then it will run when the	1310	jump occurs, that is, if it is to be run at January 1st 2011 then it will
886	system time reaches or surpasses this time.	1311	run when the system time reaches or surpasses this time.
887		1312
888	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1313	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
889		1314
890	In this mode the watcher will always be scheduled to time out at the next	1315	In this mode the watcher will always be scheduled to time out at the next
891	C<at + N * interval> time (for some integer N) and then repeat, regardless	1316	C<at + N * interval> time (for some integer N, which can also be negative)
892	of any time jumps.	1317	and then repeat, regardless of any time jumps.
893		1318
894	This can be used to create timers that do not drift with respect to system	1319	This can be used to create timers that do not drift with respect to system
895	time:	1320	time, for example, here is a C<ev_periodic> that triggers each hour, on
		1321	the hour:
896		1322
897	ev_periodic_set (&periodic, 0., 3600., 0);	1323	ev_periodic_set (&periodic, 0., 3600., 0);
898		1324
899	This doesn't mean there will always be 3600 seconds in between triggers,	1325	This doesn't mean there will always be 3600 seconds in between triggers,
900	but only that the the callback will be called when the system time shows a	1326	but only that the the callback will be called when the system time shows a
…		…
903		1329
904	Another way to think about it (for the mathematically inclined) is that	1330	Another way to think about it (for the mathematically inclined) is that
905	C<ev_periodic> will try to run the callback in this mode at the next possible	1331	C<ev_periodic> will try to run the callback in this mode at the next possible
906	time where C<time = at (mod interval)>, regardless of any time jumps.	1332	time where C<time = at (mod interval)>, regardless of any time jumps.
907		1333
		1334	For numerical stability it is preferable that the C<at> value is near
		1335	C<ev_now ()> (the current time), but there is no range requirement for
		1336	this value, and in fact is often specified as zero.
		1337
		1338	Note also that there is an upper limit to how often a timer can fire (cpu
		1339	speed for example), so if C<interval> is very small then timing stability
		1340	will of course detoriate. Libev itself tries to be exact to be about one
		1341	millisecond (if the OS supports it and the machine is fast enough).
		1342
908	=item * manual reschedule mode (reschedule_cb = callback)	1343	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
909		1344
910	In this mode the values for C<interval> and C<at> are both being	1345	In this mode the values for C<interval> and C<at> are both being
911	ignored. Instead, each time the periodic watcher gets scheduled, the	1346	ignored. Instead, each time the periodic watcher gets scheduled, the
912	reschedule callback will be called with the watcher as first, and the	1347	reschedule callback will be called with the watcher as first, and the
913	current time as second argument.	1348	current time as second argument.
914		1349
915	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1350	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
916	ever, or make any event loop modifications>. If you need to stop it,	1351	ever, or make ANY event loop modifications whatsoever>.
917	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
918	starting a prepare watcher).
919		1352
		1353	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1354	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1355	only event loop modification you are allowed to do).
		1356
920	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1357	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic
921	ev_tstamp now)>, e.g.:	1358	*w, ev_tstamp now)>, e.g.:
922		1359
923	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1360	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
924	{	1361	{
925	return now + 60.;	1362	return now + 60.;
926	}	1363	}
…		…
928	It must return the next time to trigger, based on the passed time value	1365	It must return the next time to trigger, based on the passed time value
929	(that is, the lowest time value larger than to the second argument). It	1366	(that is, the lowest time value larger than to the second argument). It
930	will usually be called just before the callback will be triggered, but	1367	will usually be called just before the callback will be triggered, but
931	might be called at other times, too.	1368	might be called at other times, too.
932		1369
933	NOTE: I<< This callback must always return a time that is later than the	1370	NOTE: I<< This callback must always return a time that is higher than or
934	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1371	equal to the passed C<now> value >>.
935		1372
936	This can be used to create very complex timers, such as a timer that	1373	This can be used to create very complex timers, such as a timer that
937	triggers on each midnight, local time. To do this, you would calculate the	1374	triggers on "next midnight, local time". To do this, you would calculate the
938	next midnight after C<now> and return the timestamp value for this. How	1375	next midnight after C<now> and return the timestamp value for this. How
939	you do this is, again, up to you (but it is not trivial, which is the main	1376	you do this is, again, up to you (but it is not trivial, which is the main
940	reason I omitted it as an example).	1377	reason I omitted it as an example).
941		1378
942	=back	1379	=back
…		…
946	Simply stops and restarts the periodic watcher again. This is only useful	1383	Simply stops and restarts the periodic watcher again. This is only useful
947	when you changed some parameters or the reschedule callback would return	1384	when you changed some parameters or the reschedule callback would return
948	a different time than the last time it was called (e.g. in a crond like	1385	a different time than the last time it was called (e.g. in a crond like
949	program when the crontabs have changed).	1386	program when the crontabs have changed).
950		1387
		1388	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1389
		1390	When active, returns the absolute time that the watcher is supposed to
		1391	trigger next.
		1392
		1393	=item ev_tstamp offset [read-write]
		1394
		1395	When repeating, this contains the offset value, otherwise this is the
		1396	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1397
		1398	Can be modified any time, but changes only take effect when the periodic
		1399	timer fires or C<ev_periodic_again> is being called.
		1400
		1401	=item ev_tstamp interval [read-write]
		1402
		1403	The current interval value. Can be modified any time, but changes only
		1404	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1405	called.
		1406
		1407	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]
		1408
		1409	The current reschedule callback, or C<0>, if this functionality is
		1410	switched off. Can be changed any time, but changes only take effect when
		1411	the periodic timer fires or C<ev_periodic_again> is being called.
		1412
951	=back	1413	=back
952		1414
		1415	=head3 Examples
		1416
953	Example: call a callback every hour, or, more precisely, whenever the	1417	Example: Call a callback every hour, or, more precisely, whenever the
954	system clock is divisible by 3600. The callback invocation times have	1418	system clock is divisible by 3600. The callback invocation times have
955	potentially a lot of jittering, but good long-term stability.	1419	potentially a lot of jittering, but good long-term stability.
956		1420
957	static void	1421	static void
958	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1422	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
…		…
962		1426
963	struct ev_periodic hourly_tick;	1427	struct ev_periodic hourly_tick;
964	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1428	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
965	ev_periodic_start (loop, &hourly_tick);	1429	ev_periodic_start (loop, &hourly_tick);
966		1430
967	Example: the same as above, but use a reschedule callback to do it:	1431	Example: The same as above, but use a reschedule callback to do it:
968		1432
969	#include <math.h>	1433	#include <math.h>
970		1434
971	static ev_tstamp	1435	static ev_tstamp
972	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1436	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
…		…
974	return fmod (now, 3600.) + 3600.;	1438	return fmod (now, 3600.) + 3600.;
975	}	1439	}
976		1440
977	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1441	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
978		1442
979	Example: call a callback every hour, starting now:	1443	Example: Call a callback every hour, starting now:
980		1444
981	struct ev_periodic hourly_tick;	1445	struct ev_periodic hourly_tick;
982	ev_periodic_init (&hourly_tick, clock_cb,	1446	ev_periodic_init (&hourly_tick, clock_cb,
983	fmod (ev_now (loop), 3600.), 3600., 0);	1447	fmod (ev_now (loop), 3600.), 3600., 0);
984	ev_periodic_start (loop, &hourly_tick);	1448	ev_periodic_start (loop, &hourly_tick);
985		1449
986		1450
987	=head2 C<ev_signal> - signal me when a signal gets signalled	1451	=head2 C<ev_signal> - signal me when a signal gets signalled!
988		1452
989	Signal watchers will trigger an event when the process receives a specific	1453	Signal watchers will trigger an event when the process receives a specific
990	signal one or more times. Even though signals are very asynchronous, libev	1454	signal one or more times. Even though signals are very asynchronous, libev
991	will try it's best to deliver signals synchronously, i.e. as part of the	1455	will try it's best to deliver signals synchronously, i.e. as part of the
992	normal event processing, like any other event.	1456	normal event processing, like any other event.
…		…
996	with the kernel (thus it coexists with your own signal handlers as long	1460	with the kernel (thus it coexists with your own signal handlers as long
997	as you don't register any with libev). Similarly, when the last signal	1461	as you don't register any with libev). Similarly, when the last signal
998	watcher for a signal is stopped libev will reset the signal handler to	1462	watcher for a signal is stopped libev will reset the signal handler to
999	SIG_DFL (regardless of what it was set to before).	1463	SIG_DFL (regardless of what it was set to before).
1000		1464
		1465	If possible and supported, libev will install its handlers with
		1466	C<SA_RESTART> behaviour enabled, so syscalls should not be unduly
		1467	interrupted. If you have a problem with syscalls getting interrupted by
		1468	signals you can block all signals in an C<ev_check> watcher and unblock
		1469	them in an C<ev_prepare> watcher.
		1470
		1471	=head3 Watcher-Specific Functions and Data Members
		1472
1001	=over 4	1473	=over 4
1002		1474
1003	=item ev_signal_init (ev_signal *, callback, int signum)	1475	=item ev_signal_init (ev_signal *, callback, int signum)
1004		1476
1005	=item ev_signal_set (ev_signal *, int signum)	1477	=item ev_signal_set (ev_signal *, int signum)
1006		1478
1007	Configures the watcher to trigger on the given signal number (usually one	1479	Configures the watcher to trigger on the given signal number (usually one
1008	of the C<SIGxxx> constants).	1480	of the C<SIGxxx> constants).
