ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.35 by root, Fri Nov 23 19:35:09 2007 UTC vs.
Revision 1.204 by root, Mon Oct 27 11:08:29 2008 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
		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_ 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_ 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	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	}
8		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<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 floating point value. Unlike the name
		119	component C<stamp> might indicate, it is also used for time differences
		120	throughout libev.
		121
		122	=head1 ERROR HANDLING
		123
		124	Libev knows three classes of errors: operating system errors, usage errors
		125	and internal errors (bugs).
		126
		127	When libev catches an operating system error it cannot handle (for example
		128	a system call indicating a condition libev cannot fix), it calls the callback
		129	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		130	abort. The default is to print a diagnostic message and to call C<abort
		131	()>.
		132
		133	When libev detects a usage error such as a negative timer interval, then
		134	it will print a diagnostic message and abort (via the C<assert> mechanism,
		135	so C<NDEBUG> will disable this checking): these are programming errors in
		136	the libev caller and need to be fixed there.
		137
		138	Libev also has a few internal error-checking C<assert>ions, and also has
		139	extensive consistency checking code. These do not trigger under normal
		140	circumstances, as they indicate either a bug in libev or worse.
52		141
53		142
54	=head1 GLOBAL FUNCTIONS	143	=head1 GLOBAL FUNCTIONS
55		144
56	These functions can be called anytime, even before initialising the	145	These functions can be called anytime, even before initialising the
…		…
62		151
63	Returns the current time as libev would use it. Please note that the	152	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	153	C<ev_now> function is usually faster and also often returns the timestamp
65	you actually want to know.	154	you actually want to know.
66		155
		156	=item ev_sleep (ev_tstamp interval)
		157
		158	Sleep for the given interval: The current thread will be blocked until
		159	either it is interrupted or the given time interval has passed. Basically
		160	this is a sub-second-resolution C<sleep ()>.
		161
67	=item int ev_version_major ()	162	=item int ev_version_major ()
68		163
69	=item int ev_version_minor ()	164	=item int ev_version_minor ()
70		165
71	You can find out the major and minor version numbers of the library	166	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	167	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	168	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	169	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
75	version of the library your program was compiled against.	170	version of the library your program was compiled against.
76		171
		172	These version numbers refer to the ABI version of the library, not the
		173	release version.
		174
77	Usually, it's a good idea to terminate if the major versions mismatch,	175	Usually, it's a good idea to terminate if the major versions mismatch,
78	as this indicates an incompatible change. Minor versions are usually	176	as this indicates an incompatible change. Minor versions are usually
79	compatible to older versions, so a larger minor version alone is usually	177	compatible to older versions, so a larger minor version alone is usually
80	not a problem.	178	not a problem.
81		179
82	Example: make sure we haven't accidentally been linked against the wrong	180	Example: Make sure we haven't accidentally been linked against the wrong
83	version:	181	version.
84		182
85	assert (("libev version mismatch",	183	assert (("libev version mismatch",
86	ev_version_major () == EV_VERSION_MAJOR	184	ev_version_major () == EV_VERSION_MAJOR
87	&& ev_version_minor () >= EV_VERSION_MINOR));	185	&& ev_version_minor () >= EV_VERSION_MINOR));
88		186
89	=item unsigned int ev_supported_backends ()	187	=item unsigned int ev_supported_backends ()
90		188
91	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	189	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
92	value) compiled into this binary of libev (independent of their	190	value) compiled into this binary of libev (independent of their
…		…
94	a description of the set values.	192	a description of the set values.
95		193
96	Example: make sure we have the epoll method, because yeah this is cool and	194	Example: make sure we have the epoll method, because yeah this is cool and
97	a must have and can we have a torrent of it please!!!11	195	a must have and can we have a torrent of it please!!!11
98		196
99	assert (("sorry, no epoll, no sex",	197	assert (("sorry, no epoll, no sex",
100	ev_supported_backends () & EVBACKEND_EPOLL));	198	ev_supported_backends () & EVBACKEND_EPOLL));
101		199
102	=item unsigned int ev_recommended_backends ()	200	=item unsigned int ev_recommended_backends ()
103		201
104	Return the set of all backends compiled into this binary of libev and also	202	Return the set of all backends compiled into this binary of libev and also
105	recommended for this platform. This set is often smaller than the one	203	recommended for this platform. This set is often smaller than the one
106	returned by C<ev_supported_backends>, as for example kqueue is broken on	204	returned by C<ev_supported_backends>, as for example kqueue is broken on
107	most BSDs and will not be autodetected unless you explicitly request it	205	most BSDs and will not be auto-detected unless you explicitly request it
108	(assuming you know what you are doing). This is the set of backends that	206	(assuming you know what you are doing). This is the set of backends that
109	libev will probe for if you specify no backends explicitly.	207	libev will probe for if you specify no backends explicitly.
110		208
111	=item unsigned int ev_embeddable_backends ()	209	=item unsigned int ev_embeddable_backends ()
112		210
…		…
116	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	214	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
117	recommended ones.	215	recommended ones.
118		216
119	See the description of C<ev_embed> watchers for more info.	217	See the description of C<ev_embed> watchers for more info.
120		218
121	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	219	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
122		220
123	Sets the allocation function to use (the prototype is similar to the	221	Sets the allocation function to use (the prototype is similar - the
124	realloc C function, the semantics are identical). It is used to allocate	222	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	223	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	224	when memory needs to be allocated (C<size != 0>), the library might abort
127	destructive action. The default is your system realloc function.	225	or take some potentially destructive action.
		226
		227	Since some systems (at least OpenBSD and Darwin) fail to implement
		228	correct C<realloc> semantics, libev will use a wrapper around the system
		229	C<realloc> and C<free> functions by default.
128		230
129	You could override this function in high-availability programs to, say,	231	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,	232	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.	233	or even to sleep a while and retry until some memory is available.
132		234
133	Example: replace the libev allocator with one that waits a bit and then	235	Example: Replace the libev allocator with one that waits a bit and then
134	retries: better than mine).	236	retries (example requires a standards-compliant C<realloc>).
135		237
136	static void *	238	static void *
137	persistent_realloc (void *ptr, long size)	239	persistent_realloc (void *ptr, size_t size)
138	{	240	{
139	for (;;)	241	for (;;)
140	{	242	{
141	void *newptr = realloc (ptr, size);	243	void *newptr = realloc (ptr, size);
142		244
…		…
148	}	250	}
149		251
150	...	252	...
151	ev_set_allocator (persistent_realloc);	253	ev_set_allocator (persistent_realloc);
152		254
153	=item ev_set_syserr_cb (void (*cb)(const char *msg));	255	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
154		256
155	Set the callback function to call on a retryable syscall error (such	257	Set the callback function to call on a retryable system call error (such
156	as failed select, poll, epoll_wait). The message is a printable string	258	as failed select, poll, epoll_wait). The message is a printable string
157	indicating the system call or subsystem causing the problem. If this	259	indicating the system call or subsystem causing the problem. If this
158	callback is set, then libev will expect it to remedy the sitution, no	260	callback is set, then libev will expect it to remedy the situation, no
159	matter what, when it returns. That is, libev will generally retry the	261	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	262	requested operation, or, if the condition doesn't go away, do bad stuff
161	(such as abort).	263	(such as abort).
162		264
163	Example: do the same thing as libev does internally:	265	Example: This is basically the same thing that libev does internally, too.
164		266
165	static void	267	static void
166	fatal_error (const char *msg)	268	fatal_error (const char *msg)
167	{	269	{
168	perror (msg);	270	perror (msg);
…		…
174		276
175	=back	277	=back
176		278
177	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
178		280
179	An event loop is described by a C<struct ev_loop *>. The library knows two	281	An event loop is described by a C<struct ev_loop *> (the C<struct>
180	types of such loops, the I<default> loop, which supports signals and child	282	is I<not> optional in this case, as there is also an C<ev_loop>
181	events, and dynamically created loops which do not.	283	I<function>).
182		284
183	If you use threads, a common model is to run the default event loop	285	The library knows two types of such loops, the I<default> loop, which
184	in your main thread (or in a separate thread) and for each thread you	286	supports signals and child events, and dynamically created loops which do
185	create, you also create another event loop. Libev itself does no locking	287	not.
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		288
190	=over 4	289	=over 4
191		290
192	=item struct ev_loop *ev_default_loop (unsigned int flags)	291	=item struct ev_loop *ev_default_loop (unsigned int flags)
193		292
…		…
197	flags. If that is troubling you, check C<ev_backend ()> afterwards).	296	flags. If that is troubling you, check C<ev_backend ()> afterwards).
198		297
199	If you don't know what event loop to use, use the one returned from this	298	If you don't know what event loop to use, use the one returned from this
200	function.	299	function.
201		300
		301	Note that this function is I<not> thread-safe, so if you want to use it
		302	from multiple threads, you have to lock (note also that this is unlikely,
		303	as loops cannot bes hared easily between threads anyway).
		304
		305	The default loop is the only loop that can handle C<ev_signal> and
		306	C<ev_child> watchers, and to do this, it always registers a handler
		307	for C<SIGCHLD>. If this is a problem for your application you can either
		308	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		309	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		310	C<ev_default_init>.
		311
202	The flags argument can be used to specify special behaviour or specific	312	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>).	313	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
204		314
205	The following flags are supported:	315	The following flags are supported:
206		316
…		…
211	The default flags value. Use this if you have no clue (it's the right	321	The default flags value. Use this if you have no clue (it's the right
212	thing, believe me).	322	thing, believe me).
213		323
214	=item C<EVFLAG_NOENV>	324	=item C<EVFLAG_NOENV>
215		325
216	If this flag bit is ored into the flag value (or the program runs setuid	326	If this flag bit is or'ed into the flag value (or the program runs setuid
217	or setgid) then libev will I<not> look at the environment variable	327	or setgid) then libev will I<not> look at the environment variable
218	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	328	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
219	override the flags completely if it is found in the environment. This is	329	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	330	useful to try out specific backends to test their performance, or to work
221	around bugs.	331	around bugs.
222		332
		333	=item C<EVFLAG_FORKCHECK>
		334
		335	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		336	a fork, you can also make libev check for a fork in each iteration by
		337	enabling this flag.
		338
		339	This works by calling C<getpid ()> on every iteration of the loop,
		340	and thus this might slow down your event loop if you do a lot of loop
		341	iterations and little real work, but is usually not noticeable (on my
		342	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
		343	without a system call and thus I<very> fast, but my GNU/Linux system also has
		344	C<pthread_atfork> which is even faster).
		345
		346	The big advantage of this flag is that you can forget about fork (and
		347	forget about forgetting to tell libev about forking) when you use this
		348	flag.
		349
		350	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
		351	environment variable.
		352
223	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	353	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
224		354
225	This is your standard select(2) backend. Not I<completely> standard, as	355	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,	356	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	357	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	358	using this backend. It doesn't scale too well (O(highest_fd)), but its
229	the fastest backend for a low number of fds.	359	usually the fastest backend for a low number of (low-numbered :) fds.
		360
		361	To get good performance out of this backend you need a high amount of
		362	parallelism (most of the file descriptors should be busy). If you are
		363	writing a server, you should C<accept ()> in a loop to accept as many
		364	connections as possible during one iteration. You might also want to have
		365	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		366	readiness notifications you get per iteration.
		367
		368	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		369	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		370	C<exceptfds> set on that platform).
230		371
231	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	372	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
232		373
233	And this is your standard poll(2) backend. It's more complicated than	374	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	375	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	376	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).	377	considerably with a lot of inactive fds). It scales similarly to select,
		378	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		379	performance tips.
		380
		381	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		382	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
237		383
238	=item C<EVBACKEND_EPOLL> (value 4, Linux)	384	=item C<EVBACKEND_EPOLL> (value 4, Linux)
239		385
240	For few fds, this backend is a bit little slower than poll and select,	386	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	387	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	388	like O(total_fds) where n is the total number of fds (or the highest fd),
243	either O(1) or O(active_fds).	389	epoll scales either O(1) or O(active_fds). The epoll design has a number
		390	of shortcomings, such as silently dropping events in some hard-to-detect
		391	cases and requiring a system call per fd change, no fork support and bad
		392	support for dup.
244		393
		394	Epoll is also notoriously buggy - embedding epoll fds should work, but
		395	of course doesn't, and epoll just loves to report events for totally
		396	I<different> file descriptors (even already closed ones) than registered
		397	in the set (especially on SMP systems). Libev tries to counter these
		398	spurious notifications by employing an additional generation counter and
		399	comparing that against the events to filter out spurious ones.
		400
245	While stopping and starting an I/O watcher in the same iteration will	401	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	402	will result in some caching, there is still a system call per such incident
247	(because the fd could point to a different file description now), so its	403	(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	404	best to avoid that. Also, C<dup ()>'ed file descriptors might not work
249	well if you register events for both fds.	405	very well if you register events for both fds.
250		406
251	Please note that epoll sometimes generates spurious notifications, so you	407	Best performance from this backend is achieved by not unregistering all
252	need to use non-blocking I/O or other means to avoid blocking when no data	408	watchers for a file descriptor until it has been closed, if possible,
253	(or space) is available.	409	i.e. keep at least one watcher active per fd at all times. Stopping and
		410	starting a watcher (without re-setting it) also usually doesn't cause
		411	extra overhead.
		412
		413	While nominally embeddable in other event loops, this feature is broken in
		414	all kernel versions tested so far.
		415
		416	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		417	C<EVBACKEND_POLL>.
254		418
255	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	419	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
256		420
257	Kqueue deserves special mention, as at the time of this writing, it	421	Kqueue deserves special mention, as at the time of this writing, it was
258	was broken on all BSDs except NetBSD (usually it doesn't work with	422	broken on all BSDs except NetBSD (usually it doesn't work reliably with
259	anything but sockets and pipes, except on Darwin, where of course its	423	anything but sockets and pipes, except on Darwin, where of course it's
260	completely useless). For this reason its not being "autodetected"	424	completely useless). For this reason it's not being "auto-detected" unless
261	unless you explicitly specify it explicitly in the flags (i.e. using	425	you explicitly specify it in the flags (i.e. using C<EVBACKEND_KQUEUE>) or
262	C<EVBACKEND_KQUEUE>).	426	libev was compiled on a known-to-be-good (-enough) system like NetBSD.
		427
		428	You still can embed kqueue into a normal poll or select backend and use it
		429	only for sockets (after having made sure that sockets work with kqueue on
		430	the target platform). See C<ev_embed> watchers for more info.
263		431
264	It scales in the same way as the epoll backend, but the interface to the	432	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	433	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	434	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	435	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
268	incident, so its best to avoid that.	436	two event changes per incident. Support for C<fork ()> is very bad and it
		437	drops fds silently in similarly hard-to-detect cases.
		438
		439	This backend usually performs well under most conditions.
		440
		441	While nominally embeddable in other event loops, this doesn't work
		442	everywhere, so you might need to test for this. And since it is broken
		443	almost everywhere, you should only use it when you have a lot of sockets
		444	(for which it usually works), by embedding it into another event loop
		445	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,
		446	using it only for sockets.
		447
		448	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		449	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		450	C<NOTE_EOF>.
269		451
270	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	452	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
271		453
272	This is not implemented yet (and might never be).	454	This is not implemented yet (and might never be, unless you send me an
		455	implementation). According to reports, C</dev/poll> only supports sockets
		456	and is not embeddable, which would limit the usefulness of this backend
		457	immensely.
273		458
274	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	459	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
275		460
276	This uses the Solaris 10 port mechanism. As with everything on Solaris,	461	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)).	462	it's really slow, but it still scales very well (O(active_fds)).
278		463
279	Please note that solaris ports can result in a lot of spurious	464	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	465	notifications, so you need to use non-blocking I/O or other means to avoid
281	blocking when no data (or space) is available.	466	blocking when no data (or space) is available.
		467
		468	While this backend scales well, it requires one system call per active
		469	file descriptor per loop iteration. For small and medium numbers of file
		470	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		471	might perform better.
		472
		473	On the positive side, with the exception of the spurious readiness
		474	notifications, this backend actually performed fully to specification
		475	in all tests and is fully embeddable, which is a rare feat among the
		476	OS-specific backends.
		477
		478	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		479	C<EVBACKEND_POLL>.
282		480
283	=item C<EVBACKEND_ALL>	481	=item C<EVBACKEND_ALL>
284		482
285	Try all backends (even potentially broken ones that wouldn't be tried	483	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	484	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
287	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	485	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
288		486
		487	It is definitely not recommended to use this flag.
		488
289	=back	489	=back
290		490
291	If one or more of these are ored into the flags value, then only these	491	If one or more of these are or'ed into the flags value, then only these
292	backends will be tried (in the reverse order as given here). If none are	492	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	493	specified, all backends in C<ev_recommended_backends ()> will be tried.
294	order of their flag values :)
295		494
296	The most typical usage is like this:	495	Example: This is the most typical usage.
297		496
298	if (!ev_default_loop (0))	497	if (!ev_default_loop (0))
299	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	498	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
300		499
301	Restrict libev to the select and poll backends, and do not allow	500	Example: Restrict libev to the select and poll backends, and do not allow
302	environment settings to be taken into account:	501	environment settings to be taken into account:
303		502
304	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	503	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
305		504
306	Use whatever libev has to offer, but make sure that kqueue is used if	505	Example: Use whatever libev has to offer, but make sure that kqueue is
307	available (warning, breaks stuff, best use only with your own private	506	used if available (warning, breaks stuff, best use only with your own
308	event loop and only if you know the OS supports your types of fds):	507	private event loop and only if you know the OS supports your types of
		508	fds):
309		509
310	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	510	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
311		511
312	=item struct ev_loop *ev_loop_new (unsigned int flags)	512	=item struct ev_loop *ev_loop_new (unsigned int flags)
313		513
314	Similar to C<ev_default_loop>, but always creates a new event loop that is	514	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	515	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	516	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).	517	undefined behaviour (or a failed assertion if assertions are enabled).
318		518
		519	Note that this function I<is> thread-safe, and the recommended way to use
		520	libev with threads is indeed to create one loop per thread, and using the
		521	default loop in the "main" or "initial" thread.
		522
319	Example: try to create a event loop that uses epoll and nothing else.	523	Example: Try to create a event loop that uses epoll and nothing else.
320		524
321	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	525	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
322	if (!epoller)	526	if (!epoller)
323	fatal ("no epoll found here, maybe it hides under your chair");	527	fatal ("no epoll found here, maybe it hides under your chair");
324		528
325	=item ev_default_destroy ()	529	=item ev_default_destroy ()
326		530
327	Destroys the default loop again (frees all memory and kernel state	531	Destroys the default loop again (frees all memory and kernel state
328	etc.). This stops all registered event watchers (by not touching them in	532	etc.). None of the active event watchers will be stopped in the normal
329	any way whatsoever, although you cannot rely on this :).	533	sense, so e.g. C<ev_is_active> might still return true. It is your
		534	responsibility to either stop all watchers cleanly yourself I<before>
		535	calling this function, or cope with the fact afterwards (which is usually
		536	the easiest thing, you can just ignore the watchers and/or C<free ()> them
		537	for example).
		538
		539	Note that certain global state, such as signal state (and installed signal
		540	handlers), will not be freed by this function, and related watchers (such
		541	as signal and child watchers) would need to be stopped manually.
		542
		543	In general it is not advisable to call this function except in the
		544	rare occasion where you really need to free e.g. the signal handling
		545	pipe fds. If you need dynamically allocated loops it is better to use
		546	C<ev_loop_new> and C<ev_loop_destroy>).
330		547
331	=item ev_loop_destroy (loop)	548	=item ev_loop_destroy (loop)
332		549
333	Like C<ev_default_destroy>, but destroys an event loop created by an	550	Like C<ev_default_destroy>, but destroys an event loop created by an
334	earlier call to C<ev_loop_new>.	551	earlier call to C<ev_loop_new>.
335		552
336	=item ev_default_fork ()	553	=item ev_default_fork ()
337		554
		555	This function sets a flag that causes subsequent C<ev_loop> iterations
338	This function reinitialises the kernel state for backends that have	556	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	557	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	558	the child process (or both child and parent, but that again makes little
341	again makes little sense).	559	sense). You I<must> call it in the child before using any of the libev
		560	functions, and it will only take effect at the next C<ev_loop> iteration.
