ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.30 by root, Fri Nov 23 04:36:03 2007 UTC vs.
Revision 1.53 by root, Tue Nov 27 20:15:02 2007 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
		7	/* this is the only header you need */
7	#include <ev.h>	8	#include <ev.h>
		9
		10	/* what follows is a fully working example program */
		11	ev_io stdin_watcher;
		12	ev_timer timeout_watcher;
		13
		14	/* called when data readable on stdin */
		15	static void
		16	stdin_cb (EV_P_ struct ev_io *w, int revents)
		17	{
		18	/* puts ("stdin ready"); */
		19	ev_io_stop (EV_A_ w); /* just a syntax example */
		20	ev_unloop (EV_A_ EVUNLOOP_ALL); /* leave all loop calls */
		21	}
		22
		23	static void
		24	timeout_cb (EV_P_ struct ev_timer *w, int revents)
		25	{
		26	/* puts ("timeout"); */
		27	ev_unloop (EV_A_ EVUNLOOP_ONE); /* leave one loop call */
		28	}
		29
		30	int
		31	main (void)
		32	{
		33	struct ev_loop *loop = ev_default_loop (0);
		34
		35	/* initialise an io watcher, then start it */
		36	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		37	ev_io_start (loop, &stdin_watcher);
		38
		39	/* simple non-repeating 5.5 second timeout */
		40	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		41	ev_timer_start (loop, &timeout_watcher);
		42
		43	/* loop till timeout or data ready */
		44	ev_loop (loop, 0);
		45
		46	return 0;
		47	}
8		48
9	=head1 DESCRIPTION	49	=head1 DESCRIPTION
10		50
11	Libev is an event loop: you register interest in certain events (such as a	51	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	52	file descriptor being readable or a timeout occuring), and it will manage
…		…
45		85
46	Libev represents time as a single floating point number, representing the	86	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	87	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	88	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	89	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the double type in C.	90	to the C<double> type in C, and when you need to do any calculations on
		91	it, you should treat it as such.
51		92
52	=head1 GLOBAL FUNCTIONS	93	=head1 GLOBAL FUNCTIONS
53		94
54	These functions can be called anytime, even before initialising the	95	These functions can be called anytime, even before initialising the
55	library in any way.	96	library in any way.
…		…
75	Usually, it's a good idea to terminate if the major versions mismatch,	116	Usually, it's a good idea to terminate if the major versions mismatch,
76	as this indicates an incompatible change. Minor versions are usually	117	as this indicates an incompatible change. Minor versions are usually
77	compatible to older versions, so a larger minor version alone is usually	118	compatible to older versions, so a larger minor version alone is usually
78	not a problem.	119	not a problem.
79		120
		121	Example: make sure we haven't accidentally been linked against the wrong
		122	version:
		123
		124	assert (("libev version mismatch",
		125	ev_version_major () == EV_VERSION_MAJOR
		126	&& ev_version_minor () >= EV_VERSION_MINOR));
		127
		128	=item unsigned int ev_supported_backends ()
		129
		130	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		131	value) compiled into this binary of libev (independent of their
		132	availability on the system you are running on). See C<ev_default_loop> for
		133	a description of the set values.
		134
		135	Example: make sure we have the epoll method, because yeah this is cool and
		136	a must have and can we have a torrent of it please!!!11
		137
		138	assert (("sorry, no epoll, no sex",
		139	ev_supported_backends () & EVBACKEND_EPOLL));
		140
		141	=item unsigned int ev_recommended_backends ()
		142
		143	Return the set of all backends compiled into this binary of libev and also
		144	recommended for this platform. This set is often smaller than the one
		145	returned by C<ev_supported_backends>, as for example kqueue is broken on
		146	most BSDs and will not be autodetected unless you explicitly request it
		147	(assuming you know what you are doing). This is the set of backends that
		148	libev will probe for if you specify no backends explicitly.
		149
		150	=item unsigned int ev_embeddable_backends ()
		151
		152	Returns the set of backends that are embeddable in other event loops. This
		153	is the theoretical, all-platform, value. To find which backends
		154	might be supported on the current system, you would need to look at
		155	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
		156	recommended ones.
		157
		158	See the description of C<ev_embed> watchers for more info.
		159
80	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	160	=item ev_set_allocator (void *(*cb)(void *ptr, size_t size))
81		161
82	Sets the allocation function to use (the prototype is similar to the	162	Sets the allocation function to use (the prototype and semantics are
83	realloc C function, the semantics are identical). It is used to allocate	163	identical to the realloc C function). It is used to allocate and free
84	and free memory (no surprises here). If it returns zero when memory	164	memory (no surprises here). If it returns zero when memory needs to be
85	needs to be allocated, the library might abort or take some potentially	165	allocated, the library might abort or take some potentially destructive
86	destructive action. The default is your system realloc function.	166	action. The default is your system realloc function.
87		167
88	You could override this function in high-availability programs to, say,	168	You could override this function in high-availability programs to, say,
89	free some memory if it cannot allocate memory, to use a special allocator,	169	free some memory if it cannot allocate memory, to use a special allocator,
90	or even to sleep a while and retry until some memory is available.	170	or even to sleep a while and retry until some memory is available.
		171
		172	Example: replace the libev allocator with one that waits a bit and then
		173	retries: better than mine).
		174
		175	static void *
		176	persistent_realloc (void *ptr, size_t size)
		177	{
		178	for (;;)
		179	{
		180	void *newptr = realloc (ptr, size);
		181
		182	if (newptr)
		183	return newptr;
		184
		185	sleep (60);
		186	}
		187	}
		188
		189	...
		190	ev_set_allocator (persistent_realloc);
91		191
92	=item ev_set_syserr_cb (void (*cb)(const char *msg));	192	=item ev_set_syserr_cb (void (*cb)(const char *msg));
93		193
94	Set the callback function to call on a retryable syscall error (such	194	Set the callback function to call on a retryable syscall error (such
95	as failed select, poll, epoll_wait). The message is a printable string	195	as failed select, poll, epoll_wait). The message is a printable string
…		…
97	callback is set, then libev will expect it to remedy the sitution, no	197	callback is set, then libev will expect it to remedy the sitution, no
98	matter what, when it returns. That is, libev will generally retry the	198	matter what, when it returns. That is, libev will generally retry the
99	requested operation, or, if the condition doesn't go away, do bad stuff	199	requested operation, or, if the condition doesn't go away, do bad stuff
100	(such as abort).	200	(such as abort).
101		201
		202	Example: do the same thing as libev does internally:
		203
		204	static void
		205	fatal_error (const char *msg)
		206	{
		207	perror (msg);
		208	abort ();
		209	}
		210
		211	...
		212	ev_set_syserr_cb (fatal_error);
		213
102	=back	214	=back
103		215
104	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	216	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
105		217
106	An event loop is described by a C<struct ev_loop *>. The library knows two	218	An event loop is described by a C<struct ev_loop *>. The library knows two
…		…
119	=item struct ev_loop *ev_default_loop (unsigned int flags)	231	=item struct ev_loop *ev_default_loop (unsigned int flags)
120		232
121	This will initialise the default event loop if it hasn't been initialised	233	This will initialise the default event loop if it hasn't been initialised
122	yet and return it. If the default loop could not be initialised, returns	234	yet and return it. If the default loop could not be initialised, returns
123	false. If it already was initialised it simply returns it (and ignores the	235	false. If it already was initialised it simply returns it (and ignores the
124	flags).	236	flags. If that is troubling you, check C<ev_backend ()> afterwards).
125		237
126	If you don't know what event loop to use, use the one returned from this	238	If you don't know what event loop to use, use the one returned from this
127	function.	239	function.
128		240
129	The flags argument can be used to specify special behaviour or specific	241	The flags argument can be used to specify special behaviour or specific
130	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	242	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
131		243
132	It supports the following flags:	244	The following flags are supported:
133		245
134	=over 4	246	=over 4
135		247
136	=item C<EVFLAG_AUTO>	248	=item C<EVFLAG_AUTO>
137		249
…		…
145	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	257	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
146	override the flags completely if it is found in the environment. This is	258	override the flags completely if it is found in the environment. This is
147	useful to try out specific backends to test their performance, or to work	259	useful to try out specific backends to test their performance, or to work
148	around bugs.	260	around bugs.
149		261
150	=item C<EVMETHOD_SELECT> (value 1, portable select backend)	262	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
151		263
152	This is your standard select(2) backend. Not I<completely> standard, as	264	This is your standard select(2) backend. Not I<completely> standard, as
153	libev tries to roll its own fd_set with no limits on the number of fds,	265	libev tries to roll its own fd_set with no limits on the number of fds,
154	but if that fails, expect a fairly low limit on the number of fds when	266	but if that fails, expect a fairly low limit on the number of fds when
155	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	267	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
156	the fastest backend for a low number of fds.	268	the fastest backend for a low number of fds.
157		269
158	=item C<EVMETHOD_POLL> (value 2, poll backend, available everywhere except on windows)	270	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
159		271
160	And this is your standard poll(2) backend. It's more complicated than	272	And this is your standard poll(2) backend. It's more complicated than
161	select, but handles sparse fds better and has no artificial limit on the	273	select, but handles sparse fds better and has no artificial limit on the
162	number of fds you can use (except it will slow down considerably with a	274	number of fds you can use (except it will slow down considerably with a
163	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	275	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
164		276
165	=item C<EVMETHOD_EPOLL> (value 4, Linux)	277	=item C<EVBACKEND_EPOLL> (value 4, Linux)
166		278
167	For few fds, this backend is a bit little slower than poll and select,	279	For few fds, this backend is a bit little slower than poll and select,
168	but it scales phenomenally better. While poll and select usually scale like	280	but it scales phenomenally better. While poll and select usually scale like
169	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	281	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
170	either O(1) or O(active_fds).	282	either O(1) or O(active_fds).