1009		1481
		1482	=item int signum [read-only]
		1483
		1484	The signal the watcher watches out for.
		1485
1010	=back	1486	=back
1011		1487
		1488	=head3 Examples
1012		1489
		1490	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1491
		1492	static void
		1493	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1494	{
		1495	ev_unloop (loop, EVUNLOOP_ALL);
		1496	}
		1497
		1498	struct ev_signal signal_watcher;
		1499	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1500	ev_signal_start (loop, &sigint_cb);
		1501
		1502
1013	=head2 C<ev_child> - wait for pid status changes	1503	=head2 C<ev_child> - watch out for process status changes
1014		1504
1015	Child watchers trigger when your process receives a SIGCHLD in response to	1505	Child watchers trigger when your process receives a SIGCHLD in response to
1016	some child status changes (most typically when a child of yours dies).	1506	some child status changes (most typically when a child of yours dies). It
		1507	is permissible to install a child watcher I<after> the child has been
		1508	forked (which implies it might have already exited), as long as the event
		1509	loop isn't entered (or is continued from a watcher).
		1510
		1511	Only the default event loop is capable of handling signals, and therefore
		1512	you can only rgeister child watchers in the default event loop.
		1513
		1514	=head3 Process Interaction
		1515
		1516	Libev grabs C<SIGCHLD> as soon as the default event loop is
		1517	initialised. This is necessary to guarantee proper behaviour even if
		1518	the first child watcher is started after the child exits. The occurance
		1519	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		1520	synchronously as part of the event loop processing. Libev always reaps all
		1521	children, even ones not watched.
		1522
		1523	=head3 Overriding the Built-In Processing
		1524
		1525	Libev offers no special support for overriding the built-in child
		1526	processing, but if your application collides with libev's default child
		1527	handler, you can override it easily by installing your own handler for
		1528	C<SIGCHLD> after initialising the default loop, and making sure the
		1529	default loop never gets destroyed. You are encouraged, however, to use an
		1530	event-based approach to child reaping and thus use libev's support for
		1531	that, so other libev users can use C<ev_child> watchers freely.
		1532
		1533	=head3 Watcher-Specific Functions and Data Members
1017		1534
1018	=over 4	1535	=over 4
1019		1536
1020	=item ev_child_init (ev_child *, callback, int pid)	1537	=item ev_child_init (ev_child *, callback, int pid, int trace)
1021		1538
1022	=item ev_child_set (ev_child *, int pid)	1539	=item ev_child_set (ev_child *, int pid, int trace)
1023		1540
1024	Configures the watcher to wait for status changes of process C<pid> (or	1541	Configures the watcher to wait for status changes of process C<pid> (or
1025	I<any> process if C<pid> is specified as C<0>). The callback can look	1542	I<any> process if C<pid> is specified as C<0>). The callback can look
1026	at the C<rstatus> member of the C<ev_child> watcher structure to see	1543	at the C<rstatus> member of the C<ev_child> watcher structure to see
1027	the status word (use the macros from C<sys/wait.h> and see your systems	1544	the status word (use the macros from C<sys/wait.h> and see your systems
1028	C<waitpid> documentation). The C<rpid> member contains the pid of the	1545	C<waitpid> documentation). The C<rpid> member contains the pid of the
1029	process causing the status change.	1546	process causing the status change. C<trace> must be either C<0> (only
		1547	activate the watcher when the process terminates) or C<1> (additionally
		1548	activate the watcher when the process is stopped or continued).
		1549
		1550	=item int pid [read-only]
		1551
		1552	The process id this watcher watches out for, or C<0>, meaning any process id.
		1553
		1554	=item int rpid [read-write]
		1555
		1556	The process id that detected a status change.
		1557
		1558	=item int rstatus [read-write]
		1559
		1560	The process exit/trace status caused by C<rpid> (see your systems
		1561	C<waitpid> and C<sys/wait.h> documentation for details).
1030		1562
1031	=back	1563	=back
1032		1564
1033	Example: try to exit cleanly on SIGINT and SIGTERM.	1565	=head3 Examples
		1566
		1567	Example: C<fork()> a new process and install a child handler to wait for
		1568	its completion.
		1569
		1570	ev_child cw;
1034		1571
1035	static void	1572	static void
1036	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	1573	child_cb (EV_P_ struct ev_child *w, int revents)
1037	{	1574	{
1038	ev_unloop (loop, EVUNLOOP_ALL);	1575	ev_child_stop (EV_A_ w);
		1576	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1039	}	1577	}
1040		1578
1041	struct ev_signal signal_watcher;	1579	pid_t pid = fork ();
1042	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1043	ev_signal_start (loop, &sigint_cb);
1044		1580
		1581	if (pid < 0)
		1582	// error
		1583	else if (pid == 0)
		1584	{
		1585	// the forked child executes here
		1586	exit (1);
		1587	}
		1588	else
		1589	{
		1590	ev_child_init (&cw, child_cb, pid, 0);
		1591	ev_child_start (EV_DEFAULT_ &cw);
		1592	}
1045		1593
		1594
		1595	=head2 C<ev_stat> - did the file attributes just change?
		1596
		1597	This watches a filesystem path for attribute changes. That is, it calls
		1598	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1599	compared to the last time, invoking the callback if it did.
		1600
		1601	The path does not need to exist: changing from "path exists" to "path does
		1602	not exist" is a status change like any other. The condition "path does
		1603	not exist" is signified by the C<st_nlink> field being zero (which is
		1604	otherwise always forced to be at least one) and all the other fields of
		1605	the stat buffer having unspecified contents.
		1606
		1607	The path I<should> be absolute and I<must not> end in a slash. If it is
		1608	relative and your working directory changes, the behaviour is undefined.
		1609
		1610	Since there is no standard to do this, the portable implementation simply
		1611	calls C<stat (2)> regularly on the path to see if it changed somehow. You
		1612	can specify a recommended polling interval for this case. If you specify
		1613	a polling interval of C<0> (highly recommended!) then a I<suitable,
		1614	unspecified default> value will be used (which you can expect to be around
		1615	five seconds, although this might change dynamically). Libev will also
		1616	impose a minimum interval which is currently around C<0.1>, but thats
		1617	usually overkill.
		1618
		1619	This watcher type is not meant for massive numbers of stat watchers,
		1620	as even with OS-supported change notifications, this can be
		1621	resource-intensive.
		1622
		1623	At the time of this writing, only the Linux inotify interface is
		1624	implemented (implementing kqueue support is left as an exercise for the
		1625	reader, note, however, that the author sees no way of implementing ev_stat
		1626	semantics with kqueue). Inotify will be used to give hints only and should
		1627	not change the semantics of C<ev_stat> watchers, which means that libev
		1628	sometimes needs to fall back to regular polling again even with inotify,
		1629	but changes are usually detected immediately, and if the file exists there
		1630	will be no polling.
		1631
		1632	=head3 ABI Issues (Largefile Support)
		1633
		1634	Libev by default (unless the user overrides this) uses the default
		1635	compilation environment, which means that on systems with optionally
		1636	disabled large file support, you get the 32 bit version of the stat
		1637	structure. When using the library from programs that change the ABI to
		1638	use 64 bit file offsets the programs will fail. In that case you have to
		1639	compile libev with the same flags to get binary compatibility. This is
		1640	obviously the case with any flags that change the ABI, but the problem is
		1641	most noticably with ev_stat and largefile support.
		1642
		1643	=head3 Inotify
		1644
		1645	When C<inotify (7)> support has been compiled into libev (generally only
		1646	available on Linux) and present at runtime, it will be used to speed up
		1647	change detection where possible. The inotify descriptor will be created lazily
		1648	when the first C<ev_stat> watcher is being started.
		1649
		1650	Inotify presence does not change the semantics of C<ev_stat> watchers
		1651	except that changes might be detected earlier, and in some cases, to avoid
		1652	making regular C<stat> calls. Even in the presence of inotify support
		1653	there are many cases where libev has to resort to regular C<stat> polling.
		1654
		1655	(There is no support for kqueue, as apparently it cannot be used to
		1656	implement this functionality, due to the requirement of having a file
		1657	descriptor open on the object at all times).
		1658
		1659	=head3 The special problem of stat time resolution
		1660
		1661	The C<stat ()> syscall only supports full-second resolution portably, and
		1662	even on systems where the resolution is higher, many filesystems still
		1663	only support whole seconds.
		1664
		1665	That means that, if the time is the only thing that changes, you can
		1666	easily miss updates: on the first update, C<ev_stat> detects a change and
		1667	calls your callback, which does something. When there is another update
		1668	within the same second, C<ev_stat> will be unable to detect it as the stat
		1669	data does not change.
		1670
		1671	The solution to this is to delay acting on a change for slightly more
		1672	than a second (or till slightly after the next full second boundary), using
		1673	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		1674	ev_timer_again (loop, w)>).
		1675
		1676	The C<.02> offset is added to work around small timing inconsistencies
		1677	of some operating systems (where the second counter of the current time
		1678	might be be delayed. One such system is the Linux kernel, where a call to
		1679	C<gettimeofday> might return a timestamp with a full second later than
		1680	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		1681	update file times then there will be a small window where the kernel uses
		1682	the previous second to update file times but libev might already execute
		1683	the timer callback).
		1684
		1685	=head3 Watcher-Specific Functions and Data Members
		1686
		1687	=over 4
		1688
		1689	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		1690
		1691	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		1692
		1693	Configures the watcher to wait for status changes of the given
		1694	C<path>. The C<interval> is a hint on how quickly a change is expected to
		1695	be detected and should normally be specified as C<0> to let libev choose
		1696	a suitable value. The memory pointed to by C<path> must point to the same
		1697	path for as long as the watcher is active.
		1698
		1699	The callback will receive C<EV_STAT> when a change was detected, relative
		1700	to the attributes at the time the watcher was started (or the last change
		1701	was detected).
		1702
		1703	=item ev_stat_stat (loop, ev_stat *)
		1704
		1705	Updates the stat buffer immediately with new values. If you change the
		1706	watched path in your callback, you could call this function to avoid
		1707	detecting this change (while introducing a race condition if you are not
		1708	the only one changing the path). Can also be useful simply to find out the
		1709	new values.
		1710
		1711	=item ev_statdata attr [read-only]
		1712
		1713	The most-recently detected attributes of the file. Although the type is
		1714	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		1715	suitable for your system, but you can only rely on the POSIX-standardised
		1716	members to be present. If the C<st_nlink> member is C<0>, then there was
		1717	some error while C<stat>ing the file.
