ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.51 by root, Tue Nov 27 19:23:31 2007 UTC vs.
Revision 1.160 by root, Thu May 22 03:06:58 2008 UTC

…		…
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
		9	=head2 EXAMPLE PROGRAM
		10
		11	// a single header file is required
		12	#include <ev.h>
		13
		14	// every watcher type has its own typedef'd struct
		15	// with the name ev_<type>
		16	ev_io stdin_watcher;
		17	ev_timer timeout_watcher;
		18
		19	// all watcher callbacks have a similar signature
		20	// this callback is called when data is readable on stdin
		21	static void
		22	stdin_cb (EV_P_ struct ev_io *w, int revents)
		23	{
		24	puts ("stdin ready");
		25	// for one-shot events, one must manually stop the watcher
		26	// with its corresponding stop function.
		27	ev_io_stop (EV_A_ w);
		28
		29	// this causes all nested ev_loop's to stop iterating
		30	ev_unloop (EV_A_ EVUNLOOP_ALL);
		31	}
		32
		33	// another callback, this time for a time-out
		34	static void
		35	timeout_cb (EV_P_ struct ev_timer *w, int revents)
		36	{
		37	puts ("timeout");
		38	// this causes the innermost ev_loop to stop iterating
		39	ev_unloop (EV_A_ EVUNLOOP_ONE);
		40	}
		41
		42	int
		43	main (void)
		44	{
		45	// use the default event loop unless you have special needs
		46	struct ev_loop *loop = ev_default_loop (0);
		47
		48	// initialise an io watcher, then start it
		49	// this one will watch for stdin to become readable
		50	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		51	ev_io_start (loop, &stdin_watcher);
		52
		53	// initialise a timer watcher, then start it
		54	// simple non-repeating 5.5 second timeout
		55	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		56	ev_timer_start (loop, &timeout_watcher);
		57
		58	// now wait for events to arrive
		59	ev_loop (loop, 0);
		60
		61	// unloop was called, so exit
		62	return 0;
		63	}
		64
9	=head1 DESCRIPTION	65	=head1 DESCRIPTION
10		66
		67	The newest version of this document is also available as an html-formatted
		68	web page you might find easier to navigate when reading it for the first
		69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		70
11	Libev is an event loop: you register interest in certain events (such as a	71	Libev is an event loop: you register interest in certain events (such as a
12	file descriptor being readable or a timeout occuring), and it will manage	72	file descriptor being readable or a timeout occurring), and it will manage
13	these event sources and provide your program with events.	73	these event sources and provide your program with events.
14		74
15	To do this, it must take more or less complete control over your process	75	To do this, it must take more or less complete control over your process
16	(or thread) by executing the I<event loop> handler, and will then	76	(or thread) by executing the I<event loop> handler, and will then
17	communicate events via a callback mechanism.	77	communicate events via a callback mechanism.
…		…
19	You register interest in certain events by registering so-called I<event	79	You register interest in certain events by registering so-called I<event
20	watchers>, which are relatively small C structures you initialise with the	80	watchers>, which are relatively small C structures you initialise with the
21	details of the event, and then hand it over to libev by I<starting> the	81	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	82	watcher.
23		83
24	=head1 FEATURES	84	=head2 FEATURES
25		85
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
29	events (related to SIGCHLD), and event watchers dealing with the event	89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	90	with customised rescheduling (C<ev_periodic>), synchronous signals
		91	(C<ev_signal>), process status change events (C<ev_child>), and event
		92	watchers dealing with the event loop mechanism itself (C<ev_idle>,
		93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as
		94	file watchers (C<ev_stat>) and even limited support for fork events
		95	(C<ev_fork>).
		96
		97	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	99	for example).
33		100
34	=head1 CONVENTIONS	101	=head2 CONVENTIONS
35		102
36	Libev is very configurable. In this manual the default configuration	103	Libev is very configurable. In this manual the default (and most common)
37	will be described, which supports multiple event loops. For more info	104	configuration will be described, which supports multiple event loops. For
38	about various configuration options please have a look at the file	105	more info about various configuration options please have a look at
39	F<README.embed> in the libev distribution. If libev was configured without	106	B<EMBED> section in this manual. If libev was configured without support
40	support for multiple event loops, then all functions taking an initial	107	for multiple event loops, then all functions taking an initial argument of
41	argument of name C<loop> (which is always of type C<struct ev_loop *>)	108	name C<loop> (which is always of type C<struct ev_loop *>) will not have
42	will not have this argument.	109	this argument.
43		110
44	=head1 TIME REPRESENTATION	111	=head2 TIME REPRESENTATION
45		112
46	Libev represents time as a single floating point number, representing the	113	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	114	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	115	the beginning of 1970, details are complicated, don't ask). This type is
49	called C<ev_tstamp>, which is what you should use too. It usually aliases	116	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the C<double> type in C, and when you need to do any calculations on	117	to the C<double> type in C, and when you need to do any calculations on
51	it, you should treat it as such.	118	it, you should treat it as some floatingpoint value. Unlike the name
		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 syscall 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.
		141
52		142
53	=head1 GLOBAL FUNCTIONS	143	=head1 GLOBAL FUNCTIONS
54		144
55	These functions can be called anytime, even before initialising the	145	These functions can be called anytime, even before initialising the
56	library in any way.	146	library in any way.
…		…
61		151
62	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
63	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
64	you actually want to know.	154	you actually want to know.
65		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 subsecond-resolution C<sleep ()>.
		161
66	=item int ev_version_major ()	162	=item int ev_version_major ()
67		163
68	=item int ev_version_minor ()	164	=item int ev_version_minor ()
69		165
70	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
71	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
72	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
73	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
74	version of the library your program was compiled against.	170	version of the library your program was compiled against.
75		171
		172	These version numbers refer to the ABI version of the library, not the
		173	release version.
		174
76	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,
77	as this indicates an incompatible change. Minor versions are usually	176	as this indicates an incompatible change. Minor versions are usually
78	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
79	not a problem.	178	not a problem.
80		179
81	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
82	version:	181	version.
83		182
84	assert (("libev version mismatch",	183	assert (("libev version mismatch",
85	ev_version_major () == EV_VERSION_MAJOR	184	ev_version_major () == EV_VERSION_MAJOR
86	&& ev_version_minor () >= EV_VERSION_MINOR));	185	&& ev_version_minor () >= EV_VERSION_MINOR));
87		186
…		…
117		216
118	See the description of C<ev_embed> watchers for more info.	217	See the description of C<ev_embed> watchers for more info.
119		218
120	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	219	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
121		220
122	Sets the allocation function to use (the prototype is similar to the	221	Sets the allocation function to use (the prototype is similar - the
123	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
124	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
125	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
126	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.
127		230
128	You could override this function in high-availability programs to, say,	231	You could override this function in high-availability programs to, say,
129	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,
130	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.
131		234
132	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
133	retries: better than mine).	236	retries (example requires a standards-compliant C<realloc>).
134		237
135	static void *	238	static void *
136	persistent_realloc (void *ptr, long size)	239	persistent_realloc (void *ptr, size_t size)
137	{	240	{
138	for (;;)	241	for (;;)
139	{	242	{
140	void *newptr = realloc (ptr, size);	243	void *newptr = realloc (ptr, size);
141		244
…		…
157	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 sitution, no
158	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
159	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
160	(such as abort).	263	(such as abort).
161		264
162	Example: do the same thing as libev does internally:	265	Example: This is basically the same thing that libev does internally, too.
163		266
164	static void	267	static void
165	fatal_error (const char *msg)	268	fatal_error (const char *msg)
166	{	269	{
167	perror (msg);	270	perror (msg);
…		…
177		280
178	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 library knows two
179	types of such loops, the I<default> loop, which supports signals and child	282	types of such loops, the I<default> loop, which supports signals and child
180	events, and dynamically created loops which do not.	283	events, and dynamically created loops which do not.
181		284
182	If you use threads, a common model is to run the default event loop
183	in your main thread (or in a separate thread) and for each thread you
184	create, you also create another event loop. Libev itself does no locking
185	whatsoever, so if you mix calls to the same event loop in different
186	threads, make sure you lock (this is usually a bad idea, though, even if
187	done correctly, because it's hideous and inefficient).
188
189	=over 4	285	=over 4
190		286
191	=item struct ev_loop *ev_default_loop (unsigned int flags)	287	=item struct ev_loop *ev_default_loop (unsigned int flags)
192		288
193	This will initialise the default event loop if it hasn't been initialised	289	This will initialise the default event loop if it hasn't been initialised
…		…
195	false. If it already was initialised it simply returns it (and ignores the	291	false. If it already was initialised it simply returns it (and ignores the
196	flags. If that is troubling you, check C<ev_backend ()> afterwards).	292	flags. If that is troubling you, check C<ev_backend ()> afterwards).
197		293
198	If you don't know what event loop to use, use the one returned from this	294	If you don't know what event loop to use, use the one returned from this
199	function.	295	function.
		296
		297	Note that this function is I<not> thread-safe, so if you want to use it
		298	from multiple threads, you have to lock (note also that this is unlikely,
		299	as loops cannot bes hared easily between threads anyway).
		300
		301	The default loop is the only loop that can handle C<ev_signal> and
		302	C<ev_child> watchers, and to do this, it always registers a handler
		303	for C<SIGCHLD>. If this is a problem for your app you can either
		304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
		305	can simply overwrite the C<SIGCHLD> signal handler I<after> calling
		306	C<ev_default_init>.
200		307
201	The flags argument can be used to specify special behaviour or specific	308	The flags argument can be used to specify special behaviour or specific
202	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	309	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
203		310
204	The following flags are supported:	311	The following flags are supported:
…		…
217	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	324	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
218	override the flags completely if it is found in the environment. This is	325	override the flags completely if it is found in the environment. This is
219	useful to try out specific backends to test their performance, or to work	326	useful to try out specific backends to test their performance, or to work
220	around bugs.	327	around bugs.
221		328
		329	=item C<EVFLAG_FORKCHECK>
		330
		331	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after
		332	a fork, you can also make libev check for a fork in each iteration by
		333	enabling this flag.
		334
		335	This works by calling C<getpid ()> on every iteration of the loop,
		336	and thus this might slow down your event loop if you do a lot of loop
		337	iterations and little real work, but is usually not noticeable (on my
		338	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
		339	without a syscall and thus I<very> fast, but my GNU/Linux system also has
		340	C<pthread_atfork> which is even faster).
		341
		342	The big advantage of this flag is that you can forget about fork (and
		343	forget about forgetting to tell libev about forking) when you use this
		344	flag.
		345
		346	This flag setting cannot be overriden or specified in the C<LIBEV_FLAGS>
		347	environment variable.
		348
222	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	349	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
223		350
224	This is your standard select(2) backend. Not I<completely> standard, as	351	This is your standard select(2) backend. Not I<completely> standard, as
225	libev tries to roll its own fd_set with no limits on the number of fds,	352	libev tries to roll its own fd_set with no limits on the number of fds,
226	but if that fails, expect a fairly low limit on the number of fds when	353	but if that fails, expect a fairly low limit on the number of fds when
227	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	354	using this backend. It doesn't scale too well (O(highest_fd)), but its
228	the fastest backend for a low number of fds.	355	usually the fastest backend for a low number of (low-numbered :) fds.
		356
		357	To get good performance out of this backend you need a high amount of
		358	parallelity (most of the file descriptors should be busy). If you are
		359	writing a server, you should C<accept ()> in a loop to accept as many
		360	connections as possible during one iteration. You might also want to have
		361	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		362	readiness notifications you get per iteration.
229		363
230	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	364	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
231		365
232	And this is your standard poll(2) backend. It's more complicated than	366	And this is your standard poll(2) backend. It's more complicated
233	select, but handles sparse fds better and has no artificial limit on the	367	than select, but handles sparse fds better and has no artificial
234	number of fds you can use (except it will slow down considerably with a	368	limit on the number of fds you can use (except it will slow down
235	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	369	considerably with a lot of inactive fds). It scales similarly to select,
		370	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		371	performance tips.
236		372
237	=item C<EVBACKEND_EPOLL> (value 4, Linux)	373	=item C<EVBACKEND_EPOLL> (value 4, Linux)
238		374
239	For few fds, this backend is a bit little slower than poll and select,	375	For few fds, this backend is a bit little slower than poll and select,
240	but it scales phenomenally better. While poll and select usually scale like	376	but it scales phenomenally better. While poll and select usually scale
241	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	377	like O(total_fds) where n is the total number of fds (or the highest fd),
242	either O(1) or O(active_fds).	378	epoll scales either O(1) or O(active_fds). The epoll design has a number
		379	of shortcomings, such as silently dropping events in some hard-to-detect
		380	cases and requiring a syscall per fd change, no fork support and bad
		381	support for dup.
243		382
244	While stopping and starting an I/O watcher in the same iteration will	383	While stopping, setting and starting an I/O watcher in the same iteration
245	result in some caching, there is still a syscall per such incident	384	will result in some caching, there is still a syscall per such incident
246	(because the fd could point to a different file description now), so its	385	(because the fd could point to a different file description now), so its
247	best to avoid that. Also, dup()ed file descriptors might not work very	386	best to avoid that. Also, C<dup ()>'ed file descriptors might not work
248	well if you register events for both fds.	387	very well if you register events for both fds.
249		388
250	Please note that epoll sometimes generates spurious notifications, so you	389	Please note that epoll sometimes generates spurious notifications, so you
251	need to use non-blocking I/O or other means to avoid blocking when no data	390	need to use non-blocking I/O or other means to avoid blocking when no data
252	(or space) is available.	391	(or space) is available.
253		392
		393	Best performance from this backend is achieved by not unregistering all
		394	watchers for a file descriptor until it has been closed, if possible, i.e.
		395	keep at least one watcher active per fd at all times.
		396
		397	While nominally embeddeble in other event loops, this feature is broken in
		398	all kernel versions tested so far.
		399
254	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	400	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
255		401
256	Kqueue deserves special mention, as at the time of this writing, it	402	Kqueue deserves special mention, as at the time of this writing, it
257	was broken on all BSDs except NetBSD (usually it doesn't work with	403	was broken on all BSDs except NetBSD (usually it doesn't work reliably
258	anything but sockets and pipes, except on Darwin, where of course its	404	with anything but sockets and pipes, except on Darwin, where of course
259	completely useless). For this reason its not being "autodetected"	405	it's completely useless). For this reason it's not being "autodetected"
260	unless you explicitly specify it explicitly in the flags (i.e. using	406	unless you explicitly specify it explicitly in the flags (i.e. using
261	C<EVBACKEND_KQUEUE>).	407	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		408	system like NetBSD.
		409
		410	You still can embed kqueue into a normal poll or select backend and use it
		411	only for sockets (after having made sure that sockets work with kqueue on
		412	the target platform). See C<ev_embed> watchers for more info.
262		413
263	It scales in the same way as the epoll backend, but the interface to the	414	It scales in the same way as the epoll backend, but the interface to the
264	kernel is more efficient (which says nothing about its actual speed, of	415	kernel is more efficient (which says nothing about its actual speed, of
265	course). While starting and stopping an I/O watcher does not cause an	416	course). While stopping, setting and starting an I/O watcher does never
266	extra syscall as with epoll, it still adds up to four event changes per	417	cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to
267	incident, so its best to avoid that.	418	two event changes per incident, support for C<fork ()> is very bad and it
		419	drops fds silently in similarly hard-to-detect cases.
		420
		421	This backend usually performs well under most conditions.
		422
		423	While nominally embeddable in other event loops, this doesn't work
		424	everywhere, so you might need to test for this. And since it is broken
		425	almost everywhere, you should only use it when you have a lot of sockets
		426	(for which it usually works), by embedding it into another event loop
		427	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for
		428	sockets.
268		429
269	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	430	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
270		431
271	This is not implemented yet (and might never be).	432	This is not implemented yet (and might never be, unless you send me an
		433	implementation). According to reports, C</dev/poll> only supports sockets
		434	and is not embeddable, which would limit the usefulness of this backend
		435	immensely.
272		436
273	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	437	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
274		438
275	This uses the Solaris 10 port mechanism. As with everything on Solaris,	439	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
276	it's really slow, but it still scales very well (O(active_fds)).	440	it's really slow, but it still scales very well (O(active_fds)).
277		441
278	Please note that solaris ports can result in a lot of spurious	442	Please note that solaris event ports can deliver a lot of spurious
279	notifications, so you need to use non-blocking I/O or other means to avoid	443	notifications, so you need to use non-blocking I/O or other means to avoid
280	blocking when no data (or space) is available.	444	blocking when no data (or space) is available.
		445
		446	While this backend scales well, it requires one system call per active
		447	file descriptor per loop iteration. For small and medium numbers of file
		448	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		449	might perform better.
		450
		451	On the positive side, ignoring the spurious readiness notifications, this
		452	backend actually performed to specification in all tests and is fully
		453	embeddable, which is a rare feat among the OS-specific backends.