…		…
173	result in some caching, there is still a syscall per such incident	285	result in some caching, there is still a syscall per such incident
174	(because the fd could point to a different file description now), so its	286	(because the fd could point to a different file description now), so its
175	best to avoid that. Also, dup()ed file descriptors might not work very	287	best to avoid that. Also, dup()ed file descriptors might not work very
176	well if you register events for both fds.	288	well if you register events for both fds.
177		289
		290	Please note that epoll sometimes generates spurious notifications, so you
		291	need to use non-blocking I/O or other means to avoid blocking when no data
		292	(or space) is available.
		293
178	=item C<EVMETHOD_KQUEUE> (value 8, most BSD clones)	294	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
179		295
180	Kqueue deserves special mention, as at the time of this writing, it	296	Kqueue deserves special mention, as at the time of this writing, it
181	was broken on all BSDs except NetBSD (usually it doesn't work with	297	was broken on all BSDs except NetBSD (usually it doesn't work with
182	anything but sockets and pipes, except on Darwin, where of course its	298	anything but sockets and pipes, except on Darwin, where of course its
183	completely useless). For this reason its not being "autodetected" unless	299	completely useless). For this reason its not being "autodetected"
184	you explicitly specify the flags (i.e. you don't use EVFLAG_AUTO).	300	unless you explicitly specify it explicitly in the flags (i.e. using
		301	C<EVBACKEND_KQUEUE>).
185		302
186	It scales in the same way as the epoll backend, but the interface to the	303	It scales in the same way as the epoll backend, but the interface to the
187	kernel is more efficient (which says nothing about its actual speed, of	304	kernel is more efficient (which says nothing about its actual speed, of
188	course). While starting and stopping an I/O watcher does not cause an	305	course). While starting and stopping an I/O watcher does not cause an
189	extra syscall as with epoll, it still adds up to four event changes per	306	extra syscall as with epoll, it still adds up to four event changes per
190	incident, so its best to avoid that.	307	incident, so its best to avoid that.
191		308
192	=item C<EVMETHOD_DEVPOLL> (value 16, Solaris 8)	309	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
193		310
194	This is not implemented yet (and might never be).	311	This is not implemented yet (and might never be).
195		312
196	=item C<EVMETHOD_PORT> (value 32, Solaris 10)	313	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
197		314
198	This uses the Solaris 10 port mechanism. As with everything on Solaris,	315	This uses the Solaris 10 port mechanism. As with everything on Solaris,
199	it's really slow, but it still scales very well (O(active_fds)).	316	it's really slow, but it still scales very well (O(active_fds)).
200		317
		318	Please note that solaris ports can result in a lot of spurious
		319	notifications, so you need to use non-blocking I/O or other means to avoid
		320	blocking when no data (or space) is available.
		321
201	=item C<EVMETHOD_ALL>	322	=item C<EVBACKEND_ALL>
202		323
203	Try all backends (even potentially broken ones that wouldn't be tried	324	Try all backends (even potentially broken ones that wouldn't be tried
204	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	325	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
205	C<EVMETHOD_ALL & ~EVMETHOD_KQUEUE>.	326	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
206		327
207	=back	328	=back
208		329
209	If one or more of these are ored into the flags value, then only these	330	If one or more of these are ored into the flags value, then only these
210	backends will be tried (in the reverse order as given here). If none are	331	backends will be tried (in the reverse order as given here). If none are
211	specified, most compiled-in backend will be tried, usually in reverse	332	specified, most compiled-in backend will be tried, usually in reverse
212	order of their flag values :)	333	order of their flag values :)
213		334
		335	The most typical usage is like this:
		336
		337	if (!ev_default_loop (0))
		338	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		339
		340	Restrict libev to the select and poll backends, and do not allow
		341	environment settings to be taken into account:
		342
		343	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		344
		345	Use whatever libev has to offer, but make sure that kqueue is used if
		346	available (warning, breaks stuff, best use only with your own private
		347	event loop and only if you know the OS supports your types of fds):
		348
		349	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
		350
214	=item struct ev_loop *ev_loop_new (unsigned int flags)	351	=item struct ev_loop *ev_loop_new (unsigned int flags)
215		352
216	Similar to C<ev_default_loop>, but always creates a new event loop that is	353	Similar to C<ev_default_loop>, but always creates a new event loop that is
217	always distinct from the default loop. Unlike the default loop, it cannot	354	always distinct from the default loop. Unlike the default loop, it cannot
218	handle signal and child watchers, and attempts to do so will be greeted by	355	handle signal and child watchers, and attempts to do so will be greeted by
219	undefined behaviour (or a failed assertion if assertions are enabled).	356	undefined behaviour (or a failed assertion if assertions are enabled).
220		357
		358	Example: try to create a event loop that uses epoll and nothing else.
		359
		360	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
		361	if (!epoller)
		362	fatal ("no epoll found here, maybe it hides under your chair");
		363
221	=item ev_default_destroy ()	364	=item ev_default_destroy ()
222		365
223	Destroys the default loop again (frees all memory and kernel state	366	Destroys the default loop again (frees all memory and kernel state
224	etc.). This stops all registered event watchers (by not touching them in	367	etc.). None of the active event watchers will be stopped in the normal
225	any way whatsoever, although you cannot rely on this :).	368	sense, so e.g. C<ev_is_active> might still return true. It is your
		369	responsibility to either stop all watchers cleanly yoursef I<before>
		370	calling this function, or cope with the fact afterwards (which is usually
		371	the easiest thing, youc na just ignore the watchers and/or C<free ()> them
		372	for example).
226		373
227	=item ev_loop_destroy (loop)	374	=item ev_loop_destroy (loop)
228		375
229	Like C<ev_default_destroy>, but destroys an event loop created by an	376	Like C<ev_default_destroy>, but destroys an event loop created by an
230	earlier call to C<ev_loop_new>.	377	earlier call to C<ev_loop_new>.
…		…
244	it just in case after a fork. To make this easy, the function will fit in	391	it just in case after a fork. To make this easy, the function will fit in
245	quite nicely into a call to C<pthread_atfork>:	392	quite nicely into a call to C<pthread_atfork>:
246		393
247	pthread_atfork (0, 0, ev_default_fork);	394	pthread_atfork (0, 0, ev_default_fork);
248		395
		396	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
		397	without calling this function, so if you force one of those backends you
		398	do not need to care.
		399
249	=item ev_loop_fork (loop)	400	=item ev_loop_fork (loop)
250		401
251	Like C<ev_default_fork>, but acts on an event loop created by	402	Like C<ev_default_fork>, but acts on an event loop created by
252	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	403	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
253	after fork, and how you do this is entirely your own problem.	404	after fork, and how you do this is entirely your own problem.
254		405
255	=item unsigned int ev_method (loop)	406	=item unsigned int ev_backend (loop)
256		407
257	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	408	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
258	use.	409	use.
259		410
260	=item ev_tstamp ev_now (loop)	411	=item ev_tstamp ev_now (loop)
261		412
262	Returns the current "event loop time", which is the time the event loop	413	Returns the current "event loop time", which is the time the event loop
263	got events and started processing them. This timestamp does not change	414	received events and started processing them. This timestamp does not
264	as long as callbacks are being processed, and this is also the base time	415	change as long as callbacks are being processed, and this is also the base
265	used for relative timers. You can treat it as the timestamp of the event	416	time used for relative timers. You can treat it as the timestamp of the
266	occuring (or more correctly, the mainloop finding out about it).	417	event occuring (or more correctly, libev finding out about it).
267		418
268	=item ev_loop (loop, int flags)	419	=item ev_loop (loop, int flags)
269		420
270	Finally, this is it, the event handler. This function usually is called	421	Finally, this is it, the event handler. This function usually is called
271	after you initialised all your watchers and you want to start handling	422	after you initialised all your watchers and you want to start handling
272	events.	423	events.
273		424
274	If the flags argument is specified as 0, it will not return until either	425	If the flags argument is specified as C<0>, it will not return until
275	no event watchers are active anymore or C<ev_unloop> was called.	426	either no event watchers are active anymore or C<ev_unloop> was called.
		427
		428	Please note that an explicit C<ev_unloop> is usually better than
		429	relying on all watchers to be stopped when deciding when a program has
		430	finished (especially in interactive programs), but having a program that
		431	automatically loops as long as it has to and no longer by virtue of
		432	relying on its watchers stopping correctly is a thing of beauty.
276		433
277	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	434	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
278	those events and any outstanding ones, but will not block your process in	435	those events and any outstanding ones, but will not block your process in
279	case there are no events and will return after one iteration of the loop.	436	case there are no events and will return after one iteration of the loop.
280		437
281	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	438	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
282	neccessary) and will handle those and any outstanding ones. It will block	439	neccessary) and will handle those and any outstanding ones. It will block
283	your process until at least one new event arrives, and will return after	440	your process until at least one new event arrives, and will return after
284	one iteration of the loop.	441	one iteration of the loop. This is useful if you are waiting for some
		442	external event in conjunction with something not expressible using other
		443	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
		444	usually a better approach for this kind of thing.