342		561
343	You I<must> call this function in the child process after forking if and	562	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	563	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.	564	you just fork+exec, you don't have to call it at all.
346		565
347	The function itself is quite fast and it's usually not a problem to call	566	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	567	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>:	568	quite nicely into a call to C<pthread_atfork>:
350		569
351	pthread_atfork (0, 0, ev_default_fork);	570	pthread_atfork (0, 0, ev_default_fork);
352		571
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)	572	=item ev_loop_fork (loop)
358		573
359	Like C<ev_default_fork>, but acts on an event loop created by	574	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	575	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.	576	after fork that you want to re-use in the child, and how you do this is
		577	entirely your own problem.
		578
		579	=item int ev_is_default_loop (loop)
		580
		581	Returns true when the given loop is, in fact, the default loop, and false
		582	otherwise.
		583
		584	=item unsigned int ev_loop_count (loop)
		585
		586	Returns the count of loop iterations for the loop, which is identical to
		587	the number of times libev did poll for new events. It starts at C<0> and
		588	happily wraps around with enough iterations.
		589
		590	This value can sometimes be useful as a generation counter of sorts (it
		591	"ticks" the number of loop iterations), as it roughly corresponds with
		592	C<ev_prepare> and C<ev_check> calls.
362		593
363	=item unsigned int ev_backend (loop)	594	=item unsigned int ev_backend (loop)
364		595
365	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	596	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
366	use.	597	use.
…		…
369		600
370	Returns the current "event loop time", which is the time the event loop	601	Returns the current "event loop time", which is the time the event loop
371	received events and started processing them. This timestamp does not	602	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	603	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	604	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).	605	event occurring (or more correctly, libev finding out about it).
		606
		607	=item ev_now_update (loop)
		608
		609	Establishes the current time by querying the kernel, updating the time
		610	returned by C<ev_now ()> in the progress. This is a costly operation and
		611	is usually done automatically within C<ev_loop ()>.
		612
		613	This function is rarely useful, but when some event callback runs for a
		614	very long time without entering the event loop, updating libev's idea of
		615	the current time is a good idea.
		616
		617	See also "The special problem of time updates" in the C<ev_timer> section.
375		618
376	=item ev_loop (loop, int flags)	619	=item ev_loop (loop, int flags)
377		620
378	Finally, this is it, the event handler. This function usually is called	621	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	622	after you initialised all your watchers and you want to start handling
…		…
382	If the flags argument is specified as C<0>, it will not return until	625	If the flags argument is specified as C<0>, it will not return until
383	either no event watchers are active anymore or C<ev_unloop> was called.	626	either no event watchers are active anymore or C<ev_unloop> was called.
384		627
385	Please note that an explicit C<ev_unloop> is usually better than	628	Please note that an explicit C<ev_unloop> is usually better than
386	relying on all watchers to be stopped when deciding when a program has	629	relying on all watchers to be stopped when deciding when a program has
387	finished (especially in interactive programs), but having a program that	630	finished (especially in interactive programs), but having a program
388	automatically loops as long as it has to and no longer by virtue of	631	that automatically loops as long as it has to and no longer by virtue
389	relying on its watchers stopping correctly is a thing of beauty.	632	of relying on its watchers stopping correctly, that is truly a thing of
		633	beauty.
390		634
391	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	635	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
392	those events and any outstanding ones, but will not block your process in	636	those events and any already outstanding ones, but will not block your
393	case there are no events and will return after one iteration of the loop.	637	process in case there are no events and will return after one iteration of
		638	the loop.
394		639
395	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	640	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
396	neccessary) and will handle those and any outstanding ones. It will block	641	necessary) and will handle those and any already outstanding ones. It
397	your process until at least one new event arrives, and will return after	642	will block your process until at least one new event arrives (which could
398	one iteration of the loop. This is useful if you are waiting for some	643	be an event internal to libev itself, so there is no guarentee that a
399	external event in conjunction with something not expressible using other	644	user-registered callback will be called), and will return after one
		645	iteration of the loop.
		646
		647	This is useful if you are waiting for some external event in conjunction
		648	with something not expressible using other libev watchers (i.e. "roll your
400	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	649	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
401	usually a better approach for this kind of thing.	650	usually a better approach for this kind of thing.
402		651
403	Here are the gory details of what C<ev_loop> does:	652	Here are the gory details of what C<ev_loop> does:
404		653
405	* If there are no active watchers (reference count is zero), return.	654	- Before the first iteration, call any pending watchers.
406	- Queue prepare watchers and then call all outstanding watchers.	655	* If EVFLAG_FORKCHECK was used, check for a fork.
		656	- If a fork was detected (by any means), queue and call all fork watchers.
		657	- Queue and call all prepare watchers.
407	- If we have been forked, recreate the kernel state.	658	- If we have been forked, detach and recreate the kernel state
		659	as to not disturb the other process.
408	- Update the kernel state with all outstanding changes.	660	- Update the kernel state with all outstanding changes.
409	- Update the "event loop time".	661	- Update the "event loop time" (ev_now ()).
410	- Calculate for how long to block.	662	- Calculate for how long to sleep or block, if at all
		663	(active idle watchers, EVLOOP_NONBLOCK or not having
		664	any active watchers at all will result in not sleeping).
		665	- Sleep if the I/O and timer collect interval say so.
411	- Block the process, waiting for any events.	666	- Block the process, waiting for any events.
412	- Queue all outstanding I/O (fd) events.	667	- Queue all outstanding I/O (fd) events.
413	- Update the "event loop time" and do time jump handling.	668	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
414	- Queue all outstanding timers.	669	- Queue all expired timers.
415	- Queue all outstanding periodics.	670	- Queue all expired periodics.
416	- If no events are pending now, queue all idle watchers.	671	- Unless any events are pending now, queue all idle watchers.
417	- Queue all check watchers.	672	- Queue all check watchers.
418	- Call all queued watchers in reverse order (i.e. check watchers first).	673	- 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	674	Signals and child watchers are implemented as I/O watchers, and will
420	be handled here by queueing them when their watcher gets executed.	675	be handled here by queueing them when their watcher gets executed.
421	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	676	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
422	were used, return, otherwise continue with step *.	677	were used, or there are no active watchers, return, otherwise
		678	continue with step *.
423		679
424	Example: queue some jobs and then loop until no events are outsanding	680	Example: Queue some jobs and then loop until no events are outstanding
425	anymore.	681	anymore.
426		682
427	... queue jobs here, make sure they register event watchers as long	683	... 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..)	684	... as they still have work to do (even an idle watcher will do..)
429	ev_loop (my_loop, 0);	685	ev_loop (my_loop, 0);
430	... jobs done. yeah!	686	... jobs done or somebody called unloop. yeah!
431		687
432	=item ev_unloop (loop, how)	688	=item ev_unloop (loop, how)
433		689
434	Can be used to make a call to C<ev_loop> return early (but only after it	690	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	691	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	692	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.	693	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
438		694
		695	This "unloop state" will be cleared when entering C<ev_loop> again.
		696
		697	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		698
439	=item ev_ref (loop)	699	=item ev_ref (loop)
440		700
441	=item ev_unref (loop)	701	=item ev_unref (loop)
442		702
443	Ref/unref can be used to add or remove a reference count on the event	703	Ref/unref can be used to add or remove a reference count on the event
444	loop: Every watcher keeps one reference, and as long as the reference	704	loop: Every watcher keeps one reference, and as long as the reference
445	count is nonzero, C<ev_loop> will not return on its own. If you have	705	count is nonzero, C<ev_loop> will not return on its own.
		706
446	a watcher you never unregister that should not keep C<ev_loop> from	707	If you have a watcher you never unregister that should not keep C<ev_loop>
447	returning, ev_unref() after starting, and ev_ref() before stopping it. For	708	from returning, call ev_unref() after starting, and ev_ref() before
		709	stopping it.
		710
448	example, libev itself uses this for its internal signal pipe: It is not	711	As an example, libev itself uses this for its internal signal pipe: It is
449	visible to the libev user and should not keep C<ev_loop> from exiting if	712	not visible to the libev user and should not keep C<ev_loop> from exiting
450	no event watchers registered by it are active. It is also an excellent	713	if 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	714	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>.	715	libraries. Just remember to I<unref after start> and I<ref before stop>
		716	(but only if the watcher wasn't active before, or was active before,
		717	respectively).
453		718
454	Example: create a signal watcher, but keep it from keeping C<ev_loop>	719	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
455	running when nothing else is active.	720	running when nothing else is active.
456		721
457	struct dv_signal exitsig;	722	ev_signal exitsig;
458	ev_signal_init (&exitsig, sig_cb, SIGINT);	723	ev_signal_init (&exitsig, sig_cb, SIGINT);
459	ev_signal_start (myloop, &exitsig);	724	ev_signal_start (loop, &exitsig);
460	evf_unref (myloop);	725	evf_unref (loop);
461		726
462	Example: for some weird reason, unregister the above signal handler again.	727	Example: For some weird reason, unregister the above signal handler again.
463		728
464	ev_ref (myloop);	729	ev_ref (loop);
465	ev_signal_stop (myloop, &exitsig);	730	ev_signal_stop (loop, &exitsig);
		731
		732	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		733
		734	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		735
		736	These advanced functions influence the time that libev will spend waiting
		737	for events. Both time intervals are by default C<0>, meaning that libev
		738	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		739	latency.
		740
		741	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		742	allows libev to delay invocation of I/O and timer/periodic callbacks
		743	to increase efficiency of loop iterations (or to increase power-saving
		744	opportunities).
		745
		746	The idea is that sometimes your program runs just fast enough to handle
		747	one (or very few) event(s) per loop iteration. While this makes the
		748	program responsive, it also wastes a lot of CPU time to poll for new
		749	events, especially with backends like C<select ()> which have a high
		750	overhead for the actual polling but can deliver many events at once.
		751
		752	By setting a higher I<io collect interval> you allow libev to spend more
		753	time collecting I/O events, so you can handle more events per iteration,
		754	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		755	C<ev_timer>) will be not affected. Setting this to a non-null value will
		756	introduce an additional C<ev_sleep ()> call into most loop iterations.
		757
		758	Likewise, by setting a higher I<timeout collect interval> you allow libev
		759	to spend more time collecting timeouts, at the expense of increased
		760	latency/jitter/inexactness (the watcher callback will be called
		761	later). C<ev_io> watchers will not be affected. Setting this to a non-null
		762	value will not introduce any overhead in libev.
		763
		764	Many (busy) programs can usually benefit by setting the I/O collect
		765	interval to a value near C<0.1> or so, which is often enough for
		766	interactive servers (of course not for games), likewise for timeouts. It
		767	usually doesn't make much sense to set it to a lower value than C<0.01>,
		768	as this approaches the timing granularity of most systems.
		769
		770	Setting the I<timeout collect interval> can improve the opportunity for
		771	saving power, as the program will "bundle" timer callback invocations that
		772	are "near" in time together, by delaying some, thus reducing the number of
		773	times the process sleeps and wakes up again. Another useful technique to
		774	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		775	they fire on, say, one-second boundaries only.
		776
		777	=item ev_loop_verify (loop)
		778
		779	This function only does something when C<EV_VERIFY> support has been
		780	compiled in, which is the default for non-minimal builds. It tries to go
		781	through all internal structures and checks them for validity. If anything
		782	is found to be inconsistent, it will print an error message to standard
		783	error and call C<abort ()>.
		784
		785	This can be used to catch bugs inside libev itself: under normal
		786	circumstances, this function will never abort as of course libev keeps its
		787	data structures consistent.
466		788
467	=back	789	=back
468		790
		791
469	=head1 ANATOMY OF A WATCHER	792	=head1 ANATOMY OF A WATCHER
		793
		794	In the following description, uppercase C<TYPE> in names stands for the
		795	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		796	watchers and C<ev_io_start> for I/O watchers.
470		797
471	A watcher is a structure that you create and register to record your	798	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	799	interest in some event. For instance, if you want to wait for STDIN to
473	become readable, you would create an C<ev_io> watcher for that:	800	become readable, you would create an C<ev_io> watcher for that:
474		801
475	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	802	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
476	{	803	{
477	ev_io_stop (w);	804	ev_io_stop (w);
478	ev_unloop (loop, EVUNLOOP_ALL);	805	ev_unloop (loop, EVUNLOOP_ALL);
479	}	806	}
480		807
481	struct ev_loop *loop = ev_default_loop (0);	808	struct ev_loop *loop = ev_default_loop (0);
		809
482	struct ev_io stdin_watcher;	810	ev_io stdin_watcher;
		811
483	ev_init (&stdin_watcher, my_cb);	812	ev_init (&stdin_watcher, my_cb);
484	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	813	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
485	ev_io_start (loop, &stdin_watcher);	814	ev_io_start (loop, &stdin_watcher);
		815
486	ev_loop (loop, 0);	816	ev_loop (loop, 0);
487		817
488	As you can see, you are responsible for allocating the memory for your	818	As you can see, you are responsible for allocating the memory for your
489	watcher structures (and it is usually a bad idea to do this on the stack,	819	watcher structures (and it is I<usually> a bad idea to do this on the
490	although this can sometimes be quite valid).	820	stack).
		821
		822	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		823	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
491		824
492	Each watcher structure must be initialised by a call to C<ev_init	825	Each watcher structure must be initialised by a call to C<ev_init
493	(watcher *, callback)>, which expects a callback to be provided. This	826	(watcher *, callback)>, which expects a callback to be provided. This
494	callback gets invoked each time the event occurs (or, in the case of io	827	callback gets invoked each time the event occurs (or, in the case of I/O
495	watchers, each time the event loop detects that the file descriptor given	828	watchers, each time the event loop detects that the file descriptor given
496	is readable and/or writable).	829	is readable and/or writable).
497		830
498	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	831	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
499	with arguments specific to this watcher type. There is also a macro	832	macro to configure it, with arguments specific to the watcher type. There
500	to combine initialisation and setting in one call: C<< ev_<type>_init	833	is also a macro to combine initialisation and setting in one call: C<<
501	(watcher *, callback, ...) >>.	834	ev_TYPE_init (watcher *, callback, ...) >>.
502		835
503	To make the watcher actually watch out for events, you have to start it	836	To make the watcher actually watch out for events, you have to start it
504	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	837	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
505	*) >>), and you can stop watching for events at any time by calling the	838	*) >>), and you can stop watching for events at any time by calling the
506	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	839	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
507		840
508	As long as your watcher is active (has been started but not stopped) you	841	As long as your watcher is active (has been started but not stopped) you
509	must not touch the values stored in it. Most specifically you must never	842	must not touch the values stored in it. Most specifically you must never
510	reinitialise it or call its set macro.	843	reinitialise it or call its C<ev_TYPE_set> macro.
511
512	You can check whether an event is active by calling the C<ev_is_active
513	(watcher *)> macro. To see whether an event is outstanding (but the
514	callback for it has not been called yet) you can use the C<ev_is_pending
515	(watcher *)> macro.
516		844
517	Each and every callback receives the event loop pointer as first, the	845	Each and every callback receives the event loop pointer as first, the
518	registered watcher structure as second, and a bitset of received events as	846	registered watcher structure as second, and a bitset of received events as
519	third argument.	847	third argument.
520		848
…		…
544	The signal specified in the C<ev_signal> watcher has been received by a thread.	872	The signal specified in the C<ev_signal> watcher has been received by a thread.
545		873
546	=item C<EV_CHILD>	874	=item C<EV_CHILD>
547		875
548	The pid specified in the C<ev_child> watcher has received a status change.	876	The pid specified in the C<ev_child> watcher has received a status change.
		877
		878	=item C<EV_STAT>
		879
		880	The path specified in the C<ev_stat> watcher changed its attributes somehow.
549		881
550	=item C<EV_IDLE>	882	=item C<EV_IDLE>
551		883
552	The C<ev_idle> watcher has determined that you have nothing better to do.	884	The C<ev_idle> watcher has determined that you have nothing better to do.
553		885
…		…
561	received events. Callbacks of both watcher types can start and stop as	893	received events. Callbacks of both watcher types can start and stop as
562	many watchers as they want, and all of them will be taken into account	894	many watchers as they want, and all of them will be taken into account
563	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	895	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
564	C<ev_loop> from blocking).	896	C<ev_loop> from blocking).
565		897
		898	=item C<EV_EMBED>
		899
		900	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		901
		902	=item C<EV_FORK>
		903
		904	The event loop has been resumed in the child process after fork (see
		905	C<ev_fork>).
		906
		907	=item C<EV_ASYNC>
		908
		909	The given async watcher has been asynchronously notified (see C<ev_async>).
		910
566	=item C<EV_ERROR>	911	=item C<EV_ERROR>
567		912
568	An unspecified error has occured, the watcher has been stopped. This might	913	An unspecified error has occurred, the watcher has been stopped. This might
569	happen because the watcher could not be properly started because libev	914	happen because the watcher could not be properly started because libev
570	ran out of memory, a file descriptor was found to be closed or any other	915	ran out of memory, a file descriptor was found to be closed or any other
		916	problem. Libev considers these application bugs.
		917
571	problem. You best act on it by reporting the problem and somehow coping	918	You best act on it by reporting the problem and somehow coping with the
572	with the watcher being stopped.	919	watcher being stopped. Note that well-written programs should not receive
		920	an error ever, so when your watcher receives it, this usually indicates a
		921	bug in your program.
573		922
574	Libev will usually signal a few "dummy" events together with an error,	923	Libev will usually signal a few "dummy" events together with an error, for
575	for example it might indicate that a fd is readable or writable, and if	924	example it might indicate that a fd is readable or writable, and if your
576	your callbacks is well-written it can just attempt the operation and cope	925	callbacks is well-written it can just attempt the operation and cope with
577	with the error from read() or write(). This will not work in multithreaded	926	the error from read() or write(). This will not work in multi-threaded
578	programs, though, so beware.	927	programs, though, as the fd could already be closed and reused for another
		928	thing, so beware.
579		929
580	=back	930	=back
581		931
		932	=head2 GENERIC WATCHER FUNCTIONS
		933
		934	=over 4
		935
		936	=item C<ev_init> (ev_TYPE *watcher, callback)
		937
		938	This macro initialises the generic portion of a watcher. The contents
		939	of the watcher object can be arbitrary (so C<malloc> will do). Only
		940	the generic parts of the watcher are initialised, you I<need> to call
		941	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		942	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		943	which rolls both calls into one.
		944
		945	You can reinitialise a watcher at any time as long as it has been stopped
		946	(or never started) and there are no pending events outstanding.
		947
		948	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
		949	int revents)>.
		950
		951	Example: Initialise an C<ev_io> watcher in two steps.
		952
		953	ev_io w;
		954	ev_init (&w, my_cb);
		955	ev_io_set (&w, STDIN_FILENO, EV_READ);
		956
		957	=item C<ev_TYPE_set> (ev_TYPE *, [args])
		958
		959	This macro initialises the type-specific parts of a watcher. You need to
		960	call C<ev_init> at least once before you call this macro, but you can
		961	call C<ev_TYPE_set> any number of times. You must not, however, call this
		962	macro on a watcher that is active (it can be pending, however, which is a
		963	difference to the C<ev_init> macro).
		964
		965	Although some watcher types do not have type-specific arguments
		966	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		967
		968	See C<ev_init>, above, for an example.
		969
		970	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		971
		972	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		973	calls into a single call. This is the most convenient method to initialise
		974	a watcher. The same limitations apply, of course.
		975
		976	Example: Initialise and set an C<ev_io> watcher in one step.
		977
		978	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		979
		980	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
		981
		982	Starts (activates) the given watcher. Only active watchers will receive
		983	events. If the watcher is already active nothing will happen.
		984
		985	Example: Start the C<ev_io> watcher that is being abused as example in this
		986	whole section.
		987
		988	ev_io_start (EV_DEFAULT_UC, &w);
		989
		990	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
		991
		992	Stops the given watcher if active, and clears the pending status (whether
		993	the watcher was active or not).
		994
		995	It is possible that stopped watchers are pending - for example,
		996	non-repeating timers are being stopped when they become pending - but
		997	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
		998	pending. If you want to free or reuse the memory used by the watcher it is
		999	therefore a good idea to always call its C<ev_TYPE_stop> function.