		1718
		1719	=item ev_statdata prev [read-only]
		1720
		1721	The previous attributes of the file. The callback gets invoked whenever
		1722	C<prev> != C<attr>, or, more precisely, one or more of these members
		1723	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		1724	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
		1725
		1726	=item ev_tstamp interval [read-only]
		1727
		1728	The specified interval.
		1729
		1730	=item const char *path [read-only]
		1731
		1732	The filesystem path that is being watched.
		1733
		1734	=back
		1735
		1736	=head3 Examples
		1737
		1738	Example: Watch C</etc/passwd> for attribute changes.
		1739
		1740	static void
		1741	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		1742	{
		1743	/* /etc/passwd changed in some way */
		1744	if (w->attr.st_nlink)
		1745	{
		1746	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1747	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1748	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1749	}
		1750	else
		1751	/* you shalt not abuse printf for puts */
		1752	puts ("wow, /etc/passwd is not there, expect problems. "
		1753	"if this is windows, they already arrived\n");
		1754	}
		1755
		1756	...
		1757	ev_stat passwd;
		1758
		1759	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
		1760	ev_stat_start (loop, &passwd);
		1761
		1762	Example: Like above, but additionally use a one-second delay so we do not
		1763	miss updates (however, frequent updates will delay processing, too, so
		1764	one might do the work both on C<ev_stat> callback invocation I<and> on
		1765	C<ev_timer> callback invocation).
		1766
		1767	static ev_stat passwd;
		1768	static ev_timer timer;
		1769
		1770	static void
		1771	timer_cb (EV_P_ ev_timer *w, int revents)
		1772	{
		1773	ev_timer_stop (EV_A_ w);
		1774
		1775	/* now it's one second after the most recent passwd change */
		1776	}
		1777
		1778	static void
		1779	stat_cb (EV_P_ ev_stat *w, int revents)
		1780	{
		1781	/* reset the one-second timer */
		1782	ev_timer_again (EV_A_ &timer);
		1783	}
		1784
		1785	...
		1786	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		1787	ev_stat_start (loop, &passwd);
		1788	ev_timer_init (&timer, timer_cb, 0., 1.02);
		1789
		1790
1046	=head2 C<ev_idle> - when you've got nothing better to do	1791	=head2 C<ev_idle> - when you've got nothing better to do...
1047		1792
1048	Idle watchers trigger events when there are no other events are pending	1793	Idle watchers trigger events when no other events of the same or higher
1049	(prepare, check and other idle watchers do not count). That is, as long	1794	priority are pending (prepare, check and other idle watchers do not
1050	as your process is busy handling sockets or timeouts (or even signals,	1795	count).
1051	imagine) it will not be triggered. But when your process is idle all idle	1796
1052	watchers are being called again and again, once per event loop iteration -	1797	That is, as long as your process is busy handling sockets or timeouts
		1798	(or even signals, imagine) of the same or higher priority it will not be
		1799	triggered. But when your process is idle (or only lower-priority watchers
		1800	are pending), the idle watchers are being called once per event loop
1053	until stopped, that is, or your process receives more events and becomes	1801	iteration - until stopped, that is, or your process receives more events
1054	busy.	1802	and becomes busy again with higher priority stuff.
1055		1803
1056	The most noteworthy effect is that as long as any idle watchers are	1804	The most noteworthy effect is that as long as any idle watchers are
1057	active, the process will not block when waiting for new events.	1805	active, the process will not block when waiting for new events.
1058		1806
1059	Apart from keeping your process non-blocking (which is a useful	1807	Apart from keeping your process non-blocking (which is a useful
1060	effect on its own sometimes), idle watchers are a good place to do	1808	effect on its own sometimes), idle watchers are a good place to do
1061	"pseudo-background processing", or delay processing stuff to after the	1809	"pseudo-background processing", or delay processing stuff to after the
1062	event loop has handled all outstanding events.	1810	event loop has handled all outstanding events.
1063		1811
		1812	=head3 Watcher-Specific Functions and Data Members
		1813
1064	=over 4	1814	=over 4
1065		1815
1066	=item ev_idle_init (ev_signal *, callback)	1816	=item ev_idle_init (ev_signal *, callback)
1067		1817
1068	Initialises and configures the idle watcher - it has no parameters of any	1818	Initialises and configures the idle watcher - it has no parameters of any
1069	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1819	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1070	believe me.	1820	believe me.
1071		1821
1072	=back	1822	=back
1073		1823
		1824	=head3 Examples
		1825
1074	Example: dynamically allocate an C<ev_idle>, start it, and in the	1826	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1075	callback, free it. Alos, use no error checking, as usual.	1827	callback, free it. Also, use no error checking, as usual.
1076		1828
1077	static void	1829	static void
1078	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	1830	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1079	{	1831	{
1080	free (w);	1832	free (w);
1081	// now do something you wanted to do when the program has	1833	// now do something you wanted to do when the program has
1082	// no longer asnything immediate to do.	1834	// no longer anything immediate to do.
1083	}	1835	}
1084		1836
1085	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	1837	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1086	ev_idle_init (idle_watcher, idle_cb);	1838	ev_idle_init (idle_watcher, idle_cb);
1087	ev_idle_start (loop, idle_cb);	1839	ev_idle_start (loop, idle_cb);
1088		1840
1089		1841
1090	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1842	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1091		1843
1092	Prepare and check watchers are usually (but not always) used in tandem:	1844	Prepare and check watchers are usually (but not always) used in tandem:
1093	prepare watchers get invoked before the process blocks and check watchers	1845	prepare watchers get invoked before the process blocks and check watchers
1094	afterwards.	1846	afterwards.
1095		1847
		1848	You I<must not> call C<ev_loop> or similar functions that enter
		1849	the current event loop from either C<ev_prepare> or C<ev_check>
		1850	watchers. Other loops than the current one are fine, however. The
		1851	rationale behind this is that you do not need to check for recursion in
		1852	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		1853	C<ev_check> so if you have one watcher of each kind they will always be
		1854	called in pairs bracketing the blocking call.
		1855
1096	Their main purpose is to integrate other event mechanisms into libev and	1856	Their main purpose is to integrate other event mechanisms into libev and
1097	their use is somewhat advanced. This could be used, for example, to track	1857	their use is somewhat advanced. This could be used, for example, to track
1098	variable changes, implement your own watchers, integrate net-snmp or a	1858	variable changes, implement your own watchers, integrate net-snmp or a
1099	coroutine library and lots more.	1859	coroutine library and lots more. They are also occasionally useful if
		1860	you cache some data and want to flush it before blocking (for example,
		1861	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		1862	watcher).
1100		1863
1101	This is done by examining in each prepare call which file descriptors need	1864	This is done by examining in each prepare call which file descriptors need
1102	to be watched by the other library, registering C<ev_io> watchers for	1865	to be watched by the other library, registering C<ev_io> watchers for
1103	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1866	them and starting an C<ev_timer> watcher for any timeouts (many libraries
1104	provide just this functionality). Then, in the check watcher you check for	1867	provide just this functionality). Then, in the check watcher you check for
…		…
1114	with priority higher than or equal to the event loop and one coroutine	1877	with priority higher than or equal to the event loop and one coroutine
1115	of lower priority, but only once, using idle watchers to keep the event	1878	of lower priority, but only once, using idle watchers to keep the event
1116	loop from blocking if lower-priority coroutines are active, thus mapping	1879	loop from blocking if lower-priority coroutines are active, thus mapping
1117	low-priority coroutines to idle/background tasks).	1880	low-priority coroutines to idle/background tasks).
1118		1881
		1882	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1883	priority, to ensure that they are being run before any other watchers
		1884	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1885	too) should not activate ("feed") events into libev. While libev fully
		1886	supports this, they might get executed before other C<ev_check> watchers
		1887	did their job. As C<ev_check> watchers are often used to embed other
		1888	(non-libev) event loops those other event loops might be in an unusable
		1889	state until their C<ev_check> watcher ran (always remind yourself to
		1890	coexist peacefully with others).
		1891
		1892	=head3 Watcher-Specific Functions and Data Members
		1893
1119	=over 4	1894	=over 4
1120		1895
1121	=item ev_prepare_init (ev_prepare *, callback)	1896	=item ev_prepare_init (ev_prepare *, callback)
1122		1897
1123	=item ev_check_init (ev_check *, callback)	1898	=item ev_check_init (ev_check *, callback)
…		…
1126	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1901	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1127	macros, but using them is utterly, utterly and completely pointless.	1902	macros, but using them is utterly, utterly and completely pointless.
1128		1903
1129	=back	1904	=back
1130		1905
1131	Example: *TODO*.	1906	=head3 Examples
1132		1907
		1908	There are a number of principal ways to embed other event loops or modules
		1909	into libev. Here are some ideas on how to include libadns into libev
		1910	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1911	use as a working example. Another Perl module named C<EV::Glib> embeds a
		1912	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		1913	Glib event loop).
1133		1914
		1915	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		1916	and in a check watcher, destroy them and call into libadns. What follows
		1917	is pseudo-code only of course. This requires you to either use a low
		1918	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1919	the callbacks for the IO/timeout watchers might not have been called yet.
		1920
		1921	static ev_io iow [nfd];
		1922	static ev_timer tw;
		1923
		1924	static void
		1925	io_cb (ev_loop *loop, ev_io *w, int revents)
		1926	{
		1927	}
		1928
		1929	// create io watchers for each fd and a timer before blocking
		1930	static void
		1931	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
		1932	{
		1933	int timeout = 3600000;
		1934	struct pollfd fds [nfd];
		1935	// actual code will need to loop here and realloc etc.