285		445
286	This flags value could be used to implement alternative looping
287	constructs, but the C<prepare> and C<check> watchers provide a better and
288	more generic mechanism.
289
290	Here are the gory details of what ev_loop does:	446	Here are the gory details of what C<ev_loop> does:
291		447
292	1. If there are no active watchers (reference count is zero), return.	448	* If there are no active watchers (reference count is zero), return.
293	2. Queue and immediately call all prepare watchers.	449	- Queue prepare watchers and then call all outstanding watchers.
294	3. If we have been forked, recreate the kernel state.	450	- If we have been forked, recreate the kernel state.
295	4. Update the kernel state with all outstanding changes.	451	- Update the kernel state with all outstanding changes.
296	5. Update the "event loop time".	452	- Update the "event loop time".
297	6. Calculate for how long to block.	453	- Calculate for how long to block.
298	7. Block the process, waiting for events.	454	- Block the process, waiting for any events.
		455	- Queue all outstanding I/O (fd) events.
299	8. Update the "event loop time" and do time jump handling.	456	- Update the "event loop time" and do time jump handling.
300	9. Queue all outstanding timers.	457	- Queue all outstanding timers.
301	10. Queue all outstanding periodics.	458	- Queue all outstanding periodics.
302	11. If no events are pending now, queue all idle watchers.	459	- If no events are pending now, queue all idle watchers.
303	12. Queue all check watchers.	460	- Queue all check watchers.
304	13. Call all queued watchers in reverse order (i.e. check watchers first).	461	- Call all queued watchers in reverse order (i.e. check watchers first).
		462	Signals and child watchers are implemented as I/O watchers, and will
		463	be handled here by queueing them when their watcher gets executed.
305	14. If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	464	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
306	was used, return, otherwise continue with step #1.	465	were used, return, otherwise continue with step *.
		466
		467	Example: queue some jobs and then loop until no events are outsanding
		468	anymore.
		469
		470	... queue jobs here, make sure they register event watchers as long
		471	... as they still have work to do (even an idle watcher will do..)
		472	ev_loop (my_loop, 0);
		473	... jobs done. yeah!
307		474
308	=item ev_unloop (loop, how)	475	=item ev_unloop (loop, how)
309		476
310	Can be used to make a call to C<ev_loop> return early (but only after it	477	Can be used to make a call to C<ev_loop> return early (but only after it
311	has processed all outstanding events). The C<how> argument must be either	478	has processed all outstanding events). The C<how> argument must be either
…		…
325	visible to the libev user and should not keep C<ev_loop> from exiting if	492	visible to the libev user and should not keep C<ev_loop> from exiting if
326	no event watchers registered by it are active. It is also an excellent	493	no event watchers registered by it are active. It is also an excellent
327	way to do this for generic recurring timers or from within third-party	494	way to do this for generic recurring timers or from within third-party
328	libraries. Just remember to I<unref after start> and I<ref before stop>.	495	libraries. Just remember to I<unref after start> and I<ref before stop>.
329		496
		497	Example: create a signal watcher, but keep it from keeping C<ev_loop>
		498	running when nothing else is active.
		499
		500	struct dv_signal exitsig;
		501	ev_signal_init (&exitsig, sig_cb, SIGINT);
		502	ev_signal_start (myloop, &exitsig);
		503	evf_unref (myloop);
		504
		505	Example: for some weird reason, unregister the above signal handler again.
		506
		507	ev_ref (myloop);
		508	ev_signal_stop (myloop, &exitsig);
		509
330	=back	510	=back
		511
331		512
332	=head1 ANATOMY OF A WATCHER	513	=head1 ANATOMY OF A WATCHER
333		514
334	A watcher is a structure that you create and register to record your	515	A watcher is a structure that you create and register to record your
335	interest in some event. For instance, if you want to wait for STDIN to	516	interest in some event. For instance, if you want to wait for STDIN to
…		…
368	*) >>), and you can stop watching for events at any time by calling the	549	*) >>), and you can stop watching for events at any time by calling the
369	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	550	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.
370		551
371	As long as your watcher is active (has been started but not stopped) you	552	As long as your watcher is active (has been started but not stopped) you
372	must not touch the values stored in it. Most specifically you must never	553	must not touch the values stored in it. Most specifically you must never
373	reinitialise it or call its set method.	554	reinitialise it or call its C<set> macro.
374
375	You can check whether an event is active by calling the C<ev_is_active
376	(watcher *)> macro. To see whether an event is outstanding (but the
377	callback for it has not been called yet) you can use the C<ev_is_pending
378	(watcher *)> macro.
379		555
380	Each and every callback receives the event loop pointer as first, the	556	Each and every callback receives the event loop pointer as first, the
381	registered watcher structure as second, and a bitset of received events as	557	registered watcher structure as second, and a bitset of received events as
382	third argument.	558	third argument.
383		559
…		…
407	The signal specified in the C<ev_signal> watcher has been received by a thread.	583	The signal specified in the C<ev_signal> watcher has been received by a thread.
408		584
409	=item C<EV_CHILD>	585	=item C<EV_CHILD>
410		586
411	The pid specified in the C<ev_child> watcher has received a status change.	587	The pid specified in the C<ev_child> watcher has received a status change.
		588
		589	=item C<EV_STAT>
		590
		591	The path specified in the C<ev_stat> watcher changed its attributes somehow.
412		592
413	=item C<EV_IDLE>	593	=item C<EV_IDLE>
414		594
415	The C<ev_idle> watcher has determined that you have nothing better to do.	595	The C<ev_idle> watcher has determined that you have nothing better to do.
416		596
…		…
424	received events. Callbacks of both watcher types can start and stop as	604	received events. Callbacks of both watcher types can start and stop as
425	many watchers as they want, and all of them will be taken into account	605	many watchers as they want, and all of them will be taken into account
426	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	606	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
427	C<ev_loop> from blocking).	607	C<ev_loop> from blocking).
428		608
		609	=item C<EV_EMBED>
		610
		611	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		612
		613	=item C<EV_FORK>
		614
		615	The event loop has been resumed in the child process after fork (see
		616	C<ev_fork>).
		617
429	=item C<EV_ERROR>	618	=item C<EV_ERROR>
430		619
431	An unspecified error has occured, the watcher has been stopped. This might	620	An unspecified error has occured, the watcher has been stopped. This might
432	happen because the watcher could not be properly started because libev	621	happen because the watcher could not be properly started because libev
433	ran out of memory, a file descriptor was found to be closed or any other	622	ran out of memory, a file descriptor was found to be closed or any other
…		…
439	your callbacks is well-written it can just attempt the operation and cope	628	your callbacks is well-written it can just attempt the operation and cope
440	with the error from read() or write(). This will not work in multithreaded	629	with the error from read() or write(). This will not work in multithreaded
441	programs, though, so beware.	630	programs, though, so beware.
442		631
443	=back	632	=back
		633
		634	=head2 GENERIC WATCHER FUNCTIONS
		635
		636	In the following description, C<TYPE> stands for the watcher type,
		637	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
		638
		639	=over 4
		640
		641	=item C<ev_init> (ev_TYPE *watcher, callback)
		642
		643	This macro initialises the generic portion of a watcher. The contents
		644	of the watcher object can be arbitrary (so C<malloc> will do). Only
		645	the generic parts of the watcher are initialised, you I<need> to call
		646	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		647	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		648	which rolls both calls into one.
		649
		650	You can reinitialise a watcher at any time as long as it has been stopped
		651	(or never started) and there are no pending events outstanding.
		652
		653	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,
		654	int revents)>.
		655
		656	=item C<ev_TYPE_set> (ev_TYPE *, [args])
		657
		658	This macro initialises the type-specific parts of a watcher. You need to
		659	call C<ev_init> at least once before you call this macro, but you can
		660	call C<ev_TYPE_set> any number of times. You must not, however, call this
		661	macro on a watcher that is active (it can be pending, however, which is a
		662	difference to the C<ev_init> macro).
		663
		664	Although some watcher types do not have type-specific arguments
		665	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		666
		667	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		668
		669	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		670	calls into a single call. This is the most convinient method to initialise
		671	a watcher. The same limitations apply, of course.
		672
		673	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)
		674
		675	Starts (activates) the given watcher. Only active watchers will receive
		676	events. If the watcher is already active nothing will happen.
		677
		678	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
		679
		680	Stops the given watcher again (if active) and clears the pending
		681	status. It is possible that stopped watchers are pending (for example,
		682	non-repeating timers are being stopped when they become pending), but
		683	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If
		684	you want to free or reuse the memory used by the watcher it is therefore a
		685	good idea to always call its C<ev_TYPE_stop> function.
		686
		687	=item bool ev_is_active (ev_TYPE *watcher)
		688
		689	Returns a true value iff the watcher is active (i.e. it has been started
		690	and not yet been stopped). As long as a watcher is active you must not modify
		691	it.
		692
		693	=item bool ev_is_pending (ev_TYPE *watcher)
		694
		695	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		696	events but its callback has not yet been invoked). As long as a watcher
		697	is pending (but not active) you must not call an init function on it (but
		698	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to
		699	libev (e.g. you cnanot C<free ()> it).
		700
		701	=item callback = ev_cb (ev_TYPE *watcher)
		702
		703	Returns the callback currently set on the watcher.
		704
		705	=item ev_cb_set (ev_TYPE *watcher, callback)
		706
		707	Change the callback. You can change the callback at virtually any time
		708	(modulo threads).