281		454
282	=item C<EVBACKEND_ALL>	455	=item C<EVBACKEND_ALL>
283		456
284	Try all backends (even potentially broken ones that wouldn't be tried	457	Try all backends (even potentially broken ones that wouldn't be tried
285	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	458	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
286	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	459	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
287		460
		461	It is definitely not recommended to use this flag.
		462
288	=back	463	=back
289		464
290	If one or more of these are ored into the flags value, then only these	465	If one or more of these are ored into the flags value, then only these
291	backends will be tried (in the reverse order as given here). If none are	466	backends will be tried (in the reverse order as listed here). If none are
292	specified, most compiled-in backend will be tried, usually in reverse	467	specified, all backends in C<ev_recommended_backends ()> will be tried.
293	order of their flag values :)
294		468
295	The most typical usage is like this:	469	The most typical usage is like this:
296		470
297	if (!ev_default_loop (0))	471	if (!ev_default_loop (0))
298	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	472	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
…		…
313	Similar to C<ev_default_loop>, but always creates a new event loop that is	487	Similar to C<ev_default_loop>, but always creates a new event loop that is
314	always distinct from the default loop. Unlike the default loop, it cannot	488	always distinct from the default loop. Unlike the default loop, it cannot
315	handle signal and child watchers, and attempts to do so will be greeted by	489	handle signal and child watchers, and attempts to do so will be greeted by
316	undefined behaviour (or a failed assertion if assertions are enabled).	490	undefined behaviour (or a failed assertion if assertions are enabled).
317		491
		492	Note that this function I<is> thread-safe, and the recommended way to use
		493	libev with threads is indeed to create one loop per thread, and using the
		494	default loop in the "main" or "initial" thread.
		495
318	Example: try to create a event loop that uses epoll and nothing else.	496	Example: Try to create a event loop that uses epoll and nothing else.
319		497
320	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	498	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
321	if (!epoller)	499	if (!epoller)
322	fatal ("no epoll found here, maybe it hides under your chair");	500	fatal ("no epoll found here, maybe it hides under your chair");
323		501
…		…
326	Destroys the default loop again (frees all memory and kernel state	504	Destroys the default loop again (frees all memory and kernel state
327	etc.). None of the active event watchers will be stopped in the normal	505	etc.). None of the active event watchers will be stopped in the normal
328	sense, so e.g. C<ev_is_active> might still return true. It is your	506	sense, so e.g. C<ev_is_active> might still return true. It is your
329	responsibility to either stop all watchers cleanly yoursef I<before>	507	responsibility to either stop all watchers cleanly yoursef I<before>
330	calling this function, or cope with the fact afterwards (which is usually	508	calling this function, or cope with the fact afterwards (which is usually
331	the easiest thing, youc na just ignore the watchers and/or C<free ()> them	509	the easiest thing, you can just ignore the watchers and/or C<free ()> them
332	for example).	510	for example).
		511
		512	Note that certain global state, such as signal state, will not be freed by
		513	this function, and related watchers (such as signal and child watchers)
		514	would need to be stopped manually.
		515
		516	In general it is not advisable to call this function except in the
		517	rare occasion where you really need to free e.g. the signal handling
		518	pipe fds. If you need dynamically allocated loops it is better to use
		519	C<ev_loop_new> and C<ev_loop_destroy>).
333		520
334	=item ev_loop_destroy (loop)	521	=item ev_loop_destroy (loop)
335		522
336	Like C<ev_default_destroy>, but destroys an event loop created by an	523	Like C<ev_default_destroy>, but destroys an event loop created by an
337	earlier call to C<ev_loop_new>.	524	earlier call to C<ev_loop_new>.
338		525
339	=item ev_default_fork ()	526	=item ev_default_fork ()
340		527
		528	This function sets a flag that causes subsequent C<ev_loop> iterations
341	This function reinitialises the kernel state for backends that have	529	to reinitialise the kernel state for backends that have one. Despite the
342	one. Despite the name, you can call it anytime, but it makes most sense	530	name, you can call it anytime, but it makes most sense after forking, in
343	after forking, in either the parent or child process (or both, but that	531	the child process (or both child and parent, but that again makes little
344	again makes little sense).	532	sense). You I<must> call it in the child before using any of the libev
		533	functions, and it will only take effect at the next C<ev_loop> iteration.
345		534
346	You I<must> call this function in the child process after forking if and	535	On the other hand, you only need to call this function in the child
347	only if you want to use the event library in both processes. If you just	536	process if and only if you want to use the event library in the child. If
348	fork+exec, you don't have to call it.	537	you just fork+exec, you don't have to call it at all.
349		538
350	The function itself is quite fast and it's usually not a problem to call	539	The function itself is quite fast and it's usually not a problem to call
351	it just in case after a fork. To make this easy, the function will fit in	540	it just in case after a fork. To make this easy, the function will fit in
352	quite nicely into a call to C<pthread_atfork>:	541	quite nicely into a call to C<pthread_atfork>:
353		542
354	pthread_atfork (0, 0, ev_default_fork);	543	pthread_atfork (0, 0, ev_default_fork);
355		544
356	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
357	without calling this function, so if you force one of those backends you
358	do not need to care.
359
360	=item ev_loop_fork (loop)	545	=item ev_loop_fork (loop)
361		546
362	Like C<ev_default_fork>, but acts on an event loop created by	547	Like C<ev_default_fork>, but acts on an event loop created by
363	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	548	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
364	after fork, and how you do this is entirely your own problem.	549	after fork, and how you do this is entirely your own problem.
		550
		551	=item int ev_is_default_loop (loop)
		552
		553	Returns true when the given loop actually is the default loop, false otherwise.
		554
		555	=item unsigned int ev_loop_count (loop)
		556
		557	Returns the count of loop iterations for the loop, which is identical to
		558	the number of times libev did poll for new events. It starts at C<0> and
		559	happily wraps around with enough iterations.
		560
		561	This value can sometimes be useful as a generation counter of sorts (it
		562	"ticks" the number of loop iterations), as it roughly corresponds with
		563	C<ev_prepare> and C<ev_check> calls.
365		564
366	=item unsigned int ev_backend (loop)	565	=item unsigned int ev_backend (loop)
367		566
368	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	567	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
369	use.	568	use.
…		…
372		571
373	Returns the current "event loop time", which is the time the event loop	572	Returns the current "event loop time", which is the time the event loop
374	received events and started processing them. This timestamp does not	573	received events and started processing them. This timestamp does not
375	change as long as callbacks are being processed, and this is also the base	574	change as long as callbacks are being processed, and this is also the base
376	time used for relative timers. You can treat it as the timestamp of the	575	time used for relative timers. You can treat it as the timestamp of the
377	event occuring (or more correctly, libev finding out about it).	576	event occurring (or more correctly, libev finding out about it).
378		577
379	=item ev_loop (loop, int flags)	578	=item ev_loop (loop, int flags)
380		579
381	Finally, this is it, the event handler. This function usually is called	580	Finally, this is it, the event handler. This function usually is called
382	after you initialised all your watchers and you want to start handling	581	after you initialised all your watchers and you want to start handling
…		…
403	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	602	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
404	usually a better approach for this kind of thing.	603	usually a better approach for this kind of thing.
405		604
406	Here are the gory details of what C<ev_loop> does:	605	Here are the gory details of what C<ev_loop> does:
407		606
408	* If there are no active watchers (reference count is zero), return.	607	- Before the first iteration, call any pending watchers.
409	- Queue prepare watchers and then call all outstanding watchers.	608	* If EVFLAG_FORKCHECK was used, check for a fork.
		609	- If a fork was detected, queue and call all fork watchers.
		610	- Queue and call all prepare watchers.
410	- If we have been forked, recreate the kernel state.	611	- If we have been forked, recreate the kernel state.
411	- Update the kernel state with all outstanding changes.	612	- Update the kernel state with all outstanding changes.
412	- Update the "event loop time".	613	- Update the "event loop time".
413	- Calculate for how long to block.	614	- Calculate for how long to sleep or block, if at all
		615	(active idle watchers, EVLOOP_NONBLOCK or not having
		616	any active watchers at all will result in not sleeping).
		617	- Sleep if the I/O and timer collect interval say so.
414	- Block the process, waiting for any events.	618	- Block the process, waiting for any events.
415	- Queue all outstanding I/O (fd) events.	619	- Queue all outstanding I/O (fd) events.
416	- Update the "event loop time" and do time jump handling.	620	- Update the "event loop time" and do time jump handling.
417	- Queue all outstanding timers.	621	- Queue all outstanding timers.
418	- Queue all outstanding periodics.	622	- Queue all outstanding periodics.
419	- If no events are pending now, queue all idle watchers.	623	- If no events are pending now, queue all idle watchers.
420	- Queue all check watchers.	624	- Queue all check watchers.
421	- Call all queued watchers in reverse order (i.e. check watchers first).	625	- Call all queued watchers in reverse order (i.e. check watchers first).
422	Signals and child watchers are implemented as I/O watchers, and will	626	Signals and child watchers are implemented as I/O watchers, and will
423	be handled here by queueing them when their watcher gets executed.	627	be handled here by queueing them when their watcher gets executed.
424	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	628	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
425	were used, return, otherwise continue with step *.	629	were used, or there are no active watchers, return, otherwise
		630	continue with step *.
426		631
427	Example: queue some jobs and then loop until no events are outsanding	632	Example: Queue some jobs and then loop until no events are outstanding
428	anymore.	633	anymore.
429		634
430	... queue jobs here, make sure they register event watchers as long	635	... queue jobs here, make sure they register event watchers as long
431	... as they still have work to do (even an idle watcher will do..)	636	... as they still have work to do (even an idle watcher will do..)
432	ev_loop (my_loop, 0);	637	ev_loop (my_loop, 0);
…		…
436		641
437	Can be used to make a call to C<ev_loop> return early (but only after it	642	Can be used to make a call to C<ev_loop> return early (but only after it
438	has processed all outstanding events). The C<how> argument must be either	643	has processed all outstanding events). The C<how> argument must be either
439	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	644	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
440	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	645	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
		646
		647	This "unloop state" will be cleared when entering C<ev_loop> again.
441		648
442	=item ev_ref (loop)	649	=item ev_ref (loop)
443		650
444	=item ev_unref (loop)	651	=item ev_unref (loop)
445		652
…		…
450	returning, ev_unref() after starting, and ev_ref() before stopping it. For	657	returning, ev_unref() after starting, and ev_ref() before stopping it. For
451	example, libev itself uses this for its internal signal pipe: It is not	658	example, libev itself uses this for its internal signal pipe: It is not
452	visible to the libev user and should not keep C<ev_loop> from exiting if	659	visible to the libev user and should not keep C<ev_loop> from exiting if
453	no event watchers registered by it are active. It is also an excellent	660	no event watchers registered by it are active. It is also an excellent
454	way to do this for generic recurring timers or from within third-party	661	way to do this for generic recurring timers or from within third-party
455	libraries. Just remember to I<unref after start> and I<ref before stop>.	662	libraries. Just remember to I<unref after start> and I<ref before stop>
		663	(but only if the watcher wasn't active before, or was active before,
		664	respectively).
456		665
457	Example: create a signal watcher, but keep it from keeping C<ev_loop>	666	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
458	running when nothing else is active.	667	running when nothing else is active.
459		668
460	struct dv_signal exitsig;	669	struct ev_signal exitsig;
461	ev_signal_init (&exitsig, sig_cb, SIGINT);	670	ev_signal_init (&exitsig, sig_cb, SIGINT);
462	ev_signal_start (myloop, &exitsig);	671	ev_signal_start (loop, &exitsig);
463	evf_unref (myloop);	672	evf_unref (loop);
464		673
465	Example: for some weird reason, unregister the above signal handler again.	674	Example: For some weird reason, unregister the above signal handler again.
466		675
467	ev_ref (myloop);	676	ev_ref (loop);
468	ev_signal_stop (myloop, &exitsig);	677	ev_signal_stop (loop, &exitsig);
		678
		679	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		680
		681	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		682
		683	These advanced functions influence the time that libev will spend waiting
		684	for events. Both are by default C<0>, meaning that libev will try to
		685	invoke timer/periodic callbacks and I/O callbacks with minimum latency.
		686
		687	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		688	allows libev to delay invocation of I/O and timer/periodic callbacks to
		689	increase efficiency of loop iterations.
		690
		691	The background is that sometimes your program runs just fast enough to
		692	handle one (or very few) event(s) per loop iteration. While this makes
		693	the program responsive, it also wastes a lot of CPU time to poll for new
		694	events, especially with backends like C<select ()> which have a high
		695	overhead for the actual polling but can deliver many events at once.
		696
		697	By setting a higher I<io collect interval> you allow libev to spend more
		698	time collecting I/O events, so you can handle more events per iteration,
		699	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		700	C<ev_timer>) will be not affected. Setting this to a non-null value will
		701	introduce an additional C<ev_sleep ()> call into most loop iterations.
		702
		703	Likewise, by setting a higher I<timeout collect interval> you allow libev
		704	to spend more time collecting timeouts, at the expense of increased
		705	latency (the watcher callback will be called later). C<ev_io> watchers
		706	will not be affected. Setting this to a non-null value will not introduce
		707	any overhead in libev.
		708
		709	Many (busy) programs can usually benefit by setting the io collect
		710	interval to a value near C<0.1> or so, which is often enough for
		711	interactive servers (of course not for games), likewise for timeouts. It
		712	usually doesn't make much sense to set it to a lower value than C<0.01>,
		713	as this approsaches the timing granularity of most systems.
		714
		715	=item ev_loop_verify (loop)
		716
		717	This function only does something when C<EV_VERIFY> support has been
		718	compiled in. It tries to go through all internal structures and checks
		719	them for validity. If anything is found to be inconsistent, it will print
		720	an error message to standard error and call C<abort ()>.
		721
		722	This can be used to catch bugs inside libev itself: under normal
		723	circumstances, this function will never abort as of course libev keeps its
		724	data structures consistent.
469		725
470	=back	726	=back
471		727
472		728
473	=head1 ANATOMY OF A WATCHER	729	=head1 ANATOMY OF A WATCHER
…		…
573	=item C<EV_FORK>	829	=item C<EV_FORK>
574		830
575	The event loop has been resumed in the child process after fork (see	831	The event loop has been resumed in the child process after fork (see
576	C<ev_fork>).	832	C<ev_fork>).
577		833
		834	=item C<EV_ASYNC>
		835
		836	The given async watcher has been asynchronously notified (see C<ev_async>).
		837
578	=item C<EV_ERROR>	838	=item C<EV_ERROR>
579		839
580	An unspecified error has occured, the watcher has been stopped. This might	840	An unspecified error has occured, the watcher has been stopped. This might
581	happen because the watcher could not be properly started because libev	841	happen because the watcher could not be properly started because libev
582	ran out of memory, a file descriptor was found to be closed or any other	842	ran out of memory, a file descriptor was found to be closed or any other
…		…
653	=item bool ev_is_pending (ev_TYPE *watcher)	913	=item bool ev_is_pending (ev_TYPE *watcher)
654		914
655	Returns a true value iff the watcher is pending, (i.e. it has outstanding	915	Returns a true value iff the watcher is pending, (i.e. it has outstanding
656	events but its callback has not yet been invoked). As long as a watcher	916	events but its callback has not yet been invoked). As long as a watcher
657	is pending (but not active) you must not call an init function on it (but	917	is pending (but not active) you must not call an init function on it (but
658	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	918	C<ev_TYPE_set> is safe), you must not change its priority, and you must
659	libev (e.g. you cnanot C<free ()> it).	919	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		920	it).
660		921
661	=item callback = ev_cb (ev_TYPE *watcher)	922	=item callback ev_cb (ev_TYPE *watcher)
662		923
663	Returns the callback currently set on the watcher.	924	Returns the callback currently set on the watcher.
664		925
665	=item ev_cb_set (ev_TYPE *watcher, callback)	926	=item ev_cb_set (ev_TYPE *watcher, callback)
666		927
667	Change the callback. You can change the callback at virtually any time	928	Change the callback. You can change the callback at virtually any time
668	(modulo threads).	929	(modulo threads).
		930
		931	=item ev_set_priority (ev_TYPE *watcher, priority)
		932
		933	=item int ev_priority (ev_TYPE *watcher)
		934
		935	Set and query the priority of the watcher. The priority is a small
		936	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		937	(default: C<-2>). Pending watchers with higher priority will be invoked
		938	before watchers with lower priority, but priority will not keep watchers
		939	from being executed (except for C<ev_idle> watchers).
		940
		941	This means that priorities are I<only> used for ordering callback
		942	invocation after new events have been received. This is useful, for
		943	example, to reduce latency after idling, or more often, to bind two
		944	watchers on the same event and make sure one is called first.
		945
		946	If you need to suppress invocation when higher priority events are pending
		947	you need to look at C<ev_idle> watchers, which provide this functionality.
		948
		949	You I<must not> change the priority of a watcher as long as it is active or
		950	pending.
		951
		952	The default priority used by watchers when no priority has been set is
		953	always C<0>, which is supposed to not be too high and not be too low :).
		954
		955	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		956	fine, as long as you do not mind that the priority value you query might
		957	or might not have been adjusted to be within valid range.