		1936	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1937
		1938	/* the callback is illegal, but won't be called as we stop during check */
		1939	ev_timer_init (&tw, 0, timeout * 1e-3);
		1940	ev_timer_start (loop, &tw);
		1941
		1942	// create one ev_io per pollfd
		1943	for (int i = 0; i < nfd; ++i)
		1944	{
		1945	ev_io_init (iow + i, io_cb, fds [i].fd,
		1946	((fds [i].events & POLLIN ? EV_READ : 0)
		1947	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1948
		1949	fds [i].revents = 0;
		1950	ev_io_start (loop, iow + i);
		1951	}
		1952	}
		1953
		1954	// stop all watchers after blocking
		1955	static void
		1956	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
		1957	{
		1958	ev_timer_stop (loop, &tw);
		1959
		1960	for (int i = 0; i < nfd; ++i)
		1961	{
		1962	// set the relevant poll flags
		1963	// could also call adns_processreadable etc. here
		1964	struct pollfd *fd = fds + i;
		1965	int revents = ev_clear_pending (iow + i);
		1966	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1967	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1968
		1969	// now stop the watcher
		1970	ev_io_stop (loop, iow + i);
		1971	}
		1972
		1973	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1974	}
		1975
		1976	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		1977	in the prepare watcher and would dispose of the check watcher.
		1978
		1979	Method 3: If the module to be embedded supports explicit event
		1980	notification (adns does), you can also make use of the actual watcher
		1981	callbacks, and only destroy/create the watchers in the prepare watcher.
		1982
		1983	static void
		1984	timer_cb (EV_P_ ev_timer *w, int revents)
		1985	{
		1986	adns_state ads = (adns_state)w->data;
		1987	update_now (EV_A);
		1988
		1989	adns_processtimeouts (ads, &tv_now);
		1990	}
		1991
		1992	static void
		1993	io_cb (EV_P_ ev_io *w, int revents)
		1994	{
		1995	adns_state ads = (adns_state)w->data;
		1996	update_now (EV_A);
		1997
		1998	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		1999	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		2000	}
		2001
		2002	// do not ever call adns_afterpoll
		2003
		2004	Method 4: Do not use a prepare or check watcher because the module you
		2005	want to embed is too inflexible to support it. Instead, youc na override
		2006	their poll function. The drawback with this solution is that the main
		2007	loop is now no longer controllable by EV. The C<Glib::EV> module does
		2008	this.
		2009
		2010	static gint
		2011	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		2012	{
		2013	int got_events = 0;
		2014
		2015	for (n = 0; n < nfds; ++n)
		2016	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		2017
		2018	if (timeout >= 0)
		2019	// create/start timer
		2020
		2021	// poll
		2022	ev_loop (EV_A_ 0);
		2023
		2024	// stop timer again
		2025	if (timeout >= 0)
		2026	ev_timer_stop (EV_A_ &to);
		2027
		2028	// stop io watchers again - their callbacks should have set
		2029	for (n = 0; n < nfds; ++n)
		2030	ev_io_stop (EV_A_ iow [n]);
		2031
		2032	return got_events;
		2033	}
		2034
		2035
1134	=head2 C<ev_embed> - when one backend isn't enough	2036	=head2 C<ev_embed> - when one backend isn't enough...
1135		2037
1136	This is a rather advanced watcher type that lets you embed one event loop	2038	This is a rather advanced watcher type that lets you embed one event loop
1137	into another (currently only C<ev_io> events are supported in the embedded	2039	into another (currently only C<ev_io> events are supported in the embedded
1138	loop, other types of watchers might be handled in a delayed or incorrect	2040	loop, other types of watchers might be handled in a delayed or incorrect
1139	fashion and must not be used).	2041	fashion and must not be used).
…		…
1178	portable one.	2080	portable one.
1179		2081
1180	So when you want to use this feature you will always have to be prepared	2082	So when you want to use this feature you will always have to be prepared
1181	that you cannot get an embeddable loop. The recommended way to get around	2083	that you cannot get an embeddable loop. The recommended way to get around
1182	this is to have a separate variables for your embeddable loop, try to	2084	this is to have a separate variables for your embeddable loop, try to
1183	create it, and if that fails, use the normal loop for everything:	2085	create it, and if that fails, use the normal loop for everything.
		2086
		2087	=head3 Watcher-Specific Functions and Data Members
		2088
		2089	=over 4
		2090
		2091	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		2092
		2093	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		2094
		2095	Configures the watcher to embed the given loop, which must be
		2096	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		2097	invoked automatically, otherwise it is the responsibility of the callback
		2098	to invoke it (it will continue to be called until the sweep has been done,
		2099	if you do not want thta, you need to temporarily stop the embed watcher).
		2100
		2101	=item ev_embed_sweep (loop, ev_embed *)
		2102
		2103	Make a single, non-blocking sweep over the embedded loop. This works
		2104	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		2105	apropriate way for embedded loops.
		2106
		2107	=item struct ev_loop *other [read-only]
		2108
		2109	The embedded event loop.
		2110
		2111	=back
		2112
		2113	=head3 Examples
		2114
		2115	Example: Try to get an embeddable event loop and embed it into the default
		2116	event loop. If that is not possible, use the default loop. The default
		2117	loop is stored in C<loop_hi>, while the mebeddable loop is stored in
		2118	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be
		2119	used).
1184		2120
1185	struct ev_loop *loop_hi = ev_default_init (0);	2121	struct ev_loop *loop_hi = ev_default_init (0);
1186	struct ev_loop *loop_lo = 0;	2122	struct ev_loop *loop_lo = 0;
1187	struct ev_embed embed;	2123	struct ev_embed embed;
1188		2124
…		…
1199	ev_embed_start (loop_hi, &embed);	2135	ev_embed_start (loop_hi, &embed);
1200	}	2136	}
1201	else	2137	else
1202	loop_lo = loop_hi;	2138	loop_lo = loop_hi;
1203		2139
		2140	Example: Check if kqueue is available but not recommended and create
		2141	a kqueue backend for use with sockets (which usually work with any
		2142	kqueue implementation). Store the kqueue/socket-only event loop in
		2143	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		2144
		2145	struct ev_loop *loop = ev_default_init (0);
		2146	struct ev_loop *loop_socket = 0;
		2147	struct ev_embed embed;
		2148
		2149	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2150	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2151	{
		2152	ev_embed_init (&embed, 0, loop_socket);
		2153	ev_embed_start (loop, &embed);
		2154	}
		2155
		2156	if (!loop_socket)
		2157	loop_socket = loop;
		2158
		2159	// now use loop_socket for all sockets, and loop for everything else
		2160
		2161
		2162	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		2163
		2164	Fork watchers are called when a C<fork ()> was detected (usually because
		2165	whoever is a good citizen cared to tell libev about it by calling
		2166	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		2167	event loop blocks next and before C<ev_check> watchers are being called,
		2168	and only in the child after the fork. If whoever good citizen calling
		2169	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		2170	handlers will be invoked, too, of course.
		2171
		2172	=head3 Watcher-Specific Functions and Data Members
		2173
1204	=over 4	2174	=over 4
1205		2175
1206	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2176	=item ev_fork_init (ev_signal *, callback)
1207		2177
1208	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	2178	Initialises and configures the fork watcher - it has no parameters of any
		2179	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		2180	believe me.
1209		2181
1210	Configures the watcher to embed the given loop, which must be	2182	=back
1211	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1212	invoked automatically, otherwise it is the responsibility of the callback
1213	to invoke it (it will continue to be called until the sweep has been done,
1214	if you do not want thta, you need to temporarily stop the embed watcher).
1215		2183
1216	=item ev_embed_sweep (loop, ev_embed *)
1217		2184
1218	Make a single, non-blocking sweep over the embedded loop. This works	2185	=head2 C<ev_async> - how to wake up another event loop
1219	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	2186
1220	apropriate way for embedded loops.	2187	In general, you cannot use an C<ev_loop> from multiple threads or other
		2188	asynchronous sources such as signal handlers (as opposed to multiple event
		2189	loops - those are of course safe to use in different threads).
		2190
		2191	Sometimes, however, you need to wake up another event loop you do not
		2192	control, for example because it belongs to another thread. This is what
		2193	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		2194	can signal it by calling C<ev_async_send>, which is thread- and signal
		2195	safe.
		2196
		2197	This functionality is very similar to C<ev_signal> watchers, as signals,
		2198	too, are asynchronous in nature, and signals, too, will be compressed
		2199	(i.e. the number of callback invocations may be less than the number of
		2200	C<ev_async_sent> calls).
		2201
		2202	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		2203	just the default loop.
		2204
		2205	=head3 Queueing
		2206
		2207	C<ev_async> does not support queueing of data in any way. The reason
		2208	is that the author does not know of a simple (or any) algorithm for a
		2209	multiple-writer-single-reader queue that works in all cases and doesn't
		2210	need elaborate support such as pthreads.
		2211
		2212	That means that if you want to queue data, you have to provide your own
		2213	queue. But at least I can tell you would implement locking around your
		2214	queue:
		2215
		2216	=over 4
		2217
		2218	=item queueing from a signal handler context
		2219
		2220	To implement race-free queueing, you simply add to the queue in the signal
		2221	handler but you block the signal handler in the watcher callback. Here is an example that does that for
		2222	some fictitiuous SIGUSR1 handler:
		2223
		2224	static ev_async mysig;
		2225
		2226	static void
		2227	sigusr1_handler (void)
		2228	{
		2229	sometype data;
		2230
		2231	// no locking etc.
		2232	queue_put (data);
		2233	ev_async_send (EV_DEFAULT_ &mysig);
		2234	}
		2235
		2236	static void
		2237	mysig_cb (EV_P_ ev_async *w, int revents)
		2238	{
		2239	sometype data;
		2240	sigset_t block, prev;
		2241
		2242	sigemptyset (&block);
		2243	sigaddset (&block, SIGUSR1);
		2244	sigprocmask (SIG_BLOCK, &block, &prev);
		2245
		2246	while (queue_get (&data))
		2247	process (data);
		2248
		2249	if (sigismember (&prev, SIGUSR1)
		2250	sigprocmask (SIG_UNBLOCK, &block, 0);
		2251	}
		2252
		2253	(Note: pthreads in theory requires you to use C<pthread_setmask>
		2254	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		2255	either...).