		709
		710	=back
		711
444		712
445	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	713	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
446		714
447	Each watcher has, by default, a member C<void *data> that you can change	715	Each watcher has, by default, a member C<void *data> that you can change
448	and read at any time, libev will completely ignore it. This can be used	716	and read at any time, libev will completely ignore it. This can be used
…		…
473		741
474		742
475	=head1 WATCHER TYPES	743	=head1 WATCHER TYPES
476		744
477	This section describes each watcher in detail, but will not repeat	745	This section describes each watcher in detail, but will not repeat
478	information given in the last section.	746	information given in the last section. Any initialisation/set macros,
		747	functions and members specific to the watcher type are explained.
479		748
		749	Members are additionally marked with either I<[read-only]>, meaning that,
		750	while the watcher is active, you can look at the member and expect some
		751	sensible content, but you must not modify it (you can modify it while the
		752	watcher is stopped to your hearts content), or I<[read-write]>, which
		753	means you can expect it to have some sensible content while the watcher
		754	is active, but you can also modify it. Modifying it may not do something
		755	sensible or take immediate effect (or do anything at all), but libev will
		756	not crash or malfunction in any way.
		757
		758
480	=head2 C<ev_io> - is this file descriptor readable or writable	759	=head2 C<ev_io> - is this file descriptor readable or writable?
481		760
482	I/O watchers check whether a file descriptor is readable or writable	761	I/O watchers check whether a file descriptor is readable or writable
483	in each iteration of the event loop (This behaviour is called	762	in each iteration of the event loop, or, more precisely, when reading
484	level-triggering because you keep receiving events as long as the	763	would not block the process and writing would at least be able to write
485	condition persists. Remember you can stop the watcher if you don't want to	764	some data. This behaviour is called level-triggering because you keep
486	act on the event and neither want to receive future events).	765	receiving events as long as the condition persists. Remember you can stop
		766	the watcher if you don't want to act on the event and neither want to
		767	receive future events.
487		768
488	In general you can register as many read and/or write event watchers per	769	In general you can register as many read and/or write event watchers per
489	fd as you want (as long as you don't confuse yourself). Setting all file	770	fd as you want (as long as you don't confuse yourself). Setting all file
490	descriptors to non-blocking mode is also usually a good idea (but not	771	descriptors to non-blocking mode is also usually a good idea (but not
491	required if you know what you are doing).	772	required if you know what you are doing).
492		773
493	You have to be careful with dup'ed file descriptors, though. Some backends	774	You have to be careful with dup'ed file descriptors, though. Some backends
494	(the linux epoll backend is a notable example) cannot handle dup'ed file	775	(the linux epoll backend is a notable example) cannot handle dup'ed file
495	descriptors correctly if you register interest in two or more fds pointing	776	descriptors correctly if you register interest in two or more fds pointing
496	to the same underlying file/socket etc. description (that is, they share	777	to the same underlying file/socket/etc. description (that is, they share
497	the same underlying "file open").	778	the same underlying "file open").
498		779
499	If you must do this, then force the use of a known-to-be-good backend	780	If you must do this, then force the use of a known-to-be-good backend
500	(at the time of this writing, this includes only EVMETHOD_SELECT and	781	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
501	EVMETHOD_POLL).	782	C<EVBACKEND_POLL>).
		783
		784	Another thing you have to watch out for is that it is quite easy to
		785	receive "spurious" readyness notifications, that is your callback might
		786	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
		787	because there is no data. Not only are some backends known to create a
		788	lot of those (for example solaris ports), it is very easy to get into
		789	this situation even with a relatively standard program structure. Thus
		790	it is best to always use non-blocking I/O: An extra C<read>(2) returning
		791	C<EAGAIN> is far preferable to a program hanging until some data arrives.
		792
		793	If you cannot run the fd in non-blocking mode (for example you should not
		794	play around with an Xlib connection), then you have to seperately re-test
		795	wether a file descriptor is really ready with a known-to-be good interface
		796	such as poll (fortunately in our Xlib example, Xlib already does this on
		797	its own, so its quite safe to use).
502		798
503	=over 4	799	=over 4
504		800
505	=item ev_io_init (ev_io *, callback, int fd, int events)	801	=item ev_io_init (ev_io *, callback, int fd, int events)
506		802
507	=item ev_io_set (ev_io *, int fd, int events)	803	=item ev_io_set (ev_io *, int fd, int events)
508		804
509	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	805	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
510	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	806	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or
511	EV_WRITE> to receive the given events.	807	C<EV_READ | EV_WRITE> to receive the given events.
		808
		809	=item int fd [read-only]
		810
		811	The file descriptor being watched.
		812
		813	=item int events [read-only]
		814
		815	The events being watched.
512		816
513	=back	817	=back
514		818
		819	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well
		820	readable, but only once. Since it is likely line-buffered, you could
		821	attempt to read a whole line in the callback:
		822
		823	static void
		824	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		825	{
		826	ev_io_stop (loop, w);
		827	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		828	}
		829
		830	...
		831	struct ev_loop *loop = ev_default_init (0);
		832	struct ev_io stdin_readable;
		833	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		834	ev_io_start (loop, &stdin_readable);
		835	ev_loop (loop, 0);
		836
		837
515	=head2 C<ev_timer> - relative and optionally recurring timeouts	838	=head2 C<ev_timer> - relative and optionally repeating timeouts
516		839
517	Timer watchers are simple relative timers that generate an event after a	840	Timer watchers are simple relative timers that generate an event after a
518	given time, and optionally repeating in regular intervals after that.	841	given time, and optionally repeating in regular intervals after that.
519		842
520	The timers are based on real time, that is, if you register an event that	843	The timers are based on real time, that is, if you register an event that
…		…
561		884
562	If the timer is repeating, either start it if necessary (with the repeat	885	If the timer is repeating, either start it if necessary (with the repeat
563	value), or reset the running timer to the repeat value.	886	value), or reset the running timer to the repeat value.
564		887
565	This sounds a bit complicated, but here is a useful and typical	888	This sounds a bit complicated, but here is a useful and typical
566	example: Imagine you have a tcp connection and you want a so-called idle	889	example: Imagine you have a tcp connection and you want a so-called
567	timeout, that is, you want to be called when there have been, say, 60	890	idle timeout, that is, you want to be called when there have been,
568	seconds of inactivity on the socket. The easiest way to do this is to	891	say, 60 seconds of inactivity on the socket. The easiest way to do
569	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	892	this is to configure an C<ev_timer> with C<after>=C<repeat>=C<60> and calling
570	time you successfully read or write some data. If you go into an idle	893	C<ev_timer_again> each time you successfully read or write some data. If
571	state where you do not expect data to travel on the socket, you can stop	894	you go into an idle state where you do not expect data to travel on the
572	the timer, and again will automatically restart it if need be.	895	socket, you can stop the timer, and again will automatically restart it if
		896	need be.
		897
		898	You can also ignore the C<after> value and C<ev_timer_start> altogether
		899	and only ever use the C<repeat> value:
		900
		901	ev_timer_init (timer, callback, 0., 5.);
		902	ev_timer_again (loop, timer);
		903	...
		904	timer->again = 17.;
		905	ev_timer_again (loop, timer);
		906	...
		907	timer->again = 10.;
		908	ev_timer_again (loop, timer);
		909
		910	This is more efficient then stopping/starting the timer eahc time you want
		911	to modify its timeout value.
		912
		913	=item ev_tstamp repeat [read-write]
		914
		915	The current C<repeat> value. Will be used each time the watcher times out
		916	or C<ev_timer_again> is called and determines the next timeout (if any),
		917	which is also when any modifications are taken into account.
573		918
574	=back	919	=back
575		920
		921	Example: create a timer that fires after 60 seconds.
		922
		923	static void
		924	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		925	{
		926	.. one minute over, w is actually stopped right here
		927	}
		928
		929	struct ev_timer mytimer;
		930	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		931	ev_timer_start (loop, &mytimer);
		932
		933	Example: create a timeout timer that times out after 10 seconds of
		934	inactivity.
		935
		936	static void
		937	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
		938	{
		939	.. ten seconds without any activity
		940	}
		941
		942	struct ev_timer mytimer;
		943	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		944	ev_timer_again (&mytimer); /* start timer */
		945	ev_loop (loop, 0);
		946
		947	// and in some piece of code that gets executed on any "activity":
		948	// reset the timeout to start ticking again at 10 seconds
		949	ev_timer_again (&mytimer);
		950
		951
576	=head2 C<ev_periodic> - to cron or not to cron	952	=head2 C<ev_periodic> - to cron or not to cron?
577		953
578	Periodic watchers are also timers of a kind, but they are very versatile	954	Periodic watchers are also timers of a kind, but they are very versatile
579	(and unfortunately a bit complex).	955	(and unfortunately a bit complex).
580		956
581	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	957	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
582	but on wallclock time (absolute time). You can tell a periodic watcher	958	but on wallclock time (absolute time). You can tell a periodic watcher
583	to trigger "at" some specific point in time. For example, if you tell a	959	to trigger "at" some specific point in time. For example, if you tell a
584	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	960	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
585	+ 10.>) and then reset your system clock to the last year, then it will	961	+ 10.>) and then reset your system clock to the last year, then it will
586	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	962	take a year to trigger the event (unlike an C<ev_timer>, which would trigger
587	roughly 10 seconds later and of course not if you reset your system time	963	roughly 10 seconds later and of course not if you reset your system time
588	again).	964	again).