		958
		959	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		960
		961	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		962	C<loop> nor C<revents> need to be valid as long as the watcher callback
		963	can deal with that fact.
		964
		965	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		966
		967	If the watcher is pending, this function returns clears its pending status
		968	and returns its C<revents> bitset (as if its callback was invoked). If the
		969	watcher isn't pending it does nothing and returns C<0>.
669		970
670	=back	971	=back
671		972
672		973
673	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	974	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
…		…
694	{	995	{
695	struct my_io *w = (struct my_io *)w_;	996	struct my_io *w = (struct my_io *)w_;
696	...	997	...
697	}	998	}
698		999
699	More interesting and less C-conformant ways of catsing your callback type	1000	More interesting and less C-conformant ways of casting your callback type
700	have been omitted....	1001	instead have been omitted.
		1002
		1003	Another common scenario is having some data structure with multiple
		1004	watchers:
		1005
		1006	struct my_biggy
		1007	{
		1008	int some_data;
		1009	ev_timer t1;
		1010	ev_timer t2;
		1011	}
		1012
		1013	In this case getting the pointer to C<my_biggy> is a bit more complicated,
		1014	you need to use C<offsetof>:
		1015
		1016	#include <stddef.h>
		1017
		1018	static void
		1019	t1_cb (EV_P_ struct ev_timer *w, int revents)
		1020	{
		1021	struct my_biggy big = (struct my_biggy *
		1022	(((char *)w) - offsetof (struct my_biggy, t1));
		1023	}
		1024
		1025	static void
		1026	t2_cb (EV_P_ struct ev_timer *w, int revents)
		1027	{
		1028	struct my_biggy big = (struct my_biggy *
		1029	(((char *)w) - offsetof (struct my_biggy, t2));
		1030	}
701		1031
702		1032
703	=head1 WATCHER TYPES	1033	=head1 WATCHER TYPES
704		1034
705	This section describes each watcher in detail, but will not repeat	1035	This section describes each watcher in detail, but will not repeat
…		…
729	In general you can register as many read and/or write event watchers per	1059	In general you can register as many read and/or write event watchers per
730	fd as you want (as long as you don't confuse yourself). Setting all file	1060	fd as you want (as long as you don't confuse yourself). Setting all file
731	descriptors to non-blocking mode is also usually a good idea (but not	1061	descriptors to non-blocking mode is also usually a good idea (but not
732	required if you know what you are doing).	1062	required if you know what you are doing).
733		1063
734	You have to be careful with dup'ed file descriptors, though. Some backends
735	(the linux epoll backend is a notable example) cannot handle dup'ed file
736	descriptors correctly if you register interest in two or more fds pointing
737	to the same underlying file/socket/etc. description (that is, they share
738	the same underlying "file open").
739
740	If you must do this, then force the use of a known-to-be-good backend	1064	If you must do this, then force the use of a known-to-be-good backend
741	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1065	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
742	C<EVBACKEND_POLL>).	1066	C<EVBACKEND_POLL>).
743		1067
744	Another thing you have to watch out for is that it is quite easy to	1068	Another thing you have to watch out for is that it is quite easy to
745	receive "spurious" readyness notifications, that is your callback might	1069	receive "spurious" readiness notifications, that is your callback might
746	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1070	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
747	because there is no data. Not only are some backends known to create a	1071	because there is no data. Not only are some backends known to create a
748	lot of those (for example solaris ports), it is very easy to get into	1072	lot of those (for example solaris ports), it is very easy to get into
749	this situation even with a relatively standard program structure. Thus	1073	this situation even with a relatively standard program structure. Thus
750	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1074	it is best to always use non-blocking I/O: An extra C<read>(2) returning
751	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1075	C<EAGAIN> is far preferable to a program hanging until some data arrives.
752		1076
753	If you cannot run the fd in non-blocking mode (for example you should not	1077	If you cannot run the fd in non-blocking mode (for example you should not
754	play around with an Xlib connection), then you have to seperately re-test	1078	play around with an Xlib connection), then you have to seperately re-test
755	wether a file descriptor is really ready with a known-to-be good interface	1079	whether a file descriptor is really ready with a known-to-be good interface
756	such as poll (fortunately in our Xlib example, Xlib already does this on	1080	such as poll (fortunately in our Xlib example, Xlib already does this on
757	its own, so its quite safe to use).	1081	its own, so its quite safe to use).
		1082
		1083	=head3 The special problem of disappearing file descriptors
		1084
		1085	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1086	descriptor (either by calling C<close> explicitly or by any other means,
		1087	such as C<dup>). The reason is that you register interest in some file
		1088	descriptor, but when it goes away, the operating system will silently drop
		1089	this interest. If another file descriptor with the same number then is
		1090	registered with libev, there is no efficient way to see that this is, in
		1091	fact, a different file descriptor.
		1092
		1093	To avoid having to explicitly tell libev about such cases, libev follows
		1094	the following policy: Each time C<ev_io_set> is being called, libev
		1095	will assume that this is potentially a new file descriptor, otherwise
		1096	it is assumed that the file descriptor stays the same. That means that
		1097	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1098	descriptor even if the file descriptor number itself did not change.
		1099
		1100	This is how one would do it normally anyway, the important point is that
		1101	the libev application should not optimise around libev but should leave
		1102	optimisations to libev.
		1103
		1104	=head3 The special problem of dup'ed file descriptors
		1105
		1106	Some backends (e.g. epoll), cannot register events for file descriptors,
		1107	but only events for the underlying file descriptions. That means when you
		1108	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1109	events for them, only one file descriptor might actually receive events.
		1110
		1111	There is no workaround possible except not registering events
		1112	for potentially C<dup ()>'ed file descriptors, or to resort to
		1113	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1114
		1115	=head3 The special problem of fork
		1116
		1117	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1118	useless behaviour. Libev fully supports fork, but needs to be told about
		1119	it in the child.
		1120
		1121	To support fork in your programs, you either have to call
		1122	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
		1123	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
		1124	C<EVBACKEND_POLL>.
		1125
		1126	=head3 The special problem of SIGPIPE
		1127
		1128	While not really specific to libev, it is easy to forget about SIGPIPE:
		1129	when reading from a pipe whose other end has been closed, your program
		1130	gets send a SIGPIPE, which, by default, aborts your program. For most
		1131	programs this is sensible behaviour, for daemons, this is usually
		1132	undesirable.
		1133
		1134	So when you encounter spurious, unexplained daemon exits, make sure you
		1135	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1136	somewhere, as that would have given you a big clue).
		1137
		1138
		1139	=head3 Watcher-Specific Functions
758		1140
759	=over 4	1141	=over 4
760		1142
761	=item ev_io_init (ev_io *, callback, int fd, int events)	1143	=item ev_io_init (ev_io *, callback, int fd, int events)
762		1144
…		…
774		1156
775	The events being watched.	1157	The events being watched.
776		1158
777	=back	1159	=back
778		1160
		1161	=head3 Examples
		1162
779	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	1163	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
780	readable, but only once. Since it is likely line-buffered, you could	1164	readable, but only once. Since it is likely line-buffered, you could
781	attempt to read a whole line in the callback:	1165	attempt to read a whole line in the callback.
782		1166
783	static void	1167	static void
784	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1168	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
785	{	1169	{
786	ev_io_stop (loop, w);	1170	ev_io_stop (loop, w);
…		…
799		1183
800	Timer watchers are simple relative timers that generate an event after a	1184	Timer watchers are simple relative timers that generate an event after a
801	given time, and optionally repeating in regular intervals after that.	1185	given time, and optionally repeating in regular intervals after that.
802		1186
803	The timers are based on real time, that is, if you register an event that	1187	The timers are based on real time, that is, if you register an event that
804	times out after an hour and you reset your system clock to last years	1188	times out after an hour and you reset your system clock to january last
805	time, it will still time out after (roughly) and hour. "Roughly" because	1189	year, it will still time out after (roughly) and hour. "Roughly" because
806	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1190	detecting time jumps is hard, and some inaccuracies are unavoidable (the
807	monotonic clock option helps a lot here).	1191	monotonic clock option helps a lot here).
808		1192
809	The relative timeouts are calculated relative to the C<ev_now ()>	1193	The relative timeouts are calculated relative to the C<ev_now ()>
810	time. This is usually the right thing as this timestamp refers to the time	1194	time. This is usually the right thing as this timestamp refers to the time
…		…
812	you suspect event processing to be delayed and you I<need> to base the timeout	1196	you suspect event processing to be delayed and you I<need> to base the timeout
813	on the current time, use something like this to adjust for this:	1197	on the current time, use something like this to adjust for this:
814		1198
815	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1199	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
816		1200
817	The callback is guarenteed to be invoked only when its timeout has passed,	1201	The callback is guarenteed to be invoked only after its timeout has passed,
818	but if multiple timers become ready during the same loop iteration then	1202	but if multiple timers become ready during the same loop iteration then
819	order of execution is undefined.	1203	order of execution is undefined.
820		1204
		1205	=head3 Watcher-Specific Functions and Data Members
		1206
821	=over 4	1207	=over 4
822		1208
823	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1209	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
824		1210
825	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1211	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
826		1212
827	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1213	Configure the timer to trigger after C<after> seconds. If C<repeat>
828	C<0.>, then it will automatically be stopped. If it is positive, then the	1214	is C<0.>, then it will automatically be stopped once the timeout is
829	timer will automatically be configured to trigger again C<repeat> seconds	1215	reached. If it is positive, then the timer will automatically be
830	later, again, and again, until stopped manually.	1216	configured to trigger again C<repeat> seconds later, again, and again,
		1217	until stopped manually.
831		1218
832	The timer itself will do a best-effort at avoiding drift, that is, if you	1219	The timer itself will do a best-effort at avoiding drift, that is, if
833	configure a timer to trigger every 10 seconds, then it will trigger at	1220	you configure a timer to trigger every 10 seconds, then it will normally
834	exactly 10 second intervals. If, however, your program cannot keep up with	1221	trigger at exactly 10 second intervals. If, however, your program cannot
835	the timer (because it takes longer than those 10 seconds to do stuff) the	1222	keep up with the timer (because it takes longer than those 10 seconds to
836	timer will not fire more than once per event loop iteration.	1223	do stuff) the timer will not fire more than once per event loop iteration.
837		1224
838	=item ev_timer_again (loop)	1225	=item ev_timer_again (loop, ev_timer *)
839		1226
840	This will act as if the timer timed out and restart it again if it is	1227	This will act as if the timer timed out and restart it again if it is
841	repeating. The exact semantics are:	1228	repeating. The exact semantics are:
842		1229
		1230	If the timer is pending, its pending status is cleared.
		1231
843	If the timer is started but nonrepeating, stop it.	1232	If the timer is started but nonrepeating, stop it (as if it timed out).
844		1233
845	If the timer is repeating, either start it if necessary (with the repeat	1234	If the timer is repeating, either start it if necessary (with the
846	value), or reset the running timer to the repeat value.	1235	C<repeat> value), or reset the running timer to the C<repeat> value.
847		1236
848	This sounds a bit complicated, but here is a useful and typical	1237	This sounds a bit complicated, but here is a useful and typical
849	example: Imagine you have a tcp connection and you want a so-called	1238	example: Imagine you have a tcp connection and you want a so-called idle
850	idle timeout, that is, you want to be called when there have been,	1239	timeout, that is, you want to be called when there have been, say, 60
851	say, 60 seconds of inactivity on the socket. The easiest way to do	1240	seconds of inactivity on the socket. The easiest way to do this is to
852	this is to configure an C<ev_timer> with C<after>=C<repeat>=C<60> and calling	1241	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
853	C<ev_timer_again> each time you successfully read or write some data. If	1242	C<ev_timer_again> each time you successfully read or write some data. If
854	you go into an idle state where you do not expect data to travel on the	1243	you go into an idle state where you do not expect data to travel on the
855	socket, you can stop the timer, and again will automatically restart it if	1244	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
856	need be.	1245	automatically restart it if need be.
857		1246
858	You can also ignore the C<after> value and C<ev_timer_start> altogether	1247	That means you can ignore the C<after> value and C<ev_timer_start>
859	and only ever use the C<repeat> value:	1248	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
860		1249
861	ev_timer_init (timer, callback, 0., 5.);	1250	ev_timer_init (timer, callback, 0., 5.);
862	ev_timer_again (loop, timer);	1251	ev_timer_again (loop, timer);
863	...	1252	...
864	timer->again = 17.;	1253	timer->again = 17.;
865	ev_timer_again (loop, timer);	1254	ev_timer_again (loop, timer);
866	...	1255	...
867	timer->again = 10.;	1256	timer->again = 10.;
868	ev_timer_again (loop, timer);	1257	ev_timer_again (loop, timer);
869		1258
870	This is more efficient then stopping/starting the timer eahc time you want	1259	This is more slightly efficient then stopping/starting the timer each time
871	to modify its timeout value.	1260	you want to modify its timeout value.
872		1261
873	=item ev_tstamp repeat [read-write]	1262	=item ev_tstamp repeat [read-write]
874		1263
875	The current C<repeat> value. Will be used each time the watcher times out	1264	The current C<repeat> value. Will be used each time the watcher times out
876	or C<ev_timer_again> is called and determines the next timeout (if any),	1265	or C<ev_timer_again> is called and determines the next timeout (if any),
877	which is also when any modifications are taken into account.	1266	which is also when any modifications are taken into account.
878		1267
879	=back	1268	=back
880		1269
		1270	=head3 Examples
		1271
881	Example: create a timer that fires after 60 seconds.	1272	Example: Create a timer that fires after 60 seconds.
882		1273
883	static void	1274	static void
884	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1275	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
885	{	1276	{
886	.. one minute over, w is actually stopped right here	1277	.. one minute over, w is actually stopped right here
…		…
888		1279
889	struct ev_timer mytimer;	1280	struct ev_timer mytimer;
890	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1281	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
891	ev_timer_start (loop, &mytimer);	1282	ev_timer_start (loop, &mytimer);
892		1283
893	Example: create a timeout timer that times out after 10 seconds of	1284	Example: Create a timeout timer that times out after 10 seconds of
894	inactivity.	1285	inactivity.
895		1286
896	static void	1287	static void
897	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1288	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
898	{	1289	{
…		…
914	Periodic watchers are also timers of a kind, but they are very versatile	1305	Periodic watchers are also timers of a kind, but they are very versatile
915	(and unfortunately a bit complex).	1306	(and unfortunately a bit complex).
916		1307
917	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1308	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
918	but on wallclock time (absolute time). You can tell a periodic watcher	1309	but on wallclock time (absolute time). You can tell a periodic watcher
919	to trigger "at" some specific point in time. For example, if you tell a	1310	to trigger after some specific point in time. For example, if you tell a
920	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1311	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
921	+ 10.>) and then reset your system clock to the last year, then it will	1312	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1313	clock to january of the previous year, then it will take more than year
922	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1314	to trigger the event (unlike an C<ev_timer>, which would still trigger
923	roughly 10 seconds later and of course not if you reset your system time	1315	roughly 10 seconds later as it uses a relative timeout).
924	again).
925		1316
926	They can also be used to implement vastly more complex timers, such as	1317	C<ev_periodic>s can also be used to implement vastly more complex timers,
927	triggering an event on eahc midnight, local time.	1318	such as triggering an event on each "midnight, local time", or other
		1319	complicated, rules.
928		1320
929	As with timers, the callback is guarenteed to be invoked only when the	1321	As with timers, the callback is guarenteed to be invoked only when the
930	time (C<at>) has been passed, but if multiple periodic timers become ready	1322	time (C<at>) has passed, but if multiple periodic timers become ready
931	during the same loop iteration then order of execution is undefined.	1323	during the same loop iteration then order of execution is undefined.
		1324
		1325	=head3 Watcher-Specific Functions and Data Members
932		1326
933	=over 4	1327	=over 4
934		1328
935	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1329	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
936		1330
…		…
939	Lots of arguments, lets sort it out... There are basically three modes of	1333	Lots of arguments, lets sort it out... There are basically three modes of
940	operation, and we will explain them from simplest to complex:	1334	operation, and we will explain them from simplest to complex:
941		1335
942	=over 4	1336	=over 4
943		1337
944	=item * absolute timer (interval = reschedule_cb = 0)	1338	=item * absolute timer (at = time, interval = reschedule_cb = 0)
945		1339
946	In this configuration the watcher triggers an event at the wallclock time	1340	In this configuration the watcher triggers an event after the wallclock
947	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1341	time C<at> has passed and doesn't repeat. It will not adjust when a time
948	that is, if it is to be run at January 1st 2011 then it will run when the	1342	jump occurs, that is, if it is to be run at January 1st 2011 then it will
949	system time reaches or surpasses this time.	1343	run when the system time reaches or surpasses this time.
950		1344
951	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	1345	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
952		1346
953	In this mode the watcher will always be scheduled to time out at the next	1347	In this mode the watcher will always be scheduled to time out at the next
954	C<at + N * interval> time (for some integer N) and then repeat, regardless	1348	C<at + N * interval> time (for some integer N, which can also be negative)
955	of any time jumps.	1349	and then repeat, regardless of any time jumps.