		2256
		2257	=item queueing from a thread context
		2258
		2259	The strategy for threads is different, as you cannot (easily) block
		2260	threads but you can easily preempt them, so to queue safely you need to
		2261	employ a traditional mutex lock, such as in this pthread example:
		2262
		2263	static ev_async mysig;
		2264	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		2265
		2266	static void
		2267	otherthread (void)
		2268	{
		2269	// only need to lock the actual queueing operation
		2270	pthread_mutex_lock (&mymutex);
		2271	queue_put (data);
		2272	pthread_mutex_unlock (&mymutex);
		2273
		2274	ev_async_send (EV_DEFAULT_ &mysig);
		2275	}
		2276
		2277	static void
		2278	mysig_cb (EV_P_ ev_async *w, int revents)
		2279	{
		2280	pthread_mutex_lock (&mymutex);
		2281
		2282	while (queue_get (&data))
		2283	process (data);
		2284
		2285	pthread_mutex_unlock (&mymutex);
		2286	}
		2287
		2288	=back
		2289
		2290
		2291	=head3 Watcher-Specific Functions and Data Members
		2292
		2293	=over 4
		2294
		2295	=item ev_async_init (ev_async *, callback)
		2296
		2297	Initialises and configures the async watcher - it has no parameters of any
		2298	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,
		2299	believe me.
		2300
		2301	=item ev_async_send (loop, ev_async *)
		2302
		2303	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		2304	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		2305	C<ev_feed_event>, this call is safe to do in other threads, signal or
		2306	similar contexts (see the dicusssion of C<EV_ATOMIC_T> in the embedding
		2307	section below on what exactly this means).
		2308
		2309	This call incurs the overhead of a syscall only once per loop iteration,
		2310	so while the overhead might be noticable, it doesn't apply to repeated
		2311	calls to C<ev_async_send>.
		2312
		2313	=item bool = ev_async_pending (ev_async *)
		2314
		2315	Returns a non-zero value when C<ev_async_send> has been called on the
		2316	watcher but the event has not yet been processed (or even noted) by the
		2317	event loop.
		2318
		2319	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2320	the loop iterates next and checks for the watcher to have become active,
		2321	it will reset the flag again. C<ev_async_pending> can be used to very
		2322	quickly check wether invoking the loop might be a good idea.
		2323
		2324	Not that this does I<not> check wether the watcher itself is pending, only
		2325	wether it has been requested to make this watcher pending.
1221		2326
1222	=back	2327	=back
1223		2328
1224		2329
1225	=head1 OTHER FUNCTIONS	2330	=head1 OTHER FUNCTIONS
…		…
1297		2402
1298	=item * Priorities are not currently supported. Initialising priorities	2403	=item * Priorities are not currently supported. Initialising priorities
1299	will fail and all watchers will have the same priority, even though there	2404	will fail and all watchers will have the same priority, even though there
1300	is an ev_pri field.	2405	is an ev_pri field.
1301		2406
		2407	=item * In libevent, the last base created gets the signals, in libev, the
		2408	first base created (== the default loop) gets the signals.
		2409
1302	=item * Other members are not supported.	2410	=item * Other members are not supported.
1303		2411
1304	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2412	=item * The libev emulation is I<not> ABI compatible to libevent, you need
1305	to use the libev header file and library.	2413	to use the libev header file and library.
1306		2414
1307	=back	2415	=back
1308		2416
1309	=head1 C++ SUPPORT	2417	=head1 C++ SUPPORT
1310		2418
1311	TBD.	2419	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		2420	you to use some convinience methods to start/stop watchers and also change
		2421	the callback model to a model using method callbacks on objects.
		2422
		2423	To use it,
		2424
		2425	#include <ev++.h>
		2426
		2427	This automatically includes F<ev.h> and puts all of its definitions (many
		2428	of them macros) into the global namespace. All C++ specific things are
		2429	put into the C<ev> namespace. It should support all the same embedding
		2430	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
		2431
		2432	Care has been taken to keep the overhead low. The only data member the C++
		2433	classes add (compared to plain C-style watchers) is the event loop pointer
		2434	that the watcher is associated with (or no additional members at all if
		2435	you disable C<EV_MULTIPLICITY> when embedding libev).
		2436
		2437	Currently, functions, and static and non-static member functions can be
		2438	used as callbacks. Other types should be easy to add as long as they only
		2439	need one additional pointer for context. If you need support for other
		2440	types of functors please contact the author (preferably after implementing
		2441	it).
		2442
		2443	Here is a list of things available in the C<ev> namespace:
		2444
		2445	=over 4
		2446
		2447	=item C<ev::READ>, C<ev::WRITE> etc.
		2448
		2449	These are just enum values with the same values as the C<EV_READ> etc.
		2450	macros from F<ev.h>.
		2451
		2452	=item C<ev::tstamp>, C<ev::now>
		2453
		2454	Aliases to the same types/functions as with the C<ev_> prefix.
		2455
		2456	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		2457
		2458	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		2459	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		2460	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		2461	defines by many implementations.
		2462
		2463	All of those classes have these methods:
		2464
		2465	=over 4
		2466
		2467	=item ev::TYPE::TYPE ()
		2468
		2469	=item ev::TYPE::TYPE (struct ev_loop *)
		2470
		2471	=item ev::TYPE::~TYPE
		2472
		2473	The constructor (optionally) takes an event loop to associate the watcher
		2474	with. If it is omitted, it will use C<EV_DEFAULT>.
		2475
		2476	The constructor calls C<ev_init> for you, which means you have to call the
		2477	C<set> method before starting it.
		2478
		2479	It will not set a callback, however: You have to call the templated C<set>
		2480	method to set a callback before you can start the watcher.
		2481
		2482	(The reason why you have to use a method is a limitation in C++ which does
		2483	not allow explicit template arguments for constructors).
		2484
		2485	The destructor automatically stops the watcher if it is active.
		2486
		2487	=item w->set<class, &class::method> (object *)
		2488
		2489	This method sets the callback method to call. The method has to have a
		2490	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		2491	first argument and the C<revents> as second. The object must be given as
		2492	parameter and is stored in the C<data> member of the watcher.
		2493
		2494	This method synthesizes efficient thunking code to call your method from
		2495	the C callback that libev requires. If your compiler can inline your
		2496	callback (i.e. it is visible to it at the place of the C<set> call and
		2497	your compiler is good :), then the method will be fully inlined into the
		2498	thunking function, making it as fast as a direct C callback.
		2499
		2500	Example: simple class declaration and watcher initialisation
		2501
		2502	struct myclass
		2503	{
		2504	void io_cb (ev::io &w, int revents) { }
		2505	}
		2506
		2507	myclass obj;
		2508	ev::io iow;
		2509	iow.set <myclass, &myclass::io_cb> (&obj);
		2510
		2511	=item w->set<function> (void *data = 0)
		2512
		2513	Also sets a callback, but uses a static method or plain function as
		2514	callback. The optional C<data> argument will be stored in the watcher's
		2515	C<data> member and is free for you to use.
		2516
		2517	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2518
		2519	See the method-C<set> above for more details.
		2520
		2521	Example:
		2522
		2523	static void io_cb (ev::io &w, int revents) { }
		2524	iow.set <io_cb> ();
		2525
		2526	=item w->set (struct ev_loop *)
		2527
		2528	Associates a different C<struct ev_loop> with this watcher. You can only
		2529	do this when the watcher is inactive (and not pending either).
		2530
		2531	=item w->set ([args])
		2532
		2533	Basically the same as C<ev_TYPE_set>, with the same args. Must be
		2534	called at least once. Unlike the C counterpart, an active watcher gets
		2535	automatically stopped and restarted when reconfiguring it with this
		2536	method.
		2537
		2538	=item w->start ()
		2539
		2540	Starts the watcher. Note that there is no C<loop> argument, as the
		2541	constructor already stores the event loop.
		2542
		2543	=item w->stop ()
		2544
		2545	Stops the watcher if it is active. Again, no C<loop> argument.
		2546
		2547	=item w->again () (C<ev::timer>, C<ev::periodic> only)
		2548
		2549	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		2550	C<ev_TYPE_again> function.
		2551
		2552	=item w->sweep () (C<ev::embed> only)
		2553
		2554	Invokes C<ev_embed_sweep>.
		2555
		2556	=item w->update () (C<ev::stat> only)
		2557
		2558	Invokes C<ev_stat_stat>.
		2559
		2560	=back
		2561
		2562	=back
		2563
		2564	Example: Define a class with an IO and idle watcher, start one of them in
		2565	the constructor.
		2566
		2567	class myclass
		2568	{
		2569	ev::io io; void io_cb (ev::io &w, int revents);
		2570	ev:idle idle void idle_cb (ev::idle &w, int revents);
		2571
		2572	myclass (int fd)
		2573	{
		2574	io .set <myclass, &myclass::io_cb > (this);
		2575	idle.set <myclass, &myclass::idle_cb> (this);
		2576
		2577	io.start (fd, ev::READ);
		2578	}
		2579	};
		2580
		2581
		2582	=head1 OTHER LANGUAGE BINDINGS
		2583
		2584	Libev does not offer other language bindings itself, but bindings for a
		2585	numbe rof languages exist in the form of third-party packages. If you know
		2586	any interesting language binding in addition to the ones listed here, drop
		2587	me a note.
		2588
		2589	=over 4
		2590
		2591	=item Perl
		2592
		2593	The EV module implements the full libev API and is actually used to test
		2594	libev. EV is developed together with libev. Apart from the EV core module,
		2595	there are additional modules that implement libev-compatible interfaces
		2596	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the
		2597	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).