589		965
…		…
673	Simply stops and restarts the periodic watcher again. This is only useful	1049	Simply stops and restarts the periodic watcher again. This is only useful
674	when you changed some parameters or the reschedule callback would return	1050	when you changed some parameters or the reschedule callback would return
675	a different time than the last time it was called (e.g. in a crond like	1051	a different time than the last time it was called (e.g. in a crond like
676	program when the crontabs have changed).	1052	program when the crontabs have changed).
677		1053
		1054	=item ev_tstamp interval [read-write]
		1055
		1056	The current interval value. Can be modified any time, but changes only
		1057	take effect when the periodic timer fires or C<ev_periodic_again> is being
		1058	called.
		1059
		1060	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]
		1061
		1062	The current reschedule callback, or C<0>, if this functionality is
		1063	switched off. Can be changed any time, but changes only take effect when
		1064	the periodic timer fires or C<ev_periodic_again> is being called.
		1065
678	=back	1066	=back
679		1067
		1068	Example: call a callback every hour, or, more precisely, whenever the
		1069	system clock is divisible by 3600. The callback invocation times have
		1070	potentially a lot of jittering, but good long-term stability.
		1071
		1072	static void
		1073	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
		1074	{
		1075	... its now a full hour (UTC, or TAI or whatever your clock follows)
		1076	}
		1077
		1078	struct ev_periodic hourly_tick;
		1079	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		1080	ev_periodic_start (loop, &hourly_tick);
		1081
		1082	Example: the same as above, but use a reschedule callback to do it:
		1083
		1084	#include <math.h>
		1085
		1086	static ev_tstamp
		1087	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		1088	{
		1089	return fmod (now, 3600.) + 3600.;
		1090	}
		1091
		1092	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		1093
		1094	Example: call a callback every hour, starting now:
		1095
		1096	struct ev_periodic hourly_tick;
		1097	ev_periodic_init (&hourly_tick, clock_cb,
		1098	fmod (ev_now (loop), 3600.), 3600., 0);
		1099	ev_periodic_start (loop, &hourly_tick);
		1100
		1101
680	=head2 C<ev_signal> - signal me when a signal gets signalled	1102	=head2 C<ev_signal> - signal me when a signal gets signalled!
681		1103
682	Signal watchers will trigger an event when the process receives a specific	1104	Signal watchers will trigger an event when the process receives a specific
683	signal one or more times. Even though signals are very asynchronous, libev	1105	signal one or more times. Even though signals are very asynchronous, libev
684	will try it's best to deliver signals synchronously, i.e. as part of the	1106	will try it's best to deliver signals synchronously, i.e. as part of the
685	normal event processing, like any other event.	1107	normal event processing, like any other event.
…		…
698	=item ev_signal_set (ev_signal *, int signum)	1120	=item ev_signal_set (ev_signal *, int signum)
699		1121
700	Configures the watcher to trigger on the given signal number (usually one	1122	Configures the watcher to trigger on the given signal number (usually one
701	of the C<SIGxxx> constants).	1123	of the C<SIGxxx> constants).
702		1124
		1125	=item int signum [read-only]
		1126
		1127	The signal the watcher watches out for.
		1128
703	=back	1129	=back
704		1130
		1131
705	=head2 C<ev_child> - wait for pid status changes	1132	=head2 C<ev_child> - watch out for process status changes
706		1133
707	Child watchers trigger when your process receives a SIGCHLD in response to	1134	Child watchers trigger when your process receives a SIGCHLD in response to
708	some child status changes (most typically when a child of yours dies).	1135	some child status changes (most typically when a child of yours dies).
709		1136
710	=over 4	1137	=over 4
…		…
718	at the C<rstatus> member of the C<ev_child> watcher structure to see	1145	at the C<rstatus> member of the C<ev_child> watcher structure to see
719	the status word (use the macros from C<sys/wait.h> and see your systems	1146	the status word (use the macros from C<sys/wait.h> and see your systems
720	C<waitpid> documentation). The C<rpid> member contains the pid of the	1147	C<waitpid> documentation). The C<rpid> member contains the pid of the
721	process causing the status change.	1148	process causing the status change.
722		1149
		1150	=item int pid [read-only]
		1151
		1152	The process id this watcher watches out for, or C<0>, meaning any process id.
		1153
		1154	=item int rpid [read-write]
		1155
		1156	The process id that detected a status change.
		1157
		1158	=item int rstatus [read-write]
		1159
		1160	The process exit/trace status caused by C<rpid> (see your systems
		1161	C<waitpid> and C<sys/wait.h> documentation for details).
		1162
723	=back	1163	=back
724		1164
		1165	Example: try to exit cleanly on SIGINT and SIGTERM.
		1166
		1167	static void
		1168	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
		1169	{
		1170	ev_unloop (loop, EVUNLOOP_ALL);
		1171	}
		1172
		1173	struct ev_signal signal_watcher;
		1174	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1175	ev_signal_start (loop, &sigint_cb);
		1176
		1177
		1178	=head2 C<ev_stat> - did the file attributes just change?
		1179
		1180	This watches a filesystem path for attribute changes. That is, it calls
		1181	C<stat> regularly (or when the OS says it changed) and sees if it changed
		1182	compared to the last time, invoking the callback if it did.
		1183
		1184	The path does not need to exist: changing from "path exists" to "path does
		1185	not exist" is a status change like any other. The condition "path does
		1186	not exist" is signified by the C<st_nlink> field being zero (which is
		1187	otherwise always forced to be at least one) and all the other fields of
		1188	the stat buffer having unspecified contents.
		1189
		1190	Since there is no standard to do this, the portable implementation simply
		1191	calls C<stat (2)> regulalry on the path to see if it changed somehow. You
		1192	can specify a recommended polling interval for this case. If you specify
		1193	a polling interval of C<0> (highly recommended!) then a I<suitable,
		1194	unspecified default> value will be used (which you can expect to be around
		1195	five seconds, although this might change dynamically). Libev will also
		1196	impose a minimum interval which is currently around C<0.1>, but thats
		1197	usually overkill.
		1198
		1199	This watcher type is not meant for massive numbers of stat watchers,
		1200	as even with OS-supported change notifications, this can be
		1201	resource-intensive.
		1202
		1203	At the time of this writing, no specific OS backends are implemented, but
		1204	if demand increases, at least a kqueue and inotify backend will be added.
		1205
		1206	=over 4
		1207
		1208	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		1209
		1210	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		1211
		1212	Configures the watcher to wait for status changes of the given
		1213	C<path>. The C<interval> is a hint on how quickly a change is expected to
		1214	be detected and should normally be specified as C<0> to let libev choose
		1215	a suitable value. The memory pointed to by C<path> must point to the same
		1216	path for as long as the watcher is active.
		1217
		1218	The callback will be receive C<EV_STAT> when a change was detected,
		1219	relative to the attributes at the time the watcher was started (or the
		1220	last change was detected).
		1221
		1222	=item ev_stat_stat (ev_stat *)
		1223
		1224	Updates the stat buffer immediately with new values. If you change the
		1225	watched path in your callback, you could call this fucntion to avoid
		1226	detecting this change (while introducing a race condition). Can also be
		1227	useful simply to find out the new values.
		1228
		1229	=item ev_statdata attr [read-only]
		1230
		1231	The most-recently detected attributes of the file. Although the type is of
		1232	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		1233	suitable for your system. If the C<st_nlink> member is C<0>, then there
		1234	was some error while C<stat>ing the file.
		1235
		1236	=item ev_statdata prev [read-only]
		1237
		1238	The previous attributes of the file. The callback gets invoked whenever
		1239	C<prev> != C<attr>.
		1240
		1241	=item ev_tstamp interval [read-only]
		1242
		1243	The specified interval.
		1244
		1245	=item const char *path [read-only]
		1246
		1247	The filesystem path that is being watched.
		1248
		1249	=back
		1250
		1251	Example: Watch C</etc/passwd> for attribute changes.
		1252
		1253	static void
		1254	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		1255	{
		1256	/* /etc/passwd changed in some way */
		1257	if (w->attr.st_nlink)
		1258	{
		1259	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1260	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1261	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1262	}
		1263	else
		1264	/* you shalt not abuse printf for puts */
		1265	puts ("wow, /etc/passwd is not there, expect problems. "
		1266	"if this is windows, they already arrived\n");
		1267	}
		1268
		1269	...
		1270	ev_stat passwd;
		1271
		1272	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
		1273	ev_stat_start (loop, &passwd);
		1274
		1275
725	=head2 C<ev_idle> - when you've got nothing better to do	1276	=head2 C<ev_idle> - when you've got nothing better to do...
726		1277
727	Idle watchers trigger events when there are no other events are pending	1278	Idle watchers trigger events when there are no other events are pending
728	(prepare, check and other idle watchers do not count). That is, as long	1279	(prepare, check and other idle watchers do not count). That is, as long
729	as your process is busy handling sockets or timeouts (or even signals,	1280	as your process is busy handling sockets or timeouts (or even signals,
730	imagine) it will not be triggered. But when your process is idle all idle	1281	imagine) it will not be triggered. But when your process is idle all idle
…		…
748	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	1299	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
749	believe me.	1300	believe me.
750		1301
751	=back	1302	=back
752		1303
		1304	Example: dynamically allocate an C<ev_idle>, start it, and in the
		1305	callback, free it. Alos, use no error checking, as usual.
		1306
		1307	static void
		1308	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
		1309	{
		1310	free (w);
		1311	// now do something you wanted to do when the program has
		1312	// no longer asnything immediate to do.