956		1350
957	This can be used to create timers that do not drift with respect to system	1351	This can be used to create timers that do not drift with respect to system
958	time:	1352	time, for example, here is a C<ev_periodic> that triggers each hour, on
		1353	the hour:
959		1354
960	ev_periodic_set (&periodic, 0., 3600., 0);	1355	ev_periodic_set (&periodic, 0., 3600., 0);
961		1356
962	This doesn't mean there will always be 3600 seconds in between triggers,	1357	This doesn't mean there will always be 3600 seconds in between triggers,
963	but only that the the callback will be called when the system time shows a	1358	but only that the the callback will be called when the system time shows a
…		…
966		1361
967	Another way to think about it (for the mathematically inclined) is that	1362	Another way to think about it (for the mathematically inclined) is that
968	C<ev_periodic> will try to run the callback in this mode at the next possible	1363	C<ev_periodic> will try to run the callback in this mode at the next possible
969	time where C<time = at (mod interval)>, regardless of any time jumps.	1364	time where C<time = at (mod interval)>, regardless of any time jumps.
970		1365
		1366	For numerical stability it is preferable that the C<at> value is near
		1367	C<ev_now ()> (the current time), but there is no range requirement for
		1368	this value, and in fact is often specified as zero.
		1369
		1370	Note also that there is an upper limit to how often a timer can fire (cpu
		1371	speed for example), so if C<interval> is very small then timing stability
		1372	will of course detoriate. Libev itself tries to be exact to be about one
		1373	millisecond (if the OS supports it and the machine is fast enough).
		1374
971	=item * manual reschedule mode (reschedule_cb = callback)	1375	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
972		1376
973	In this mode the values for C<interval> and C<at> are both being	1377	In this mode the values for C<interval> and C<at> are both being
974	ignored. Instead, each time the periodic watcher gets scheduled, the	1378	ignored. Instead, each time the periodic watcher gets scheduled, the
975	reschedule callback will be called with the watcher as first, and the	1379	reschedule callback will be called with the watcher as first, and the
976	current time as second argument.	1380	current time as second argument.
977		1381
978	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1382	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
979	ever, or make any event loop modifications>. If you need to stop it,	1383	ever, or make ANY event loop modifications whatsoever>.
980	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
981	starting a prepare watcher).
982		1384
		1385	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1386	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1387	only event loop modification you are allowed to do).
		1388
983	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	1389	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic
984	ev_tstamp now)>, e.g.:	1390	*w, ev_tstamp now)>, e.g.:
985		1391
986	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1392	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
987	{	1393	{
988	return now + 60.;	1394	return now + 60.;
989	}	1395	}
…		…
991	It must return the next time to trigger, based on the passed time value	1397	It must return the next time to trigger, based on the passed time value
992	(that is, the lowest time value larger than to the second argument). It	1398	(that is, the lowest time value larger than to the second argument). It
993	will usually be called just before the callback will be triggered, but	1399	will usually be called just before the callback will be triggered, but
994	might be called at other times, too.	1400	might be called at other times, too.
995		1401
996	NOTE: I<< This callback must always return a time that is later than the	1402	NOTE: I<< This callback must always return a time that is higher than or
997	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1403	equal to the passed C<now> value >>.
998		1404
999	This can be used to create very complex timers, such as a timer that	1405	This can be used to create very complex timers, such as a timer that
1000	triggers on each midnight, local time. To do this, you would calculate the	1406	triggers on "next midnight, local time". To do this, you would calculate the
1001	next midnight after C<now> and return the timestamp value for this. How	1407	next midnight after C<now> and return the timestamp value for this. How
1002	you do this is, again, up to you (but it is not trivial, which is the main	1408	you do this is, again, up to you (but it is not trivial, which is the main
1003	reason I omitted it as an example).	1409	reason I omitted it as an example).
1004		1410
1005	=back	1411	=back
…		…
1009	Simply stops and restarts the periodic watcher again. This is only useful	1415	Simply stops and restarts the periodic watcher again. This is only useful
1010	when you changed some parameters or the reschedule callback would return	1416	when you changed some parameters or the reschedule callback would return
1011	a different time than the last time it was called (e.g. in a crond like	1417	a different time than the last time it was called (e.g. in a crond like
1012	program when the crontabs have changed).	1418	program when the crontabs have changed).
1013		1419
		1420	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1421
		1422	When active, returns the absolute time that the watcher is supposed to
		1423	trigger next.
		1424
		1425	=item ev_tstamp offset [read-write]
		1426
		1427	When repeating, this contains the offset value, otherwise this is the
		1428	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
		1429
		1430	Can be modified any time, but changes only take effect when the periodic
		1431	timer fires or C<ev_periodic_again> is being called.
		1432
1014	=item ev_tstamp interval [read-write]	1433	=item ev_tstamp interval [read-write]
1015		1434
1016	The current interval value. Can be modified any time, but changes only	1435	The current interval value. Can be modified any time, but changes only
1017	take effect when the periodic timer fires or C<ev_periodic_again> is being	1436	take effect when the periodic timer fires or C<ev_periodic_again> is being
1018	called.	1437	called.
…		…
1023	switched off. Can be changed any time, but changes only take effect when	1442	switched off. Can be changed any time, but changes only take effect when
1024	the periodic timer fires or C<ev_periodic_again> is being called.	1443	the periodic timer fires or C<ev_periodic_again> is being called.
1025		1444
1026	=back	1445	=back
1027		1446
		1447	=head3 Examples
		1448
1028	Example: call a callback every hour, or, more precisely, whenever the	1449	Example: Call a callback every hour, or, more precisely, whenever the
1029	system clock is divisible by 3600. The callback invocation times have	1450	system clock is divisible by 3600. The callback invocation times have
1030	potentially a lot of jittering, but good long-term stability.	1451	potentially a lot of jittering, but good long-term stability.
1031		1452
1032	static void	1453	static void
1033	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1454	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
…		…
1037		1458
1038	struct ev_periodic hourly_tick;	1459	struct ev_periodic hourly_tick;
1039	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	1460	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1040	ev_periodic_start (loop, &hourly_tick);	1461	ev_periodic_start (loop, &hourly_tick);
1041		1462
1042	Example: the same as above, but use a reschedule callback to do it:	1463	Example: The same as above, but use a reschedule callback to do it:
1043		1464
1044	#include <math.h>	1465	#include <math.h>
1045		1466
1046	static ev_tstamp	1467	static ev_tstamp
1047	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	1468	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
…		…
1049	return fmod (now, 3600.) + 3600.;	1470	return fmod (now, 3600.) + 3600.;
1050	}	1471	}
1051		1472
1052	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	1473	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1053		1474
1054	Example: call a callback every hour, starting now:	1475	Example: Call a callback every hour, starting now:
1055		1476
1056	struct ev_periodic hourly_tick;	1477	struct ev_periodic hourly_tick;
1057	ev_periodic_init (&hourly_tick, clock_cb,	1478	ev_periodic_init (&hourly_tick, clock_cb,
1058	fmod (ev_now (loop), 3600.), 3600., 0);	1479	fmod (ev_now (loop), 3600.), 3600., 0);
1059	ev_periodic_start (loop, &hourly_tick);	1480	ev_periodic_start (loop, &hourly_tick);
…		…
1071	with the kernel (thus it coexists with your own signal handlers as long	1492	with the kernel (thus it coexists with your own signal handlers as long
1072	as you don't register any with libev). Similarly, when the last signal	1493	as you don't register any with libev). Similarly, when the last signal
1073	watcher for a signal is stopped libev will reset the signal handler to	1494	watcher for a signal is stopped libev will reset the signal handler to
1074	SIG_DFL (regardless of what it was set to before).	1495	SIG_DFL (regardless of what it was set to before).
1075		1496
		1497	If possible and supported, libev will install its handlers with
		1498	C<SA_RESTART> behaviour enabled, so syscalls should not be unduly
		1499	interrupted. If you have a problem with syscalls getting interrupted by
		1500	signals you can block all signals in an C<ev_check> watcher and unblock
		1501	them in an C<ev_prepare> watcher.
		1502
		1503	=head3 Watcher-Specific Functions and Data Members
		1504
1076	=over 4	1505	=over 4
1077		1506
1078	=item ev_signal_init (ev_signal *, callback, int signum)	1507	=item ev_signal_init (ev_signal *, callback, int signum)
1079		1508
1080	=item ev_signal_set (ev_signal *, int signum)	1509	=item ev_signal_set (ev_signal *, int signum)
…		…
1086		1515
1087	The signal the watcher watches out for.	1516	The signal the watcher watches out for.
1088		1517
1089	=back	1518	=back
1090		1519
		1520	=head3 Examples
		1521
		1522	Example: Try to exit cleanly on SIGINT and SIGTERM.
		1523
		1524	static void
		1525	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1526	{
		1527	ev_unloop (loop, EVUNLOOP_ALL);
		1528	}
		1529
		1530	struct ev_signal signal_watcher;
		1531	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1532	ev_signal_start (loop, &sigint_cb);
		1533
1091		1534
1092	=head2 C<ev_child> - watch out for process status changes	1535	=head2 C<ev_child> - watch out for process status changes
1093		1536
1094	Child watchers trigger when your process receives a SIGCHLD in response to	1537	Child watchers trigger when your process receives a SIGCHLD in response to
1095	some child status changes (most typically when a child of yours dies).	1538	some child status changes (most typically when a child of yours dies). It
		1539	is permissible to install a child watcher I<after> the child has been
		1540	forked (which implies it might have already exited), as long as the event
		1541	loop isn't entered (or is continued from a watcher).
		1542
		1543	Only the default event loop is capable of handling signals, and therefore
		1544	you can only rgeister child watchers in the default event loop.
		1545
		1546	=head3 Process Interaction
		1547
		1548	Libev grabs C<SIGCHLD> as soon as the default event loop is
		1549	initialised. This is necessary to guarantee proper behaviour even if
		1550	the first child watcher is started after the child exits. The occurance
		1551	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		1552	synchronously as part of the event loop processing. Libev always reaps all
		1553	children, even ones not watched.
		1554
		1555	=head3 Overriding the Built-In Processing
		1556
		1557	Libev offers no special support for overriding the built-in child
		1558	processing, but if your application collides with libev's default child
		1559	handler, you can override it easily by installing your own handler for
		1560	C<SIGCHLD> after initialising the default loop, and making sure the
		1561	default loop never gets destroyed. You are encouraged, however, to use an
		1562	event-based approach to child reaping and thus use libev's support for
		1563	that, so other libev users can use C<ev_child> watchers freely.
		1564
		1565	=head3 Watcher-Specific Functions and Data Members
1096		1566
1097	=over 4	1567	=over 4
1098		1568
1099	=item ev_child_init (ev_child *, callback, int pid)	1569	=item ev_child_init (ev_child *, callback, int pid, int trace)
1100		1570
1101	=item ev_child_set (ev_child *, int pid)	1571	=item ev_child_set (ev_child *, int pid, int trace)
1102		1572
1103	Configures the watcher to wait for status changes of process C<pid> (or	1573	Configures the watcher to wait for status changes of process C<pid> (or
1104	I<any> process if C<pid> is specified as C<0>). The callback can look	1574	I<any> process if C<pid> is specified as C<0>). The callback can look
1105	at the C<rstatus> member of the C<ev_child> watcher structure to see	1575	at the C<rstatus> member of the C<ev_child> watcher structure to see
1106	the status word (use the macros from C<sys/wait.h> and see your systems	1576	the status word (use the macros from C<sys/wait.h> and see your systems
1107	C<waitpid> documentation). The C<rpid> member contains the pid of the	1577	C<waitpid> documentation). The C<rpid> member contains the pid of the
1108	process causing the status change.	1578	process causing the status change. C<trace> must be either C<0> (only
		1579	activate the watcher when the process terminates) or C<1> (additionally
		1580	activate the watcher when the process is stopped or continued).
1109		1581
1110	=item int pid [read-only]	1582	=item int pid [read-only]
1111		1583
1112	The process id this watcher watches out for, or C<0>, meaning any process id.	1584	The process id this watcher watches out for, or C<0>, meaning any process id.
1113		1585
…		…
1120	The process exit/trace status caused by C<rpid> (see your systems	1592	The process exit/trace status caused by C<rpid> (see your systems
1121	C<waitpid> and C<sys/wait.h> documentation for details).	1593	C<waitpid> and C<sys/wait.h> documentation for details).
1122		1594
1123	=back	1595	=back
1124		1596
1125	Example: try to exit cleanly on SIGINT and SIGTERM.	1597	=head3 Examples
		1598
		1599	Example: C<fork()> a new process and install a child handler to wait for
		1600	its completion.
		1601
		1602	ev_child cw;
1126		1603
1127	static void	1604	static void
1128	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	1605	child_cb (EV_P_ struct ev_child *w, int revents)
1129	{	1606	{
1130	ev_unloop (loop, EVUNLOOP_ALL);	1607	ev_child_stop (EV_A_ w);
		1608	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1131	}	1609	}
1132		1610
1133	struct ev_signal signal_watcher;	1611	pid_t pid = fork ();
1134	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	1612
1135	ev_signal_start (loop, &sigint_cb);	1613	if (pid < 0)
		1614	// error
		1615	else if (pid == 0)
		1616	{
		1617	// the forked child executes here
		1618	exit (1);
		1619	}
		1620	else
		1621	{
		1622	ev_child_init (&cw, child_cb, pid, 0);
		1623	ev_child_start (EV_DEFAULT_ &cw);
		1624	}
1136		1625
1137		1626
1138	=head2 C<ev_stat> - did the file attributes just change?	1627	=head2 C<ev_stat> - did the file attributes just change?
1139		1628
1140	This watches a filesystem path for attribute changes. That is, it calls	1629	This watches a filesystem path for attribute changes. That is, it calls
…		…
1145	not exist" is a status change like any other. The condition "path does	1634	not exist" is a status change like any other. The condition "path does
1146	not exist" is signified by the C<st_nlink> field being zero (which is	1635	not exist" is signified by the C<st_nlink> field being zero (which is
1147	otherwise always forced to be at least one) and all the other fields of	1636	otherwise always forced to be at least one) and all the other fields of
1148	the stat buffer having unspecified contents.	1637	the stat buffer having unspecified contents.
1149		1638
		1639	The path I<should> be absolute and I<must not> end in a slash. If it is
		1640	relative and your working directory changes, the behaviour is undefined.
		1641
1150	Since there is no standard to do this, the portable implementation simply	1642	Since there is no standard to do this, the portable implementation simply
1151	calls C<stat (2)> regulalry on the path to see if it changed somehow. You	1643	calls C<stat (2)> regularly on the path to see if it changed somehow. You
1152	can specify a recommended polling interval for this case. If you specify	1644	can specify a recommended polling interval for this case. If you specify
1153	a polling interval of C<0> (highly recommended!) then a I<suitable,	1645	a polling interval of C<0> (highly recommended!) then a I<suitable,
1154	unspecified default> value will be used (which you can expect to be around	1646	unspecified default> value will be used (which you can expect to be around
1155	five seconds, although this might change dynamically). Libev will also	1647	five seconds, although this might change dynamically). Libev will also
1156	impose a minimum interval which is currently around C<0.1>, but thats	1648	impose a minimum interval which is currently around C<0.1>, but thats
…		…
1158		1650
1159	This watcher type is not meant for massive numbers of stat watchers,	1651	This watcher type is not meant for massive numbers of stat watchers,
1160	as even with OS-supported change notifications, this can be	1652	as even with OS-supported change notifications, this can be
1161	resource-intensive.	1653	resource-intensive.
1162		1654
1163	At the time of this writing, no specific OS backends are implemented, but	1655	At the time of this writing, only the Linux inotify interface is
1164	if demand increases, at least a kqueue and inotify backend will be added.	1656	implemented (implementing kqueue support is left as an exercise for the
		1657	reader, note, however, that the author sees no way of implementing ev_stat
		1658	semantics with kqueue). Inotify will be used to give hints only and should
		1659	not change the semantics of C<ev_stat> watchers, which means that libev
		1660	sometimes needs to fall back to regular polling again even with inotify,
		1661	but changes are usually detected immediately, and if the file exists there
		1662	will be no polling.
		1663
		1664	=head3 ABI Issues (Largefile Support)
		1665
		1666	Libev by default (unless the user overrides this) uses the default
		1667	compilation environment, which means that on systems with optionally
		1668	disabled large file support, you get the 32 bit version of the stat
		1669	structure. When using the library from programs that change the ABI to
		1670	use 64 bit file offsets the programs will fail. In that case you have to
		1671	compile libev with the same flags to get binary compatibility. This is
		1672	obviously the case with any flags that change the ABI, but the problem is
		1673	most noticably with ev_stat and largefile support.
		1674
		1675	=head3 Inotify
		1676
		1677	When C<inotify (7)> support has been compiled into libev (generally only
		1678	available on Linux) and present at runtime, it will be used to speed up
		1679	change detection where possible. The inotify descriptor will be created lazily
		1680	when the first C<ev_stat> watcher is being started.