		1000
		1001	=item bool ev_is_active (ev_TYPE *watcher)
		1002
		1003	Returns a true value iff the watcher is active (i.e. it has been started
		1004	and not yet been stopped). As long as a watcher is active you must not modify
		1005	it.
		1006
		1007	=item bool ev_is_pending (ev_TYPE *watcher)
		1008
		1009	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		1010	events but its callback has not yet been invoked). As long as a watcher
		1011	is pending (but not active) you must not call an init function on it (but
		1012	C<ev_TYPE_set> is safe), you must not change its priority, and you must
		1013	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		1014	it).
		1015
		1016	=item callback ev_cb (ev_TYPE *watcher)
		1017
		1018	Returns the callback currently set on the watcher.
		1019
		1020	=item ev_cb_set (ev_TYPE *watcher, callback)
		1021
		1022	Change the callback. You can change the callback at virtually any time
		1023	(modulo threads).
		1024
		1025	=item ev_set_priority (ev_TYPE *watcher, priority)
		1026
		1027	=item int ev_priority (ev_TYPE *watcher)
		1028
		1029	Set and query the priority of the watcher. The priority is a small
		1030	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		1031	(default: C<-2>). Pending watchers with higher priority will be invoked
		1032	before watchers with lower priority, but priority will not keep watchers
		1033	from being executed (except for C<ev_idle> watchers).
		1034
		1035	This means that priorities are I<only> used for ordering callback
		1036	invocation after new events have been received. This is useful, for
		1037	example, to reduce latency after idling, or more often, to bind two
		1038	watchers on the same event and make sure one is called first.
		1039
		1040	If you need to suppress invocation when higher priority events are pending
		1041	you need to look at C<ev_idle> watchers, which provide this functionality.
		1042
		1043	You I<must not> change the priority of a watcher as long as it is active or
		1044	pending.
		1045
		1046	The default priority used by watchers when no priority has been set is
		1047	always C<0>, which is supposed to not be too high and not be too low :).
		1048
		1049	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1050	fine, as long as you do not mind that the priority value you query might
		1051	or might not have been clamped to the valid range.
		1052
		1053	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		1054
		1055	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		1056	C<loop> nor C<revents> need to be valid as long as the watcher callback
		1057	can deal with that fact, as both are simply passed through to the
		1058	callback.
		1059
		1060	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		1061
		1062	If the watcher is pending, this function clears its pending status and
		1063	returns its C<revents> bitset (as if its callback was invoked). If the
		1064	watcher isn't pending it does nothing and returns C<0>.
		1065
		1066	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1067	callback to be invoked, which can be accomplished with this function.
		1068
		1069	=back
		1070
		1071
582	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1072	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
583		1073
584	Each watcher has, by default, a member C<void *data> that you can change	1074	Each watcher has, by default, a member C<void *data> that you can change
585	and read at any time, libev will completely ignore it. This can be used	1075	and read at any time: libev will completely ignore it. This can be used
586	to associate arbitrary data with your watcher. If you need more data and	1076	to associate arbitrary data with your watcher. If you need more data and
587	don't want to allocate memory and store a pointer to it in that data	1077	don't want to allocate memory and store a pointer to it in that data
588	member, you can also "subclass" the watcher type and provide your own	1078	member, you can also "subclass" the watcher type and provide your own
589	data:	1079	data:
590		1080
591	struct my_io	1081	struct my_io
592	{	1082	{
593	struct ev_io io;	1083	ev_io io;
594	int otherfd;	1084	int otherfd;
595	void *somedata;	1085	void *somedata;
596	struct whatever *mostinteresting;	1086	struct whatever *mostinteresting;
597	}	1087	};
		1088
		1089	...
		1090	struct my_io w;
		1091	ev_io_init (&w.io, my_cb, fd, EV_READ);
598		1092
599	And since your callback will be called with a pointer to the watcher, you	1093	And since your callback will be called with a pointer to the watcher, you
600	can cast it back to your own type:	1094	can cast it back to your own type:
601		1095
602	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1096	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
603	{	1097	{
604	struct my_io *w = (struct my_io *)w_;	1098	struct my_io *w = (struct my_io *)w_;
605	...	1099	...
606	}	1100	}
607		1101
608	More interesting and less C-conformant ways of catsing your callback type	1102	More interesting and less C-conformant ways of casting your callback type
609	have been omitted....	1103	instead have been omitted.
		1104
		1105	Another common scenario is to use some data structure with multiple
		1106	embedded watchers:
		1107
		1108	struct my_biggy
		1109	{
		1110	int some_data;
		1111	ev_timer t1;
		1112	ev_timer t2;
		1113	}
		1114
		1115	In this case getting the pointer to C<my_biggy> is a bit more
		1116	complicated: Either you store the address of your C<my_biggy> struct
		1117	in the C<data> member of the watcher (for woozies), or you need to use
		1118	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1119	programmers):
		1120
		1121	#include <stddef.h>
		1122
		1123	static void
		1124	t1_cb (EV_P_ ev_timer *w, int revents)
		1125	{
		1126	struct my_biggy big = (struct my_biggy *
		1127	(((char *)w) - offsetof (struct my_biggy, t1));
		1128	}
		1129
		1130	static void
		1131	t2_cb (EV_P_ ev_timer *w, int revents)
		1132	{
		1133	struct my_biggy big = (struct my_biggy *
		1134	(((char *)w) - offsetof (struct my_biggy, t2));
		1135	}
610		1136
611		1137
612	=head1 WATCHER TYPES	1138	=head1 WATCHER TYPES
613		1139
614	This section describes each watcher in detail, but will not repeat	1140	This section describes each watcher in detail, but will not repeat
615	information given in the last section.	1141	information given in the last section. Any initialisation/set macros,
		1142	functions and members specific to the watcher type are explained.
616		1143
		1144	Members are additionally marked with either I<[read-only]>, meaning that,
		1145	while the watcher is active, you can look at the member and expect some
		1146	sensible content, but you must not modify it (you can modify it while the
		1147	watcher is stopped to your hearts content), or I<[read-write]>, which
		1148	means you can expect it to have some sensible content while the watcher
		1149	is active, but you can also modify it. Modifying it may not do something
		1150	sensible or take immediate effect (or do anything at all), but libev will
		1151	not crash or malfunction in any way.
617		1152
		1153
618	=head2 C<ev_io> - is this file descriptor readable or writable	1154	=head2 C<ev_io> - is this file descriptor readable or writable?
619		1155
620	I/O watchers check whether a file descriptor is readable or writable	1156	I/O watchers check whether a file descriptor is readable or writable
621	in each iteration of the event loop (This behaviour is called	1157	in each iteration of the event loop, or, more precisely, when reading
622	level-triggering because you keep receiving events as long as the	1158	would not block the process and writing would at least be able to write
623	condition persists. Remember you can stop the watcher if you don't want to	1159	some data. This behaviour is called level-triggering because you keep
624	act on the event and neither want to receive future events).	1160	receiving events as long as the condition persists. Remember you can stop
		1161	the watcher if you don't want to act on the event and neither want to
		1162	receive future events.
625		1163
626	In general you can register as many read and/or write event watchers per	1164	In general you can register as many read and/or write event watchers per
627	fd as you want (as long as you don't confuse yourself). Setting all file	1165	fd as you want (as long as you don't confuse yourself). Setting all file
628	descriptors to non-blocking mode is also usually a good idea (but not	1166	descriptors to non-blocking mode is also usually a good idea (but not
629	required if you know what you are doing).	1167	required if you know what you are doing).
630		1168
631	You have to be careful with dup'ed file descriptors, though. Some backends	1169	If you cannot use non-blocking mode, then force the use of a
632	(the linux epoll backend is a notable example) cannot handle dup'ed file	1170	known-to-be-good backend (at the time of this writing, this includes only
633	descriptors correctly if you register interest in two or more fds pointing	1171	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>).
634	to the same underlying file/socket etc. description (that is, they share
635	the same underlying "file open").
636		1172
637	If you must do this, then force the use of a known-to-be-good backend	1173	Another thing you have to watch out for is that it is quite easy to
638	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1174	receive "spurious" readiness notifications, that is your callback might
		1175	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		1176	because there is no data. Not only are some backends known to create a
		1177	lot of those (for example Solaris ports), it is very easy to get into
		1178	this situation even with a relatively standard program structure. Thus
		1179	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		1180	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		1181
		1182	If you cannot run the fd in non-blocking mode (for example you should
		1183	not play around with an Xlib connection), then you have to separately
		1184	re-test whether a file descriptor is really ready with a known-to-be good
		1185	interface such as poll (fortunately in our Xlib example, Xlib already
		1186	does this on its own, so its quite safe to use). Some people additionally
		1187	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1188	indefinitely.
		1189
		1190	But really, best use non-blocking mode.
		1191
		1192	=head3 The special problem of disappearing file descriptors
		1193
		1194	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1195	descriptor (either due to calling C<close> explicitly or any other means,
		1196	such as C<dup2>). The reason is that you register interest in some file
		1197	descriptor, but when it goes away, the operating system will silently drop
		1198	this interest. If another file descriptor with the same number then is
		1199	registered with libev, there is no efficient way to see that this is, in
		1200	fact, a different file descriptor.
		1201
		1202	To avoid having to explicitly tell libev about such cases, libev follows
		1203	the following policy: Each time C<ev_io_set> is being called, libev
		1204	will assume that this is potentially a new file descriptor, otherwise
		1205	it is assumed that the file descriptor stays the same. That means that
		1206	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1207	descriptor even if the file descriptor number itself did not change.
		1208
		1209	This is how one would do it normally anyway, the important point is that
		1210	the libev application should not optimise around libev but should leave
		1211	optimisations to libev.
		1212
		1213	=head3 The special problem of dup'ed file descriptors
		1214
		1215	Some backends (e.g. epoll), cannot register events for file descriptors,
		1216	but only events for the underlying file descriptions. That means when you
		1217	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1218	events for them, only one file descriptor might actually receive events.
		1219
		1220	There is no workaround possible except not registering events
		1221	for potentially C<dup ()>'ed file descriptors, or to resort to
		1222	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1223
		1224	=head3 The special problem of fork
		1225
		1226	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1227	useless behaviour. Libev fully supports fork, but needs to be told about
		1228	it in the child.
		1229
		1230	To support fork in your programs, you either have to call
		1231	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
		1232	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
639	C<EVBACKEND_POLL>).	1233	C<EVBACKEND_POLL>.
		1234
		1235	=head3 The special problem of SIGPIPE
		1236
		1237	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1238	when writing to a pipe whose other end has been closed, your program gets
		1239	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1240	this is sensible behaviour, for daemons, this is usually undesirable.
		1241
		1242	So when you encounter spurious, unexplained daemon exits, make sure you
		1243	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1244	somewhere, as that would have given you a big clue).
		1245
		1246
		1247	=head3 Watcher-Specific Functions
640		1248
641	=over 4	1249	=over 4
642		1250
643	=item ev_io_init (ev_io *, callback, int fd, int events)	1251	=item ev_io_init (ev_io *, callback, int fd, int events)
644		1252
645	=item ev_io_set (ev_io *, int fd, int events)	1253	=item ev_io_set (ev_io *, int fd, int events)
646		1254
647	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	1255	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
648	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	1256	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
649	EV_WRITE> to receive the given events.	1257	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
650		1258
651	Please note that most of the more scalable backend mechanisms (for example	1259	=item int fd [read-only]
652	epoll and solaris ports) can result in spurious readyness notifications	1260
653	for file descriptors, so you practically need to use non-blocking I/O (and	1261	The file descriptor being watched.
654	treat callback invocation as hint only), or retest separately with a safe	1262
655	interface before doing I/O (XLib can do this), or force the use of either	1263	=item int events [read-only]
656	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>, which don't suffer from this	1264
657	problem. Also note that it is quite easy to have your callback invoked	1265	The events being watched.
658	when the readyness condition is no longer valid even when employing
659	typical ways of handling events, so its a good idea to use non-blocking
660	I/O unconditionally.
661		1266
662	=back	1267	=back
663		1268
		1269	=head3 Examples
		1270
664	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	1271	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
665	readable, but only once. Since it is likely line-buffered, you could	1272	readable, but only once. Since it is likely line-buffered, you could
666	attempt to read a whole line in the callback:	1273	attempt to read a whole line in the callback.
667		1274
668	static void	1275	static void
669	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1276	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
670	{	1277	{
671	ev_io_stop (loop, w);	1278	ev_io_stop (loop, w);
672	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1279	.. read from stdin here (or from w->fd) and handle any I/O errors
673	}	1280	}
674		1281
675	...	1282	...
676	struct ev_loop *loop = ev_default_init (0);	1283	struct ev_loop *loop = ev_default_init (0);
677	struct ev_io stdin_readable;	1284	ev_io stdin_readable;
678	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1285	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
679	ev_io_start (loop, &stdin_readable);	1286	ev_io_start (loop, &stdin_readable);
680	ev_loop (loop, 0);	1287	ev_loop (loop, 0);
681		1288
682		1289
683	=head2 C<ev_timer> - relative and optionally recurring timeouts	1290	=head2 C<ev_timer> - relative and optionally repeating timeouts
684		1291
685	Timer watchers are simple relative timers that generate an event after a	1292	Timer watchers are simple relative timers that generate an event after a
686	given time, and optionally repeating in regular intervals after that.	1293	given time, and optionally repeating in regular intervals after that.
687		1294
688	The timers are based on real time, that is, if you register an event that	1295	The timers are based on real time, that is, if you register an event that
689	times out after an hour and you reset your system clock to last years	1296	times out after an hour and you reset your system clock to January last
690	time, it will still time out after (roughly) and hour. "Roughly" because	1297	year, it will still time out after (roughly) one hour. "Roughly" because
691	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1298	detecting time jumps is hard, and some inaccuracies are unavoidable (the
692	monotonic clock option helps a lot here).	1299	monotonic clock option helps a lot here).
		1300
		1301	The callback is guaranteed to be invoked only I<after> its timeout has
		1302	passed, but if multiple timers become ready during the same loop iteration
		1303	then order of execution is undefined.
		1304
		1305	=head3 Be smart about timeouts
		1306
		1307	Many real-world problems involve some kind of timeout, usually for error
		1308	recovery. A typical example is an HTTP request - if the other side hangs,
		1309	you want to raise some error after a while.
		1310
		1311	What follows are some ways to handle this problem, from obvious and
		1312	inefficient to smart and efficient.
		1313
		1314	In the following, a 60 second activity timeout is assumed - a timeout that
		1315	gets reset to 60 seconds each time there is activity (e.g. each time some
		1316	data or other life sign was received).
		1317
		1318	=over 4
		1319
		1320	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1321
		1322	This is the most obvious, but not the most simple way: In the beginning,
		1323	start the watcher:
		1324
		1325	ev_timer_init (timer, callback, 60., 0.);
		1326	ev_timer_start (loop, timer);
		1327
		1328	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1329	and start it again:
		1330
		1331	ev_timer_stop (loop, timer);
		1332	ev_timer_set (timer, 60., 0.);
		1333	ev_timer_start (loop, timer);
		1334
		1335	This is relatively simple to implement, but means that each time there is
		1336	some activity, libev will first have to remove the timer from its internal
		1337	data structure and then add it again. Libev tries to be fast, but it's
		1338	still not a constant-time operation.
		1339
		1340	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1341
		1342	This is the easiest way, and involves using C<ev_timer_again> instead of
		1343	C<ev_timer_start>.
		1344
		1345	To implement this, configure an C<ev_timer> with a C<repeat> value
		1346	of C<60> and then call C<ev_timer_again> at start and each time you
		1347	successfully read or write some data. If you go into an idle state where
		1348	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1349	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1350
		1351	That means you can ignore both the C<ev_timer_start> function and the
		1352	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1353	member and C<ev_timer_again>.
		1354
		1355	At start:
		1356
		1357	ev_timer_init (timer, callback);
		1358	timer->repeat = 60.;
		1359	ev_timer_again (loop, timer);
		1360
		1361	Each time there is some activity:
		1362
		1363	ev_timer_again (loop, timer);
		1364
		1365	It is even possible to change the time-out on the fly, regardless of
		1366	whether the watcher is active or not:
		1367
		1368	timer->repeat = 30.;
		1369	ev_timer_again (loop, timer);
		1370
		1371	This is slightly more efficient then stopping/starting the timer each time
		1372	you want to modify its timeout value, as libev does not have to completely
		1373	remove and re-insert the timer from/into its internal data structure.
		1374
		1375	It is, however, even simpler than the "obvious" way to do it.
		1376
		1377	=item 3. Let the timer time out, but then re-arm it as required.
		1378
		1379	This method is more tricky, but usually most efficient: Most timeouts are
		1380	relatively long compared to the intervals between other activity - in
		1381	our example, within 60 seconds, there are usually many I/O events with
		1382	associated activity resets.
		1383
		1384	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1385	but remember the time of last activity, and check for a real timeout only
		1386	within the callback:
		1387
		1388	ev_tstamp last_activity; // time of last activity
		1389
		1390	static void
		1391	callback (EV_P_ ev_timer *w, int revents)
		1392	{
		1393	ev_tstamp now = ev_now (EV_A);
		1394	ev_tstamp timeout = last_activity + 60.;
		1395
		1396	// if last_activity + 60. is older than now, we did time out
		1397	if (timeout < now)
		1398	{
		1399	// timeout occured, take action
		1400	}
		1401	else
		1402	{
		1403	// callback was invoked, but there was some activity, re-arm
		1404	// the watcher to fire in last_activity + 60, which is
		1405	// guaranteed to be in the future, so "again" is positive:
		1406	w->again = timeout - now;
		1407	ev_timer_again (EV_A_ w);
		1408	}
		1409	}
		1410
		1411	To summarise the callback: first calculate the real timeout (defined
		1412	as "60 seconds after the last activity"), then check if that time has
		1413	been reached, which means something I<did>, in fact, time out. Otherwise
		1414	the callback was invoked too early (C<timeout> is in the future), so
		1415	re-schedule the timer to fire at that future time, to see if maybe we have
		1416	a timeout then.
		1417
		1418	Note how C<ev_timer_again> is used, taking advantage of the
		1419	C<ev_timer_again> optimisation when the timer is already running.
		1420
		1421	This scheme causes more callback invocations (about one every 60 seconds
		1422	minus half the average time between activity), but virtually no calls to
		1423	libev to change the timeout.
		1424
		1425	To start the timer, simply initialise the watcher and set C<last_activity>
		1426	to the current time (meaning we just have some activity :), then call the
		1427	callback, which will "do the right thing" and start the timer:
		1428
		1429	ev_timer_init (timer, callback);
		1430	last_activity = ev_now (loop);
		1431	callback (loop, timer, EV_TIMEOUT);
		1432
		1433	And when there is some activity, simply store the current time in
		1434	C<last_activity>, no libev calls at all:
		1435
		1436	last_actiivty = ev_now (loop);
		1437
		1438	This technique is slightly more complex, but in most cases where the
		1439	time-out is unlikely to be triggered, much more efficient.
		1440
		1441	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1442	callback :) - just change the timeout and invoke the callback, which will
		1443	fix things for you.
		1444
		1445	=item 4. Wee, just use a double-linked list for your timeouts.
		1446
		1447	If there is not one request, but many thousands (millions...), all
		1448	employing some kind of timeout with the same timeout value, then one can
		1449	do even better:
		1450
		1451	When starting the timeout, calculate the timeout value and put the timeout
		1452	at the I<end> of the list.
		1453
		1454	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1455	the list is expected to fire (for example, using the technique #3).
		1456
		1457	When there is some activity, remove the timer from the list, recalculate
		1458	the timeout, append it to the end of the list again, and make sure to
		1459	update the C<ev_timer> if it was taken from the beginning of the list.
		1460
		1461	This way, one can manage an unlimited number of timeouts in O(1) time for
		1462	starting, stopping and updating the timers, at the expense of a major
		1463	complication, and having to use a constant timeout. The constant timeout
		1464	ensures that the list stays sorted.
		1465
		1466	=back
		1467
		1468	So which method the best?
		1469
		1470	Method #2 is a simple no-brain-required solution that is adequate in most
		1471	situations. Method #3 requires a bit more thinking, but handles many cases
		1472	better, and isn't very complicated either. In most case, choosing either
		1473	one is fine, with #3 being better in typical situations.