		1313	}
		1314
		1315	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1316	ev_idle_init (idle_watcher, idle_cb);
		1317	ev_idle_start (loop, idle_cb);
		1318
		1319
753	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1320	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
754		1321
755	Prepare and check watchers are usually (but not always) used in tandem:	1322	Prepare and check watchers are usually (but not always) used in tandem:
756	prepare watchers get invoked before the process blocks and check watchers	1323	prepare watchers get invoked before the process blocks and check watchers
757	afterwards.	1324	afterwards.
758		1325
		1326	You I<must not> call C<ev_loop> or similar functions that enter
		1327	the current event loop from either C<ev_prepare> or C<ev_check>
		1328	watchers. Other loops than the current one are fine, however. The
		1329	rationale behind this is that you do not need to check for recursion in
		1330	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		1331	C<ev_check> so if you have one watcher of each kind they will always be
		1332	called in pairs bracketing the blocking call.
		1333
759	Their main purpose is to integrate other event mechanisms into libev. This	1334	Their main purpose is to integrate other event mechanisms into libev and
760	could be used, for example, to track variable changes, implement your own	1335	their use is somewhat advanced. This could be used, for example, to track
761	watchers, integrate net-snmp or a coroutine library and lots more.	1336	variable changes, implement your own watchers, integrate net-snmp or a
		1337	coroutine library and lots more. They are also occasionally useful if
		1338	you cache some data and want to flush it before blocking (for example,
		1339	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		1340	watcher).
762		1341
763	This is done by examining in each prepare call which file descriptors need	1342	This is done by examining in each prepare call which file descriptors need
764	to be watched by the other library, registering C<ev_io> watchers for	1343	to be watched by the other library, registering C<ev_io> watchers for
765	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1344	them and starting an C<ev_timer> watcher for any timeouts (many libraries
766	provide just this functionality). Then, in the check watcher you check for	1345	provide just this functionality). Then, in the check watcher you check for
…		…
788	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1367	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
789	macros, but using them is utterly, utterly and completely pointless.	1368	macros, but using them is utterly, utterly and completely pointless.
790		1369
791	=back	1370	=back
792		1371
		1372	Example: To include a library such as adns, you would add IO watchers
		1373	and a timeout watcher in a prepare handler, as required by libadns, and
		1374	in a check watcher, destroy them and call into libadns. What follows is
		1375	pseudo-code only of course:
		1376
		1377	static ev_io iow [nfd];
		1378	static ev_timer tw;
		1379
		1380	static void
		1381	io_cb (ev_loop *loop, ev_io *w, int revents)
		1382	{
		1383	// set the relevant poll flags
		1384	// could also call adns_processreadable etc. here
		1385	struct pollfd *fd = (struct pollfd *)w->data;
		1386	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		1387	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		1388	}
		1389
		1390	// create io watchers for each fd and a timer before blocking
		1391	static void
		1392	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
		1393	{
		1394	int timeout = 3600000;truct pollfd fds [nfd];
		1395	// actual code will need to loop here and realloc etc.
		1396	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1397
		1398	/* the callback is illegal, but won't be called as we stop during check */
		1399	ev_timer_init (&tw, 0, timeout * 1e-3);
		1400	ev_timer_start (loop, &tw);
		1401
		1402	// create on ev_io per pollfd
		1403	for (int i = 0; i < nfd; ++i)
		1404	{
		1405	ev_io_init (iow + i, io_cb, fds [i].fd,
		1406	((fds [i].events & POLLIN ? EV_READ : 0)
		1407	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1408
		1409	fds [i].revents = 0;
		1410	iow [i].data = fds + i;
		1411	ev_io_start (loop, iow + i);
		1412	}
		1413	}
		1414
		1415	// stop all watchers after blocking
		1416	static void
		1417	adns_check_cb (ev_loop *loop, ev_check *w, int revents)
		1418	{
		1419	ev_timer_stop (loop, &tw);
		1420
		1421	for (int i = 0; i < nfd; ++i)
		1422	ev_io_stop (loop, iow + i);
		1423
		1424	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1425	}
		1426
		1427
		1428	=head2 C<ev_embed> - when one backend isn't enough...
		1429
		1430	This is a rather advanced watcher type that lets you embed one event loop
		1431	into another (currently only C<ev_io> events are supported in the embedded
		1432	loop, other types of watchers might be handled in a delayed or incorrect
		1433	fashion and must not be used).
		1434
		1435	There are primarily two reasons you would want that: work around bugs and
		1436	prioritise I/O.
		1437
		1438	As an example for a bug workaround, the kqueue backend might only support
		1439	sockets on some platform, so it is unusable as generic backend, but you
		1440	still want to make use of it because you have many sockets and it scales
		1441	so nicely. In this case, you would create a kqueue-based loop and embed it
		1442	into your default loop (which might use e.g. poll). Overall operation will
		1443	be a bit slower because first libev has to poll and then call kevent, but
		1444	at least you can use both at what they are best.
		1445
		1446	As for prioritising I/O: rarely you have the case where some fds have
		1447	to be watched and handled very quickly (with low latency), and even
		1448	priorities and idle watchers might have too much overhead. In this case
		1449	you would put all the high priority stuff in one loop and all the rest in
		1450	a second one, and embed the second one in the first.
		1451
		1452	As long as the watcher is active, the callback will be invoked every time
		1453	there might be events pending in the embedded loop. The callback must then
		1454	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke
		1455	their callbacks (you could also start an idle watcher to give the embedded
		1456	loop strictly lower priority for example). You can also set the callback
		1457	to C<0>, in which case the embed watcher will automatically execute the
		1458	embedded loop sweep.
		1459
		1460	As long as the watcher is started it will automatically handle events. The
		1461	callback will be invoked whenever some events have been handled. You can
		1462	set the callback to C<0> to avoid having to specify one if you are not
		1463	interested in that.
		1464
		1465	Also, there have not currently been made special provisions for forking:
		1466	when you fork, you not only have to call C<ev_loop_fork> on both loops,
		1467	but you will also have to stop and restart any C<ev_embed> watchers
		1468	yourself.
		1469
		1470	Unfortunately, not all backends are embeddable, only the ones returned by
		1471	C<ev_embeddable_backends> are, which, unfortunately, does not include any
		1472	portable one.
		1473
		1474	So when you want to use this feature you will always have to be prepared
		1475	that you cannot get an embeddable loop. The recommended way to get around
		1476	this is to have a separate variables for your embeddable loop, try to
		1477	create it, and if that fails, use the normal loop for everything:
		1478
		1479	struct ev_loop *loop_hi = ev_default_init (0);
		1480	struct ev_loop *loop_lo = 0;
		1481	struct ev_embed embed;
		1482
		1483	// see if there is a chance of getting one that works
		1484	// (remember that a flags value of 0 means autodetection)
		1485	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		1486	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		1487	: 0;
		1488
		1489	// if we got one, then embed it, otherwise default to loop_hi
		1490	if (loop_lo)
		1491	{
		1492	ev_embed_init (&embed, 0, loop_lo);
		1493	ev_embed_start (loop_hi, &embed);
		1494	}
		1495	else
		1496	loop_lo = loop_hi;
		1497
		1498	=over 4
		1499
		1500	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		1501
		1502	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		1503
		1504	Configures the watcher to embed the given loop, which must be
		1505	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		1506	invoked automatically, otherwise it is the responsibility of the callback
		1507	to invoke it (it will continue to be called until the sweep has been done,
		1508	if you do not want thta, you need to temporarily stop the embed watcher).
		1509
		1510	=item ev_embed_sweep (loop, ev_embed *)
		1511
		1512	Make a single, non-blocking sweep over the embedded loop. This works
		1513	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
		1514	apropriate way for embedded loops.
		1515
		1516	=item struct ev_loop *loop [read-only]
		1517
		1518	The embedded event loop.
		1519
		1520	=back
		1521
		1522
		1523	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		1524
		1525	Fork watchers are called when a C<fork ()> was detected (usually because
		1526	whoever is a good citizen cared to tell libev about it by calling
		1527	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		1528	event loop blocks next and before C<ev_check> watchers are being called,
		1529	and only in the child after the fork. If whoever good citizen calling
		1530	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		1531	handlers will be invoked, too, of course.
		1532
		1533	=over 4
		1534
		1535	=item ev_fork_init (ev_signal *, callback)
		1536
		1537	Initialises and configures the fork watcher - it has no parameters of any
		1538	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		1539	believe me.
		1540
		1541	=back
		1542
		1543
793	=head1 OTHER FUNCTIONS	1544	=head1 OTHER FUNCTIONS
794		1545
795	There are some other functions of possible interest. Described. Here. Now.	1546	There are some other functions of possible interest. Described. Here. Now.
796		1547
797	=over 4	1548	=over 4
…		…
826	/* stdin might have data for us, joy! */;	1577	/* stdin might have data for us, joy! */;
827	}	1578	}
828		1579
829	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	1580	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
830		1581
831	=item ev_feed_event (loop, watcher, int events)	1582	=item ev_feed_event (ev_loop *, watcher *, int revents)
832		1583
833	Feeds the given event set into the event loop, as if the specified event	1584	Feeds the given event set into the event loop, as if the specified event
834	had happened for the specified watcher (which must be a pointer to an	1585	had happened for the specified watcher (which must be a pointer to an
835	initialised but not necessarily started event watcher).	1586	initialised but not necessarily started event watcher).