		1681
		1682	Inotify presence does not change the semantics of C<ev_stat> watchers
		1683	except that changes might be detected earlier, and in some cases, to avoid
		1684	making regular C<stat> calls. Even in the presence of inotify support
		1685	there are many cases where libev has to resort to regular C<stat> polling.
		1686
		1687	(There is no support for kqueue, as apparently it cannot be used to
		1688	implement this functionality, due to the requirement of having a file
		1689	descriptor open on the object at all times).
		1690
		1691	=head3 The special problem of stat time resolution
		1692
		1693	The C<stat ()> syscall only supports full-second resolution portably, and
		1694	even on systems where the resolution is higher, many filesystems still
		1695	only support whole seconds.
		1696
		1697	That means that, if the time is the only thing that changes, you can
		1698	easily miss updates: on the first update, C<ev_stat> detects a change and
		1699	calls your callback, which does something. When there is another update
		1700	within the same second, C<ev_stat> will be unable to detect it as the stat
		1701	data does not change.
		1702
		1703	The solution to this is to delay acting on a change for slightly more
		1704	than a second (or till slightly after the next full second boundary), using
		1705	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		1706	ev_timer_again (loop, w)>).
		1707
		1708	The C<.02> offset is added to work around small timing inconsistencies
		1709	of some operating systems (where the second counter of the current time
		1710	might be be delayed. One such system is the Linux kernel, where a call to
		1711	C<gettimeofday> might return a timestamp with a full second later than
		1712	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		1713	update file times then there will be a small window where the kernel uses
		1714	the previous second to update file times but libev might already execute
		1715	the timer callback).
		1716
		1717	=head3 Watcher-Specific Functions and Data Members
1165		1718
1166	=over 4	1719	=over 4
1167		1720
1168	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)	1721	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
1169		1722
…		…
1173	C<path>. The C<interval> is a hint on how quickly a change is expected to	1726	C<path>. The C<interval> is a hint on how quickly a change is expected to
1174	be detected and should normally be specified as C<0> to let libev choose	1727	be detected and should normally be specified as C<0> to let libev choose
1175	a suitable value. The memory pointed to by C<path> must point to the same	1728	a suitable value. The memory pointed to by C<path> must point to the same
1176	path for as long as the watcher is active.	1729	path for as long as the watcher is active.
1177		1730
1178	The callback will be receive C<EV_STAT> when a change was detected,	1731	The callback will receive C<EV_STAT> when a change was detected, relative
1179	relative to the attributes at the time the watcher was started (or the	1732	to the attributes at the time the watcher was started (or the last change
1180	last change was detected).	1733	was detected).
1181		1734
1182	=item ev_stat_stat (ev_stat *)	1735	=item ev_stat_stat (loop, ev_stat *)
1183		1736
1184	Updates the stat buffer immediately with new values. If you change the	1737	Updates the stat buffer immediately with new values. If you change the
1185	watched path in your callback, you could call this fucntion to avoid	1738	watched path in your callback, you could call this function to avoid
1186	detecting this change (while introducing a race condition). Can also be	1739	detecting this change (while introducing a race condition if you are not
1187	useful simply to find out the new values.	1740	the only one changing the path). Can also be useful simply to find out the
		1741	new values.
1188		1742
1189	=item ev_statdata attr [read-only]	1743	=item ev_statdata attr [read-only]
1190		1744
1191	The most-recently detected attributes of the file. Although the type is of	1745	The most-recently detected attributes of the file. Although the type is
1192	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types	1746	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
1193	suitable for your system. If the C<st_nlink> member is C<0>, then there	1747	suitable for your system, but you can only rely on the POSIX-standardised
		1748	members to be present. If the C<st_nlink> member is C<0>, then there was
1194	was some error while C<stat>ing the file.	1749	some error while C<stat>ing the file.
1195		1750
1196	=item ev_statdata prev [read-only]	1751	=item ev_statdata prev [read-only]
1197		1752
1198	The previous attributes of the file. The callback gets invoked whenever	1753	The previous attributes of the file. The callback gets invoked whenever
1199	C<prev> != C<attr>.	1754	C<prev> != C<attr>, or, more precisely, one or more of these members
		1755	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		1756	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1200		1757
1201	=item ev_tstamp interval [read-only]	1758	=item ev_tstamp interval [read-only]
1202		1759
1203	The specified interval.	1760	The specified interval.
1204		1761
1205	=item const char *path [read-only]	1762	=item const char *path [read-only]
1206		1763
1207	The filesystem path that is being watched.	1764	The filesystem path that is being watched.
1208		1765
1209	=back	1766	=back
		1767
		1768	=head3 Examples
1210		1769
1211	Example: Watch C</etc/passwd> for attribute changes.	1770	Example: Watch C</etc/passwd> for attribute changes.
1212		1771
1213	static void	1772	static void
1214	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)	1773	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
…		…
1227	}	1786	}
1228		1787
1229	...	1788	...
1230	ev_stat passwd;	1789	ev_stat passwd;
1231		1790
1232	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");	1791	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1233	ev_stat_start (loop, &passwd);	1792	ev_stat_start (loop, &passwd);
1234		1793
		1794	Example: Like above, but additionally use a one-second delay so we do not
		1795	miss updates (however, frequent updates will delay processing, too, so
		1796	one might do the work both on C<ev_stat> callback invocation I<and> on
		1797	C<ev_timer> callback invocation).
		1798
		1799	static ev_stat passwd;
		1800	static ev_timer timer;
		1801
		1802	static void
		1803	timer_cb (EV_P_ ev_timer *w, int revents)
		1804	{
		1805	ev_timer_stop (EV_A_ w);
		1806
		1807	/* now it's one second after the most recent passwd change */
		1808	}
		1809
		1810	static void
		1811	stat_cb (EV_P_ ev_stat *w, int revents)
		1812	{
		1813	/* reset the one-second timer */
		1814	ev_timer_again (EV_A_ &timer);
		1815	}
		1816
		1817	...
		1818	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		1819	ev_stat_start (loop, &passwd);
		1820	ev_timer_init (&timer, timer_cb, 0., 1.02);
		1821
1235		1822
1236	=head2 C<ev_idle> - when you've got nothing better to do...	1823	=head2 C<ev_idle> - when you've got nothing better to do...
1237		1824
1238	Idle watchers trigger events when there are no other events are pending	1825	Idle watchers trigger events when no other events of the same or higher
1239	(prepare, check and other idle watchers do not count). That is, as long	1826	priority are pending (prepare, check and other idle watchers do not
1240	as your process is busy handling sockets or timeouts (or even signals,	1827	count).
1241	imagine) it will not be triggered. But when your process is idle all idle	1828
1242	watchers are being called again and again, once per event loop iteration -	1829	That is, as long as your process is busy handling sockets or timeouts
		1830	(or even signals, imagine) of the same or higher priority it will not be
		1831	triggered. But when your process is idle (or only lower-priority watchers
		1832	are pending), the idle watchers are being called once per event loop
1243	until stopped, that is, or your process receives more events and becomes	1833	iteration - until stopped, that is, or your process receives more events
1244	busy.	1834	and becomes busy again with higher priority stuff.
1245		1835
1246	The most noteworthy effect is that as long as any idle watchers are	1836	The most noteworthy effect is that as long as any idle watchers are
1247	active, the process will not block when waiting for new events.	1837	active, the process will not block when waiting for new events.
1248		1838
1249	Apart from keeping your process non-blocking (which is a useful	1839	Apart from keeping your process non-blocking (which is a useful
1250	effect on its own sometimes), idle watchers are a good place to do	1840	effect on its own sometimes), idle watchers are a good place to do
1251	"pseudo-background processing", or delay processing stuff to after the	1841	"pseudo-background processing", or delay processing stuff to after the
1252	event loop has handled all outstanding events.	1842	event loop has handled all outstanding events.
1253		1843
		1844	=head3 Watcher-Specific Functions and Data Members
		1845
1254	=over 4	1846	=over 4
1255		1847
1256	=item ev_idle_init (ev_signal *, callback)	1848	=item ev_idle_init (ev_signal *, callback)
1257		1849
1258	Initialises and configures the idle watcher - it has no parameters of any	1850	Initialises and configures the idle watcher - it has no parameters of any
1259	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1851	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1260	believe me.	1852	believe me.
1261		1853
1262	=back	1854	=back
1263		1855
		1856	=head3 Examples
		1857
1264	Example: dynamically allocate an C<ev_idle>, start it, and in the	1858	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1265	callback, free it. Alos, use no error checking, as usual.	1859	callback, free it. Also, use no error checking, as usual.
1266		1860
1267	static void	1861	static void
1268	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	1862	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1269	{	1863	{
1270	free (w);	1864	free (w);
1271	// now do something you wanted to do when the program has	1865	// now do something you wanted to do when the program has
1272	// no longer asnything immediate to do.	1866	// no longer anything immediate to do.
1273	}	1867	}
1274		1868
1275	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	1869	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1276	ev_idle_init (idle_watcher, idle_cb);	1870	ev_idle_init (idle_watcher, idle_cb);
1277	ev_idle_start (loop, idle_cb);	1871	ev_idle_start (loop, idle_cb);
…		…
1315	with priority higher than or equal to the event loop and one coroutine	1909	with priority higher than or equal to the event loop and one coroutine
1316	of lower priority, but only once, using idle watchers to keep the event	1910	of lower priority, but only once, using idle watchers to keep the event
1317	loop from blocking if lower-priority coroutines are active, thus mapping	1911	loop from blocking if lower-priority coroutines are active, thus mapping
1318	low-priority coroutines to idle/background tasks).	1912	low-priority coroutines to idle/background tasks).
1319		1913
		1914	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		1915	priority, to ensure that they are being run before any other watchers
		1916	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
		1917	too) should not activate ("feed") events into libev. While libev fully
		1918	supports this, they might get executed before other C<ev_check> watchers
		1919	did their job. As C<ev_check> watchers are often used to embed other
		1920	(non-libev) event loops those other event loops might be in an unusable
		1921	state until their C<ev_check> watcher ran (always remind yourself to
		1922	coexist peacefully with others).
		1923
		1924	=head3 Watcher-Specific Functions and Data Members
		1925
1320	=over 4	1926	=over 4
1321		1927
1322	=item ev_prepare_init (ev_prepare *, callback)	1928	=item ev_prepare_init (ev_prepare *, callback)
1323		1929
1324	=item ev_check_init (ev_check *, callback)	1930	=item ev_check_init (ev_check *, callback)
…		…
1327	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1933	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1328	macros, but using them is utterly, utterly and completely pointless.	1934	macros, but using them is utterly, utterly and completely pointless.
1329		1935
1330	=back	1936	=back
1331		1937
1332	Example: To include a library such as adns, you would add IO watchers	1938	=head3 Examples
1333	and a timeout watcher in a prepare handler, as required by libadns, and	1939
		1940	There are a number of principal ways to embed other event loops or modules
		1941	into libev. Here are some ideas on how to include libadns into libev
		1942	(there is a Perl module named C<EV::ADNS> that does this, which you could
		1943	use as a working example. Another Perl module named C<EV::Glib> embeds a
		1944	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		1945	Glib event loop).
		1946
		1947	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1334	in a check watcher, destroy them and call into libadns. What follows is	1948	and in a check watcher, destroy them and call into libadns. What follows
1335	pseudo-code only of course:	1949	is pseudo-code only of course. This requires you to either use a low
		1950	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		1951	the callbacks for the IO/timeout watchers might not have been called yet.
1336		1952
1337	static ev_io iow [nfd];	1953	static ev_io iow [nfd];
1338	static ev_timer tw;	1954	static ev_timer tw;
1339		1955
1340	static void	1956	static void
1341	io_cb (ev_loop *loop, ev_io *w, int revents)	1957	io_cb (ev_loop *loop, ev_io *w, int revents)
1342	{	1958	{
1343	// set the relevant poll flags
1344	// could also call adns_processreadable etc. here
1345	struct pollfd *fd = (struct pollfd *)w->data;
1346	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1347	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1348	}	1959	}
1349		1960
1350	// create io watchers for each fd and a timer before blocking	1961	// create io watchers for each fd and a timer before blocking
1351	static void	1962	static void
1352	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	1963	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
1353	{	1964	{
1354	int timeout = 3600000;truct pollfd fds [nfd];	1965	int timeout = 3600000;
		1966	struct pollfd fds [nfd];
1355	// actual code will need to loop here and realloc etc.	1967	// actual code will need to loop here and realloc etc.
1356	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	1968	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1357		1969
1358	/* the callback is illegal, but won't be called as we stop during check */	1970	/* the callback is illegal, but won't be called as we stop during check */
1359	ev_timer_init (&tw, 0, timeout * 1e-3);	1971	ev_timer_init (&tw, 0, timeout * 1e-3);
1360	ev_timer_start (loop, &tw);	1972	ev_timer_start (loop, &tw);
1361		1973
1362	// create on ev_io per pollfd	1974	// create one ev_io per pollfd
1363	for (int i = 0; i < nfd; ++i)	1975	for (int i = 0; i < nfd; ++i)
1364	{	1976	{
1365	ev_io_init (iow + i, io_cb, fds [i].fd,	1977	ev_io_init (iow + i, io_cb, fds [i].fd,
1366	((fds [i].events & POLLIN ? EV_READ : 0)	1978	((fds [i].events & POLLIN ? EV_READ : 0)
1367	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));	1979	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1368		1980
1369	fds [i].revents = 0;	1981	fds [i].revents = 0;
1370	iow [i].data = fds + i;
1371	ev_io_start (loop, iow + i);	1982	ev_io_start (loop, iow + i);
1372	}	1983	}
1373	}	1984	}
1374		1985
1375	// stop all watchers after blocking	1986	// stop all watchers after blocking
…		…
1377	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	1988	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
1378	{	1989	{
1379	ev_timer_stop (loop, &tw);	1990	ev_timer_stop (loop, &tw);
1380		1991
1381	for (int i = 0; i < nfd; ++i)	1992	for (int i = 0; i < nfd; ++i)
		1993	{
		1994	// set the relevant poll flags
		1995	// could also call adns_processreadable etc. here
		1996	struct pollfd *fd = fds + i;
		1997	int revents = ev_clear_pending (iow + i);
		1998	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1999	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		2000
		2001	// now stop the watcher
1382	ev_io_stop (loop, iow + i);	2002	ev_io_stop (loop, iow + i);
		2003	}
1383		2004
1384	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));	2005	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		2006	}
		2007
		2008	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		2009	in the prepare watcher and would dispose of the check watcher.
		2010
		2011	Method 3: If the module to be embedded supports explicit event
		2012	notification (adns does), you can also make use of the actual watcher
		2013	callbacks, and only destroy/create the watchers in the prepare watcher.
		2014
		2015	static void
		2016	timer_cb (EV_P_ ev_timer *w, int revents)
		2017	{
		2018	adns_state ads = (adns_state)w->data;
		2019	update_now (EV_A);
		2020
		2021	adns_processtimeouts (ads, &tv_now);
		2022	}
		2023
		2024	static void
		2025	io_cb (EV_P_ ev_io *w, int revents)
		2026	{
		2027	adns_state ads = (adns_state)w->data;
		2028	update_now (EV_A);
		2029
		2030	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		2031	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		2032	}
		2033
		2034	// do not ever call adns_afterpoll
		2035
		2036	Method 4: Do not use a prepare or check watcher because the module you
		2037	want to embed is too inflexible to support it. Instead, youc na override
		2038	their poll function. The drawback with this solution is that the main
		2039	loop is now no longer controllable by EV. The C<Glib::EV> module does
		2040	this.
		2041
		2042	static gint
		2043	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		2044	{
		2045	int got_events = 0;
		2046
		2047	for (n = 0; n < nfds; ++n)
		2048	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		2049
		2050	if (timeout >= 0)
		2051	// create/start timer
		2052
		2053	// poll
		2054	ev_loop (EV_A_ 0);
		2055
		2056	// stop timer again
		2057	if (timeout >= 0)
		2058	ev_timer_stop (EV_A_ &to);
		2059
		2060	// stop io watchers again - their callbacks should have set
		2061	for (n = 0; n < nfds; ++n)
		2062	ev_io_stop (EV_A_ iow [n]);
		2063
		2064	return got_events;
1385	}	2065	}
1386		2066
1387		2067
1388	=head2 C<ev_embed> - when one backend isn't enough...	2068	=head2 C<ev_embed> - when one backend isn't enough...
1389		2069
…		…
1432	portable one.	2112	portable one.
1433		2113
1434	So when you want to use this feature you will always have to be prepared	2114	So when you want to use this feature you will always have to be prepared
1435	that you cannot get an embeddable loop. The recommended way to get around	2115	that you cannot get an embeddable loop. The recommended way to get around
1436	this is to have a separate variables for your embeddable loop, try to	2116	this is to have a separate variables for your embeddable loop, try to
1437	create it, and if that fails, use the normal loop for everything:	2117	create it, and if that fails, use the normal loop for everything.