		2598
		2599	It can be found and installed via CPAN, its homepage is found at
		2600	L<http://software.schmorp.de/pkg/EV>.
		2601
		2602	=item Ruby
		2603
		2604	Tony Arcieri has written a ruby extension that offers access to a subset
		2605	of the libev API and adds filehandle abstractions, asynchronous DNS and
		2606	more on top of it. It can be found via gem servers. Its homepage is at
		2607	L<http://rev.rubyforge.org/>.
		2608
		2609	=item D
		2610
		2611	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		2612	be found at L<http://git.llucax.com.ar/?p=software/ev.d.git;a=summary>.
		2613
		2614	=back
		2615
		2616
		2617	=head1 MACRO MAGIC
		2618
		2619	Libev can be compiled with a variety of options, the most fundamantal
		2620	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		2621	functions and callbacks have an initial C<struct ev_loop *> argument.
		2622
		2623	To make it easier to write programs that cope with either variant, the
		2624	following macros are defined:
		2625
		2626	=over 4
		2627
		2628	=item C<EV_A>, C<EV_A_>
		2629
		2630	This provides the loop I<argument> for functions, if one is required ("ev
		2631	loop argument"). The C<EV_A> form is used when this is the sole argument,
		2632	C<EV_A_> is used when other arguments are following. Example:
		2633
		2634	ev_unref (EV_A);
		2635	ev_timer_add (EV_A_ watcher);
		2636	ev_loop (EV_A_ 0);
		2637
		2638	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		2639	which is often provided by the following macro.
		2640
		2641	=item C<EV_P>, C<EV_P_>
		2642
		2643	This provides the loop I<parameter> for functions, if one is required ("ev
		2644	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		2645	C<EV_P_> is used when other parameters are following. Example:
		2646
		2647	// this is how ev_unref is being declared
		2648	static void ev_unref (EV_P);
		2649
		2650	// this is how you can declare your typical callback
		2651	static void cb (EV_P_ ev_timer *w, int revents)
		2652
		2653	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		2654	suitable for use with C<EV_A>.
		2655
		2656	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		2657
		2658	Similar to the other two macros, this gives you the value of the default
		2659	loop, if multiple loops are supported ("ev loop default").
		2660
		2661	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		2662
		2663	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		2664	default loop has been initialised (C<UC> == unchecked). Their behaviour
		2665	is undefined when the default loop has not been initialised by a previous
		2666	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		2667
		2668	It is often prudent to use C<EV_DEFAULT> when initialising the first
		2669	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		2670
		2671	=back
		2672
		2673	Example: Declare and initialise a check watcher, utilising the above
		2674	macros so it will work regardless of whether multiple loops are supported
		2675	or not.
		2676
		2677	static void
		2678	check_cb (EV_P_ ev_timer *w, int revents)
		2679	{
		2680	ev_check_stop (EV_A_ w);
		2681	}
		2682
		2683	ev_check check;
		2684	ev_check_init (&check, check_cb);
		2685	ev_check_start (EV_DEFAULT_ &check);
		2686	ev_loop (EV_DEFAULT_ 0);
		2687
		2688	=head1 EMBEDDING
		2689
		2690	Libev can (and often is) directly embedded into host
		2691	applications. Examples of applications that embed it include the Deliantra
		2692	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		2693	and rxvt-unicode.
		2694
		2695	The goal is to enable you to just copy the necessary files into your
		2696	source directory without having to change even a single line in them, so
		2697	you can easily upgrade by simply copying (or having a checked-out copy of
		2698	libev somewhere in your source tree).
		2699
		2700	=head2 FILESETS
		2701
		2702	Depending on what features you need you need to include one or more sets of files
		2703	in your app.
		2704
		2705	=head3 CORE EVENT LOOP
		2706
		2707	To include only the libev core (all the C<ev_*> functions), with manual
		2708	configuration (no autoconf):
		2709
		2710	#define EV_STANDALONE 1
		2711	#include "ev.c"
		2712
		2713	This will automatically include F<ev.h>, too, and should be done in a
		2714	single C source file only to provide the function implementations. To use
		2715	it, do the same for F<ev.h> in all files wishing to use this API (best
		2716	done by writing a wrapper around F<ev.h> that you can include instead and
		2717	where you can put other configuration options):
		2718
		2719	#define EV_STANDALONE 1
		2720	#include "ev.h"
		2721
		2722	Both header files and implementation files can be compiled with a C++
		2723	compiler (at least, thats a stated goal, and breakage will be treated
		2724	as a bug).
		2725
		2726	You need the following files in your source tree, or in a directory
		2727	in your include path (e.g. in libev/ when using -Ilibev):
		2728
		2729	ev.h
		2730	ev.c
		2731	ev_vars.h
		2732	ev_wrap.h
		2733
		2734	ev_win32.c required on win32 platforms only
		2735
		2736	ev_select.c only when select backend is enabled (which is enabled by default)
		2737	ev_poll.c only when poll backend is enabled (disabled by default)
		2738	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		2739	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		2740	ev_port.c only when the solaris port backend is enabled (disabled by default)
		2741
		2742	F<ev.c> includes the backend files directly when enabled, so you only need
		2743	to compile this single file.
		2744
		2745	=head3 LIBEVENT COMPATIBILITY API
		2746
		2747	To include the libevent compatibility API, also include:
		2748
		2749	#include "event.c"
		2750
		2751	in the file including F<ev.c>, and:
		2752
		2753	#include "event.h"
		2754
		2755	in the files that want to use the libevent API. This also includes F<ev.h>.
		2756
		2757	You need the following additional files for this:
		2758
		2759	event.h
		2760	event.c
		2761
		2762	=head3 AUTOCONF SUPPORT
		2763
		2764	Instead of using C<EV_STANDALONE=1> and providing your config in
		2765	whatever way you want, you can also C<m4_include([libev.m4])> in your
		2766	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		2767	include F<config.h> and configure itself accordingly.
		2768
		2769	For this of course you need the m4 file:
		2770
		2771	libev.m4
		2772
		2773	=head2 PREPROCESSOR SYMBOLS/MACROS
		2774
		2775	Libev can be configured via a variety of preprocessor symbols you have to
		2776	define before including any of its files. The default in the absense of
		2777	autoconf is noted for every option.
		2778
		2779	=over 4
		2780
		2781	=item EV_STANDALONE
		2782
		2783	Must always be C<1> if you do not use autoconf configuration, which
		2784	keeps libev from including F<config.h>, and it also defines dummy
		2785	implementations for some libevent functions (such as logging, which is not
		2786	supported). It will also not define any of the structs usually found in
		2787	F<event.h> that are not directly supported by the libev core alone.
		2788
		2789	=item EV_USE_MONOTONIC
		2790
		2791	If defined to be C<1>, libev will try to detect the availability of the
		2792	monotonic clock option at both compiletime and runtime. Otherwise no use
		2793	of the monotonic clock option will be attempted. If you enable this, you
		2794	usually have to link against librt or something similar. Enabling it when
		2795	the functionality isn't available is safe, though, although you have
		2796	to make sure you link against any libraries where the C<clock_gettime>
		2797	function is hiding in (often F<-lrt>).
		2798
		2799	=item EV_USE_REALTIME
		2800
		2801	If defined to be C<1>, libev will try to detect the availability of the
		2802	realtime clock option at compiletime (and assume its availability at
		2803	runtime if successful). Otherwise no use of the realtime clock option will
		2804	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		2805	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
		2806	note about libraries in the description of C<EV_USE_MONOTONIC>, though.
		2807
		2808	=item EV_USE_NANOSLEEP
		2809
		2810	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		2811	and will use it for delays. Otherwise it will use C<select ()>.
		2812
		2813	=item EV_USE_EVENTFD
		2814
		2815	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		2816	available and will probe for kernel support at runtime. This will improve
		2817	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		2818	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		2819	2.7 or newer, otherwise disabled.
		2820
		2821	=item EV_USE_SELECT
		2822
		2823	If undefined or defined to be C<1>, libev will compile in support for the
		2824	C<select>(2) backend. No attempt at autodetection will be done: if no
		2825	other method takes over, select will be it. Otherwise the select backend
		2826	will not be compiled in.
		2827
		2828	=item EV_SELECT_USE_FD_SET
		2829
		2830	If defined to C<1>, then the select backend will use the system C<fd_set>
		2831	structure. This is useful if libev doesn't compile due to a missing
		2832	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on
		2833	exotic systems. This usually limits the range of file descriptors to some
		2834	low limit such as 1024 or might have other limitations (winsocket only
		2835	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		2836	influence the size of the C<fd_set> used.
		2837
		2838	=item EV_SELECT_IS_WINSOCKET
		2839
		2840	When defined to C<1>, the select backend will assume that
		2841	select/socket/connect etc. don't understand file descriptors but
		2842	wants osf handles on win32 (this is the case when the select to
		2843	be used is the winsock select). This means that it will call
		2844	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		2845	it is assumed that all these functions actually work on fds, even
		2846	on win32. Should not be defined on non-win32 platforms.
		2847
		2848	=item EV_FD_TO_WIN32_HANDLE
		2849
		2850	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		2851	file descriptors to socket handles. When not defining this symbol (the
		2852	default), then libev will call C<_get_osfhandle>, which is usually
		2853	correct. In some cases, programs use their own file descriptor management,
		2854	in which case they can provide this function to map fds to socket handles.
		2855
		2856	=item EV_USE_POLL
		2857
		2858	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		2859	backend. Otherwise it will be enabled on non-win32 platforms. It
		2860	takes precedence over select.
		2861
		2862	=item EV_USE_EPOLL
		2863
		2864	If defined to be C<1>, libev will compile in support for the Linux
		2865	C<epoll>(7) backend. Its availability will be detected at runtime,
		2866	otherwise another method will be used as fallback. This is the preferred
		2867	backend for GNU/Linux systems. If undefined, it will be enabled if the
		2868	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		2869
		2870	=item EV_USE_KQUEUE
		2871
		2872	If defined to be C<1>, libev will compile in support for the BSD style
		2873	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		2874	otherwise another method will be used as fallback. This is the preferred
		2875	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		2876	supports some types of fds correctly (the only platform we found that
		2877	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		2878	not be used unless explicitly requested. The best way to use it is to find
		2879	out whether kqueue supports your type of fd properly and use an embedded
		2880	kqueue loop.