836		1587
837	=item ev_feed_fd_event (loop, int fd, int revents)	1588	=item ev_feed_fd_event (ev_loop *, int fd, int revents)
838		1589
839	Feed an event on the given fd, as if a file descriptor backend detected	1590	Feed an event on the given fd, as if a file descriptor backend detected
840	the given events it.	1591	the given events it.
841		1592
842	=item ev_feed_signal_event (loop, int signum)	1593	=item ev_feed_signal_event (ev_loop *loop, int signum)
843		1594
844	Feed an event as if the given signal occured (loop must be the default loop!).	1595	Feed an event as if the given signal occured (C<loop> must be the default
		1596	loop!).
845		1597
846	=back	1598	=back
		1599
847		1600
848	=head1 LIBEVENT EMULATION	1601	=head1 LIBEVENT EMULATION
849		1602
850	Libev offers a compatibility emulation layer for libevent. It cannot	1603	Libev offers a compatibility emulation layer for libevent. It cannot
851	emulate the internals of libevent, so here are some usage hints:	1604	emulate the internals of libevent, so here are some usage hints:
…		…
872		1625
873	=back	1626	=back
874		1627
875	=head1 C++ SUPPORT	1628	=head1 C++ SUPPORT
876		1629
877	TBD.	1630	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		1631	you to use some convinience methods to start/stop watchers and also change
		1632	the callback model to a model using method callbacks on objects.
		1633
		1634	To use it,
		1635
		1636	#include <ev++.h>
		1637
		1638	(it is not installed by default). This automatically includes F<ev.h>
		1639	and puts all of its definitions (many of them macros) into the global
		1640	namespace. All C++ specific things are put into the C<ev> namespace.
		1641
		1642	It should support all the same embedding options as F<ev.h>, most notably
		1643	C<EV_MULTIPLICITY>.
		1644
		1645	Here is a list of things available in the C<ev> namespace:
		1646
		1647	=over 4
		1648
		1649	=item C<ev::READ>, C<ev::WRITE> etc.
		1650
		1651	These are just enum values with the same values as the C<EV_READ> etc.
		1652	macros from F<ev.h>.
		1653
		1654	=item C<ev::tstamp>, C<ev::now>
		1655
		1656	Aliases to the same types/functions as with the C<ev_> prefix.
		1657
		1658	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		1659
		1660	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		1661	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		1662	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		1663	defines by many implementations.
		1664
		1665	All of those classes have these methods:
		1666
		1667	=over 4
		1668
		1669	=item ev::TYPE::TYPE (object *, object::method *)
		1670
		1671	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)
		1672
		1673	=item ev::TYPE::~TYPE
		1674
		1675	The constructor takes a pointer to an object and a method pointer to
		1676	the event handler callback to call in this class. The constructor calls
		1677	C<ev_init> for you, which means you have to call the C<set> method
		1678	before starting it. If you do not specify a loop then the constructor
		1679	automatically associates the default loop with this watcher.
		1680
		1681	The destructor automatically stops the watcher if it is active.
		1682
		1683	=item w->set (struct ev_loop *)
		1684
		1685	Associates a different C<struct ev_loop> with this watcher. You can only
		1686	do this when the watcher is inactive (and not pending either).
		1687
		1688	=item w->set ([args])
		1689
		1690	Basically the same as C<ev_TYPE_set>, with the same args. Must be
		1691	called at least once. Unlike the C counterpart, an active watcher gets
		1692	automatically stopped and restarted.
		1693
		1694	=item w->start ()
		1695
		1696	Starts the watcher. Note that there is no C<loop> argument as the
		1697	constructor already takes the loop.
		1698
		1699	=item w->stop ()
		1700
		1701	Stops the watcher if it is active. Again, no C<loop> argument.
		1702
		1703	=item w->again () C<ev::timer>, C<ev::periodic> only
		1704
		1705	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		1706	C<ev_TYPE_again> function.
		1707
		1708	=item w->sweep () C<ev::embed> only
		1709
		1710	Invokes C<ev_embed_sweep>.
		1711
		1712	=item w->update () C<ev::stat> only
		1713
		1714	Invokes C<ev_stat_stat>.
		1715
		1716	=back
		1717
		1718	=back
		1719
		1720	Example: Define a class with an IO and idle watcher, start one of them in
		1721	the constructor.
		1722
		1723	class myclass
		1724	{
		1725	ev_io io; void io_cb (ev::io &w, int revents);
		1726	ev_idle idle void idle_cb (ev::idle &w, int revents);
		1727
		1728	myclass ();
		1729	}
		1730
		1731	myclass::myclass (int fd)
		1732	: io (this, &myclass::io_cb),
		1733	idle (this, &myclass::idle_cb)
		1734	{
		1735	io.start (fd, ev::READ);
		1736	}
		1737
		1738
		1739	=head1 MACRO MAGIC
		1740
		1741	Libev can be compiled with a variety of options, the most fundemantal is
		1742	C<EV_MULTIPLICITY>. This option determines wether (most) functions and
		1743	callbacks have an initial C<struct ev_loop *> argument.
		1744
		1745	To make it easier to write programs that cope with either variant, the
		1746	following macros are defined:
		1747
		1748	=over 4
		1749
		1750	=item C<EV_A>, C<EV_A_>
		1751
		1752	This provides the loop I<argument> for functions, if one is required ("ev
		1753	loop argument"). The C<EV_A> form is used when this is the sole argument,
		1754	C<EV_A_> is used when other arguments are following. Example:
		1755
		1756	ev_unref (EV_A);
		1757	ev_timer_add (EV_A_ watcher);
		1758	ev_loop (EV_A_ 0);
		1759
		1760	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		1761	which is often provided by the following macro.
		1762
		1763	=item C<EV_P>, C<EV_P_>
		1764
		1765	This provides the loop I<parameter> for functions, if one is required ("ev
		1766	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		1767	C<EV_P_> is used when other parameters are following. Example:
		1768
		1769	// this is how ev_unref is being declared
		1770	static void ev_unref (EV_P);
		1771
		1772	// this is how you can declare your typical callback
		1773	static void cb (EV_P_ ev_timer *w, int revents)
		1774
		1775	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		1776	suitable for use with C<EV_A>.
		1777
		1778	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		1779
		1780	Similar to the other two macros, this gives you the value of the default
		1781	loop, if multiple loops are supported ("ev loop default").
		1782
		1783	=back
		1784
		1785	Example: Declare and initialise a check watcher, working regardless of
		1786	wether multiple loops are supported or not.
		1787
		1788	static void
		1789	check_cb (EV_P_ ev_timer *w, int revents)
		1790	{
		1791	ev_check_stop (EV_A_ w);
		1792	}
		1793
		1794	ev_check check;
		1795	ev_check_init (&check, check_cb);
		1796	ev_check_start (EV_DEFAULT_ &check);
		1797	ev_loop (EV_DEFAULT_ 0);
		1798
		1799
		1800	=head1 EMBEDDING
		1801
		1802	Libev can (and often is) directly embedded into host
		1803	applications. Examples of applications that embed it include the Deliantra
		1804	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		1805	and rxvt-unicode.
		1806
		1807	The goal is to enable you to just copy the neecssary files into your
		1808	source directory without having to change even a single line in them, so
		1809	you can easily upgrade by simply copying (or having a checked-out copy of
		1810	libev somewhere in your source tree).
		1811
		1812	=head2 FILESETS
		1813
		1814	Depending on what features you need you need to include one or more sets of files
		1815	in your app.
		1816
		1817	=head3 CORE EVENT LOOP
		1818
		1819	To include only the libev core (all the C<ev_*> functions), with manual
		1820	configuration (no autoconf):
		1821
		1822	#define EV_STANDALONE 1
		1823	#include "ev.c"
		1824
		1825	This will automatically include F<ev.h>, too, and should be done in a
		1826	single C source file only to provide the function implementations. To use
		1827	it, do the same for F<ev.h> in all files wishing to use this API (best
		1828	done by writing a wrapper around F<ev.h> that you can include instead and
		1829	where you can put other configuration options):
		1830
		1831	#define EV_STANDALONE 1
		1832	#include "ev.h"
		1833
		1834	Both header files and implementation files can be compiled with a C++
		1835	compiler (at least, thats a stated goal, and breakage will be treated
		1836	as a bug).
		1837
		1838	You need the following files in your source tree, or in a directory
		1839	in your include path (e.g. in libev/ when using -Ilibev):
		1840
		1841	ev.h
		1842	ev.c
		1843	ev_vars.h
		1844	ev_wrap.h
		1845
		1846	ev_win32.c required on win32 platforms only
		1847
		1848	ev_select.c only when select backend is enabled (which is by default)
		1849	ev_poll.c only when poll backend is enabled (disabled by default)
		1850	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		1851	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		1852	ev_port.c only when the solaris port backend is enabled (disabled by default)
		1853
		1854	F<ev.c> includes the backend files directly when enabled, so you only need
		1855	to compile this single file.
		1856
		1857	=head3 LIBEVENT COMPATIBILITY API
		1858
		1859	To include the libevent compatibility API, also include:
		1860
		1861	#include "event.c"
		1862
		1863	in the file including F<ev.c>, and:
		1864
		1865	#include "event.h"
		1866
		1867	in the files that want to use the libevent API. This also includes F<ev.h>.
