[ViewVC] Diff of: cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.18 by root, Mon Nov 12 09:01:12 2007 UTC vs.
Revision 1.34 by root, Fri Nov 23 16:17:12 2007 UTC

…		…
45		45
46	Libev represents time as a single floating point number, representing the	46	Libev represents time as a single floating point number, representing the
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	47	(fractional) number of seconds since the (POSIX) epoch (somewhere near
48	the beginning of 1970, details are complicated, don't ask). This type is	48	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	49	called C<ev_tstamp>, which is what you should use too. It usually aliases
50	to the double type in C.	50	to the C<double> type in C, and when you need to do any calculations on
		51	it, you should treat it as such.
		52
51		53
52	=head1 GLOBAL FUNCTIONS	54	=head1 GLOBAL FUNCTIONS
53		55
54	These functions can be called anytime, even before initialising the	56	These functions can be called anytime, even before initialising the
55	library in any way.	57	library in any way.
56		58
57	=over 4	59	=over 4
58		60
59	=item ev_tstamp ev_time ()	61	=item ev_tstamp ev_time ()
60		62
61	Returns the current time as libev would use it.	63	Returns the current time as libev would use it. Please note that the
		64	C<ev_now> function is usually faster and also often returns the timestamp
		65	you actually want to know.
62		66
63	=item int ev_version_major ()	67	=item int ev_version_major ()
64		68
65	=item int ev_version_minor ()	69	=item int ev_version_minor ()
66		70
…		…
73	Usually, it's a good idea to terminate if the major versions mismatch,	77	Usually, it's a good idea to terminate if the major versions mismatch,
74	as this indicates an incompatible change. Minor versions are usually	78	as this indicates an incompatible change. Minor versions are usually
75	compatible to older versions, so a larger minor version alone is usually	79	compatible to older versions, so a larger minor version alone is usually
76	not a problem.	80	not a problem.
77		81
		82	Example: make sure we haven't accidentally been linked against the wrong
		83	version:
		84
		85	assert (("libev version mismatch",
		86	ev_version_major () == EV_VERSION_MAJOR
		87	&& ev_version_minor () >= EV_VERSION_MINOR));
		88
		89	=item unsigned int ev_supported_backends ()
		90
		91	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		92	value) compiled into this binary of libev (independent of their
		93	availability on the system you are running on). See C<ev_default_loop> for
		94	a description of the set values.
		95
		96	Example: make sure we have the epoll method, because yeah this is cool and
		97	a must have and can we have a torrent of it please!!!11
		98
		99	assert (("sorry, no epoll, no sex",
		100	ev_supported_backends () & EVBACKEND_EPOLL));
		101
		102	=item unsigned int ev_recommended_backends ()
		103
		104	Return the set of all backends compiled into this binary of libev and also
		105	recommended for this platform. This set is often smaller than the one
		106	returned by C<ev_supported_backends>, as for example kqueue is broken on
		107	most BSDs and will not be autodetected unless you explicitly request it
		108	(assuming you know what you are doing). This is the set of backends that
		109	libev will probe for if you specify no backends explicitly.
		110
78	=item ev_set_allocator (void (cb)(void *ptr, long size))	111	=item ev_set_allocator (void (cb)(void *ptr, long size))
79		112
80	Sets the allocation function to use (the prototype is similar to the	113	Sets the allocation function to use (the prototype is similar to the
81	realloc C function, the semantics are identical). It is used to allocate	114	realloc C function, the semantics are identical). It is used to allocate
82	and free memory (no surprises here). If it returns zero when memory	115	and free memory (no surprises here). If it returns zero when memory
…		…
84	destructive action. The default is your system realloc function.	117	destructive action. The default is your system realloc function.
85		118
86	You could override this function in high-availability programs to, say,	119	You could override this function in high-availability programs to, say,
87	free some memory if it cannot allocate memory, to use a special allocator,	120	free some memory if it cannot allocate memory, to use a special allocator,
88	or even to sleep a while and retry until some memory is available.	121	or even to sleep a while and retry until some memory is available.
		122
		123	Example: replace the libev allocator with one that waits a bit and then
		124	retries: better than mine).