		2881
		2882	=item EV_USE_PORT
		2883
		2884	If defined to be C<1>, libev will compile in support for the Solaris
		2885	10 port style backend. Its availability will be detected at runtime,
		2886	otherwise another method will be used as fallback. This is the preferred
		2887	backend for Solaris 10 systems.
		2888
		2889	=item EV_USE_DEVPOLL
		2890
		2891	reserved for future expansion, works like the USE symbols above.
		2892
		2893	=item EV_USE_INOTIFY
		2894
		2895	If defined to be C<1>, libev will compile in support for the Linux inotify
		2896	interface to speed up C<ev_stat> watchers. Its actual availability will
		2897	be detected at runtime. If undefined, it will be enabled if the headers
		2898	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		2899
		2900	=item EV_ATOMIC_T
		2901
		2902	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		2903	access is atomic with respect to other threads or signal contexts. No such
		2904	type is easily found in the C language, so you can provide your own type
		2905	that you know is safe for your purposes. It is used both for signal handler "locking"
		2906	as well as for signal and thread safety in C<ev_async> watchers.
		2907
		2908	In the absense of this define, libev will use C<sig_atomic_t volatile>
		2909	(from F<signal.h>), which is usually good enough on most platforms.
		2910
		2911	=item EV_H
		2912
		2913	The name of the F<ev.h> header file used to include it. The default if
		2914	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
		2915	used to virtually rename the F<ev.h> header file in case of conflicts.
		2916
		2917	=item EV_CONFIG_H
		2918
		2919	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		2920	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		2921	C<EV_H>, above.
		2922
		2923	=item EV_EVENT_H
		2924
		2925	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		2926	of how the F<event.h> header can be found, the default is C<"event.h">.
		2927
		2928	=item EV_PROTOTYPES
		2929
		2930	If defined to be C<0>, then F<ev.h> will not define any function
		2931	prototypes, but still define all the structs and other symbols. This is
		2932	occasionally useful if you want to provide your own wrapper functions
		2933	around libev functions.
		2934
		2935	=item EV_MULTIPLICITY
		2936
		2937	If undefined or defined to C<1>, then all event-loop-specific functions
		2938	will have the C<struct ev_loop *> as first argument, and you can create
		2939	additional independent event loops. Otherwise there will be no support
		2940	for multiple event loops and there is no first event loop pointer
		2941	argument. Instead, all functions act on the single default loop.
		2942
		2943	=item EV_MINPRI
		2944
		2945	=item EV_MAXPRI
		2946
		2947	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2948	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2949	provide for more priorities by overriding those symbols (usually defined
		2950	to be C<-2> and C<2>, respectively).
		2951
		2952	When doing priority-based operations, libev usually has to linearly search
		2953	all the priorities, so having many of them (hundreds) uses a lot of space
		2954	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2955	fine.
		2956
		2957	If your embedding app does not need any priorities, defining these both to
		2958	C<0> will save some memory and cpu.
		2959
		2960	=item EV_PERIODIC_ENABLE
		2961
		2962	If undefined or defined to be C<1>, then periodic timers are supported. If
		2963	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2964	code.
		2965
		2966	=item EV_IDLE_ENABLE
		2967
		2968	If undefined or defined to be C<1>, then idle watchers are supported. If
		2969	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2970	code.
		2971
		2972	=item EV_EMBED_ENABLE
		2973
		2974	If undefined or defined to be C<1>, then embed watchers are supported. If
		2975	defined to be C<0>, then they are not.
		2976
		2977	=item EV_STAT_ENABLE
		2978
		2979	If undefined or defined to be C<1>, then stat watchers are supported. If
		2980	defined to be C<0>, then they are not.
		2981
		2982	=item EV_FORK_ENABLE
		2983
		2984	If undefined or defined to be C<1>, then fork watchers are supported. If
		2985	defined to be C<0>, then they are not.
		2986
		2987	=item EV_ASYNC_ENABLE
		2988
		2989	If undefined or defined to be C<1>, then async watchers are supported. If
		2990	defined to be C<0>, then they are not.
		2991
		2992	=item EV_MINIMAL
		2993
		2994	If you need to shave off some kilobytes of code at the expense of some
		2995	speed, define this symbol to C<1>. Currently this is used to override some
		2996	inlining decisions, saves roughly 30% codesize of amd64. It also selects a
		2997	much smaller 2-heap for timer management over the default 4-heap.
		2998
		2999	=item EV_PID_HASHSIZE
		3000
		3001	C<ev_child> watchers use a small hash table to distribute workload by
		3002	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		3003	than enough. If you need to manage thousands of children you might want to
		3004	increase this value (I<must> be a power of two).
		3005
		3006	=item EV_INOTIFY_HASHSIZE
		3007
		3008	C<ev_stat> watchers use a small hash table to distribute workload by
		3009	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		3010	usually more than enough. If you need to manage thousands of C<ev_stat>
		3011	watchers you might want to increase this value (I<must> be a power of
		3012	two).
		3013
		3014	=item EV_USE_4HEAP
		3015
		3016	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3017	timer and periodics heap, libev uses a 4-heap when this symbol is defined
		3018	to C<1>. The 4-heap uses more complicated (longer) code but has
		3019	noticably faster performance with many (thousands) of watchers.
		3020
		3021	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3022	(disabled).
		3023
		3024	=item EV_HEAP_CACHE_AT
		3025
		3026	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3027	timer and periodics heap, libev can cache the timestamp (I<at>) within
		3028	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3029	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3030	but avoids random read accesses on heap changes. This improves performance
		3031	noticably with with many (hundreds) of watchers.
		3032
		3033	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3034	(disabled).
		3035
		3036	=item EV_COMMON
		3037
		3038	By default, all watchers have a C<void *data> member. By redefining
		3039	this macro to a something else you can include more and other types of
		3040	members. You have to define it each time you include one of the files,
		3041	though, and it must be identical each time.
		3042
		3043	For example, the perl EV module uses something like this:
		3044
		3045	#define EV_COMMON \
		3046	SV *self; /* contains this struct */ \
		3047	SV *cb_sv, *fh /* note no trailing ";" */
		3048
		3049	=item EV_CB_DECLARE (type)
		3050
		3051	=item EV_CB_INVOKE (watcher, revents)
		3052
		3053	=item ev_set_cb (ev, cb)
		3054
		3055	Can be used to change the callback member declaration in each watcher,
		3056	and the way callbacks are invoked and set. Must expand to a struct member
		3057	definition and a statement, respectively. See the F<ev.h> header file for
		3058	their default definitions. One possible use for overriding these is to
		3059	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		3060	method calls instead of plain function calls in C++.
		3061
		3062	=head2 EXPORTED API SYMBOLS
		3063
		3064	If you need to re-export the API (e.g. via a dll) and you need a list of
		3065	exported symbols, you can use the provided F<Symbol.*> files which list
		3066	all public symbols, one per line:
		3067
		3068	Symbols.ev for libev proper
		3069	Symbols.event for the libevent emulation
		3070
		3071	This can also be used to rename all public symbols to avoid clashes with
		3072	multiple versions of libev linked together (which is obviously bad in
		3073	itself, but sometimes it is inconvinient to avoid this).
		3074
		3075	A sed command like this will create wrapper C<#define>'s that you need to
		3076	include before including F<ev.h>:
		3077
		3078	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		3079
		3080	This would create a file F<wrap.h> which essentially looks like this:
		3081
		3082	#define ev_backend myprefix_ev_backend
		3083	#define ev_check_start myprefix_ev_check_start
		3084	#define ev_check_stop myprefix_ev_check_stop
		3085	...
		3086
		3087	=head2 EXAMPLES
		3088
		3089	For a real-world example of a program the includes libev
		3090	verbatim, you can have a look at the EV perl module
		3091	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		3092	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		3093	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		3094	will be compiled. It is pretty complex because it provides its own header
		3095	file.
		3096
		3097	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		3098	that everybody includes and which overrides some configure choices:
		3099
		3100	#define EV_MINIMAL 1
		3101	#define EV_USE_POLL 0
		3102	#define EV_MULTIPLICITY 0
		3103	#define EV_PERIODIC_ENABLE 0
		3104	#define EV_STAT_ENABLE 0
		3105	#define EV_FORK_ENABLE 0
		3106	#define EV_CONFIG_H <config.h>
		3107	#define EV_MINPRI 0
		3108	#define EV_MAXPRI 0
		3109
		3110	#include "ev++.h"
		3111
		3112	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		3113
		3114	#include "ev_cpp.h"
		3115	#include "ev.c"
		3116
		3117
		3118	=head1 THREADS AND COROUTINES
		3119
		3120	=head2 THREADS
		3121
		3122	Libev itself is completely threadsafe, but it uses no locking. This
		3123	means that you can use as many loops as you want in parallel, as long as
		3124	only one thread ever calls into one libev function with the same loop
		3125	parameter.
		3126
		3127	Or put differently: calls with different loop parameters can be done in
		3128	parallel from multiple threads, calls with the same loop parameter must be
		3129	done serially (but can be done from different threads, as long as only one
		3130	thread ever is inside a call at any point in time, e.g. by using a mutex
		3131	per loop).
		3132
		3133	If you want to know which design is best for your problem, then I cannot
		3134	help you but by giving some generic advice:
		3135
		3136	=over 4
		3137
		3138	=item * most applications have a main thread: use the default libev loop
		3139	in that thread, or create a seperate thread running only the default loop.
		3140
		3141	This helps integrating other libraries or software modules that use libev
		3142	themselves and don't care/know about threading.
		3143
		3144	=item * one loop per thread is usually a good model.