		2118
		2119	=head3 Watcher-Specific Functions and Data Members
		2120
		2121	=over 4
		2122
		2123	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		2124
		2125	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		2126
		2127	Configures the watcher to embed the given loop, which must be
		2128	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		2129	invoked automatically, otherwise it is the responsibility of the callback
		2130	to invoke it (it will continue to be called until the sweep has been done,
		2131	if you do not want thta, you need to temporarily stop the embed watcher).
		2132
		2133	=item ev_embed_sweep (loop, ev_embed *)
		2134
		2135	Make a single, non-blocking sweep over the embedded loop. This works
		2136	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		2137	apropriate way for embedded loops.
		2138
		2139	=item struct ev_loop *other [read-only]
		2140
		2141	The embedded event loop.
		2142
		2143	=back
		2144
		2145	=head3 Examples
		2146
		2147	Example: Try to get an embeddable event loop and embed it into the default
		2148	event loop. If that is not possible, use the default loop. The default
		2149	loop is stored in C<loop_hi>, while the mebeddable loop is stored in
		2150	C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be
		2151	used).
1438		2152
1439	struct ev_loop *loop_hi = ev_default_init (0);	2153	struct ev_loop *loop_hi = ev_default_init (0);
1440	struct ev_loop *loop_lo = 0;	2154	struct ev_loop *loop_lo = 0;
1441	struct ev_embed embed;	2155	struct ev_embed embed;
1442		2156
…		…
1453	ev_embed_start (loop_hi, &embed);	2167	ev_embed_start (loop_hi, &embed);
1454	}	2168	}
1455	else	2169	else
1456	loop_lo = loop_hi;	2170	loop_lo = loop_hi;
1457		2171
1458	=over 4	2172	Example: Check if kqueue is available but not recommended and create
		2173	a kqueue backend for use with sockets (which usually work with any
		2174	kqueue implementation). Store the kqueue/socket-only event loop in
		2175	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
1459		2176
1460	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	2177	struct ev_loop *loop = ev_default_init (0);
		2178	struct ev_loop *loop_socket = 0;
		2179	struct ev_embed embed;
		2180
		2181	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		2182	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		2183	{
		2184	ev_embed_init (&embed, 0, loop_socket);
		2185	ev_embed_start (loop, &embed);
		2186	}
1461		2187
1462	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	2188	if (!loop_socket)
		2189	loop_socket = loop;
1463		2190
1464	Configures the watcher to embed the given loop, which must be	2191	// now use loop_socket for all sockets, and loop for everything else
1465	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1466	invoked automatically, otherwise it is the responsibility of the callback
1467	to invoke it (it will continue to be called until the sweep has been done,
1468	if you do not want thta, you need to temporarily stop the embed watcher).
1469
1470	=item ev_embed_sweep (loop, ev_embed *)
1471
1472	Make a single, non-blocking sweep over the embedded loop. This works
1473	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1474	apropriate way for embedded loops.
1475
1476	=item struct ev_loop *loop [read-only]
1477
1478	The embedded event loop.
1479
1480	=back
1481		2192
1482		2193
1483	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	2194	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
1484		2195
1485	Fork watchers are called when a C<fork ()> was detected (usually because	2196	Fork watchers are called when a C<fork ()> was detected (usually because
…		…
1488	event loop blocks next and before C<ev_check> watchers are being called,	2199	event loop blocks next and before C<ev_check> watchers are being called,
1489	and only in the child after the fork. If whoever good citizen calling	2200	and only in the child after the fork. If whoever good citizen calling
1490	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2201	C<ev_default_fork> cheats and calls it in the wrong process, the fork
1491	handlers will be invoked, too, of course.	2202	handlers will be invoked, too, of course.
1492		2203
		2204	=head3 Watcher-Specific Functions and Data Members
		2205
1493	=over 4	2206	=over 4
1494		2207
1495	=item ev_fork_init (ev_signal *, callback)	2208	=item ev_fork_init (ev_signal *, callback)
1496		2209
1497	Initialises and configures the fork watcher - it has no parameters of any	2210	Initialises and configures the fork watcher - it has no parameters of any
1498	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	2211	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
1499	believe me.	2212	believe me.
		2213
		2214	=back
		2215
		2216
		2217	=head2 C<ev_async> - how to wake up another event loop
		2218
		2219	In general, you cannot use an C<ev_loop> from multiple threads or other
		2220	asynchronous sources such as signal handlers (as opposed to multiple event
		2221	loops - those are of course safe to use in different threads).
		2222
		2223	Sometimes, however, you need to wake up another event loop you do not
		2224	control, for example because it belongs to another thread. This is what
		2225	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you
		2226	can signal it by calling C<ev_async_send>, which is thread- and signal
		2227	safe.
		2228
		2229	This functionality is very similar to C<ev_signal> watchers, as signals,
		2230	too, are asynchronous in nature, and signals, too, will be compressed
		2231	(i.e. the number of callback invocations may be less than the number of
		2232	C<ev_async_sent> calls).
		2233
		2234	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		2235	just the default loop.
		2236
		2237	=head3 Queueing
		2238
		2239	C<ev_async> does not support queueing of data in any way. The reason
		2240	is that the author does not know of a simple (or any) algorithm for a
		2241	multiple-writer-single-reader queue that works in all cases and doesn't
		2242	need elaborate support such as pthreads.
		2243
		2244	That means that if you want to queue data, you have to provide your own
		2245	queue. But at least I can tell you would implement locking around your
		2246	queue:
		2247
		2248	=over 4
		2249
		2250	=item queueing from a signal handler context
		2251
		2252	To implement race-free queueing, you simply add to the queue in the signal
		2253	handler but you block the signal handler in the watcher callback. Here is an example that does that for
		2254	some fictitiuous SIGUSR1 handler:
		2255
		2256	static ev_async mysig;
		2257
		2258	static void
		2259	sigusr1_handler (void)
		2260	{
		2261	sometype data;
		2262
		2263	// no locking etc.
		2264	queue_put (data);
		2265	ev_async_send (EV_DEFAULT_ &mysig);
		2266	}
		2267
		2268	static void
		2269	mysig_cb (EV_P_ ev_async *w, int revents)
		2270	{
		2271	sometype data;
		2272	sigset_t block, prev;
		2273
		2274	sigemptyset (&block);
		2275	sigaddset (&block, SIGUSR1);
		2276	sigprocmask (SIG_BLOCK, &block, &prev);
		2277
		2278	while (queue_get (&data))
		2279	process (data);
		2280
		2281	if (sigismember (&prev, SIGUSR1)
		2282	sigprocmask (SIG_UNBLOCK, &block, 0);
		2283	}
		2284
		2285	(Note: pthreads in theory requires you to use C<pthread_setmask>
		2286	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		2287	either...).
		2288
		2289	=item queueing from a thread context
		2290
		2291	The strategy for threads is different, as you cannot (easily) block
		2292	threads but you can easily preempt them, so to queue safely you need to
		2293	employ a traditional mutex lock, such as in this pthread example:
		2294
		2295	static ev_async mysig;
		2296	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		2297
		2298	static void
		2299	otherthread (void)
		2300	{
		2301	// only need to lock the actual queueing operation
		2302	pthread_mutex_lock (&mymutex);
		2303	queue_put (data);
		2304	pthread_mutex_unlock (&mymutex);
		2305
		2306	ev_async_send (EV_DEFAULT_ &mysig);
		2307	}
		2308
		2309	static void
		2310	mysig_cb (EV_P_ ev_async *w, int revents)
		2311	{
		2312	pthread_mutex_lock (&mymutex);
		2313
		2314	while (queue_get (&data))
		2315	process (data);
		2316
		2317	pthread_mutex_unlock (&mymutex);
		2318	}
		2319
		2320	=back
		2321
		2322
		2323	=head3 Watcher-Specific Functions and Data Members
		2324
		2325	=over 4
		2326
		2327	=item ev_async_init (ev_async *, callback)
		2328
		2329	Initialises and configures the async watcher - it has no parameters of any
		2330	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,
		2331	believe me.
		2332
		2333	=item ev_async_send (loop, ev_async *)
		2334
		2335	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		2336	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
		2337	C<ev_feed_event>, this call is safe to do in other threads, signal or
		2338	similar contexts (see the dicusssion of C<EV_ATOMIC_T> in the embedding
		2339	section below on what exactly this means).
		2340
		2341	This call incurs the overhead of a syscall only once per loop iteration,
		2342	so while the overhead might be noticable, it doesn't apply to repeated
		2343	calls to C<ev_async_send>.
		2344
		2345	=item bool = ev_async_pending (ev_async *)
		2346
		2347	Returns a non-zero value when C<ev_async_send> has been called on the
		2348	watcher but the event has not yet been processed (or even noted) by the
		2349	event loop.
		2350
		2351	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2352	the loop iterates next and checks for the watcher to have become active,
		2353	it will reset the flag again. C<ev_async_pending> can be used to very
		2354	quickly check wether invoking the loop might be a good idea.
		2355
		2356	Not that this does I<not> check wether the watcher itself is pending, only
		2357	wether it has been requested to make this watcher pending.
1500		2358
1501	=back	2359	=back
1502		2360
1503		2361
1504	=head1 OTHER FUNCTIONS	2362	=head1 OTHER FUNCTIONS
…		…
1576		2434
1577	=item * Priorities are not currently supported. Initialising priorities	2435	=item * Priorities are not currently supported. Initialising priorities
1578	will fail and all watchers will have the same priority, even though there	2436	will fail and all watchers will have the same priority, even though there
1579	is an ev_pri field.	2437	is an ev_pri field.
1580		2438
		2439	=item * In libevent, the last base created gets the signals, in libev, the
		2440	first base created (== the default loop) gets the signals.
		2441
1581	=item * Other members are not supported.	2442	=item * Other members are not supported.
1582		2443
1583	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2444	=item * The libev emulation is I<not> ABI compatible to libevent, you need
1584	to use the libev header file and library.	2445	to use the libev header file and library.
1585		2446
…		…
1593		2454
1594	To use it,	2455	To use it,
1595		2456
1596	#include <ev++.h>	2457	#include <ev++.h>
1597		2458
1598	(it is not installed by default). This automatically includes F<ev.h>	2459	This automatically includes F<ev.h> and puts all of its definitions (many
1599	and puts all of its definitions (many of them macros) into the global	2460	of them macros) into the global namespace. All C++ specific things are
1600	namespace. All C++ specific things are put into the C<ev> namespace.	2461	put into the C<ev> namespace. It should support all the same embedding
		2462	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1601		2463
1602	It should support all the same embedding options as F<ev.h>, most notably	2464	Care has been taken to keep the overhead low. The only data member the C++
1603	C<EV_MULTIPLICITY>.	2465	classes add (compared to plain C-style watchers) is the event loop pointer
		2466	that the watcher is associated with (or no additional members at all if
		2467	you disable C<EV_MULTIPLICITY> when embedding libev).
		2468
		2469	Currently, functions, and static and non-static member functions can be
		2470	used as callbacks. Other types should be easy to add as long as they only
		2471	need one additional pointer for context. If you need support for other
		2472	types of functors please contact the author (preferably after implementing
		2473	it).
1604		2474
1605	Here is a list of things available in the C<ev> namespace:	2475	Here is a list of things available in the C<ev> namespace:
1606		2476
1607	=over 4	2477	=over 4
1608		2478
…		…
1624		2494
1625	All of those classes have these methods:	2495	All of those classes have these methods:
1626		2496
1627	=over 4	2497	=over 4
1628		2498
1629	=item ev::TYPE::TYPE (object *, object::method *)	2499	=item ev::TYPE::TYPE ()
1630		2500
1631	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)	2501	=item ev::TYPE::TYPE (struct ev_loop *)
1632		2502
1633	=item ev::TYPE::~TYPE	2503	=item ev::TYPE::~TYPE
1634		2504
1635	The constructor takes a pointer to an object and a method pointer to	2505	The constructor (optionally) takes an event loop to associate the watcher
1636	the event handler callback to call in this class. The constructor calls	2506	with. If it is omitted, it will use C<EV_DEFAULT>.
1637	C<ev_init> for you, which means you have to call the C<set> method	2507
1638	before starting it. If you do not specify a loop then the constructor	2508	The constructor calls C<ev_init> for you, which means you have to call the
1639	automatically associates the default loop with this watcher.	2509	C<set> method before starting it.
		2510
		2511	It will not set a callback, however: You have to call the templated C<set>
		2512	method to set a callback before you can start the watcher.
		2513
		2514	(The reason why you have to use a method is a limitation in C++ which does
		2515	not allow explicit template arguments for constructors).
1640		2516
1641	The destructor automatically stops the watcher if it is active.	2517	The destructor automatically stops the watcher if it is active.
		2518
		2519	=item w->set<class, &class::method> (object *)
		2520
		2521	This method sets the callback method to call. The method has to have a
		2522	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		2523	first argument and the C<revents> as second. The object must be given as
		2524	parameter and is stored in the C<data> member of the watcher.
		2525
		2526	This method synthesizes efficient thunking code to call your method from
		2527	the C callback that libev requires. If your compiler can inline your
		2528	callback (i.e. it is visible to it at the place of the C<set> call and
		2529	your compiler is good :), then the method will be fully inlined into the
		2530	thunking function, making it as fast as a direct C callback.
		2531
		2532	Example: simple class declaration and watcher initialisation
		2533
		2534	struct myclass
		2535	{
		2536	void io_cb (ev::io &w, int revents) { }
		2537	}
		2538
		2539	myclass obj;
		2540	ev::io iow;
		2541	iow.set <myclass, &myclass::io_cb> (&obj);
		2542
		2543	=item w->set<function> (void *data = 0)
		2544
		2545	Also sets a callback, but uses a static method or plain function as
		2546	callback. The optional C<data> argument will be stored in the watcher's
		2547	C<data> member and is free for you to use.
		2548
		2549	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		2550
		2551	See the method-C<set> above for more details.
		2552
		2553	Example:
		2554
		2555	static void io_cb (ev::io &w, int revents) { }
		2556	iow.set <io_cb> ();
1642		2557
1643	=item w->set (struct ev_loop *)	2558	=item w->set (struct ev_loop *)
1644		2559
1645	Associates a different C<struct ev_loop> with this watcher. You can only	2560	Associates a different C<struct ev_loop> with this watcher. You can only
1646	do this when the watcher is inactive (and not pending either).	2561	do this when the watcher is inactive (and not pending either).
1647		2562
1648	=item w->set ([args])	2563	=item w->set ([args])
1649		2564
1650	Basically the same as C<ev_TYPE_set>, with the same args. Must be	2565	Basically the same as C<ev_TYPE_set>, with the same args. Must be
1651	called at least once. Unlike the C counterpart, an active watcher gets	2566	called at least once. Unlike the C counterpart, an active watcher gets
1652	automatically stopped and restarted.	2567	automatically stopped and restarted when reconfiguring it with this
		2568	method.
1653		2569
1654	=item w->start ()	2570	=item w->start ()
1655		2571
1656	Starts the watcher. Note that there is no C<loop> argument as the	2572	Starts the watcher. Note that there is no C<loop> argument, as the
1657	constructor already takes the loop.	2573	constructor already stores the event loop.
1658		2574
1659	=item w->stop ()	2575	=item w->stop ()
1660		2576
1661	Stops the watcher if it is active. Again, no C<loop> argument.	2577	Stops the watcher if it is active. Again, no C<loop> argument.
1662		2578
1663	=item w->again () C<ev::timer>, C<ev::periodic> only	2579	=item w->again () (C<ev::timer>, C<ev::periodic> only)
1664		2580
1665	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding	2581	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
1666	C<ev_TYPE_again> function.	2582	C<ev_TYPE_again> function.
1667		2583
1668	=item w->sweep () C<ev::embed> only	2584	=item w->sweep () (C<ev::embed> only)
1669		2585
1670	Invokes C<ev_embed_sweep>.	2586	Invokes C<ev_embed_sweep>.
1671		2587
1672	=item w->update () C<ev::stat> only	2588	=item w->update () (C<ev::stat> only)
1673		2589
1674	Invokes C<ev_stat_stat>.	2590	Invokes C<ev_stat_stat>.
1675		2591
1676	=back	2592	=back
1677		2593
…		…
1680	Example: Define a class with an IO and idle watcher, start one of them in	2596	Example: Define a class with an IO and idle watcher, start one of them in
1681	the constructor.	2597	the constructor.
1682		2598
1683	class myclass	2599	class myclass
1684	{	2600	{
1685	ev_io io; void io_cb (ev::io &w, int revents);	2601	ev::io io; void io_cb (ev::io &w, int revents);
1686	ev_idle idle void idle_cb (ev::idle &w, int revents);	2602	ev:idle idle void idle_cb (ev::idle &w, int revents);
1687		2603
1688	myclass ();	2604	myclass (int fd)
1689	}
1690
1691	myclass::myclass (int fd)
1692	: io (this, &myclass::io_cb),
1693	idle (this, &myclass::idle_cb)
1694	{	2605	{
		2606	io .set <myclass, &myclass::io_cb > (this);
		2607	idle.set <myclass, &myclass::idle_cb> (this);
		2608
1695	io.start (fd, ev::READ);	2609	io.start (fd, ev::READ);
		2610	}
1696	}	2611	};
		2612
		2613
		2614	=head1 OTHER LANGUAGE BINDINGS
		2615
		2616	Libev does not offer other language bindings itself, but bindings for a
		2617	numbe rof languages exist in the form of third-party packages. If you know
		2618	any interesting language binding in addition to the ones listed here, drop
		2619	me a note.