		1474
		1475	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1476	rather complicated, but extremely efficient, something that really pays
		1477	off after the first million or so of active timers, i.e. it's usually
		1478	overkill :)
		1479
		1480	=head3 The special problem of time updates
		1481
		1482	Establishing the current time is a costly operation (it usually takes at
		1483	least two system calls): EV therefore updates its idea of the current
		1484	time only before and after C<ev_loop> collects new events, which causes a
		1485	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1486	lots of events in one iteration.
693		1487
694	The relative timeouts are calculated relative to the C<ev_now ()>	1488	The relative timeouts are calculated relative to the C<ev_now ()>
695	time. This is usually the right thing as this timestamp refers to the time	1489	time. This is usually the right thing as this timestamp refers to the time
696	of the event triggering whatever timeout you are modifying/starting. If	1490	of the event triggering whatever timeout you are modifying/starting. If
697	you suspect event processing to be delayed and you I<need> to base the timeout	1491	you suspect event processing to be delayed and you I<need> to base the
698	on the current time, use something like this to adjust for this:	1492	timeout on the current time, use something like this to adjust for this:
699		1493
700	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1494	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
701		1495
702	The callback is guarenteed to be invoked only when its timeout has passed,	1496	If the event loop is suspended for a long time, you can also force an
703	but if multiple timers become ready during the same loop iteration then	1497	update of the time returned by C<ev_now ()> by calling C<ev_now_update
704	order of execution is undefined.	1498	()>.
		1499
		1500	=head3 Watcher-Specific Functions and Data Members
705		1501
706	=over 4	1502	=over 4
707		1503
708	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1504	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
709		1505
710	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1506	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
711		1507
712	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1508	Configure the timer to trigger after C<after> seconds. If C<repeat>
713	C<0.>, then it will automatically be stopped. If it is positive, then the	1509	is C<0.>, then it will automatically be stopped once the timeout is
714	timer will automatically be configured to trigger again C<repeat> seconds	1510	reached. If it is positive, then the timer will automatically be
715	later, again, and again, until stopped manually.	1511	configured to trigger again C<repeat> seconds later, again, and again,
		1512	until stopped manually.
716		1513
717	The timer itself will do a best-effort at avoiding drift, that is, if you	1514	The timer itself will do a best-effort at avoiding drift, that is, if
718	configure a timer to trigger every 10 seconds, then it will trigger at	1515	you configure a timer to trigger every 10 seconds, then it will normally
719	exactly 10 second intervals. If, however, your program cannot keep up with	1516	trigger at exactly 10 second intervals. If, however, your program cannot
720	the timer (because it takes longer than those 10 seconds to do stuff) the	1517	keep up with the timer (because it takes longer than those 10 seconds to
721	timer will not fire more than once per event loop iteration.	1518	do stuff) the timer will not fire more than once per event loop iteration.
722		1519
723	=item ev_timer_again (loop)	1520	=item ev_timer_again (loop, ev_timer *)
724		1521
725	This will act as if the timer timed out and restart it again if it is	1522	This will act as if the timer timed out and restart it again if it is
726	repeating. The exact semantics are:	1523	repeating. The exact semantics are:
727		1524
		1525	If the timer is pending, its pending status is cleared.
		1526
728	If the timer is started but nonrepeating, stop it.	1527	If the timer is started but non-repeating, stop it (as if it timed out).
729		1528
730	If the timer is repeating, either start it if necessary (with the repeat	1529	If the timer is repeating, either start it if necessary (with the
731	value), or reset the running timer to the repeat value.	1530	C<repeat> value), or reset the running timer to the C<repeat> value.
732		1531
733	This sounds a bit complicated, but here is a useful and typical	1532	This sounds a bit complicated, see "Be smart about timeouts", above, for a
734	example: Imagine you have a tcp connection and you want a so-called idle	1533	usage example.
735	timeout, that is, you want to be called when there have been, say, 60	1534
736	seconds of inactivity on the socket. The easiest way to do this is to	1535	=item ev_tstamp repeat [read-write]
737	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	1536
738	time you successfully read or write some data. If you go into an idle	1537	The current C<repeat> value. Will be used each time the watcher times out
739	state where you do not expect data to travel on the socket, you can stop	1538	or C<ev_timer_again> is called, and determines the next timeout (if any),
740	the timer, and again will automatically restart it if need be.	1539	which is also when any modifications are taken into account.
741		1540
742	=back	1541	=back
743		1542
		1543	=head3 Examples
		1544
744	Example: create a timer that fires after 60 seconds.	1545	Example: Create a timer that fires after 60 seconds.
745		1546
746	static void	1547	static void
747	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1548	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
748	{	1549	{
749	.. one minute over, w is actually stopped right here	1550	.. one minute over, w is actually stopped right here
750	}	1551	}
751		1552
752	struct ev_timer mytimer;	1553	ev_timer mytimer;
753	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1554	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
754	ev_timer_start (loop, &mytimer);	1555	ev_timer_start (loop, &mytimer);
755		1556
756	Example: create a timeout timer that times out after 10 seconds of	1557	Example: Create a timeout timer that times out after 10 seconds of
757	inactivity.	1558	inactivity.
758		1559
759	static void	1560	static void
760	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1561	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
761	{	1562	{
762	.. ten seconds without any activity	1563	.. ten seconds without any activity
763	}	1564	}
764		1565
765	struct ev_timer mytimer;	1566	ev_timer mytimer;
766	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1567	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
767	ev_timer_again (&mytimer); /* start timer */	1568	ev_timer_again (&mytimer); /* start timer */
768	ev_loop (loop, 0);	1569	ev_loop (loop, 0);
769		1570
770	// and in some piece of code that gets executed on any "activity":	1571	// and in some piece of code that gets executed on any "activity":
771	// reset the timeout to start ticking again at 10 seconds	1572	// reset the timeout to start ticking again at 10 seconds
772	ev_timer_again (&mytimer);	1573	ev_timer_again (&mytimer);
773		1574
774		1575
775	=head2 C<ev_periodic> - to cron or not to cron	1576	=head2 C<ev_periodic> - to cron or not to cron?
776		1577
777	Periodic watchers are also timers of a kind, but they are very versatile	1578	Periodic watchers are also timers of a kind, but they are very versatile
778	(and unfortunately a bit complex).	1579	(and unfortunately a bit complex).
779		1580
780	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1581	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
781	but on wallclock time (absolute time). You can tell a periodic watcher	1582	but on wall clock time (absolute time). You can tell a periodic watcher
782	to trigger "at" some specific point in time. For example, if you tell a	1583	to trigger after some specific point in time. For example, if you tell a
783	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	1584	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()
784	+ 10.>) and then reset your system clock to the last year, then it will	1585	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1586	clock to January of the previous year, then it will take more than year
785	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1587	to trigger the event (unlike an C<ev_timer>, which would still trigger
786	roughly 10 seconds later and of course not if you reset your system time	1588	roughly 10 seconds later as it uses a relative timeout).
787	again).
788		1589
789	They can also be used to implement vastly more complex timers, such as	1590	C<ev_periodic>s can also be used to implement vastly more complex timers,
790	triggering an event on eahc midnight, local time.	1591	such as triggering an event on each "midnight, local time", or other
		1592	complicated rules.
791		1593
792	As with timers, the callback is guarenteed to be invoked only when the	1594	As with timers, the callback is guaranteed to be invoked only when the
793	time (C<at>) has been passed, but if multiple periodic timers become ready	1595	time (C<at>) has passed, but if multiple periodic timers become ready
794	during the same loop iteration then order of execution is undefined.	1596	during the same loop iteration, then order of execution is undefined.
		1597
		1598	=head3 Watcher-Specific Functions and Data Members
795		1599
796	=over 4	1600	=over 4
797		1601
798	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1602	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
799		1603
800	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1604	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
801		1605
802	Lots of arguments, lets sort it out... There are basically three modes of	1606	Lots of arguments, lets sort it out... There are basically three modes of
803	operation, and we will explain them from simplest to complex:	1607	operation, and we will explain them from simplest to most complex:
804		1608
805	=over 4	1609	=over 4
806		1610
807	=item * absolute timer (interval = reschedule_cb = 0)	1611	=item * absolute timer (at = time, interval = reschedule_cb = 0)
808		1612
809	In this configuration the watcher triggers an event at the wallclock time	1613	In this configuration the watcher triggers an event after the wall clock
810	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1614	time C<at> has passed. It will not repeat and will not adjust when a time
811	that is, if it is to be run at January 1st 2011 then it will run when the	1615	jump occurs, that is, if it is to be run at January 1st 2011 then it will
812	system time reaches or surpasses this time.	1616	only run when the system clock reaches or surpasses this time.
813		1617
814	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1618	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
815		1619
816	In this mode the watcher will always be scheduled to time out at the next	1620	In this mode the watcher will always be scheduled to time out at the next
817	C<at + N * interval> time (for some integer N) and then repeat, regardless	1621	C<at + N * interval> time (for some integer N, which can also be negative)
818	of any time jumps.	1622	and then repeat, regardless of any time jumps.
819		1623
820	This can be used to create timers that do not drift with respect to system	1624	This can be used to create timers that do not drift with respect to the
821	time:	1625	system clock, for example, here is a C<ev_periodic> that triggers each
		1626	hour, on the hour:
822		1627
823	ev_periodic_set (&periodic, 0., 3600., 0);	1628	ev_periodic_set (&periodic, 0., 3600., 0);
824		1629
825	This doesn't mean there will always be 3600 seconds in between triggers,	1630	This doesn't mean there will always be 3600 seconds in between triggers,
826	but only that the the callback will be called when the system time shows a	1631	but only that the callback will be called when the system time shows a
827	full hour (UTC), or more correctly, when the system time is evenly divisible	1632	full hour (UTC), or more correctly, when the system time is evenly divisible
828	by 3600.	1633	by 3600.
829		1634
830	Another way to think about it (for the mathematically inclined) is that	1635	Another way to think about it (for the mathematically inclined) is that
831	C<ev_periodic> will try to run the callback in this mode at the next possible	1636	C<ev_periodic> will try to run the callback in this mode at the next possible
832	time where C<time = at (mod interval)>, regardless of any time jumps.	1637	time where C<time = at (mod interval)>, regardless of any time jumps.
833		1638
		1639	For numerical stability it is preferable that the C<at> value is near
		1640	C<ev_now ()> (the current time), but there is no range requirement for
		1641	this value, and in fact is often specified as zero.
		1642
		1643	Note also that there is an upper limit to how often a timer can fire (CPU
		1644	speed for example), so if C<interval> is very small then timing stability
		1645	will of course deteriorate. Libev itself tries to be exact to be about one
		1646	millisecond (if the OS supports it and the machine is fast enough).
		1647
834	=item * manual reschedule mode (reschedule_cb = callback)	1648	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
835		1649
836	In this mode the values for C<interval> and C<at> are both being	1650	In this mode the values for C<interval> and C<at> are both being
837	ignored. Instead, each time the periodic watcher gets scheduled, the	1651	ignored. Instead, each time the periodic watcher gets scheduled, the
838	reschedule callback will be called with the watcher as first, and the	1652	reschedule callback will be called with the watcher as first, and the
839	current time as second argument.	1653	current time as second argument.
840		1654
841	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1655	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
842	ever, or make any event loop modifications>. If you need to stop it,	1656	ever, or make ANY event loop modifications whatsoever>.
843	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
844	starting a prepare watcher).
845		1657
		1658	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1659	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1660	only event loop modification you are allowed to do).
		1661
846	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1662	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
847	ev_tstamp now)>, e.g.:	1663	*w, ev_tstamp now)>, e.g.:
848		1664
		1665	static ev_tstamp
849	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1666	my_rescheduler (ev_periodic *w, ev_tstamp now)
850	{	1667	{
851	return now + 60.;	1668	return now + 60.;
852	}	1669	}
853		1670
854	It must return the next time to trigger, based on the passed time value	1671	It must return the next time to trigger, based on the passed time value
855	(that is, the lowest time value larger than to the second argument). It	1672	(that is, the lowest time value larger than to the second argument). It
856	will usually be called just before the callback will be triggered, but	1673	will usually be called just before the callback will be triggered, but
857	might be called at other times, too.	1674	might be called at other times, too.
858		1675
859	NOTE: I<< This callback must always return a time that is later than the	1676	NOTE: I<< This callback must always return a time that is higher than or
860	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1677	equal to the passed C<now> value >>.
861		1678
862	This can be used to create very complex timers, such as a timer that	1679	This can be used to create very complex timers, such as a timer that
863	triggers on each midnight, local time. To do this, you would calculate the	1680	triggers on "next midnight, local time". To do this, you would calculate the
864	next midnight after C<now> and return the timestamp value for this. How	1681	next midnight after C<now> and return the timestamp value for this. How
865	you do this is, again, up to you (but it is not trivial, which is the main	1682	you do this is, again, up to you (but it is not trivial, which is the main
866	reason I omitted it as an example).	1683	reason I omitted it as an example).
867		1684
868	=back	1685	=back
…		…
872	Simply stops and restarts the periodic watcher again. This is only useful	1689	Simply stops and restarts the periodic watcher again. This is only useful
873	when you changed some parameters or the reschedule callback would return	1690	when you changed some parameters or the reschedule callback would return
874	a different time than the last time it was called (e.g. in a crond like	1691	a different time than the last time it was called (e.g. in a crond like
875	program when the crontabs have changed).	1692	program when the crontabs have changed).
876		1693
		1694	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1695
		1696	When active, returns the absolute time that the watcher is supposed to
		1697	trigger next.
		1698
		1699	=item ev_tstamp offset [read-write]
		1700
		1701	When repeating, this contains the offset value, otherwise this is the
		1702	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1703
		1704	Can be modified any time, but changes only take effect when the periodic
		1705	timer fires or C<ev_periodic_again> is being called.
		1706
		1707	=item ev_tstamp interval [read-write]
		1708
		1709	The current interval value. Can be modified any time, but changes only
		1710	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1711	called.
		1712
		1713	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
		1714
		1715	The current reschedule callback, or C<0>, if this functionality is
		1716	switched off. Can be changed any time, but changes only take effect when
		1717	the periodic timer fires or C<ev_periodic_again> is being called.
		1718
877	=back	1719	=back
878		1720
		1721	=head3 Examples
		1722
879	Example: call a callback every hour, or, more precisely, whenever the	1723	Example: Call a callback every hour, or, more precisely, whenever the
880	system clock is divisible by 3600. The callback invocation times have	1724	system time is divisible by 3600. The callback invocation times have
881	potentially a lot of jittering, but good long-term stability.	1725	potentially a lot of jitter, but good long-term stability.
882		1726
883	static void	1727	static void
884	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1728	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
885	{	1729	{
886	... its now a full hour (UTC, or TAI or whatever your clock follows)	1730	... its now a full hour (UTC, or TAI or whatever your clock follows)
887	}	1731	}
888		1732
889	struct ev_periodic hourly_tick;	1733	ev_periodic hourly_tick;
890	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1734	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
891	ev_periodic_start (loop, &hourly_tick);	1735	ev_periodic_start (loop, &hourly_tick);
892		1736
893	Example: the same as above, but use a reschedule callback to do it:	1737	Example: The same as above, but use a reschedule callback to do it:
894		1738
895	#include <math.h>	1739	#include <math.h>
896		1740
897	static ev_tstamp	1741	static ev_tstamp
898	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1742	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
899	{	1743	{
900	return fmod (now, 3600.) + 3600.;	1744	return now + (3600. - fmod (now, 3600.));
901	}	1745	}
902		1746
903	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1747	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
904		1748
905	Example: call a callback every hour, starting now:	1749	Example: Call a callback every hour, starting now:
906		1750
907	struct ev_periodic hourly_tick;	1751	ev_periodic hourly_tick;
908	ev_periodic_init (&hourly_tick, clock_cb,	1752	ev_periodic_init (&hourly_tick, clock_cb,
909	fmod (ev_now (loop), 3600.), 3600., 0);	1753	fmod (ev_now (loop), 3600.), 3600., 0);
910	ev_periodic_start (loop, &hourly_tick);	1754	ev_periodic_start (loop, &hourly_tick);
911		1755
912		1756
913	=head2 C<ev_signal> - signal me when a signal gets signalled	1757	=head2 C<ev_signal> - signal me when a signal gets signalled!
914		1758
915	Signal watchers will trigger an event when the process receives a specific	1759	Signal watchers will trigger an event when the process receives a specific
916	signal one or more times. Even though signals are very asynchronous, libev	1760	signal one or more times. Even though signals are very asynchronous, libev
917	will try it's best to deliver signals synchronously, i.e. as part of the	1761	will try it's best to deliver signals synchronously, i.e. as part of the
918	normal event processing, like any other event.	1762	normal event processing, like any other event.
919		1763
		1764	If you want signals asynchronously, just use C<sigaction> as you would
		1765	do without libev and forget about sharing the signal. You can even use
		1766	C<ev_async> from a signal handler to synchronously wake up an event loop.
		1767
920	You can configure as many watchers as you like per signal. Only when the	1768	You can configure as many watchers as you like per signal. Only when the
921	first watcher gets started will libev actually register a signal watcher	1769	first watcher gets started will libev actually register a signal handler
922	with the kernel (thus it coexists with your own signal handlers as long	1770	with the kernel (thus it coexists with your own signal handlers as long as
923	as you don't register any with libev). Similarly, when the last signal	1771	you don't register any with libev for the same signal). Similarly, when
924	watcher for a signal is stopped libev will reset the signal handler to	1772	the last signal watcher for a signal is stopped, libev will reset the
925	SIG_DFL (regardless of what it was set to before).	1773	signal handler to SIG_DFL (regardless of what it was set to before).
		1774
		1775	If possible and supported, libev will install its handlers with
		1776	C<SA_RESTART> behaviour enabled, so system calls should not be unduly
		1777	interrupted. If you have a problem with system calls getting interrupted by
		1778	signals you can block all signals in an C<ev_check> watcher and unblock
		1779	them in an C<ev_prepare> watcher.
		1780
		1781	=head3 Watcher-Specific Functions and Data Members
926		1782
927	=over 4	1783	=over 4
928		1784
929	=item ev_signal_init (ev_signal *, callback, int signum)	1785	=item ev_signal_init (ev_signal *, callback, int signum)
930		1786
931	=item ev_signal_set (ev_signal *, int signum)	1787	=item ev_signal_set (ev_signal *, int signum)
932		1788
933	Configures the watcher to trigger on the given signal number (usually one	1789	Configures the watcher to trigger on the given signal number (usually one
934	of the C<SIGxxx> constants).	1790	of the C<SIGxxx> constants).
935		1791
		1792	=item int signum [read-only]
		1793
		1794	The signal the watcher watches out for.
		1795
936	=back	1796	=back
937		1797
		1798	=head3 Examples
938		1799
		1800	Example: Try to exit cleanly on SIGINT.
		1801
		1802	static void
		1803	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		1804	{
		1805	ev_unloop (loop, EVUNLOOP_ALL);
		1806	}
		1807
		1808	ev_signal signal_watcher;
		1809	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1810	ev_signal_start (loop, &signal_watcher);
		1811
		1812
939	=head2 C<ev_child> - wait for pid status changes	1813	=head2 C<ev_child> - watch out for process status changes
940		1814
941	Child watchers trigger when your process receives a SIGCHLD in response to	1815	Child watchers trigger when your process receives a SIGCHLD in response to
942	some child status changes (most typically when a child of yours dies).	1816	some child status changes (most typically when a child of yours dies or
		1817	exits). It is permissible to install a child watcher I<after> the child
		1818	has been forked (which implies it might have already exited), as long
		1819	as the event loop isn't entered (or is continued from a watcher), i.e.,
		1820	forking and then immediately registering a watcher for the child is fine,
		1821	but forking and registering a watcher a few event loop iterations later is
		1822	not.
		1823
		1824	Only the default event loop is capable of handling signals, and therefore
		1825	you can only register child watchers in the default event loop.
		1826
		1827	=head3 Process Interaction
		1828
		1829	Libev grabs C<SIGCHLD> as soon as the default event loop is
		1830	initialised. This is necessary to guarantee proper behaviour even if
		1831	the first child watcher is started after the child exits. The occurrence
		1832	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		1833	synchronously as part of the event loop processing. Libev always reaps all
		1834	children, even ones not watched.