		1868
		1869	You need the following additional files for this:
		1870
		1871	event.h
		1872	event.c
		1873
		1874	=head3 AUTOCONF SUPPORT
		1875
		1876	Instead of using C<EV_STANDALONE=1> and providing your config in
		1877	whatever way you want, you can also C<m4_include([libev.m4])> in your
		1878	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		1879	include F<config.h> and configure itself accordingly.
		1880
		1881	For this of course you need the m4 file:
		1882
		1883	libev.m4
		1884
		1885	=head2 PREPROCESSOR SYMBOLS/MACROS
		1886
		1887	Libev can be configured via a variety of preprocessor symbols you have to define
		1888	before including any of its files. The default is not to build for multiplicity
		1889	and only include the select backend.
		1890
		1891	=over 4
		1892
		1893	=item EV_STANDALONE
		1894
		1895	Must always be C<1> if you do not use autoconf configuration, which
		1896	keeps libev from including F<config.h>, and it also defines dummy
		1897	implementations for some libevent functions (such as logging, which is not
		1898	supported). It will also not define any of the structs usually found in
		1899	F<event.h> that are not directly supported by the libev core alone.
		1900
		1901	=item EV_USE_MONOTONIC
		1902
		1903	If defined to be C<1>, libev will try to detect the availability of the
		1904	monotonic clock option at both compiletime and runtime. Otherwise no use
		1905	of the monotonic clock option will be attempted. If you enable this, you
		1906	usually have to link against librt or something similar. Enabling it when
		1907	the functionality isn't available is safe, though, althoguh you have
		1908	to make sure you link against any libraries where the C<clock_gettime>
		1909	function is hiding in (often F<-lrt>).
		1910
		1911	=item EV_USE_REALTIME
		1912
		1913	If defined to be C<1>, libev will try to detect the availability of the
		1914	realtime clock option at compiletime (and assume its availability at
		1915	runtime if successful). Otherwise no use of the realtime clock option will
		1916	be attempted. This effectively replaces C<gettimeofday> by C<clock_get
		1917	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries
		1918	in the description of C<EV_USE_MONOTONIC>, though.
		1919
		1920	=item EV_USE_SELECT
		1921
		1922	If undefined or defined to be C<1>, libev will compile in support for the
		1923	C<select>(2) backend. No attempt at autodetection will be done: if no
		1924	other method takes over, select will be it. Otherwise the select backend
		1925	will not be compiled in.
		1926
		1927	=item EV_SELECT_USE_FD_SET
		1928
		1929	If defined to C<1>, then the select backend will use the system C<fd_set>
		1930	structure. This is useful if libev doesn't compile due to a missing
		1931	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on
		1932	exotic systems. This usually limits the range of file descriptors to some
		1933	low limit such as 1024 or might have other limitations (winsocket only
		1934	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might
		1935	influence the size of the C<fd_set> used.
		1936
		1937	=item EV_SELECT_IS_WINSOCKET
		1938
		1939	When defined to C<1>, the select backend will assume that
		1940	select/socket/connect etc. don't understand file descriptors but
		1941	wants osf handles on win32 (this is the case when the select to
		1942	be used is the winsock select). This means that it will call
		1943	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		1944	it is assumed that all these functions actually work on fds, even
		1945	on win32. Should not be defined on non-win32 platforms.
		1946
		1947	=item EV_USE_POLL
		1948
		1949	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		1950	backend. Otherwise it will be enabled on non-win32 platforms. It
		1951	takes precedence over select.
		1952
		1953	=item EV_USE_EPOLL
		1954
		1955	If defined to be C<1>, libev will compile in support for the Linux
		1956	C<epoll>(7) backend. Its availability will be detected at runtime,
		1957	otherwise another method will be used as fallback. This is the
		1958	preferred backend for GNU/Linux systems.
		1959
		1960	=item EV_USE_KQUEUE
		1961
		1962	If defined to be C<1>, libev will compile in support for the BSD style
		1963	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		1964	otherwise another method will be used as fallback. This is the preferred
		1965	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		1966	supports some types of fds correctly (the only platform we found that
		1967	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		1968	not be used unless explicitly requested. The best way to use it is to find
		1969	out whether kqueue supports your type of fd properly and use an embedded
		1970	kqueue loop.
		1971
		1972	=item EV_USE_PORT
		1973
		1974	If defined to be C<1>, libev will compile in support for the Solaris
		1975	10 port style backend. Its availability will be detected at runtime,
		1976	otherwise another method will be used as fallback. This is the preferred
		1977	backend for Solaris 10 systems.
		1978
		1979	=item EV_USE_DEVPOLL
		1980
		1981	reserved for future expansion, works like the USE symbols above.
		1982
		1983	=item EV_H
		1984
		1985	The name of the F<ev.h> header file used to include it. The default if
		1986	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This
		1987	can be used to virtually rename the F<ev.h> header file in case of conflicts.
		1988
		1989	=item EV_CONFIG_H
		1990
		1991	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		1992	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		1993	C<EV_H>, above.
		1994
		1995	=item EV_EVENT_H
		1996
		1997	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		1998	of how the F<event.h> header can be found.
		1999
		2000	=item EV_PROTOTYPES
		2001
		2002	If defined to be C<0>, then F<ev.h> will not define any function
		2003	prototypes, but still define all the structs and other symbols. This is
		2004	occasionally useful if you want to provide your own wrapper functions
		2005	around libev functions.
		2006
		2007	=item EV_MULTIPLICITY
		2008
		2009	If undefined or defined to C<1>, then all event-loop-specific functions
		2010	will have the C<struct ev_loop *> as first argument, and you can create
		2011	additional independent event loops. Otherwise there will be no support
		2012	for multiple event loops and there is no first event loop pointer
		2013	argument. Instead, all functions act on the single default loop.
		2014
		2015	=item EV_PERIODIC_ENABLE
		2016
		2017	If undefined or defined to be C<1>, then periodic timers are supported. If
		2018	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2019	code.
		2020
		2021	=item EV_EMBED_ENABLE
		2022
		2023	If undefined or defined to be C<1>, then embed watchers are supported. If
		2024	defined to be C<0>, then they are not.
		2025
		2026	=item EV_STAT_ENABLE
		2027
		2028	If undefined or defined to be C<1>, then stat watchers are supported. If
		2029	defined to be C<0>, then they are not.
		2030
		2031	=item EV_FORK_ENABLE
		2032
		2033	If undefined or defined to be C<1>, then fork watchers are supported. If
		2034	defined to be C<0>, then they are not.
		2035
		2036	=item EV_MINIMAL
		2037
		2038	If you need to shave off some kilobytes of code at the expense of some
		2039	speed, define this symbol to C<1>. Currently only used for gcc to override
		2040	some inlining decisions, saves roughly 30% codesize of amd64.
		2041
		2042	=item EV_PID_HASHSIZE
		2043
		2044	C<ev_child> watchers use a small hash table to distribute workload by
		2045	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
		2046	than enough. If you need to manage thousands of children you might want to
		2047	increase this value.
		2048
		2049	=item EV_COMMON
		2050
		2051	By default, all watchers have a C<void *data> member. By redefining
		2052	this macro to a something else you can include more and other types of
		2053	members. You have to define it each time you include one of the files,
		2054	though, and it must be identical each time.
		2055
		2056	For example, the perl EV module uses something like this:
		2057
		2058	#define EV_COMMON \
		2059	SV *self; /* contains this struct */ \
		2060	SV *cb_sv, *fh /* note no trailing ";" */
		2061
		2062	=item EV_CB_DECLARE (type)
		2063
		2064	=item EV_CB_INVOKE (watcher, revents)
		2065
		2066	=item ev_set_cb (ev, cb)
		2067
		2068	Can be used to change the callback member declaration in each watcher,
		2069	and the way callbacks are invoked and set. Must expand to a struct member
		2070	definition and a statement, respectively. See the F<ev.v> header file for
		2071	their default definitions. One possible use for overriding these is to
		2072	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		2073	method calls instead of plain function calls in C++.
		2074
		2075	=head2 EXAMPLES
		2076
		2077	For a real-world example of a program the includes libev
		2078	verbatim, you can have a look at the EV perl module
		2079	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		2080	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		2081	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		2082	will be compiled. It is pretty complex because it provides its own header
		2083	file.
		2084
		2085	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		2086	that everybody includes and which overrides some autoconf choices:
		2087
		2088	#define EV_USE_POLL 0
		2089	#define EV_MULTIPLICITY 0
		2090	#define EV_PERIODICS 0
		2091	#define EV_CONFIG_H <config.h>
		2092
		2093	#include "ev++.h"
		2094
		2095	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		2096
		2097	#include "ev_cpp.h"
		2098	#include "ev.c"
		2099
		2100
		2101	=head1 COMPLEXITIES
		2102
		2103	In this section the complexities of (many of) the algorithms used inside
		2104	libev will be explained. For complexity discussions about backends see the
		2105	documentation for C<ev_default_init>.
		2106
		2107	=over 4
		2108
		2109	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		2110
		2111	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
		2112
		2113	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
		2114
		2115	=item Stopping check/prepare/idle watchers: O(1)
		2116
		2117	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))
		2118
		2119	=item Finding the next timer per loop iteration: O(1)
		2120
		2121	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		2122
		2123	=item Activating one watcher: O(1)
		2124
		2125	=back
		2126
878		2127
879	=head1 AUTHOR	2128	=head1 AUTHOR
880		2129
881	Marc Lehmann <libev@schmorp.de>.	2130	Marc Lehmann <libev@schmorp.de>.
882		2131

Diff Legend

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