		2620
		2621	=over 4
		2622
		2623	=item Perl
		2624
		2625	The EV module implements the full libev API and is actually used to test
		2626	libev. EV is developed together with libev. Apart from the EV core module,
		2627	there are additional modules that implement libev-compatible interfaces
		2628	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the
		2629	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).
		2630
		2631	It can be found and installed via CPAN, its homepage is found at
		2632	L<http://software.schmorp.de/pkg/EV>.
		2633
		2634	=item Ruby
		2635
		2636	Tony Arcieri has written a ruby extension that offers access to a subset
		2637	of the libev API and adds filehandle abstractions, asynchronous DNS and
		2638	more on top of it. It can be found via gem servers. Its homepage is at
		2639	L<http://rev.rubyforge.org/>.
		2640
		2641	=item D
		2642
		2643	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		2644	be found at L<http://git.llucax.com.ar/?p=software/ev.d.git;a=summary>.
		2645
		2646	=back
1697		2647
1698		2648
1699	=head1 MACRO MAGIC	2649	=head1 MACRO MAGIC
1700		2650
1701	Libev can be compiled with a variety of options, the most fundemantal is	2651	Libev can be compiled with a variety of options, the most fundamantal
1702	C<EV_MULTIPLICITY>. This option determines wether (most) functions and	2652	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
1703	callbacks have an initial C<struct ev_loop *> argument.	2653	functions and callbacks have an initial C<struct ev_loop *> argument.
1704		2654
1705	To make it easier to write programs that cope with either variant, the	2655	To make it easier to write programs that cope with either variant, the
1706	following macros are defined:	2656	following macros are defined:
1707		2657
1708	=over 4	2658	=over 4
…		…
1738	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	2688	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
1739		2689
1740	Similar to the other two macros, this gives you the value of the default	2690	Similar to the other two macros, this gives you the value of the default
1741	loop, if multiple loops are supported ("ev loop default").	2691	loop, if multiple loops are supported ("ev loop default").
1742		2692
		2693	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		2694
		2695	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		2696	default loop has been initialised (C<UC> == unchecked). Their behaviour
		2697	is undefined when the default loop has not been initialised by a previous
		2698	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		2699
		2700	It is often prudent to use C<EV_DEFAULT> when initialising the first
		2701	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		2702
1743	=back	2703	=back
1744		2704
1745	Example: Declare and initialise a check watcher, working regardless of	2705	Example: Declare and initialise a check watcher, utilising the above
1746	wether multiple loops are supported or not.	2706	macros so it will work regardless of whether multiple loops are supported
		2707	or not.
1747		2708
1748	static void	2709	static void
1749	check_cb (EV_P_ ev_timer *w, int revents)	2710	check_cb (EV_P_ ev_timer *w, int revents)
1750	{	2711	{
1751	ev_check_stop (EV_A_ w);	2712	ev_check_stop (EV_A_ w);
…		…
1754	ev_check check;	2715	ev_check check;
1755	ev_check_init (&check, check_cb);	2716	ev_check_init (&check, check_cb);
1756	ev_check_start (EV_DEFAULT_ &check);	2717	ev_check_start (EV_DEFAULT_ &check);
1757	ev_loop (EV_DEFAULT_ 0);	2718	ev_loop (EV_DEFAULT_ 0);
1758		2719
1759
1760	=head1 EMBEDDING	2720	=head1 EMBEDDING
1761		2721
1762	Libev can (and often is) directly embedded into host	2722	Libev can (and often is) directly embedded into host
1763	applications. Examples of applications that embed it include the Deliantra	2723	applications. Examples of applications that embed it include the Deliantra
1764	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)	2724	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
1765	and rxvt-unicode.	2725	and rxvt-unicode.
1766		2726
1767	The goal is to enable you to just copy the neecssary files into your	2727	The goal is to enable you to just copy the necessary files into your
1768	source directory without having to change even a single line in them, so	2728	source directory without having to change even a single line in them, so
1769	you can easily upgrade by simply copying (or having a checked-out copy of	2729	you can easily upgrade by simply copying (or having a checked-out copy of
1770	libev somewhere in your source tree).	2730	libev somewhere in your source tree).
1771		2731
1772	=head2 FILESETS	2732	=head2 FILESETS
…		…
1803	ev_vars.h	2763	ev_vars.h
1804	ev_wrap.h	2764	ev_wrap.h
1805		2765
1806	ev_win32.c required on win32 platforms only	2766	ev_win32.c required on win32 platforms only
1807		2767
1808	ev_select.c only when select backend is enabled (which is by default)	2768	ev_select.c only when select backend is enabled (which is enabled by default)
1809	ev_poll.c only when poll backend is enabled (disabled by default)	2769	ev_poll.c only when poll backend is enabled (disabled by default)
1810	ev_epoll.c only when the epoll backend is enabled (disabled by default)	2770	ev_epoll.c only when the epoll backend is enabled (disabled by default)
1811	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	2771	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1812	ev_port.c only when the solaris port backend is enabled (disabled by default)	2772	ev_port.c only when the solaris port backend is enabled (disabled by default)
1813		2773
…		…
1842		2802
1843	libev.m4	2803	libev.m4
1844		2804
1845	=head2 PREPROCESSOR SYMBOLS/MACROS	2805	=head2 PREPROCESSOR SYMBOLS/MACROS
1846		2806
1847	Libev can be configured via a variety of preprocessor symbols you have to define	2807	Libev can be configured via a variety of preprocessor symbols you have to
1848	before including any of its files. The default is not to build for multiplicity	2808	define before including any of its files. The default in the absense of
1849	and only include the select backend.	2809	autoconf is noted for every option.
1850		2810
1851	=over 4	2811	=over 4
1852		2812
1853	=item EV_STANDALONE	2813	=item EV_STANDALONE
1854		2814
…		…
1862		2822
1863	If defined to be C<1>, libev will try to detect the availability of the	2823	If defined to be C<1>, libev will try to detect the availability of the
1864	monotonic clock option at both compiletime and runtime. Otherwise no use	2824	monotonic clock option at both compiletime and runtime. Otherwise no use
1865	of the monotonic clock option will be attempted. If you enable this, you	2825	of the monotonic clock option will be attempted. If you enable this, you
1866	usually have to link against librt or something similar. Enabling it when	2826	usually have to link against librt or something similar. Enabling it when
1867	the functionality isn't available is safe, though, althoguh you have	2827	the functionality isn't available is safe, though, although you have
1868	to make sure you link against any libraries where the C<clock_gettime>	2828	to make sure you link against any libraries where the C<clock_gettime>
1869	function is hiding in (often F<-lrt>).	2829	function is hiding in (often F<-lrt>).
1870		2830
1871	=item EV_USE_REALTIME	2831	=item EV_USE_REALTIME
1872		2832
1873	If defined to be C<1>, libev will try to detect the availability of the	2833	If defined to be C<1>, libev will try to detect the availability of the
1874	realtime clock option at compiletime (and assume its availability at	2834	realtime clock option at compiletime (and assume its availability at
1875	runtime if successful). Otherwise no use of the realtime clock option will	2835	runtime if successful). Otherwise no use of the realtime clock option will
1876	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	2836	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
1877	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries	2837	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
1878	in the description of C<EV_USE_MONOTONIC>, though.	2838	note about libraries in the description of C<EV_USE_MONOTONIC>, though.
		2839
		2840	=item EV_USE_NANOSLEEP
		2841
		2842	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		2843	and will use it for delays. Otherwise it will use C<select ()>.
		2844
		2845	=item EV_USE_EVENTFD
		2846
		2847	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		2848	available and will probe for kernel support at runtime. This will improve
		2849	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		2850	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		2851	2.7 or newer, otherwise disabled.
1879		2852
1880	=item EV_USE_SELECT	2853	=item EV_USE_SELECT
1881		2854
1882	If undefined or defined to be C<1>, libev will compile in support for the	2855	If undefined or defined to be C<1>, libev will compile in support for the
1883	C<select>(2) backend. No attempt at autodetection will be done: if no	2856	C<select>(2) backend. No attempt at autodetection will be done: if no
…		…
1902	be used is the winsock select). This means that it will call	2875	be used is the winsock select). This means that it will call
1903	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	2876	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
1904	it is assumed that all these functions actually work on fds, even	2877	it is assumed that all these functions actually work on fds, even
1905	on win32. Should not be defined on non-win32 platforms.	2878	on win32. Should not be defined on non-win32 platforms.
1906		2879
		2880	=item EV_FD_TO_WIN32_HANDLE
		2881
		2882	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		2883	file descriptors to socket handles. When not defining this symbol (the
		2884	default), then libev will call C<_get_osfhandle>, which is usually
		2885	correct. In some cases, programs use their own file descriptor management,
		2886	in which case they can provide this function to map fds to socket handles.
		2887
1907	=item EV_USE_POLL	2888	=item EV_USE_POLL
1908		2889
1909	If defined to be C<1>, libev will compile in support for the C<poll>(2)	2890	If defined to be C<1>, libev will compile in support for the C<poll>(2)
1910	backend. Otherwise it will be enabled on non-win32 platforms. It	2891	backend. Otherwise it will be enabled on non-win32 platforms. It
1911	takes precedence over select.	2892	takes precedence over select.
1912		2893
1913	=item EV_USE_EPOLL	2894	=item EV_USE_EPOLL
1914		2895
1915	If defined to be C<1>, libev will compile in support for the Linux	2896	If defined to be C<1>, libev will compile in support for the Linux
1916	C<epoll>(7) backend. Its availability will be detected at runtime,	2897	C<epoll>(7) backend. Its availability will be detected at runtime,
1917	otherwise another method will be used as fallback. This is the	2898	otherwise another method will be used as fallback. This is the preferred
1918	preferred backend for GNU/Linux systems.	2899	backend for GNU/Linux systems. If undefined, it will be enabled if the
		2900	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
1919		2901
1920	=item EV_USE_KQUEUE	2902	=item EV_USE_KQUEUE
1921		2903
1922	If defined to be C<1>, libev will compile in support for the BSD style	2904	If defined to be C<1>, libev will compile in support for the BSD style
1923	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	2905	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
1938		2920
1939	=item EV_USE_DEVPOLL	2921	=item EV_USE_DEVPOLL
1940		2922
1941	reserved for future expansion, works like the USE symbols above.	2923	reserved for future expansion, works like the USE symbols above.
1942		2924
		2925	=item EV_USE_INOTIFY
		2926
		2927	If defined to be C<1>, libev will compile in support for the Linux inotify
		2928	interface to speed up C<ev_stat> watchers. Its actual availability will
		2929	be detected at runtime. If undefined, it will be enabled if the headers
		2930	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		2931
		2932	=item EV_ATOMIC_T
		2933
		2934	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		2935	access is atomic with respect to other threads or signal contexts. No such
		2936	type is easily found in the C language, so you can provide your own type
		2937	that you know is safe for your purposes. It is used both for signal handler "locking"
		2938	as well as for signal and thread safety in C<ev_async> watchers.
		2939
		2940	In the absense of this define, libev will use C<sig_atomic_t volatile>
		2941	(from F<signal.h>), which is usually good enough on most platforms.
		2942
1943	=item EV_H	2943	=item EV_H
1944		2944
1945	The name of the F<ev.h> header file used to include it. The default if	2945	The name of the F<ev.h> header file used to include it. The default if
1946	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This	2946	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
1947	can be used to virtually rename the F<ev.h> header file in case of conflicts.	2947	used to virtually rename the F<ev.h> header file in case of conflicts.
1948		2948
1949	=item EV_CONFIG_H	2949	=item EV_CONFIG_H
1950		2950
1951	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	2951	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
1952	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	2952	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
1953	C<EV_H>, above.	2953	C<EV_H>, above.
1954		2954
1955	=item EV_EVENT_H	2955	=item EV_EVENT_H
1956		2956
1957	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	2957	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
1958	of how the F<event.h> header can be found.	2958	of how the F<event.h> header can be found, the default is C<"event.h">.
1959		2959
1960	=item EV_PROTOTYPES	2960	=item EV_PROTOTYPES
1961		2961
1962	If defined to be C<0>, then F<ev.h> will not define any function	2962	If defined to be C<0>, then F<ev.h> will not define any function
1963	prototypes, but still define all the structs and other symbols. This is	2963	prototypes, but still define all the structs and other symbols. This is
…		…
1970	will have the C<struct ev_loop *> as first argument, and you can create	2970	will have the C<struct ev_loop *> as first argument, and you can create
1971	additional independent event loops. Otherwise there will be no support	2971	additional independent event loops. Otherwise there will be no support
1972	for multiple event loops and there is no first event loop pointer	2972	for multiple event loops and there is no first event loop pointer
1973	argument. Instead, all functions act on the single default loop.	2973	argument. Instead, all functions act on the single default loop.
1974		2974
		2975	=item EV_MINPRI
		2976
		2977	=item EV_MAXPRI
		2978
		2979	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2980	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2981	provide for more priorities by overriding those symbols (usually defined
		2982	to be C<-2> and C<2>, respectively).
		2983
		2984	When doing priority-based operations, libev usually has to linearly search
		2985	all the priorities, so having many of them (hundreds) uses a lot of space
		2986	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2987	fine.
		2988
		2989	If your embedding app does not need any priorities, defining these both to
		2990	C<0> will save some memory and cpu.
		2991
1975	=item EV_PERIODIC_ENABLE	2992	=item EV_PERIODIC_ENABLE
1976		2993
1977	If undefined or defined to be C<1>, then periodic timers are supported. If	2994	If undefined or defined to be C<1>, then periodic timers are supported. If
1978	defined to be C<0>, then they are not. Disabling them saves a few kB of	2995	defined to be C<0>, then they are not. Disabling them saves a few kB of
1979	code.	2996	code.
1980		2997
		2998	=item EV_IDLE_ENABLE
		2999
		3000	If undefined or defined to be C<1>, then idle watchers are supported. If
		3001	defined to be C<0>, then they are not. Disabling them saves a few kB of
		3002	code.
		3003
1981	=item EV_EMBED_ENABLE	3004	=item EV_EMBED_ENABLE
1982		3005
1983	If undefined or defined to be C<1>, then embed watchers are supported. If	3006	If undefined or defined to be C<1>, then embed watchers are supported. If
1984	defined to be C<0>, then they are not.	3007	defined to be C<0>, then they are not.
1985		3008
…		…
1991	=item EV_FORK_ENABLE	3014	=item EV_FORK_ENABLE
1992		3015
1993	If undefined or defined to be C<1>, then fork watchers are supported. If	3016	If undefined or defined to be C<1>, then fork watchers are supported. If
1994	defined to be C<0>, then they are not.	3017	defined to be C<0>, then they are not.
1995		3018
		3019	=item EV_ASYNC_ENABLE
		3020
		3021	If undefined or defined to be C<1>, then async watchers are supported. If
		3022	defined to be C<0>, then they are not.
		3023
1996	=item EV_MINIMAL	3024	=item EV_MINIMAL
1997		3025
1998	If you need to shave off some kilobytes of code at the expense of some	3026	If you need to shave off some kilobytes of code at the expense of some
1999	speed, define this symbol to C<1>. Currently only used for gcc to override	3027	speed, define this symbol to C<1>. Currently this is used to override some
2000	some inlining decisions, saves roughly 30% codesize of amd64.	3028	inlining decisions, saves roughly 30% codesize of amd64. It also selects a
		3029	much smaller 2-heap for timer management over the default 4-heap.
2001		3030
2002	=item EV_PID_HASHSIZE	3031	=item EV_PID_HASHSIZE
2003		3032
2004	C<ev_child> watchers use a small hash table to distribute workload by	3033	C<ev_child> watchers use a small hash table to distribute workload by
2005	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3034	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
2006	than enough. If you need to manage thousands of children you might want to	3035	than enough. If you need to manage thousands of children you might want to
2007	increase this value.	3036	increase this value (I<must> be a power of two).
		3037
		3038	=item EV_INOTIFY_HASHSIZE
		3039
		3040	C<ev_stat> watchers use a small hash table to distribute workload by
		3041	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
		3042	usually more than enough. If you need to manage thousands of C<ev_stat>
		3043	watchers you might want to increase this value (I<must> be a power of
		3044	two).
		3045
		3046	=item EV_USE_4HEAP
		3047
		3048	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3049	timer and periodics heap, libev uses a 4-heap when this symbol is defined
		3050	to C<1>. The 4-heap uses more complicated (longer) code but has
		3051	noticably faster performance with many (thousands) of watchers.
		3052
		3053	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3054	(disabled).