		1835
		1836	=head3 Overriding the Built-In Processing
		1837
		1838	Libev offers no special support for overriding the built-in child
		1839	processing, but if your application collides with libev's default child
		1840	handler, you can override it easily by installing your own handler for
		1841	C<SIGCHLD> after initialising the default loop, and making sure the
		1842	default loop never gets destroyed. You are encouraged, however, to use an
		1843	event-based approach to child reaping and thus use libev's support for
		1844	that, so other libev users can use C<ev_child> watchers freely.
		1845
		1846	=head3 Stopping the Child Watcher
		1847
		1848	Currently, the child watcher never gets stopped, even when the
		1849	child terminates, so normally one needs to stop the watcher in the
		1850	callback. Future versions of libev might stop the watcher automatically
		1851	when a child exit is detected.
		1852
		1853	=head3 Watcher-Specific Functions and Data Members
943		1854
944	=over 4	1855	=over 4
945		1856
946	=item ev_child_init (ev_child *, callback, int pid)	1857	=item ev_child_init (ev_child *, callback, int pid, int trace)
947		1858
948	=item ev_child_set (ev_child *, int pid)	1859	=item ev_child_set (ev_child *, int pid, int trace)
949		1860
950	Configures the watcher to wait for status changes of process C<pid> (or	1861	Configures the watcher to wait for status changes of process C<pid> (or
951	I<any> process if C<pid> is specified as C<0>). The callback can look	1862	I<any> process if C<pid> is specified as C<0>). The callback can look
952	at the C<rstatus> member of the C<ev_child> watcher structure to see	1863	at the C<rstatus> member of the C<ev_child> watcher structure to see
953	the status word (use the macros from C<sys/wait.h> and see your systems	1864	the status word (use the macros from C<sys/wait.h> and see your systems
954	C<waitpid> documentation). The C<rpid> member contains the pid of the	1865	C<waitpid> documentation). The C<rpid> member contains the pid of the
955	process causing the status change.	1866	process causing the status change. C<trace> must be either C<0> (only
		1867	activate the watcher when the process terminates) or C<1> (additionally
		1868	activate the watcher when the process is stopped or continued).
		1869
		1870	=item int pid [read-only]
		1871
		1872	The process id this watcher watches out for, or C<0>, meaning any process id.
		1873
		1874	=item int rpid [read-write]
		1875
		1876	The process id that detected a status change.
		1877
		1878	=item int rstatus [read-write]
		1879
		1880	The process exit/trace status caused by C<rpid> (see your systems
		1881	C<waitpid> and C<sys/wait.h> documentation for details).
956		1882
957	=back	1883	=back
958		1884
959	Example: try to exit cleanly on SIGINT and SIGTERM.	1885	=head3 Examples
960		1886
		1887	Example: C<fork()> a new process and install a child handler to wait for
		1888	its completion.
		1889
		1890	ev_child cw;
		1891
961	static void	1892	static void
962	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	1893	child_cb (EV_P_ ev_child *w, int revents)
963	{	1894	{
964	ev_unloop (loop, EVUNLOOP_ALL);	1895	ev_child_stop (EV_A_ w);
		1896	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
965	}	1897	}
966		1898
967	struct ev_signal signal_watcher;	1899	pid_t pid = fork ();
968	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
969	ev_signal_start (loop, &sigint_cb);
970		1900
		1901	if (pid < 0)
		1902	// error
		1903	else if (pid == 0)
		1904	{
		1905	// the forked child executes here
		1906	exit (1);
		1907	}
		1908	else
		1909	{
		1910	ev_child_init (&cw, child_cb, pid, 0);
		1911	ev_child_start (EV_DEFAULT_ &cw);
		1912	}
971		1913
		1914
		1915	=head2 C<ev_stat> - did the file attributes just change?
		1916
		1917	This watches a file system path for attribute changes. That is, it calls
		1918	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1919	compared to the last time, invoking the callback if it did.
		1920
		1921	The path does not need to exist: changing from "path exists" to "path does
		1922	not exist" is a status change like any other. The condition "path does
		1923	not exist" is signified by the C<st_nlink> field being zero (which is
		1924	otherwise always forced to be at least one) and all the other fields of
		1925	the stat buffer having unspecified contents.
		1926
		1927	The path I<should> be absolute and I<must not> end in a slash. If it is
		1928	relative and your working directory changes, the behaviour is undefined.
		1929
		1930	Since there is no standard kernel interface to do this, the portable
		1931	implementation simply calls C<stat (2)> regularly on the path to see if
		1932	it changed somehow. You can specify a recommended polling interval for
		1933	this case. If you specify a polling interval of C<0> (highly recommended!)
		1934	then a I<suitable, unspecified default> value will be used (which
		1935	you can expect to be around five seconds, although this might change
		1936	dynamically). Libev will also impose a minimum interval which is currently
		1937	around C<0.1>, but thats usually overkill.
		1938
		1939	This watcher type is not meant for massive numbers of stat watchers,
		1940	as even with OS-supported change notifications, this can be
		1941	resource-intensive.
		1942
		1943	At the time of this writing, the only OS-specific interface implemented
		1944	is the Linux inotify interface (implementing kqueue support is left as
		1945	an exercise for the reader. Note, however, that the author sees no way
		1946	of implementing C<ev_stat> semantics with kqueue).
		1947
		1948	=head3 ABI Issues (Largefile Support)
		1949
		1950	Libev by default (unless the user overrides this) uses the default
		1951	compilation environment, which means that on systems with large file
		1952	support disabled by default, you get the 32 bit version of the stat
		1953	structure. When using the library from programs that change the ABI to
		1954	use 64 bit file offsets the programs will fail. In that case you have to
		1955	compile libev with the same flags to get binary compatibility. This is
		1956	obviously the case with any flags that change the ABI, but the problem is
		1957	most noticeably disabled with ev_stat and large file support.
		1958
		1959	The solution for this is to lobby your distribution maker to make large
		1960	file interfaces available by default (as e.g. FreeBSD does) and not
		1961	optional. Libev cannot simply switch on large file support because it has
		1962	to exchange stat structures with application programs compiled using the
		1963	default compilation environment.
		1964
		1965	=head3 Inotify and Kqueue
		1966
		1967	When C<inotify (7)> support has been compiled into libev (generally
		1968	only available with Linux 2.6.25 or above due to bugs in earlier
		1969	implementations) and present at runtime, it will be used to speed up
		1970	change detection where possible. The inotify descriptor will be created
		1971	lazily when the first C<ev_stat> watcher is being started.
		1972
		1973	Inotify presence does not change the semantics of C<ev_stat> watchers
		1974	except that changes might be detected earlier, and in some cases, to avoid
		1975	making regular C<stat> calls. Even in the presence of inotify support
		1976	there are many cases where libev has to resort to regular C<stat> polling,
		1977	but as long as the path exists, libev usually gets away without polling.
		1978
		1979	There is no support for kqueue, as apparently it cannot be used to
		1980	implement this functionality, due to the requirement of having a file
		1981	descriptor open on the object at all times, and detecting renames, unlinks
		1982	etc. is difficult.
		1983
		1984	=head3 The special problem of stat time resolution
		1985
		1986	The C<stat ()> system call only supports full-second resolution portably, and
		1987	even on systems where the resolution is higher, most file systems still
		1988	only support whole seconds.
		1989
		1990	That means that, if the time is the only thing that changes, you can
		1991	easily miss updates: on the first update, C<ev_stat> detects a change and
		1992	calls your callback, which does something. When there is another update
		1993	within the same second, C<ev_stat> will be unable to detect unless the
		1994	stat data does change in other ways (e.g. file size).
		1995
		1996	The solution to this is to delay acting on a change for slightly more
		1997	than a second (or till slightly after the next full second boundary), using
		1998	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		1999	ev_timer_again (loop, w)>).
		2000
		2001	The C<.02> offset is added to work around small timing inconsistencies
		2002	of some operating systems (where the second counter of the current time
		2003	might be be delayed. One such system is the Linux kernel, where a call to
		2004	C<gettimeofday> might return a timestamp with a full second later than
		2005	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2006	update file times then there will be a small window where the kernel uses
		2007	the previous second to update file times but libev might already execute
		2008	the timer callback).
		2009
		2010	=head3 Watcher-Specific Functions and Data Members
		2011
		2012	=over 4
		2013
		2014	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		2015
		2016	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		2017
		2018	Configures the watcher to wait for status changes of the given
		2019	C<path>. The C<interval> is a hint on how quickly a change is expected to
		2020	be detected and should normally be specified as C<0> to let libev choose
		2021	a suitable value. The memory pointed to by C<path> must point to the same
		2022	path for as long as the watcher is active.
		2023
		2024	The callback will receive an C<EV_STAT> event when a change was detected,
		2025	relative to the attributes at the time the watcher was started (or the
		2026	last change was detected).
		2027
		2028	=item ev_stat_stat (loop, ev_stat *)
		2029
		2030	Updates the stat buffer immediately with new values. If you change the
		2031	watched path in your callback, you could call this function to avoid
		2032	detecting this change (while introducing a race condition if you are not
		2033	the only one changing the path). Can also be useful simply to find out the
		2034	new values.
		2035
		2036	=item ev_statdata attr [read-only]
		2037
		2038	The most-recently detected attributes of the file. Although the type is
		2039	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		2040	suitable for your system, but you can only rely on the POSIX-standardised
		2041	members to be present. If the C<st_nlink> member is C<0>, then there was
		2042	some error while C<stat>ing the file.
		2043
		2044	=item ev_statdata prev [read-only]
		2045
		2046	The previous attributes of the file. The callback gets invoked whenever
		2047	C<prev> != C<attr>, or, more precisely, one or more of these members
		2048	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2049	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
		2050
		2051	=item ev_tstamp interval [read-only]
		2052
		2053	The specified interval.
		2054
		2055	=item const char *path [read-only]
		2056
		2057	The file system path that is being watched.
		2058
		2059	=back
		2060
		2061	=head3 Examples
		2062
		2063	Example: Watch C</etc/passwd> for attribute changes.
		2064
		2065	static void
		2066	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		2067	{
		2068	/* /etc/passwd changed in some way */
		2069	if (w->attr.st_nlink)
		2070	{
		2071	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		2072	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		2073	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		2074	}
		2075	else
		2076	/* you shalt not abuse printf for puts */
		2077	puts ("wow, /etc/passwd is not there, expect problems. "
		2078	"if this is windows, they already arrived\n");
		2079	}
		2080
		2081	...
		2082	ev_stat passwd;
		2083
		2084	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
		2085	ev_stat_start (loop, &passwd);
		2086
		2087	Example: Like above, but additionally use a one-second delay so we do not
		2088	miss updates (however, frequent updates will delay processing, too, so
		2089	one might do the work both on C<ev_stat> callback invocation I<and> on
		2090	C<ev_timer> callback invocation).
		2091
		2092	static ev_stat passwd;
		2093	static ev_timer timer;
		2094
		2095	static void
		2096	timer_cb (EV_P_ ev_timer *w, int revents)
		2097	{
		2098	ev_timer_stop (EV_A_ w);
		2099
		2100	/* now it's one second after the most recent passwd change */
		2101	}
		2102
		2103	static void
		2104	stat_cb (EV_P_ ev_stat *w, int revents)
		2105	{
		2106	/* reset the one-second timer */
		2107	ev_timer_again (EV_A_ &timer);
		2108	}
		2109
		2110	...
		2111	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		2112	ev_stat_start (loop, &passwd);
		2113	ev_timer_init (&timer, timer_cb, 0., 1.02);
		2114
		2115
972	=head2 C<ev_idle> - when you've got nothing better to do	2116	=head2 C<ev_idle> - when you've got nothing better to do...
973		2117
974	Idle watchers trigger events when there are no other events are pending	2118	Idle watchers trigger events when no other events of the same or higher
975	(prepare, check and other idle watchers do not count). That is, as long	2119	priority are pending (prepare, check and other idle watchers do not count
976	as your process is busy handling sockets or timeouts (or even signals,	2120	as receiving "events").
977	imagine) it will not be triggered. But when your process is idle all idle	2121
978	watchers are being called again and again, once per event loop iteration -	2122	That is, as long as your process is busy handling sockets or timeouts
		2123	(or even signals, imagine) of the same or higher priority it will not be
		2124	triggered. But when your process is idle (or only lower-priority watchers
		2125	are pending), the idle watchers are being called once per event loop
979	until stopped, that is, or your process receives more events and becomes	2126	iteration - until stopped, that is, or your process receives more events
980	busy.	2127	and becomes busy again with higher priority stuff.
981		2128
982	The most noteworthy effect is that as long as any idle watchers are	2129	The most noteworthy effect is that as long as any idle watchers are
983	active, the process will not block when waiting for new events.	2130	active, the process will not block when waiting for new events.
984		2131
985	Apart from keeping your process non-blocking (which is a useful	2132	Apart from keeping your process non-blocking (which is a useful
986	effect on its own sometimes), idle watchers are a good place to do	2133	effect on its own sometimes), idle watchers are a good place to do
987	"pseudo-background processing", or delay processing stuff to after the	2134	"pseudo-background processing", or delay processing stuff to after the
988	event loop has handled all outstanding events.	2135	event loop has handled all outstanding events.
989		2136
		2137	=head3 Watcher-Specific Functions and Data Members
		2138
990	=over 4	2139	=over 4
991		2140
992	=item ev_idle_init (ev_signal *, callback)	2141	=item ev_idle_init (ev_signal *, callback)
993		2142
994	Initialises and configures the idle watcher - it has no parameters of any	2143	Initialises and configures the idle watcher - it has no parameters of any
995	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2144	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
996	believe me.	2145	believe me.
997		2146
998	=back	2147	=back
999		2148
		2149	=head3 Examples
		2150
1000	Example: dynamically allocate an C<ev_idle>, start it, and in the	2151	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1001	callback, free it. Alos, use no error checking, as usual.	2152	callback, free it. Also, use no error checking, as usual.
1002		2153
1003	static void	2154	static void
1004	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2155	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1005	{	2156	{
1006	free (w);	2157	free (w);
1007	// now do something you wanted to do when the program has	2158	// now do something you wanted to do when the program has
1008	// no longer asnything immediate to do.	2159	// no longer anything immediate to do.
1009	}	2160	}
1010		2161
1011	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2162	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1012	ev_idle_init (idle_watcher, idle_cb);	2163	ev_idle_init (idle_watcher, idle_cb);
1013	ev_idle_start (loop, idle_cb);	2164	ev_idle_start (loop, idle_cb);
1014		2165
1015		2166
1016	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	2167	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1017		2168
1018	Prepare and check watchers are usually (but not always) used in tandem:	2169	Prepare and check watchers are usually (but not always) used in pairs:
1019	prepare watchers get invoked before the process blocks and check watchers	2170	prepare watchers get invoked before the process blocks and check watchers
1020	afterwards.	2171	afterwards.
1021		2172
		2173	You I<must not> call C<ev_loop> or similar functions that enter
		2174	the current event loop from either C<ev_prepare> or C<ev_check>
		2175	watchers. Other loops than the current one are fine, however. The
		2176	rationale behind this is that you do not need to check for recursion in
		2177	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		2178	C<ev_check> so if you have one watcher of each kind they will always be
		2179	called in pairs bracketing the blocking call.
		2180
1022	Their main purpose is to integrate other event mechanisms into libev and	2181	Their main purpose is to integrate other event mechanisms into libev and
1023	their use is somewhat advanced. This could be used, for example, to track	2182	their use is somewhat advanced. They could be used, for example, to track
1024	variable changes, implement your own watchers, integrate net-snmp or a	2183	variable changes, implement your own watchers, integrate net-snmp or a
1025	coroutine library and lots more.	2184	coroutine library and lots more. They are also occasionally useful if
		2185	you cache some data and want to flush it before blocking (for example,
		2186	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		2187	watcher).
1026		2188
1027	This is done by examining in each prepare call which file descriptors need	2189	This is done by examining in each prepare call which file descriptors
1028	to be watched by the other library, registering C<ev_io> watchers for	2190	need to be watched by the other library, registering C<ev_io> watchers
1029	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2191	for them and starting an C<ev_timer> watcher for any timeouts (many
1030	provide just this functionality). Then, in the check watcher you check for	2192	libraries provide exactly this functionality). Then, in the check watcher,
1031	any events that occured (by checking the pending status of all watchers	2193	you check for any events that occurred (by checking the pending status
1032	and stopping them) and call back into the library. The I/O and timer	2194	of all watchers and stopping them) and call back into the library. The
1033	callbacks will never actually be called (but must be valid nevertheless,	2195	I/O and timer callbacks will never actually be called (but must be valid
1034	because you never know, you know?).	2196	nevertheless, because you never know, you know?).
1035		2197
1036	As another example, the Perl Coro module uses these hooks to integrate	2198	As another example, the Perl Coro module uses these hooks to integrate
1037	coroutines into libev programs, by yielding to other active coroutines	2199	coroutines into libev programs, by yielding to other active coroutines
1038	during each prepare and only letting the process block if no coroutines	2200	during each prepare and only letting the process block if no coroutines
1039	are ready to run (it's actually more complicated: it only runs coroutines	2201	are ready to run (it's actually more complicated: it only runs coroutines
1040	with priority higher than or equal to the event loop and one coroutine	2202	with priority higher than or equal to the event loop and one coroutine
1041	of lower priority, but only once, using idle watchers to keep the event	2203	of lower priority, but only once, using idle watchers to keep the event
1042	loop from blocking if lower-priority coroutines are active, thus mapping	2204	loop from blocking if lower-priority coroutines are active, thus mapping
1043	low-priority coroutines to idle/background tasks).	2205	low-priority coroutines to idle/background tasks).
1044		2206
		2207	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		2208	priority, to ensure that they are being run before any other watchers
		2209	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2210
		2211	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
		2212	activate ("feed") events into libev. While libev fully supports this, they
		2213	might get executed before other C<ev_check> watchers did their job. As
		2214	C<ev_check> watchers are often used to embed other (non-libev) event
		2215	loops those other event loops might be in an unusable state until their
		2216	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		2217	others).
		2218
		2219	=head3 Watcher-Specific Functions and Data Members
		2220
1045	=over 4	2221	=over 4
1046		2222
1047	=item ev_prepare_init (ev_prepare *, callback)	2223	=item ev_prepare_init (ev_prepare *, callback)
1048		2224
1049	=item ev_check_init (ev_check *, callback)	2225	=item ev_check_init (ev_check *, callback)
1050		2226
1051	Initialises and configures the prepare or check watcher - they have no	2227	Initialises and configures the prepare or check watcher - they have no
1052	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2228	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1053	macros, but using them is utterly, utterly and completely pointless.	2229	macros, but using them is utterly, utterly, utterly and completely
		2230	pointless.
1054		2231
1055	=back	2232	=back
1056		2233
1057	Example: *TODO*.	2234	=head3 Examples
1058		2235
		2236	There are a number of principal ways to embed other event loops or modules
		2237	into libev. Here are some ideas on how to include libadns into libev
		2238	(there is a Perl module named C<EV::ADNS> that does this, which you could
		2239	use as a working example. Another Perl module named C<EV::Glib> embeds a
		2240	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		2241	Glib event loop).
1059		2242
		2243	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		2244	and in a check watcher, destroy them and call into libadns. What follows
		2245	is pseudo-code only of course. This requires you to either use a low
		2246	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		2247	the callbacks for the IO/timeout watchers might not have been called yet.
		2248
		2249	static ev_io iow [nfd];
		2250	static ev_timer tw;
		2251
		2252	static void
		2253	io_cb (struct ev_loop *loop, ev_io *w, int revents)
		2254	{
		2255	}
		2256
		2257	// create io watchers for each fd and a timer before blocking
		2258	static void
		2259	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
		2260	{
		2261	int timeout = 3600000;
		2262	struct pollfd fds [nfd];
		2263	// actual code will need to loop here and realloc etc.