		3145
		3146	Doing this is almost never wrong, sometimes a better-performance model
		3147	exists, but it is always a good start.
		3148
		3149	=item * other models exist, such as the leader/follower pattern, where one
		3150	loop is handed through multiple threads in a kind of round-robbin fashion.
		3151
		3152	Chosing a model is hard - look around, learn, know that usually you cna do
		3153	better than you currently do :-)
		3154
		3155	=item * often you need to talk to some other thread which blocks in the
		3156	event loop - C<ev_async> watchers can be used to wake them up from other
		3157	threads safely (or from signal contexts...).
		3158
		3159	=back
		3160
		3161	=head2 COROUTINES
		3162
		3163	Libev is much more accomodating to coroutines ("cooperative threads"):
		3164	libev fully supports nesting calls to it's functions from different
		3165	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3166	different coroutines and switch freely between both coroutines running the
		3167	loop, as long as you don't confuse yourself). The only exception is that
		3168	you must not do this from C<ev_periodic> reschedule callbacks.
		3169
		3170	Care has been invested into making sure that libev does not keep local
		3171	state inside C<ev_loop>, and other calls do not usually allow coroutine
		3172	switches.
		3173
		3174
		3175	=head1 COMPLEXITIES
		3176
		3177	In this section the complexities of (many of) the algorithms used inside
		3178	libev will be explained. For complexity discussions about backends see the
		3179	documentation for C<ev_default_init>.
		3180
		3181	All of the following are about amortised time: If an array needs to be
		3182	extended, libev needs to realloc and move the whole array, but this
		3183	happens asymptotically never with higher number of elements, so O(1) might
		3184	mean it might do a lengthy realloc operation in rare cases, but on average
		3185	it is much faster and asymptotically approaches constant time.
		3186
		3187	=over 4
		3188
		3189	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		3190
		3191	This means that, when you have a watcher that triggers in one hour and
		3192	there are 100 watchers that would trigger before that then inserting will
		3193	have to skip roughly seven (C<ld 100>) of these watchers.
		3194
		3195	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		3196
		3197	That means that changing a timer costs less than removing/adding them
		3198	as only the relative motion in the event queue has to be paid for.
		3199
		3200	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		3201
		3202	These just add the watcher into an array or at the head of a list.
		3203
		3204	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		3205
		3206	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		3207
		3208	These watchers are stored in lists then need to be walked to find the
		3209	correct watcher to remove. The lists are usually short (you don't usually
		3210	have many watchers waiting for the same fd or signal).
		3211
		3212	=item Finding the next timer in each loop iteration: O(1)
		3213
		3214	By virtue of using a binary or 4-heap, the next timer is always found at a
		3215	fixed position in the storage array.
		3216
		3217	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		3218
		3219	A change means an I/O watcher gets started or stopped, which requires
		3220	libev to recalculate its status (and possibly tell the kernel, depending
		3221	on backend and wether C<ev_io_set> was used).
		3222
		3223	=item Activating one watcher (putting it into the pending state): O(1)
		3224
		3225	=item Priority handling: O(number_of_priorities)
		3226
		3227	Priorities are implemented by allocating some space for each
		3228	priority. When doing priority-based operations, libev usually has to
		3229	linearly search all the priorities, but starting/stopping and activating
		3230	watchers becomes O(1) w.r.t. priority handling.
		3231
		3232	=item Sending an ev_async: O(1)
		3233
		3234	=item Processing ev_async_send: O(number_of_async_watchers)
		3235
		3236	=item Processing signals: O(max_signal_number)
		3237
		3238	Sending involves a syscall I<iff> there were no other C<ev_async_send>
		3239	calls in the current loop iteration. Checking for async and signal events
		3240	involves iterating over all running async watchers or all signal numbers.
		3241
		3242	=back
		3243
		3244
		3245	=head1 Win32 platform limitations and workarounds
		3246
		3247	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		3248	requires, and its I/O model is fundamentally incompatible with the POSIX
		3249	model. Libev still offers limited functionality on this platform in
		3250	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		3251	descriptors. This only applies when using Win32 natively, not when using
		3252	e.g. cygwin.
		3253
		3254	Lifting these limitations would basically require the full
		3255	re-implementation of the I/O system. If you are into these kinds of
		3256	things, then note that glib does exactly that for you in a very portable
		3257	way (note also that glib is the slowest event library known to man).
		3258
		3259	There is no supported compilation method available on windows except
		3260	embedding it into other applications.
		3261
		3262	Due to the many, low, and arbitrary limits on the win32 platform and
		3263	the abysmal performance of winsockets, using a large number of sockets
		3264	is not recommended (and not reasonable). If your program needs to use
		3265	more than a hundred or so sockets, then likely it needs to use a totally
		3266	different implementation for windows, as libev offers the POSIX readiness
		3267	notification model, which cannot be implemented efficiently on windows
		3268	(microsoft monopoly games).
		3269
		3270	=over 4
		3271
		3272	=item The winsocket select function
		3273
		3274	The winsocket C<select> function doesn't follow POSIX in that it requires
		3275	socket I<handles> and not socket I<file descriptors>. This makes select
		3276	very inefficient, and also requires a mapping from file descriptors
		3277	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,
		3278	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor
		3279	symbols for more info.
		3280
		3281	The configuration for a "naked" win32 using the microsoft runtime
		3282	libraries and raw winsocket select is:
		3283
		3284	#define EV_USE_SELECT 1
		3285	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		3286
		3287	Note that winsockets handling of fd sets is O(n), so you can easily get a
		3288	complexity in the O(n²) range when using win32.
		3289
		3290	=item Limited number of file descriptors
		3291
		3292	Windows has numerous arbitrary (and low) limits on things.
		3293
		3294	Early versions of winsocket's select only supported waiting for a maximum
		3295	of C<64> handles (probably owning to the fact that all windows kernels
		3296	can only wait for C<64> things at the same time internally; microsoft
		3297	recommends spawning a chain of threads and wait for 63 handles and the
		3298	previous thread in each. Great).
		3299
		3300	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		3301	to some high number (e.g. C<2048>) before compiling the winsocket select
		3302	call (which might be in libev or elsewhere, for example, perl does its own
		3303	select emulation on windows).
		3304
		3305	Another limit is the number of file descriptors in the microsoft runtime
		3306	libraries, which by default is C<64> (there must be a hidden I<64> fetish
		3307	or something like this inside microsoft). You can increase this by calling
		3308	C<_setmaxstdio>, which can increase this limit to C<2048> (another
		3309	arbitrary limit), but is broken in many versions of the microsoft runtime
		3310	libraries.
		3311
		3312	This might get you to about C<512> or C<2048> sockets (depending on
		3313	windows version and/or the phase of the moon). To get more, you need to
		3314	wrap all I/O functions and provide your own fd management, but the cost of
		3315	calling select (O(n²)) will likely make this unworkable.
		3316
		3317	=back
		3318
		3319
		3320	=head1 PORTABILITY REQUIREMENTS
		3321
		3322	In addition to a working ISO-C implementation, libev relies on a few
		3323	additional extensions:
		3324
		3325	=over 4
		3326
		3327	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3328
		3329	The type C<sig_atomic_t volatile> (or whatever is defined as
		3330	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different
		3331	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3332	believed to be sufficiently portable.
		3333
		3334	=item C<sigprocmask> must work in a threaded environment
		3335
		3336	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3337	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3338	pthread implementations will either allow C<sigprocmask> in the "main
		3339	thread" or will block signals process-wide, both behaviours would
		3340	be compatible with libev. Interaction between C<sigprocmask> and
		3341	C<pthread_sigmask> could complicate things, however.
		3342
		3343	The most portable way to handle signals is to block signals in all threads
		3344	except the initial one, and run the default loop in the initial thread as
		3345	well.
		3346
		3347	=item C<long> must be large enough for common memory allocation sizes
		3348
		3349	To improve portability and simplify using libev, libev uses C<long>
		3350	internally instead of C<size_t> when allocating its data structures. On
		3351	non-POSIX systems (Microsoft...) this might be unexpectedly low, but
		3352	is still at least 31 bits everywhere, which is enough for hundreds of
		3353	millions of watchers.
		3354
		3355	=item C<double> must hold a time value in seconds with enough accuracy
		3356
		3357	The type C<double> is used to represent timestamps. It is required to
		3358	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3359	enough for at least into the year 4000. This requirement is fulfilled by
		3360	implementations implementing IEEE 754 (basically all existing ones).
		3361
		3362	=back
		3363
		3364	If you know of other additional requirements drop me a note.
		3365
		3366
		3367	=head1 VALGRIND
		3368
		3369	Valgrind has a special section here because it is a popular tool that is
		3370	highly useful, but valgrind reports are very hard to interpret.
		3371
		3372	If you think you found a bug (memory leak, uninitialised data access etc.)
		3373	in libev, then check twice: If valgrind reports something like:
		3374
		3375	==2274== definitely lost: 0 bytes in 0 blocks.
		3376	==2274== possibly lost: 0 bytes in 0 blocks.
		3377	==2274== still reachable: 256 bytes in 1 blocks.
		3378
		3379	then there is no memory leak. Similarly, under some circumstances,
		3380	valgrind might report kernel bugs as if it were a bug in libev, or it
		3381	might be confused (it is a very good tool, but only a tool).
		3382
		3383	If you are unsure about something, feel free to contact the mailing list
		3384	with the full valgrind report and an explanation on why you think this is
		3385	a bug in libev. However, don't be annoyed when you get a brisk "this is
		3386	no bug" answer and take the chance of learning how to interpret valgrind
		3387	properly.
		3388
		3389	If you need, for some reason, empty reports from valgrind for your project
		3390	I suggest using suppression lists.
		3391
1312		3392
1313	=head1 AUTHOR	3393	=head1 AUTHOR
1314		3394
1315	Marc Lehmann <libev@schmorp.de>.	3395	Marc Lehmann <libev@schmorp.de>.
1316		3396

Diff Legend

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