		3055
		3056	=item EV_HEAP_CACHE_AT
		3057
		3058	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3059	timer and periodics heap, libev can cache the timestamp (I<at>) within
		3060	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3061	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3062	but avoids random read accesses on heap changes. This improves performance
		3063	noticably with with many (hundreds) of watchers.
		3064
		3065	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3066	(disabled).
		3067
		3068	=item EV_VERIFY
		3069
		3070	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3071	be done: If set to C<0>, no internal verification code will be compiled
		3072	in. If set to C<1>, then verification code will be compiled in, but not
		3073	called. If set to C<2>, then the internal verification code will be
		3074	called once per loop, which can slow down libev. If set to C<3>, then the
		3075	verification code will be called very frequently, which will slow down
		3076	libev considerably.
		3077
		3078	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3079	C<0.>
2008		3080
2009	=item EV_COMMON	3081	=item EV_COMMON
2010		3082
2011	By default, all watchers have a C<void *data> member. By redefining	3083	By default, all watchers have a C<void *data> member. By redefining
2012	this macro to a something else you can include more and other types of	3084	this macro to a something else you can include more and other types of
…		…
2025		3097
2026	=item ev_set_cb (ev, cb)	3098	=item ev_set_cb (ev, cb)
2027		3099
2028	Can be used to change the callback member declaration in each watcher,	3100	Can be used to change the callback member declaration in each watcher,
2029	and the way callbacks are invoked and set. Must expand to a struct member	3101	and the way callbacks are invoked and set. Must expand to a struct member
2030	definition and a statement, respectively. See the F<ev.v> header file for	3102	definition and a statement, respectively. See the F<ev.h> header file for
2031	their default definitions. One possible use for overriding these is to	3103	their default definitions. One possible use for overriding these is to
2032	avoid the C<struct ev_loop *> as first argument in all cases, or to use	3104	avoid the C<struct ev_loop *> as first argument in all cases, or to use
2033	method calls instead of plain function calls in C++.	3105	method calls instead of plain function calls in C++.
		3106
		3107	=head2 EXPORTED API SYMBOLS
		3108
		3109	If you need to re-export the API (e.g. via a dll) and you need a list of
		3110	exported symbols, you can use the provided F<Symbol.*> files which list
		3111	all public symbols, one per line:
		3112
		3113	Symbols.ev for libev proper
		3114	Symbols.event for the libevent emulation
		3115
		3116	This can also be used to rename all public symbols to avoid clashes with
		3117	multiple versions of libev linked together (which is obviously bad in
		3118	itself, but sometimes it is inconvinient to avoid this).
		3119
		3120	A sed command like this will create wrapper C<#define>'s that you need to
		3121	include before including F<ev.h>:
		3122
		3123	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		3124
		3125	This would create a file F<wrap.h> which essentially looks like this:
		3126
		3127	#define ev_backend myprefix_ev_backend
		3128	#define ev_check_start myprefix_ev_check_start
		3129	#define ev_check_stop myprefix_ev_check_stop
		3130	...
2034		3131
2035	=head2 EXAMPLES	3132	=head2 EXAMPLES
2036		3133
2037	For a real-world example of a program the includes libev	3134	For a real-world example of a program the includes libev
2038	verbatim, you can have a look at the EV perl module	3135	verbatim, you can have a look at the EV perl module
…		…
2041	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file	3138	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
2042	will be compiled. It is pretty complex because it provides its own header	3139	will be compiled. It is pretty complex because it provides its own header
2043	file.	3140	file.
2044		3141
2045	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	3142	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
2046	that everybody includes and which overrides some autoconf choices:	3143	that everybody includes and which overrides some configure choices:
2047		3144
		3145	#define EV_MINIMAL 1
2048	#define EV_USE_POLL 0	3146	#define EV_USE_POLL 0
2049	#define EV_MULTIPLICITY 0	3147	#define EV_MULTIPLICITY 0
2050	#define EV_PERIODICS 0	3148	#define EV_PERIODIC_ENABLE 0
		3149	#define EV_STAT_ENABLE 0
		3150	#define EV_FORK_ENABLE 0
2051	#define EV_CONFIG_H <config.h>	3151	#define EV_CONFIG_H <config.h>
		3152	#define EV_MINPRI 0
		3153	#define EV_MAXPRI 0
2052		3154
2053	#include "ev++.h"	3155	#include "ev++.h"
2054		3156
2055	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	3157	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
2056		3158
2057	#include "ev_cpp.h"	3159	#include "ev_cpp.h"
2058	#include "ev.c"	3160	#include "ev.c"
		3161
		3162
		3163	=head1 THREADS AND COROUTINES
		3164
		3165	=head2 THREADS
		3166
		3167	Libev itself is completely threadsafe, but it uses no locking. This
		3168	means that you can use as many loops as you want in parallel, as long as
		3169	only one thread ever calls into one libev function with the same loop
		3170	parameter.
		3171
		3172	Or put differently: calls with different loop parameters can be done in
		3173	parallel from multiple threads, calls with the same loop parameter must be
		3174	done serially (but can be done from different threads, as long as only one
		3175	thread ever is inside a call at any point in time, e.g. by using a mutex
		3176	per loop).
		3177
		3178	If you want to know which design is best for your problem, then I cannot
		3179	help you but by giving some generic advice:
		3180
		3181	=over 4
		3182
		3183	=item * most applications have a main thread: use the default libev loop
		3184	in that thread, or create a seperate thread running only the default loop.
		3185
		3186	This helps integrating other libraries or software modules that use libev
		3187	themselves and don't care/know about threading.
		3188
		3189	=item * one loop per thread is usually a good model.
		3190
		3191	Doing this is almost never wrong, sometimes a better-performance model
		3192	exists, but it is always a good start.
		3193
		3194	=item * other models exist, such as the leader/follower pattern, where one
		3195	loop is handed through multiple threads in a kind of round-robbin fashion.
		3196
		3197	Chosing a model is hard - look around, learn, know that usually you cna do
		3198	better than you currently do :-)
		3199
		3200	=item * often you need to talk to some other thread which blocks in the
		3201	event loop - C<ev_async> watchers can be used to wake them up from other
		3202	threads safely (or from signal contexts...).
		3203
		3204	=back
		3205
		3206	=head2 COROUTINES
		3207
		3208	Libev is much more accomodating to coroutines ("cooperative threads"):
		3209	libev fully supports nesting calls to it's functions from different
		3210	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3211	different coroutines and switch freely between both coroutines running the
		3212	loop, as long as you don't confuse yourself). The only exception is that
		3213	you must not do this from C<ev_periodic> reschedule callbacks.
		3214
		3215	Care has been invested into making sure that libev does not keep local
		3216	state inside C<ev_loop>, and other calls do not usually allow coroutine
		3217	switches.
2059		3218
2060		3219
2061	=head1 COMPLEXITIES	3220	=head1 COMPLEXITIES
2062		3221
2063	In this section the complexities of (many of) the algorithms used inside	3222	In this section the complexities of (many of) the algorithms used inside
2064	libev will be explained. For complexity discussions about backends see the	3223	libev will be explained. For complexity discussions about backends see the
2065	documentation for C<ev_default_init>.	3224	documentation for C<ev_default_init>.
2066		3225
		3226	All of the following are about amortised time: If an array needs to be
		3227	extended, libev needs to realloc and move the whole array, but this
		3228	happens asymptotically never with higher number of elements, so O(1) might
		3229	mean it might do a lengthy realloc operation in rare cases, but on average
		3230	it is much faster and asymptotically approaches constant time.
		3231
2067	=over 4	3232	=over 4
2068		3233
2069	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	3234	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2070		3235
		3236	This means that, when you have a watcher that triggers in one hour and
		3237	there are 100 watchers that would trigger before that then inserting will
		3238	have to skip roughly seven (C<ld 100>) of these watchers.
		3239
2071	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)	3240	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
2072		3241
		3242	That means that changing a timer costs less than removing/adding them
		3243	as only the relative motion in the event queue has to be paid for.
		3244
2073	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	3245	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
2074		3246
		3247	These just add the watcher into an array or at the head of a list.
		3248
2075	=item Stopping check/prepare/idle watchers: O(1)	3249	=item Stopping check/prepare/idle/fork/async watchers: O(1)
2076		3250
2077	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))	3251	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
2078		3252
		3253	These watchers are stored in lists then need to be walked to find the
		3254	correct watcher to remove. The lists are usually short (you don't usually
		3255	have many watchers waiting for the same fd or signal).
		3256
2079	=item Finding the next timer per loop iteration: O(1)	3257	=item Finding the next timer in each loop iteration: O(1)
		3258
		3259	By virtue of using a binary or 4-heap, the next timer is always found at a
		3260	fixed position in the storage array.
2080		3261
2081	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	3262	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2082		3263
2083	=item Activating one watcher: O(1)	3264	A change means an I/O watcher gets started or stopped, which requires
		3265	libev to recalculate its status (and possibly tell the kernel, depending
		3266	on backend and wether C<ev_io_set> was used).
		3267
		3268	=item Activating one watcher (putting it into the pending state): O(1)
		3269
		3270	=item Priority handling: O(number_of_priorities)
		3271
		3272	Priorities are implemented by allocating some space for each
		3273	priority. When doing priority-based operations, libev usually has to
		3274	linearly search all the priorities, but starting/stopping and activating
		3275	watchers becomes O(1) w.r.t. priority handling.
		3276
		3277	=item Sending an ev_async: O(1)
		3278
		3279	=item Processing ev_async_send: O(number_of_async_watchers)
		3280
		3281	=item Processing signals: O(max_signal_number)
		3282
		3283	Sending involves a syscall I<iff> there were no other C<ev_async_send>
		3284	calls in the current loop iteration. Checking for async and signal events
		3285	involves iterating over all running async watchers or all signal numbers.
2084		3286
2085	=back	3287	=back
2086		3288
2087		3289
		3290	=head1 Win32 platform limitations and workarounds
		3291
		3292	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		3293	requires, and its I/O model is fundamentally incompatible with the POSIX
		3294	model. Libev still offers limited functionality on this platform in
		3295	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		3296	descriptors. This only applies when using Win32 natively, not when using
		3297	e.g. cygwin.
		3298
		3299	Lifting these limitations would basically require the full
		3300	re-implementation of the I/O system. If you are into these kinds of
		3301	things, then note that glib does exactly that for you in a very portable
		3302	way (note also that glib is the slowest event library known to man).
		3303
		3304	There is no supported compilation method available on windows except
		3305	embedding it into other applications.
		3306
		3307	Due to the many, low, and arbitrary limits on the win32 platform and
		3308	the abysmal performance of winsockets, using a large number of sockets
		3309	is not recommended (and not reasonable). If your program needs to use
		3310	more than a hundred or so sockets, then likely it needs to use a totally
		3311	different implementation for windows, as libev offers the POSIX readiness
		3312	notification model, which cannot be implemented efficiently on windows
		3313	(microsoft monopoly games).
		3314
		3315	=over 4
		3316
		3317	=item The winsocket select function
		3318
		3319	The winsocket C<select> function doesn't follow POSIX in that it
		3320	requires socket I<handles> and not socket I<file descriptors> (it is
		3321	also extremely buggy). This makes select very inefficient, and also
		3322	requires a mapping from file descriptors to socket handles. See the
		3323	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		3324	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		3325
		3326	The configuration for a "naked" win32 using the microsoft runtime
		3327	libraries and raw winsocket select is:
		3328
		3329	#define EV_USE_SELECT 1
		3330	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		3331
		3332	Note that winsockets handling of fd sets is O(n), so you can easily get a
		3333	complexity in the O(n²) range when using win32.
		3334
		3335	=item Limited number of file descriptors
		3336
		3337	Windows has numerous arbitrary (and low) limits on things.
		3338
		3339	Early versions of winsocket's select only supported waiting for a maximum
		3340	of C<64> handles (probably owning to the fact that all windows kernels
		3341	can only wait for C<64> things at the same time internally; microsoft
		3342	recommends spawning a chain of threads and wait for 63 handles and the
		3343	previous thread in each. Great).
		3344
		3345	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		3346	to some high number (e.g. C<2048>) before compiling the winsocket select
		3347	call (which might be in libev or elsewhere, for example, perl does its own
		3348	select emulation on windows).
		3349
		3350	Another limit is the number of file descriptors in the microsoft runtime
		3351	libraries, which by default is C<64> (there must be a hidden I<64> fetish
		3352	or something like this inside microsoft). You can increase this by calling
		3353	C<_setmaxstdio>, which can increase this limit to C<2048> (another
		3354	arbitrary limit), but is broken in many versions of the microsoft runtime
		3355	libraries.
		3356
		3357	This might get you to about C<512> or C<2048> sockets (depending on
		3358	windows version and/or the phase of the moon). To get more, you need to
		3359	wrap all I/O functions and provide your own fd management, but the cost of
		3360	calling select (O(n²)) will likely make this unworkable.
		3361
		3362	=back
		3363
		3364
		3365	=head1 PORTABILITY REQUIREMENTS
		3366
		3367	In addition to a working ISO-C implementation, libev relies on a few
		3368	additional extensions:
		3369
		3370	=over 4
		3371
		3372	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3373
		3374	The type C<sig_atomic_t volatile> (or whatever is defined as
		3375	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different
		3376	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3377	believed to be sufficiently portable.
		3378
		3379	=item C<sigprocmask> must work in a threaded environment
		3380
		3381	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3382	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3383	pthread implementations will either allow C<sigprocmask> in the "main
		3384	thread" or will block signals process-wide, both behaviours would
		3385	be compatible with libev. Interaction between C<sigprocmask> and
		3386	C<pthread_sigmask> could complicate things, however.
		3387
		3388	The most portable way to handle signals is to block signals in all threads
		3389	except the initial one, and run the default loop in the initial thread as
		3390	well.
		3391
		3392	=item C<long> must be large enough for common memory allocation sizes
		3393
		3394	To improve portability and simplify using libev, libev uses C<long>
		3395	internally instead of C<size_t> when allocating its data structures. On
		3396	non-POSIX systems (Microsoft...) this might be unexpectedly low, but
		3397	is still at least 31 bits everywhere, which is enough for hundreds of
		3398	millions of watchers.
		3399
		3400	=item C<double> must hold a time value in seconds with enough accuracy
		3401
		3402	The type C<double> is used to represent timestamps. It is required to
		3403	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3404	enough for at least into the year 4000. This requirement is fulfilled by
		3405	implementations implementing IEEE 754 (basically all existing ones).
		3406
		3407	=back
		3408
		3409	If you know of other additional requirements drop me a note.
		3410
		3411
		3412	=head1 COMPILER WARNINGS
		3413
		3414	Depending on your compiler and compiler settings, you might get no or a
		3415	lot of warnings when compiling libev code. Some people are apparently
		3416	scared by this.
		3417
		3418	However, these are unavoidable for many reasons. For one, each compiler
		3419	has different warnings, and each user has different tastes regarding
		3420	warning options. "Warn-free" code therefore cannot be a goal except when
		3421	targetting a specific compiler and compiler-version.
		3422
		3423	Another reason is that some compiler warnings require elaborate
		3424	workarounds, or other changes to the code that make it less clear and less
		3425	maintainable.
		3426
		3427	And of course, some compiler warnings are just plain stupid, or simply
		3428	wrong (because they don't actually warn about the cindition their message
		3429	seems to warn about).
		3430
		3431	While libev is written to generate as few warnings as possible,
		3432	"warn-free" code is not a goal, and it is recommended not to build libev
		3433	with any compiler warnings enabled unless you are prepared to cope with
		3434	them (e.g. by ignoring them). Remember that warnings are just that:
		3435	warnings, not errors, or proof of bugs.
		3436
		3437
		3438	=head1 VALGRIND
		3439
		3440	Valgrind has a special section here because it is a popular tool that is
		3441	highly useful, but valgrind reports are very hard to interpret.
		3442
		3443	If you think you found a bug (memory leak, uninitialised data access etc.)
		3444	in libev, then check twice: If valgrind reports something like:
		3445
		3446	==2274== definitely lost: 0 bytes in 0 blocks.
		3447	==2274== possibly lost: 0 bytes in 0 blocks.
		3448	==2274== still reachable: 256 bytes in 1 blocks.
		3449
		3450	then there is no memory leak. Similarly, under some circumstances,
		3451	valgrind might report kernel bugs as if it were a bug in libev, or it
		3452	might be confused (it is a very good tool, but only a tool).
		3453
		3454	If you are unsure about something, feel free to contact the mailing list
		3455	with the full valgrind report and an explanation on why you think this is
		3456	a bug in libev. However, don't be annoyed when you get a brisk "this is
		3457	no bug" answer and take the chance of learning how to interpret valgrind
		3458	properly.
		3459
		3460	If you need, for some reason, empty reports from valgrind for your project
		3461	I suggest using suppression lists.
		3462
		3463
2088	=head1 AUTHOR	3464	=head1 AUTHOR
2089		3465
2090	Marc Lehmann <libev@schmorp.de>.	3466	Marc Lehmann <libev@schmorp.de>.
2091		3467

Diff Legend

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