		2264	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		2265
		2266	/* the callback is illegal, but won't be called as we stop during check */
		2267	ev_timer_init (&tw, 0, timeout * 1e-3);
		2268	ev_timer_start (loop, &tw);
		2269
		2270	// create one ev_io per pollfd
		2271	for (int i = 0; i < nfd; ++i)
		2272	{
		2273	ev_io_init (iow + i, io_cb, fds [i].fd,
		2274	((fds [i].events & POLLIN ? EV_READ : 0)
		2275	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		2276
		2277	fds [i].revents = 0;
		2278	ev_io_start (loop, iow + i);
		2279	}
		2280	}
		2281
		2282	// stop all watchers after blocking
		2283	static void
		2284	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
		2285	{
		2286	ev_timer_stop (loop, &tw);
		2287
		2288	for (int i = 0; i < nfd; ++i)
		2289	{
		2290	// set the relevant poll flags
		2291	// could also call adns_processreadable etc. here
		2292	struct pollfd *fd = fds + i;
		2293	int revents = ev_clear_pending (iow + i);
		2294	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		2295	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		2296
		2297	// now stop the watcher
		2298	ev_io_stop (loop, iow + i);
		2299	}
		2300
		2301	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		2302	}
		2303
		2304	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		2305	in the prepare watcher and would dispose of the check watcher.
		2306
		2307	Method 3: If the module to be embedded supports explicit event
		2308	notification (libadns does), you can also make use of the actual watcher
		2309	callbacks, and only destroy/create the watchers in the prepare watcher.
		2310
		2311	static void
		2312	timer_cb (EV_P_ ev_timer *w, int revents)
		2313	{
		2314	adns_state ads = (adns_state)w->data;
		2315	update_now (EV_A);
		2316
		2317	adns_processtimeouts (ads, &tv_now);
		2318	}
		2319
		2320	static void
		2321	io_cb (EV_P_ ev_io *w, int revents)
		2322	{
		2323	adns_state ads = (adns_state)w->data;
		2324	update_now (EV_A);
		2325
		2326	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		2327	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		2328	}
		2329
		2330	// do not ever call adns_afterpoll
		2331
		2332	Method 4: Do not use a prepare or check watcher because the module you
		2333	want to embed is not flexible enough to support it. Instead, you can
		2334	override their poll function. The drawback with this solution is that the
		2335	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
		2336	this approach, effectively embedding EV as a client into the horrible
		2337	libglib event loop.
		2338
		2339	static gint
		2340	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		2341	{
		2342	int got_events = 0;
		2343
		2344	for (n = 0; n < nfds; ++n)
		2345	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		2346
		2347	if (timeout >= 0)
		2348	// create/start timer
		2349
		2350	// poll
		2351	ev_loop (EV_A_ 0);
		2352
		2353	// stop timer again
		2354	if (timeout >= 0)
		2355	ev_timer_stop (EV_A_ &to);
		2356
		2357	// stop io watchers again - their callbacks should have set
		2358	for (n = 0; n < nfds; ++n)
		2359	ev_io_stop (EV_A_ iow [n]);
		2360
		2361	return got_events;
		2362	}
		2363
		2364
1060	=head2 C<ev_embed> - when one backend isn't enough	2365	=head2 C<ev_embed> - when one backend isn't enough...
1061		2366
1062	This is a rather advanced watcher type that lets you embed one event loop	2367	This is a rather advanced watcher type that lets you embed one event loop
1063	into another.	2368	into another (currently only C<ev_io> events are supported in the embedded
		2369	loop, other types of watchers might be handled in a delayed or incorrect
		2370	fashion and must not be used).
1064		2371
1065	There are primarily two reasons you would want that: work around bugs and	2372	There are primarily two reasons you would want that: work around bugs and
1066	prioritise I/O.	2373	prioritise I/O.
1067		2374
1068	As an example for a bug workaround, the kqueue backend might only support	2375	As an example for a bug workaround, the kqueue backend might only support
1069	sockets on some platform, so it is unusable as generic backend, but you	2376	sockets on some platform, so it is unusable as generic backend, but you
1070	still want to make use of it because you have many sockets and it scales	2377	still want to make use of it because you have many sockets and it scales
1071	so nicely. In this case, you would create a kqueue-based loop and embed it	2378	so nicely. In this case, you would create a kqueue-based loop and embed
1072	into your default loop (which might use e.g. poll). Overall operation will	2379	it into your default loop (which might use e.g. poll). Overall operation
1073	be a bit slower because first libev has to poll and then call kevent, but	2380	will be a bit slower because first libev has to call C<poll> and then
1074	at least you can use both at what they are best.	2381	C<kevent>, but at least you can use both mechanisms for what they are
		2382	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1075		2383
1076	As for prioritising I/O: rarely you have the case where some fds have	2384	As for prioritising I/O: under rare circumstances you have the case where
1077	to be watched and handled very quickly (with low latency), and even	2385	some fds have to be watched and handled very quickly (with low latency),
1078	priorities and idle watchers might have too much overhead. In this case	2386	and even priorities and idle watchers might have too much overhead. In
1079	you would put all the high priority stuff in one loop and all the rest in	2387	this case you would put all the high priority stuff in one loop and all
1080	a second one, and embed the second one in the first.	2388	the rest in a second one, and embed the second one in the first.
		2389
		2390	As long as the watcher is active, the callback will be invoked every time
		2391	there might be events pending in the embedded loop. The callback must then
		2392	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
		2393	their callbacks (you could also start an idle watcher to give the embedded
		2394	loop strictly lower priority for example). You can also set the callback
		2395	to C<0>, in which case the embed watcher will automatically execute the
		2396	embedded loop sweep.
1081		2397
1082	As long as the watcher is started it will automatically handle events. The	2398	As long as the watcher is started it will automatically handle events. The
1083	callback will be invoked whenever some events have been handled. You can	2399	callback will be invoked whenever some events have been handled. You can
1084	set the callback to C<0> to avoid having to specify one if you are not	2400	set the callback to C<0> to avoid having to specify one if you are not
1085	interested in that.	2401	interested in that.
1086		2402
1087	Also, there have not currently been made special provisions for forking:	2403	Also, there have not currently been made special provisions for forking:
1088	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2404	when you fork, you not only have to call C<ev_loop_fork> on both loops,
1089	but you will also have to stop and restart any C<ev_embed> watchers	2405	but you will also have to stop and restart any C<ev_embed> watchers
1090	yourself.	2406	yourself - but you can use a fork watcher to handle this automatically,
		2407	and future versions of libev might do just that.
1091		2408
1092	Unfortunately, not all backends are embeddable, only the ones returned by	2409	Unfortunately, not all backends are embeddable: only the ones returned by
1093	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2410	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1094	portable one.	2411	portable one.
1095		2412
1096	So when you want to use this feature you will always have to be prepared	2413	So when you want to use this feature you will always have to be prepared
1097	that you cannot get an embeddable loop. The recommended way to get around	2414	that you cannot get an embeddable loop. The recommended way to get around
1098	this is to have a separate variables for your embeddable loop, try to	2415	this is to have a separate variables for your embeddable loop, try to
1099	create it, and if that fails, use the normal loop for everything:	2416	create it, and if that fails, use the normal loop for everything.
1100		2417
1101	struct ev_loop *loop_hi = ev_default_init (0);	2418	=head3 C<ev_embed> and fork
1102	struct ev_loop *loop_lo = 0;
1103	struct ev_embed embed;
1104
1105	// see if there is a chance of getting one that works
1106	// (remember that a flags value of 0 means autodetection)
1107	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1108	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1109	: 0;
1110		2419
1111	// if we got one, then embed it, otherwise default to loop_hi	2420	While the C<ev_embed> watcher is running, forks in the embedding loop will
1112	if (loop_lo)	2421	automatically be applied to the embedded loop as well, so no special
1113	{	2422	fork handling is required in that case. When the watcher is not running,
1114	ev_embed_init (&embed, 0, loop_lo);	2423	however, it is still the task of the libev user to call C<ev_loop_fork ()>
1115	ev_embed_start (loop_hi, &embed);	2424	as applicable.
1116	}	2425
1117	else	2426	=head3 Watcher-Specific Functions and Data Members
1118	loop_lo = loop_hi;
1119		2427
1120	=over 4	2428	=over 4
1121		2429
1122	=item ev_embed_init (ev_embed *, callback, struct ev_loop *loop)	2430	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
1123		2431
1124	=item ev_embed_set (ev_embed *, callback, struct ev_loop *loop)	2432	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
1125		2433
1126	Configures the watcher to embed the given loop, which must be embeddable.	2434	Configures the watcher to embed the given loop, which must be
		2435	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		2436	invoked automatically, otherwise it is the responsibility of the callback
		2437	to invoke it (it will continue to be called until the sweep has been done,
		2438	if you do not want that, you need to temporarily stop the embed watcher).
		2439
		2440	=item ev_embed_sweep (loop, ev_embed *)
		2441
		2442	Make a single, non-blocking sweep over the embedded loop. This works
		2443	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		2444	appropriate way for embedded loops.
		2445
		2446	=item struct ev_loop *other [read-only]
		2447
		2448	The embedded event loop.
1127		2449
1128	=back	2450	=back
1129		2451
		2452	=head3 Examples
		2453
		2454	Example: Try to get an embeddable event loop and embed it into the default
		2455	event loop. If that is not possible, use the default loop. The default
		2456	loop is stored in C<loop_hi>, while the embeddable loop is stored in
		2457	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
		2458	used).
		2459
		2460	struct ev_loop *loop_hi = ev_default_init (0);
		2461	struct ev_loop *loop_lo = 0;
		2462	ev_embed embed;
		2463
		2464	// see if there is a chance of getting one that works
		2465	// (remember that a flags value of 0 means autodetection)
		2466	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		2467	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		2468	: 0;
		2469
		2470	// if we got one, then embed it, otherwise default to loop_hi
		2471	if (loop_lo)
		2472	{
		2473	ev_embed_init (&embed, 0, loop_lo);
		2474	ev_embed_start (loop_hi, &embed);
		2475	}
		2476	else
		2477	loop_lo = loop_hi;
		2478
		2479	Example: Check if kqueue is available but not recommended and create
		2480	a kqueue backend for use with sockets (which usually work with any
		2481	kqueue implementation). Store the kqueue/socket-only event loop in
		2482	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		2483
		2484	struct ev_loop *loop = ev_default_init (0);
		2485	struct ev_loop *loop_socket = 0;
		2486	ev_embed embed;
		2487
		2488	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2489	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2490	{
		2491	ev_embed_init (&embed, 0, loop_socket);
		2492	ev_embed_start (loop, &embed);
		2493	}
		2494
		2495	if (!loop_socket)
		2496	loop_socket = loop;
		2497
		2498	// now use loop_socket for all sockets, and loop for everything else
		2499
		2500
		2501	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		2502
		2503	Fork watchers are called when a C<fork ()> was detected (usually because
		2504	whoever is a good citizen cared to tell libev about it by calling
		2505	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		2506	event loop blocks next and before C<ev_check> watchers are being called,
		2507	and only in the child after the fork. If whoever good citizen calling
		2508	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		2509	handlers will be invoked, too, of course.
		2510
		2511	=head3 Watcher-Specific Functions and Data Members
		2512
		2513	=over 4
		2514
		2515	=item ev_fork_init (ev_signal *, callback)
		2516
		2517	Initialises and configures the fork watcher - it has no parameters of any
		2518	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		2519	believe me.
		2520
		2521	=back
		2522
		2523
		2524	=head2 C<ev_async> - how to wake up another event loop
		2525
		2526	In general, you cannot use an C<ev_loop> from multiple threads or other
		2527	asynchronous sources such as signal handlers (as opposed to multiple event
		2528	loops - those are of course safe to use in different threads).
		2529
		2530	Sometimes, however, you need to wake up another event loop you do not
		2531	control, for example because it belongs to another thread. This is what
		2532	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		2533	can signal it by calling C<ev_async_send>, which is thread- and signal
		2534	safe.
		2535
		2536	This functionality is very similar to C<ev_signal> watchers, as signals,
		2537	too, are asynchronous in nature, and signals, too, will be compressed
		2538	(i.e. the number of callback invocations may be less than the number of
		2539	C<ev_async_sent> calls).
		2540
		2541	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		2542	just the default loop.
		2543
		2544	=head3 Queueing
		2545
		2546	C<ev_async> does not support queueing of data in any way. The reason
		2547	is that the author does not know of a simple (or any) algorithm for a
		2548	multiple-writer-single-reader queue that works in all cases and doesn't
		2549	need elaborate support such as pthreads.
		2550
		2551	That means that if you want to queue data, you have to provide your own
		2552	queue. But at least I can tell you how to implement locking around your
		2553	queue:
		2554
		2555	=over 4
		2556
		2557	=item queueing from a signal handler context
		2558
		2559	To implement race-free queueing, you simply add to the queue in the signal
		2560	handler but you block the signal handler in the watcher callback. Here is
		2561	an example that does that for some fictitious SIGUSR1 handler:
		2562
		2563	static ev_async mysig;
		2564
		2565	static void
		2566	sigusr1_handler (void)
		2567	{
		2568	sometype data;
		2569
		2570	// no locking etc.
		2571	queue_put (data);
		2572	ev_async_send (EV_DEFAULT_ &mysig);
		2573	}
		2574
		2575	static void
		2576	mysig_cb (EV_P_ ev_async *w, int revents)
		2577	{
		2578	sometype data;
		2579	sigset_t block, prev;
		2580
		2581	sigemptyset (&block);
		2582	sigaddset (&block, SIGUSR1);
		2583	sigprocmask (SIG_BLOCK, &block, &prev);
		2584
		2585	while (queue_get (&data))
		2586	process (data);
		2587
		2588	if (sigismember (&prev, SIGUSR1)
		2589	sigprocmask (SIG_UNBLOCK, &block, 0);
		2590	}
		2591
		2592	(Note: pthreads in theory requires you to use C<pthread_setmask>
		2593	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		2594	either...).
		2595
		2596	=item queueing from a thread context
		2597
		2598	The strategy for threads is different, as you cannot (easily) block
		2599	threads but you can easily preempt them, so to queue safely you need to
		2600	employ a traditional mutex lock, such as in this pthread example:
		2601
		2602	static ev_async mysig;
		2603	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		2604
		2605	static void
		2606	otherthread (void)
		2607	{
		2608	// only need to lock the actual queueing operation
		2609	pthread_mutex_lock (&mymutex);
		2610	queue_put (data);
		2611	pthread_mutex_unlock (&mymutex);
		2612
		2613	ev_async_send (EV_DEFAULT_ &mysig);
		2614	}
		2615
		2616	static void
		2617	mysig_cb (EV_P_ ev_async *w, int revents)
		2618	{
		2619	pthread_mutex_lock (&mymutex);
		2620
		2621	while (queue_get (&data))
		2622	process (data);
		2623
		2624	pthread_mutex_unlock (&mymutex);
		2625	}
		2626
		2627	=back
		2628
		2629
		2630	=head3 Watcher-Specific Functions and Data Members
		2631
		2632	=over 4
		2633
		2634	=item ev_async_init (ev_async *, callback)
		2635
		2636	Initialises and configures the async watcher - it has no parameters of any
		2637	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,
		2638	trust me.
		2639
		2640	=item ev_async_send (loop, ev_async *)
		2641
		2642	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		2643	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		2644	C<ev_feed_event>, this call is safe to do from other threads, signal or
		2645	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
		2646	section below on what exactly this means).
		2647
		2648	This call incurs the overhead of a system call only once per loop iteration,
		2649	so while the overhead might be noticeable, it doesn't apply to repeated
		2650	calls to C<ev_async_send>.
		2651
		2652	=item bool = ev_async_pending (ev_async *)
		2653
		2654	Returns a non-zero value when C<ev_async_send> has been called on the
		2655	watcher but the event has not yet been processed (or even noted) by the
		2656	event loop.
		2657
		2658	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2659	the loop iterates next and checks for the watcher to have become active,
		2660	it will reset the flag again. C<ev_async_pending> can be used to very
		2661	quickly check whether invoking the loop might be a good idea.
		2662
		2663	Not that this does I<not> check whether the watcher itself is pending, only
		2664	whether it has been requested to make this watcher pending.
		2665
		2666	=back
		2667
1130		2668
1131	=head1 OTHER FUNCTIONS	2669	=head1 OTHER FUNCTIONS
1132		2670
1133	There are some other functions of possible interest. Described. Here. Now.	2671	There are some other functions of possible interest. Described. Here. Now.
1134		2672
1135	=over 4	2673	=over 4
1136		2674
1137	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	2675	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
1138		2676
1139	This function combines a simple timer and an I/O watcher, calls your	2677	This function combines a simple timer and an I/O watcher, calls your
1140	callback on whichever event happens first and automatically stop both	2678	callback on whichever event happens first and automatically stops both
1141	watchers. This is useful if you want to wait for a single event on an fd	2679	watchers. This is useful if you want to wait for a single event on an fd
1142	or timeout without having to allocate/configure/start/stop/free one or	2680	or timeout without having to allocate/configure/start/stop/free one or
1143	more watchers yourself.	2681	more watchers yourself.
1144		2682
1145	If C<fd> is less than 0, then no I/O watcher will be started and events	2683	If C<fd> is less than 0, then no I/O watcher will be started and the
1146	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	2684	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
1147	C<events> set will be craeted and started.	2685	the given C<fd> and C<events> set will be created and started.
1148		2686
1149	If C<timeout> is less than 0, then no timeout watcher will be	2687	If C<timeout> is less than 0, then no timeout watcher will be
1150	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	2688	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
1151	repeat = 0) will be started. While C<0> is a valid timeout, it is of	2689	repeat = 0) will be started. C<0> is a valid timeout.
1152	dubious value.
1153		2690
1154	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	2691	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
1155	passed an C<revents> set like normal event callbacks (a combination of	2692	passed an C<revents> set like normal event callbacks (a combination of
1156	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	2693	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
1157	value passed to C<ev_once>:	2694	value passed to C<ev_once>. Note that it is possible to receive I<both>
		2695	a timeout and an io event at the same time - you probably should give io
		2696	events precedence.
1158		2697
		2698	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		2699
1159	static void stdin_ready (int revents, void *arg)	2700	static void stdin_ready (int revents, void *arg)
1160	{	2701	{
1161	if (revents & EV_TIMEOUT)
1162	/* doh, nothing entered */;
1163	else if (revents & EV_READ)	2702	if (revents & EV_READ)
1164	/* stdin might have data for us, joy! */;	2703	/* stdin might have data for us, joy! */;
		2704	else if (revents & EV_TIMEOUT)
		2705	/* doh, nothing entered */;
1165	}	2706	}
1166		2707
1167	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	2708	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
1168		2709
1169	=item ev_feed_event (loop, watcher, int events)	2710	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
1170		2711
1171	Feeds the given event set into the event loop, as if the specified event	2712	Feeds the given event set into the event loop, as if the specified event
1172	had happened for the specified watcher (which must be a pointer to an	2713	had happened for the specified watcher (which must be a pointer to an
1173	initialised but not necessarily started event watcher).	2714	initialised but not necessarily started event watcher).
1174		2715
1175	=item ev_feed_fd_event (loop, int fd, int revents)	2716	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)
1176		2717
1177	Feed an event on the given fd, as if a file descriptor backend detected	2718	Feed an event on the given fd, as if a file descriptor backend detected
1178	the given events it.	2719	the given events it.
1179		2720
1180	=item ev_feed_signal_event (loop, int signum)	2721	=item ev_feed_signal_event (struct ev_loop *loop, int signum)
1181		2722
1182	Feed an event as if the given signal occured (loop must be the default loop!).	2723	Feed an event as if the given signal occurred (C<loop> must be the default
		2724	loop!).
1183		2725
1184	=back	2726	=back
1185		2727
1186		2728
1187	=head1 LIBEVENT EMULATION	2729	=head1 LIBEVENT EMULATION
…		…
1202		2744
1203	=item * Priorities are not currently supported. Initialising priorities	2745	=item * Priorities are not currently supported. Initialising priorities
1204	will fail and all watchers will have the same priority, even though there	2746	will fail and all watchers will have the same priority, even though there
1205	is an ev_pri field.	2747	is an ev_pri field.
1206		2748
		2749	=item * In libevent, the last base created gets the signals, in libev, the
		2750	first base created (== the default loop) gets the signals.
		2751
1207	=item * Other members are not supported.	2752	=item * Other members are not supported.
1208		2753
1209	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2754	=item * The libev emulation is I<not> ABI compatible to libevent, you need
1210	to use the libev header file and library.	2755	to use the libev header file and library.
1211		2756
1212	=back	2757	=back
1213		2758
1214	=head1 C++ SUPPORT	2759	=head1 C++ SUPPORT
1215		2760
1216	TBD.	2761	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		2762	you to use some convenience methods to start/stop watchers and also change
		2763	the callback model to a model using method callbacks on objects.
		2764
		2765	To use it,
		2766
		2767	#include <ev++.h>
		2768
		2769	This automatically includes F<ev.h> and puts all of its definitions (many
		2770	of them macros) into the global namespace. All C++ specific things are
		2771	put into the C<ev> namespace. It should support all the same embedding
		2772	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
		2773
		2774	Care has been taken to keep the overhead low. The only data member the C++
		2775	classes add (compared to plain C-style watchers) is the event loop pointer
		2776	that the watcher is associated with (or no additional members at all if
		2777	you disable C<EV_MULTIPLICITY> when embedding libev).
		2778
		2779	Currently, functions, and static and non-static member functions can be
		2780	used as callbacks. Other types should be easy to add as long as they only
		2781	need one additional pointer for context. If you need support for other
		2782	types of functors please contact the author (preferably after implementing
		2783	it).
		2784
		2785	Here is a list of things available in the C<ev> namespace:
		2786
		2787	=over 4
		2788
		2789	=item C<ev::READ>, C<ev::WRITE> etc.
		2790
		2791	These are just enum values with the same values as the C<EV_READ> etc.
		2792	macros from F<ev.h>.
		2793
		2794	=item C<ev::tstamp>, C<ev::now>
		2795
		2796	Aliases to the same types/functions as with the C<ev_> prefix.
		2797
		2798	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		2799
		2800	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		2801	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		2802	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		2803	defines by many implementations.
		2804
		2805	All of those classes have these methods:
		2806
		2807	=over 4
		2808
		2809	=item ev::TYPE::TYPE ()
		2810
		2811	=item ev::TYPE::TYPE (struct ev_loop *)
		2812
		2813	=item ev::TYPE::~TYPE
		2814
		2815	The constructor (optionally) takes an event loop to associate the watcher
		2816	with. If it is omitted, it will use C<EV_DEFAULT>.
		2817
		2818	The constructor calls C<ev_init> for you, which means you have to call the
		2819	C<set> method before starting it.
		2820
		2821	It will not set a callback, however: You have to call the templated C<set>
		2822	method to set a callback before you can start the watcher.
		2823
		2824	(The reason why you have to use a method is a limitation in C++ which does
		2825	not allow explicit template arguments for constructors).
		2826
		2827	The destructor automatically stops the watcher if it is active.
		2828
		2829	=item w->set<class, &class::method> (object *)
		2830
		2831	This method sets the callback method to call. The method has to have a
		2832	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		2833	first argument and the C<revents> as second. The object must be given as
		2834	parameter and is stored in the C<data> member of the watcher.
		2835
		2836	This method synthesizes efficient thunking code to call your method from
		2837	the C callback that libev requires. If your compiler can inline your
		2838	callback (i.e. it is visible to it at the place of the C<set> call and
		2839	your compiler is good :), then the method will be fully inlined into the
		2840	thunking function, making it as fast as a direct C callback.
		2841
		2842	Example: simple class declaration and watcher initialisation
		2843
		2844	struct myclass
		2845	{
		2846	void io_cb (ev::io &w, int revents) { }
		2847	}
		2848
		2849	myclass obj;
		2850	ev::io iow;
		2851	iow.set <myclass, &myclass::io_cb> (&obj);
		2852
		2853	=item w->set<function> (void *data = 0)
		2854
		2855	Also sets a callback, but uses a static method or plain function as
		2856	callback. The optional C<data> argument will be stored in the watcher's
		2857	C<data> member and is free for you to use.
		2858
		2859	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2860
		2861	See the method-C<set> above for more details.
		2862
		2863	Example: Use a plain function as callback.
		2864
		2865	static void io_cb (ev::io &w, int revents) { }
		2866	iow.set <io_cb> ();
		2867
		2868	=item w->set (struct ev_loop *)
		2869
		2870	Associates a different C<struct ev_loop> with this watcher. You can only
		2871	do this when the watcher is inactive (and not pending either).
		2872
		2873	=item w->set ([arguments])
		2874
		2875	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be
		2876	called at least once. Unlike the C counterpart, an active watcher gets
		2877	automatically stopped and restarted when reconfiguring it with this
		2878	method.
		2879
		2880	=item w->start ()
		2881
		2882	Starts the watcher. Note that there is no C<loop> argument, as the
		2883	constructor already stores the event loop.
		2884
		2885	=item w->stop ()
		2886
		2887	Stops the watcher if it is active. Again, no C<loop> argument.
		2888
		2889	=item w->again () (C<ev::timer>, C<ev::periodic> only)
		2890
		2891	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		2892	C<ev_TYPE_again> function.
		2893
		2894	=item w->sweep () (C<ev::embed> only)
		2895
		2896	Invokes C<ev_embed_sweep>.
		2897
		2898	=item w->update () (C<ev::stat> only)
		2899
		2900	Invokes C<ev_stat_stat>.
		2901
		2902	=back
		2903
		2904	=back
		2905
		2906	Example: Define a class with an IO and idle watcher, start one of them in
		2907	the constructor.
		2908
		2909	class myclass
		2910	{
		2911	ev::io io ; void io_cb (ev::io &w, int revents);
		2912	ev::idle idle; void idle_cb (ev::idle &w, int revents);
		2913
		2914	myclass (int fd)
		2915	{
		2916	io .set <myclass, &myclass::io_cb > (this);
		2917	idle.set <myclass, &myclass::idle_cb> (this);
		2918
		2919	io.start (fd, ev::READ);
		2920	}
		2921	};
		2922
		2923
		2924	=head1 OTHER LANGUAGE BINDINGS
		2925
		2926	Libev does not offer other language bindings itself, but bindings for a
		2927	number of languages exist in the form of third-party packages. If you know
		2928	any interesting language binding in addition to the ones listed here, drop
		2929	me a note.
		2930
		2931	=over 4
		2932
		2933	=item Perl
		2934
		2935	The EV module implements the full libev API and is actually used to test
		2936	libev. EV is developed together with libev. Apart from the EV core module,
		2937	there are additional modules that implement libev-compatible interfaces
		2938	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		2939	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		2940	and C<EV::Glib>).
		2941
		2942	It can be found and installed via CPAN, its homepage is at
		2943	L<http://software.schmorp.de/pkg/EV>.
		2944
		2945	=item Python
		2946
		2947	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		2948	seems to be quite complete and well-documented. Note, however, that the
		2949	patch they require for libev is outright dangerous as it breaks the ABI
		2950	for everybody else, and therefore, should never be applied in an installed
		2951	libev (if python requires an incompatible ABI then it needs to embed
		2952	libev).
		2953
		2954	=item Ruby
		2955
		2956	Tony Arcieri has written a ruby extension that offers access to a subset
		2957	of the libev API and adds file handle abstractions, asynchronous DNS and
		2958	more on top of it. It can be found via gem servers. Its homepage is at
		2959	L<http://rev.rubyforge.org/>.
		2960
		2961	=item D
		2962
		2963	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		2964	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		2965
		2966	=item Ocaml
		2967
		2968	Erkki Seppala has written Ocaml bindings for libev, to be found at
		2969	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		2970
		2971	=back
		2972
		2973
		2974	=head1 MACRO MAGIC
		2975
		2976	Libev can be compiled with a variety of options, the most fundamental
		2977	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		2978	functions and callbacks have an initial C<struct ev_loop *> argument.
		2979
		2980	To make it easier to write programs that cope with either variant, the
		2981	following macros are defined:
		2982
		2983	=over 4
		2984
		2985	=item C<EV_A>, C<EV_A_>
		2986
		2987	This provides the loop I<argument> for functions, if one is required ("ev
		2988	loop argument"). The C<EV_A> form is used when this is the sole argument,
		2989	C<EV_A_> is used when other arguments are following. Example:
		2990
		2991	ev_unref (EV_A);
		2992	ev_timer_add (EV_A_ watcher);
		2993	ev_loop (EV_A_ 0);
		2994
		2995	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		2996	which is often provided by the following macro.
		2997
		2998	=item C<EV_P>, C<EV_P_>
		2999
		3000	This provides the loop I<parameter> for functions, if one is required ("ev
		3001	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		3002	C<EV_P_> is used when other parameters are following. Example:
		3003
		3004	// this is how ev_unref is being declared
		3005	static void ev_unref (EV_P);
		3006
		3007	// this is how you can declare your typical callback
		3008	static void cb (EV_P_ ev_timer *w, int revents)
		3009
		3010	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		3011	suitable for use with C<EV_A>.
		3012
		3013	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		3014
		3015	Similar to the other two macros, this gives you the value of the default
		3016	loop, if multiple loops are supported ("ev loop default").
		3017
		3018	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		3019
		3020	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		3021	default loop has been initialised (C<UC> == unchecked). Their behaviour
		3022	is undefined when the default loop has not been initialised by a previous
		3023	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		3024
		3025	It is often prudent to use C<EV_DEFAULT> when initialising the first
		3026	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		3027
		3028	=back
		3029
		3030	Example: Declare and initialise a check watcher, utilising the above
		3031	macros so it will work regardless of whether multiple loops are supported
		3032	or not.
		3033
		3034	static void
		3035	check_cb (EV_P_ ev_timer *w, int revents)
		3036	{
		3037	ev_check_stop (EV_A_ w);
		3038	}
		3039
		3040	ev_check check;
		3041	ev_check_init (&check, check_cb);
		3042	ev_check_start (EV_DEFAULT_ &check);
		3043	ev_loop (EV_DEFAULT_ 0);
		3044
		3045	=head1 EMBEDDING
		3046
		3047	Libev can (and often is) directly embedded into host
		3048	applications. Examples of applications that embed it include the Deliantra
		3049	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		3050	and rxvt-unicode.
		3051
		3052	The goal is to enable you to just copy the necessary files into your
		3053	source directory without having to change even a single line in them, so
		3054	you can easily upgrade by simply copying (or having a checked-out copy of
		3055	libev somewhere in your source tree).
		3056
		3057	=head2 FILESETS
		3058
		3059	Depending on what features you need you need to include one or more sets of files
		3060	in your application.
		3061
		3062	=head3 CORE EVENT LOOP
		3063
		3064	To include only the libev core (all the C<ev_*> functions), with manual
		3065	configuration (no autoconf):
		3066
		3067	#define EV_STANDALONE 1
		3068	#include "ev.c"
		3069
		3070	This will automatically include F<ev.h>, too, and should be done in a
		3071	single C source file only to provide the function implementations. To use
		3072	it, do the same for F<ev.h> in all files wishing to use this API (best
		3073	done by writing a wrapper around F<ev.h> that you can include instead and
		3074	where you can put other configuration options):
		3075
		3076	#define EV_STANDALONE 1
		3077	#include "ev.h"
		3078
		3079	Both header files and implementation files can be compiled with a C++
		3080	compiler (at least, thats a stated goal, and breakage will be treated
		3081	as a bug).
		3082
		3083	You need the following files in your source tree, or in a directory
		3084	in your include path (e.g. in libev/ when using -Ilibev):
		3085
		3086	ev.h
		3087	ev.c
		3088	ev_vars.h
		3089	ev_wrap.h
		3090
		3091	ev_win32.c required on win32 platforms only
		3092
		3093	ev_select.c only when select backend is enabled (which is enabled by default)
		3094	ev_poll.c only when poll backend is enabled (disabled by default)
		3095	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		3096	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		3097	ev_port.c only when the solaris port backend is enabled (disabled by default)
		3098
		3099	F<ev.c> includes the backend files directly when enabled, so you only need
		3100	to compile this single file.
		3101
		3102	=head3 LIBEVENT COMPATIBILITY API
		3103
		3104	To include the libevent compatibility API, also include:
		3105
		3106	#include "event.c"
		3107
		3108	in the file including F<ev.c>, and:
		3109
		3110	#include "event.h"
		3111
		3112	in the files that want to use the libevent API. This also includes F<ev.h>.
		3113
		3114	You need the following additional files for this:
		3115
		3116	event.h
		3117	event.c
		3118
		3119	=head3 AUTOCONF SUPPORT
		3120
		3121	Instead of using C<EV_STANDALONE=1> and providing your configuration in
		3122	whatever way you want, you can also C<m4_include([libev.m4])> in your
		3123	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		3124	include F<config.h> and configure itself accordingly.
		3125
		3126	For this of course you need the m4 file:
		3127
		3128	libev.m4
		3129
		3130	=head2 PREPROCESSOR SYMBOLS/MACROS
		3131
		3132	Libev can be configured via a variety of preprocessor symbols you have to
		3133	define before including any of its files. The default in the absence of
		3134	autoconf is documented for every option.
		3135
		3136	=over 4
		3137
		3138	=item EV_STANDALONE
		3139
		3140	Must always be C<1> if you do not use autoconf configuration, which
		3141	keeps libev from including F<config.h>, and it also defines dummy
		3142	implementations for some libevent functions (such as logging, which is not
		3143	supported). It will also not define any of the structs usually found in
		3144	F<event.h> that are not directly supported by the libev core alone.
		3145
		3146	=item EV_USE_MONOTONIC
		3147
		3148	If defined to be C<1>, libev will try to detect the availability of the
		3149	monotonic clock option at both compile time and runtime. Otherwise no use
		3150	of the monotonic clock option will be attempted. If you enable this, you
		3151	usually have to link against librt or something similar. Enabling it when
		3152	the functionality isn't available is safe, though, although you have
		3153	to make sure you link against any libraries where the C<clock_gettime>
		3154	function is hiding in (often F<-lrt>).
		3155
		3156	=item EV_USE_REALTIME
		3157
		3158	If defined to be C<1>, libev will try to detect the availability of the
		3159	real-time clock option at compile time (and assume its availability at
		3160	runtime if successful). Otherwise no use of the real-time clock option will
		3161	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		3162	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
		3163	note about libraries in the description of C<EV_USE_MONOTONIC>, though.
		3164
		3165	=item EV_USE_NANOSLEEP
		3166
		3167	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		3168	and will use it for delays. Otherwise it will use C<select ()>.
		3169
		3170	=item EV_USE_EVENTFD
		3171
		3172	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		3173	available and will probe for kernel support at runtime. This will improve
		3174	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		3175	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		3176	2.7 or newer, otherwise disabled.
		3177
		3178	=item EV_USE_SELECT
		3179
		3180	If undefined or defined to be C<1>, libev will compile in support for the
		3181	C<select>(2) backend. No attempt at auto-detection will be done: if no
		3182	other method takes over, select will be it. Otherwise the select backend
		3183	will not be compiled in.
		3184
		3185	=item EV_SELECT_USE_FD_SET
		3186
		3187	If defined to C<1>, then the select backend will use the system C<fd_set>
		3188	structure. This is useful if libev doesn't compile due to a missing
		3189	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on
		3190	exotic systems. This usually limits the range of file descriptors to some
		3191	low limit such as 1024 or might have other limitations (winsocket only
		3192	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		3193	influence the size of the C<fd_set> used.
		3194
		3195	=item EV_SELECT_IS_WINSOCKET
		3196
		3197	When defined to C<1>, the select backend will assume that
		3198	select/socket/connect etc. don't understand file descriptors but
		3199	wants osf handles on win32 (this is the case when the select to
		3200	be used is the winsock select). This means that it will call
		3201	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		3202	it is assumed that all these functions actually work on fds, even
		3203	on win32. Should not be defined on non-win32 platforms.
		3204
		3205	=item EV_FD_TO_WIN32_HANDLE
		3206
		3207	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		3208	file descriptors to socket handles. When not defining this symbol (the
		3209	default), then libev will call C<_get_osfhandle>, which is usually
		3210	correct. In some cases, programs use their own file descriptor management,
		3211	in which case they can provide this function to map fds to socket handles.
		3212
		3213	=item EV_USE_POLL
		3214
		3215	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		3216	backend. Otherwise it will be enabled on non-win32 platforms. It
		3217	takes precedence over select.
		3218
		3219	=item EV_USE_EPOLL
		3220
		3221	If defined to be C<1>, libev will compile in support for the Linux
		3222	C<epoll>(7) backend. Its availability will be detected at runtime,
		3223	otherwise another method will be used as fallback. This is the preferred
		3224	backend for GNU/Linux systems. If undefined, it will be enabled if the
		3225	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		3226
		3227	=item EV_USE_KQUEUE
		3228
		3229	If defined to be C<1>, libev will compile in support for the BSD style
		3230	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		3231	otherwise another method will be used as fallback. This is the preferred
		3232	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		3233	supports some types of fds correctly (the only platform we found that
		3234	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		3235	not be used unless explicitly requested. The best way to use it is to find
		3236	out whether kqueue supports your type of fd properly and use an embedded
		3237	kqueue loop.
		3238
		3239	=item EV_USE_PORT
		3240
		3241	If defined to be C<1>, libev will compile in support for the Solaris
		3242	10 port style backend. Its availability will be detected at runtime,
		3243	otherwise another method will be used as fallback. This is the preferred
		3244	backend for Solaris 10 systems.
		3245
		3246	=item EV_USE_DEVPOLL
		3247
		3248	Reserved for future expansion, works like the USE symbols above.
		3249
		3250	=item EV_USE_INOTIFY
		3251
		3252	If defined to be C<1>, libev will compile in support for the Linux inotify
		3253	interface to speed up C<ev_stat> watchers. Its actual availability will
		3254	be detected at runtime. If undefined, it will be enabled if the headers
		3255	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		3256
		3257	=item EV_ATOMIC_T
		3258
		3259	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		3260	access is atomic with respect to other threads or signal contexts. No such
		3261	type is easily found in the C language, so you can provide your own type
		3262	that you know is safe for your purposes. It is used both for signal handler "locking"
		3263	as well as for signal and thread safety in C<ev_async> watchers.
		3264
		3265	In the absence of this define, libev will use C<sig_atomic_t volatile>
		3266	(from F<signal.h>), which is usually good enough on most platforms.
		3267
		3268	=item EV_H
		3269
		3270	The name of the F<ev.h> header file used to include it. The default if
		3271	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
		3272	used to virtually rename the F<ev.h> header file in case of conflicts.
		3273
		3274	=item EV_CONFIG_H
		3275
		3276	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		3277	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		3278	C<EV_H>, above.
		3279
		3280	=item EV_EVENT_H
		3281
		3282	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		3283	of how the F<event.h> header can be found, the default is C<"event.h">.
		3284
		3285	=item EV_PROTOTYPES
		3286
		3287	If defined to be C<0>, then F<ev.h> will not define any function
		3288	prototypes, but still define all the structs and other symbols. This is
		3289	occasionally useful if you want to provide your own wrapper functions
		3290	around libev functions.
		3291
		3292	=item EV_MULTIPLICITY
		3293
		3294	If undefined or defined to C<1>, then all event-loop-specific functions
		3295	will have the C<struct ev_loop *> as first argument, and you can create
		3296	additional independent event loops. Otherwise there will be no support
		3297	for multiple event loops and there is no first event loop pointer
		3298	argument. Instead, all functions act on the single default loop.
		3299
		3300	=item EV_MINPRI
		3301
		3302	=item EV_MAXPRI
		3303
		3304	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		3305	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		3306	provide for more priorities by overriding those symbols (usually defined
		3307	to be C<-2> and C<2>, respectively).
		3308
		3309	When doing priority-based operations, libev usually has to linearly search
		3310	all the priorities, so having many of them (hundreds) uses a lot of space
		3311	and time, so using the defaults of five priorities (-2 .. +2) is usually
		3312	fine.
		3313
		3314	If your embedding application does not need any priorities, defining these
		3315	both to C<0> will save some memory and CPU.
		3316
		3317	=item EV_PERIODIC_ENABLE
		3318
		3319	If undefined or defined to be C<1>, then periodic timers are supported. If
		3320	defined to be C<0>, then they are not. Disabling them saves a few kB of
		3321	code.
		3322
		3323	=item EV_IDLE_ENABLE
		3324
		3325	If undefined or defined to be C<1>, then idle watchers are supported. If
		3326	defined to be C<0>, then they are not. Disabling them saves a few kB of
		3327	code.
		3328
		3329	=item EV_EMBED_ENABLE
		3330
		3331	If undefined or defined to be C<1>, then embed watchers are supported. If
		3332	defined to be C<0>, then they are not. Embed watchers rely on most other
		3333	watcher types, which therefore must not be disabled.
		3334
		3335	=item EV_STAT_ENABLE
		3336
		3337	If undefined or defined to be C<1>, then stat watchers are supported. If
		3338	defined to be C<0>, then they are not.
		3339
		3340	=item EV_FORK_ENABLE
		3341
		3342	If undefined or defined to be C<1>, then fork watchers are supported. If
		3343	defined to be C<0>, then they are not.
		3344
		3345	=item EV_ASYNC_ENABLE
		3346
		3347	If undefined or defined to be C<1>, then async watchers are supported. If
		3348	defined to be C<0>, then they are not.
		3349
		3350	=item EV_MINIMAL
		3351
		3352	If you need to shave off some kilobytes of code at the expense of some
		3353	speed, define this symbol to C<1>. Currently this is used to override some
		3354	inlining decisions, saves roughly 30% code size on amd64. It also selects a
		3355	much smaller 2-heap for timer management over the default 4-heap.
		3356
		3357	=item EV_PID_HASHSIZE
		3358
		3359	C<ev_child> watchers use a small hash table to distribute workload by
		3360	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		3361	than enough. If you need to manage thousands of children you might want to
		3362	increase this value (I<must> be a power of two).
		3363
		3364	=item EV_INOTIFY_HASHSIZE
		3365
		3366	C<ev_stat> watchers use a small hash table to distribute workload by
		3367	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		3368	usually more than enough. If you need to manage thousands of C<ev_stat>
		3369	watchers you might want to increase this value (I<must> be a power of
		3370	two).
		3371
		3372	=item EV_USE_4HEAP
		3373
		3374	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3375	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		3376	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		3377	faster performance with many (thousands) of watchers.
		3378
		3379	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3380	(disabled).
		3381
		3382	=item EV_HEAP_CACHE_AT
		3383
		3384	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3385	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		3386	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3387	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3388	but avoids random read accesses on heap changes. This improves performance
		3389	noticeably with many (hundreds) of watchers.
		3390
		3391	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3392	(disabled).
		3393
		3394	=item EV_VERIFY
		3395
		3396	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3397	be done: If set to C<0>, no internal verification code will be compiled
		3398	in. If set to C<1>, then verification code will be compiled in, but not
		3399	called. If set to C<2>, then the internal verification code will be
		3400	called once per loop, which can slow down libev. If set to C<3>, then the
		3401	verification code will be called very frequently, which will slow down
		3402	libev considerably.
		3403
		3404	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3405	C<0>.
		3406
		3407	=item EV_COMMON
		3408
		3409	By default, all watchers have a C<void *data> member. By redefining
		3410	this macro to a something else you can include more and other types of
		3411	members. You have to define it each time you include one of the files,
		3412	though, and it must be identical each time.
		3413
		3414	For example, the perl EV module uses something like this:
		3415
		3416	#define EV_COMMON \
		3417	SV *self; /* contains this struct */ \
		3418	SV *cb_sv, *fh /* note no trailing ";" */
		3419
		3420	=item EV_CB_DECLARE (type)
		3421
		3422	=item EV_CB_INVOKE (watcher, revents)
		3423
		3424	=item ev_set_cb (ev, cb)
		3425
		3426	Can be used to change the callback member declaration in each watcher,
		3427	and the way callbacks are invoked and set. Must expand to a struct member
		3428	definition and a statement, respectively. See the F<ev.h> header file for
		3429	their default definitions. One possible use for overriding these is to
		3430	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		3431	method calls instead of plain function calls in C++.
		3432
		3433	=back
		3434
		3435	=head2 EXPORTED API SYMBOLS
		3436
		3437	If you need to re-export the API (e.g. via a DLL) and you need a list of
		3438	exported symbols, you can use the provided F<Symbol.*> files which list
		3439	all public symbols, one per line:
		3440
		3441	Symbols.ev for libev proper
		3442	Symbols.event for the libevent emulation
		3443
		3444	This can also be used to rename all public symbols to avoid clashes with
		3445	multiple versions of libev linked together (which is obviously bad in
		3446	itself, but sometimes it is inconvenient to avoid this).
		3447
		3448	A sed command like this will create wrapper C<#define>'s that you need to
		3449	include before including F<ev.h>:
		3450
		3451	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		3452
		3453	This would create a file F<wrap.h> which essentially looks like this:
		3454
		3455	#define ev_backend myprefix_ev_backend
		3456	#define ev_check_start myprefix_ev_check_start
		3457	#define ev_check_stop myprefix_ev_check_stop
		3458	...
		3459
		3460	=head2 EXAMPLES
		3461
		3462	For a real-world example of a program the includes libev
		3463	verbatim, you can have a look at the EV perl module
		3464	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		3465	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		3466	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		3467	will be compiled. It is pretty complex because it provides its own header
		3468	file.
		3469
		3470	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		3471	that everybody includes and which overrides some configure choices:
		3472
		3473	#define EV_MINIMAL 1
		3474	#define EV_USE_POLL 0
		3475	#define EV_MULTIPLICITY 0
		3476	#define EV_PERIODIC_ENABLE 0
		3477	#define EV_STAT_ENABLE 0
		3478	#define EV_FORK_ENABLE 0
		3479	#define EV_CONFIG_H <config.h>
		3480	#define EV_MINPRI 0
		3481	#define EV_MAXPRI 0
		3482
		3483	#include "ev++.h"
		3484
		3485	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		3486
		3487	#include "ev_cpp.h"
		3488	#include "ev.c"
		3489
		3490	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
		3491
		3492	=head2 THREADS AND COROUTINES
		3493
		3494	=head3 THREADS
		3495
		3496	All libev functions are reentrant and thread-safe unless explicitly
		3497	documented otherwise, but libev implements no locking itself. This means
		3498	that you can use as many loops as you want in parallel, as long as there
		3499	are no concurrent calls into any libev function with the same loop
		3500	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		3501	of course): libev guarantees that different event loops share no data
		3502	structures that need any locking.
		3503
		3504	Or to put it differently: calls with different loop parameters can be done
		3505	concurrently from multiple threads, calls with the same loop parameter
		3506	must be done serially (but can be done from different threads, as long as
		3507	only one thread ever is inside a call at any point in time, e.g. by using
		3508	a mutex per loop).
		3509
		3510	Specifically to support threads (and signal handlers), libev implements
		3511	so-called C<ev_async> watchers, which allow some limited form of
		3512	concurrency on the same event loop, namely waking it up "from the
		3513	outside".
		3514
		3515	If you want to know which design (one loop, locking, or multiple loops
		3516	without or something else still) is best for your problem, then I cannot
		3517	help you, but here is some generic advice:
		3518
		3519	=over 4
		3520
		3521	=item * most applications have a main thread: use the default libev loop
		3522	in that thread, or create a separate thread running only the default loop.
		3523
		3524	This helps integrating other libraries or software modules that use libev
		3525	themselves and don't care/know about threading.
		3526
		3527	=item * one loop per thread is usually a good model.
		3528
		3529	Doing this is almost never wrong, sometimes a better-performance model
		3530	exists, but it is always a good start.
		3531
		3532	=item * other models exist, such as the leader/follower pattern, where one
		3533	loop is handed through multiple threads in a kind of round-robin fashion.
		3534
		3535	Choosing a model is hard - look around, learn, know that usually you can do
		3536	better than you currently do :-)
		3537
		3538	=item * often you need to talk to some other thread which blocks in the
		3539	event loop.
		3540
		3541	C<ev_async> watchers can be used to wake them up from other threads safely
		3542	(or from signal contexts...).
		3543
		3544	An example use would be to communicate signals or other events that only
		3545	work in the default loop by registering the signal watcher with the
		3546	default loop and triggering an C<ev_async> watcher from the default loop
		3547	watcher callback into the event loop interested in the signal.
		3548
		3549	=back
		3550
		3551	=head3 COROUTINES
		3552
		3553	Libev is very accommodating to coroutines ("cooperative threads"):
		3554	libev fully supports nesting calls to its functions from different
		3555	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3556	different coroutines, and switch freely between both coroutines running the
		3557	loop, as long as you don't confuse yourself). The only exception is that
		3558	you must not do this from C<ev_periodic> reschedule callbacks.
		3559
		3560	Care has been taken to ensure that libev does not keep local state inside
		3561	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		3562	they do not clal any callbacks.
		3563
		3564	=head2 COMPILER WARNINGS
		3565
		3566	Depending on your compiler and compiler settings, you might get no or a
		3567	lot of warnings when compiling libev code. Some people are apparently
		3568	scared by this.
		3569
		3570	However, these are unavoidable for many reasons. For one, each compiler
		3571	has different warnings, and each user has different tastes regarding
		3572	warning options. "Warn-free" code therefore cannot be a goal except when
		3573	targeting a specific compiler and compiler-version.
		3574
		3575	Another reason is that some compiler warnings require elaborate
		3576	workarounds, or other changes to the code that make it less clear and less
		3577	maintainable.
		3578
		3579	And of course, some compiler warnings are just plain stupid, or simply
		3580	wrong (because they don't actually warn about the condition their message
		3581	seems to warn about). For example, certain older gcc versions had some
		3582	warnings that resulted an extreme number of false positives. These have
		3583	been fixed, but some people still insist on making code warn-free with
		3584	such buggy versions.
		3585
		3586	While libev is written to generate as few warnings as possible,
		3587	"warn-free" code is not a goal, and it is recommended not to build libev
		3588	with any compiler warnings enabled unless you are prepared to cope with
		3589	them (e.g. by ignoring them). Remember that warnings are just that:
		3590	warnings, not errors, or proof of bugs.
		3591
		3592
		3593	=head2 VALGRIND
		3594
		3595	Valgrind has a special section here because it is a popular tool that is
		3596	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		3597
		3598	If you think you found a bug (memory leak, uninitialised data access etc.)
		3599	in libev, then check twice: If valgrind reports something like:
		3600
		3601	==2274== definitely lost: 0 bytes in 0 blocks.
		3602	==2274== possibly lost: 0 bytes in 0 blocks.
		3603	==2274== still reachable: 256 bytes in 1 blocks.
		3604
		3605	Then there is no memory leak, just as memory accounted to global variables
		3606	is not a memleak - the memory is still being refernced, and didn't leak.
		3607
		3608	Similarly, under some circumstances, valgrind might report kernel bugs
		3609	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		3610	although an acceptable workaround has been found here), or it might be
		3611	confused.
		3612
		3613	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		3614	make it into some kind of religion.
		3615
		3616	If you are unsure about something, feel free to contact the mailing list
		3617	with the full valgrind report and an explanation on why you think this
		3618	is a bug in libev (best check the archives, too :). However, don't be
		3619	annoyed when you get a brisk "this is no bug" answer and take the chance
		3620	of learning how to interpret valgrind properly.
		3621
		3622	If you need, for some reason, empty reports from valgrind for your project
		3623	I suggest using suppression lists.
		3624
		3625
		3626	=head1 PORTABILITY NOTES
		3627
		3628	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		3629
		3630	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		3631	requires, and its I/O model is fundamentally incompatible with the POSIX
		3632	model. Libev still offers limited functionality on this platform in
		3633	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		3634	descriptors. This only applies when using Win32 natively, not when using
		3635	e.g. cygwin.
		3636
		3637	Lifting these limitations would basically require the full
		3638	re-implementation of the I/O system. If you are into these kinds of
		3639	things, then note that glib does exactly that for you in a very portable
		3640	way (note also that glib is the slowest event library known to man).
		3641
		3642	There is no supported compilation method available on windows except
		3643	embedding it into other applications.
		3644
		3645	Not a libev limitation but worth mentioning: windows apparently doesn't
		3646	accept large writes: instead of resulting in a partial write, windows will
		3647	either accept everything or return C<ENOBUFS> if the buffer is too large,
		3648	so make sure you only write small amounts into your sockets (less than a
		3649	megabyte seems safe, but this apparently depends on the amount of memory
		3650	available).
		3651
		3652	Due to the many, low, and arbitrary limits on the win32 platform and
		3653	the abysmal performance of winsockets, using a large number of sockets
		3654	is not recommended (and not reasonable). If your program needs to use
		3655	more than a hundred or so sockets, then likely it needs to use a totally
		3656	different implementation for windows, as libev offers the POSIX readiness
		3657	notification model, which cannot be implemented efficiently on windows
		3658	(Microsoft monopoly games).
		3659
		3660	A typical way to use libev under windows is to embed it (see the embedding
		3661	section for details) and use the following F<evwrap.h> header file instead
		3662	of F<ev.h>:
		3663
		3664	#define EV_STANDALONE /* keeps ev from requiring config.h */
		3665	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		3666
		3667	#include "ev.h"
		3668
		3669	And compile the following F<evwrap.c> file into your project (make sure
		3670	you do I<not> compile the F<ev.c> or any other embedded source files!):
		3671
		3672	#include "evwrap.h"
		3673	#include "ev.c"
		3674
		3675	=over 4
		3676
		3677	=item The winsocket select function
		3678
		3679	The winsocket C<select> function doesn't follow POSIX in that it
		3680	requires socket I<handles> and not socket I<file descriptors> (it is
		3681	also extremely buggy). This makes select very inefficient, and also
		3682	requires a mapping from file descriptors to socket handles (the Microsoft
		3683	C runtime provides the function C<_open_osfhandle> for this). See the
		3684	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		3685	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		3686
		3687	The configuration for a "naked" win32 using the Microsoft runtime
		3688	libraries and raw winsocket select is:
		3689
		3690	#define EV_USE_SELECT 1
		3691	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		3692
		3693	Note that winsockets handling of fd sets is O(n), so you can easily get a
		3694	complexity in the O(n²) range when using win32.
		3695
		3696	=item Limited number of file descriptors
		3697
		3698	Windows has numerous arbitrary (and low) limits on things.
		3699
		3700	Early versions of winsocket's select only supported waiting for a maximum
		3701	of C<64> handles (probably owning to the fact that all windows kernels
		3702	can only wait for C<64> things at the same time internally; Microsoft
		3703	recommends spawning a chain of threads and wait for 63 handles and the
		3704	previous thread in each. Great).
		3705
		3706	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		3707	to some high number (e.g. C<2048>) before compiling the winsocket select
		3708	call (which might be in libev or elsewhere, for example, perl does its own
		3709	select emulation on windows).
		3710
		3711	Another limit is the number of file descriptors in the Microsoft runtime
		3712	libraries, which by default is C<64> (there must be a hidden I<64> fetish
		3713	or something like this inside Microsoft). You can increase this by calling
		3714	C<_setmaxstdio>, which can increase this limit to C<2048> (another
		3715	arbitrary limit), but is broken in many versions of the Microsoft runtime
		3716	libraries.
		3717
		3718	This might get you to about C<512> or C<2048> sockets (depending on
		3719	windows version and/or the phase of the moon). To get more, you need to
		3720	wrap all I/O functions and provide your own fd management, but the cost of
		3721	calling select (O(n²)) will likely make this unworkable.
		3722
		3723	=back
		3724
		3725	=head2 PORTABILITY REQUIREMENTS
		3726
		3727	In addition to a working ISO-C implementation and of course the
		3728	backend-specific APIs, libev relies on a few additional extensions:
		3729
		3730	=over 4
		3731
		3732	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		3733	calling conventions regardless of C<ev_watcher_type *>.
		3734
		3735	Libev assumes not only that all watcher pointers have the same internal
		3736	structure (guaranteed by POSIX but not by ISO C for example), but it also
		3737	assumes that the same (machine) code can be used to call any watcher
		3738	callback: The watcher callbacks have different type signatures, but libev
		3739	calls them using an C<ev_watcher *> internally.
		3740
		3741	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3742
		3743	The type C<sig_atomic_t volatile> (or whatever is defined as
		3744	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		3745	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3746	believed to be sufficiently portable.
		3747
		3748	=item C<sigprocmask> must work in a threaded environment
		3749
		3750	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3751	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3752	pthread implementations will either allow C<sigprocmask> in the "main
		3753	thread" or will block signals process-wide, both behaviours would
		3754	be compatible with libev. Interaction between C<sigprocmask> and
		3755	C<pthread_sigmask> could complicate things, however.
		3756
		3757	The most portable way to handle signals is to block signals in all threads
		3758	except the initial one, and run the default loop in the initial thread as
		3759	well.
		3760
		3761	=item C<long> must be large enough for common memory allocation sizes
		3762
		3763	To improve portability and simplify its API, libev uses C<long> internally
		3764	instead of C<size_t> when allocating its data structures. On non-POSIX
		3765	systems (Microsoft...) this might be unexpectedly low, but is still at
		3766	least 31 bits everywhere, which is enough for hundreds of millions of
		3767	watchers.
		3768
		3769	=item C<double> must hold a time value in seconds with enough accuracy
		3770
		3771	The type C<double> is used to represent timestamps. It is required to
		3772	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3773	enough for at least into the year 4000. This requirement is fulfilled by
		3774	implementations implementing IEEE 754 (basically all existing ones).
		3775
		3776	=back
		3777
		3778	If you know of other additional requirements drop me a note.
		3779
		3780
		3781	=head1 ALGORITHMIC COMPLEXITIES
		3782
		3783	In this section the complexities of (many of) the algorithms used inside
		3784	libev will be documented. For complexity discussions about backends see
		3785	the documentation for C<ev_default_init>.
		3786
		3787	All of the following are about amortised time: If an array needs to be
		3788	extended, libev needs to realloc and move the whole array, but this
		3789	happens asymptotically rarer with higher number of elements, so O(1) might
		3790	mean that libev does a lengthy realloc operation in rare cases, but on
		3791	average it is much faster and asymptotically approaches constant time.
		3792
		3793	=over 4
		3794
		3795	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		3796
		3797	This means that, when you have a watcher that triggers in one hour and
		3798	there are 100 watchers that would trigger before that, then inserting will
		3799	have to skip roughly seven (C<ld 100>) of these watchers.
		3800
		3801	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		3802
		3803	That means that changing a timer costs less than removing/adding them,
		3804	as only the relative motion in the event queue has to be paid for.
		3805
		3806	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		3807
		3808	These just add the watcher into an array or at the head of a list.
		3809
		3810	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		3811
		3812	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		3813
		3814	These watchers are stored in lists, so they need to be walked to find the
		3815	correct watcher to remove. The lists are usually short (you don't usually
		3816	have many watchers waiting for the same fd or signal: one is typical, two
		3817	is rare).
		3818
		3819	=item Finding the next timer in each loop iteration: O(1)
		3820
		3821	By virtue of using a binary or 4-heap, the next timer is always found at a
		3822	fixed position in the storage array.
		3823
		3824	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		3825
		3826	A change means an I/O watcher gets started or stopped, which requires
		3827	libev to recalculate its status (and possibly tell the kernel, depending
		3828	on backend and whether C<ev_io_set> was used).
		3829
		3830	=item Activating one watcher (putting it into the pending state): O(1)
		3831
		3832	=item Priority handling: O(number_of_priorities)
		3833
		3834	Priorities are implemented by allocating some space for each
		3835	priority. When doing priority-based operations, libev usually has to
		3836	linearly search all the priorities, but starting/stopping and activating
		3837	watchers becomes O(1) with respect to priority handling.
		3838
		3839	=item Sending an ev_async: O(1)
		3840
		3841	=item Processing ev_async_send: O(number_of_async_watchers)
		3842
		3843	=item Processing signals: O(max_signal_number)
		3844
		3845	Sending involves a system call I<iff> there were no other C<ev_async_send>
		3846	calls in the current loop iteration. Checking for async and signal events
		3847	involves iterating over all running async watchers or all signal numbers.
		3848
		3849	=back
		3850
1217		3851
1218	=head1 AUTHOR	3852	=head1 AUTHOR
1219		3853
1220	Marc Lehmann <libev@schmorp.de>.	3854	Marc Lehmann <libev@schmorp.de>.
1221		3855

Diff Legend

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