		125
		126	static void *
		127	persistent_realloc (void *ptr, long size)
		128	{
		129	for (;;)
		130	{
		131	void *newptr = realloc (ptr, size);
		132
		133	if (newptr)
		134	return newptr;
		135
		136	sleep (60);
		137	}
		138	}
		139
		140	...
		141	ev_set_allocator (persistent_realloc);
89		142
90	=item ev_set_syserr_cb (void (cb)(const char msg));	143	=item ev_set_syserr_cb (void (cb)(const char msg));
91		144
92	Set the callback function to call on a retryable syscall error (such	145	Set the callback function to call on a retryable syscall error (such
93	as failed select, poll, epoll_wait). The message is a printable string	146	as failed select, poll, epoll_wait). The message is a printable string
…		…
95	callback is set, then libev will expect it to remedy the sitution, no	148	callback is set, then libev will expect it to remedy the sitution, no
96	matter what, when it returns. That is, libev will generally retry the	149	matter what, when it returns. That is, libev will generally retry the
97	requested operation, or, if the condition doesn't go away, do bad stuff	150	requested operation, or, if the condition doesn't go away, do bad stuff
98	(such as abort).	151	(such as abort).
99		152
		153	Example: do the same thing as libev does internally:
		154
		155	static void
		156	fatal_error (const char *msg)
		157	{
		158	perror (msg);
		159	abort ();
		160	}
		161
		162	...
		163	ev_set_syserr_cb (fatal_error);
		164
100	=back	165	=back
101		166
102	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	167	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
103		168
104	An event loop is described by a C<struct ev_loop *>. The library knows two	169	An event loop is described by a C<struct ev_loop *>. The library knows two
…		…
117	=item struct ev_loop *ev_default_loop (unsigned int flags)	182	=item struct ev_loop *ev_default_loop (unsigned int flags)
118		183
119	This will initialise the default event loop if it hasn't been initialised	184	This will initialise the default event loop if it hasn't been initialised
120	yet and return it. If the default loop could not be initialised, returns	185	yet and return it. If the default loop could not be initialised, returns
121	false. If it already was initialised it simply returns it (and ignores the	186	false. If it already was initialised it simply returns it (and ignores the
122	flags).	187	flags. If that is troubling you, check C<ev_backend ()> afterwards).
123		188
124	If you don't know what event loop to use, use the one returned from this	189	If you don't know what event loop to use, use the one returned from this
125	function.	190	function.
126		191
127	The flags argument can be used to specify special behaviour or specific	192	The flags argument can be used to specify special behaviour or specific
128	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	193	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
129		194
130	It supports the following flags:	195	The following flags are supported:
131		196
132	=over 4	197	=over 4
133		198
134	=item C<EVFLAG_AUTO>	199	=item C<EVFLAG_AUTO>
135		200
…		…
143	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	208	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
144	override the flags completely if it is found in the environment. This is	209	override the flags completely if it is found in the environment. This is
145	useful to try out specific backends to test their performance, or to work	210	useful to try out specific backends to test their performance, or to work
146	around bugs.	211	around bugs.
147		212
148	=item C<EVMETHOD_SELECT> (portable select backend)	213	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
149		214
		215	This is your standard select(2) backend. Not I<completely> standard, as
		216	libev tries to roll its own fd_set with no limits on the number of fds,
		217	but if that fails, expect a fairly low limit on the number of fds when
		218	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
		219	the fastest backend for a low number of fds.
		220
150	=item C<EVMETHOD_POLL> (poll backend, available everywhere except on windows)	221	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
151		222
152	=item C<EVMETHOD_EPOLL> (linux only)	223	And this is your standard poll(2) backend. It's more complicated than
		224	select, but handles sparse fds better and has no artificial limit on the
		225	number of fds you can use (except it will slow down considerably with a
		226	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
153		227
154	=item C<EVMETHOD_KQUEUE> (some bsds only)	228	=item C<EVBACKEND_EPOLL> (value 4, Linux)
155		229
156	=item C<EVMETHOD_DEVPOLL> (solaris 8 only)	230	For few fds, this backend is a bit little slower than poll and select,
		231	but it scales phenomenally better. While poll and select usually scale like
		232	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
		233	either O(1) or O(active_fds).
157		234
158	=item C<EVMETHOD_PORT> (solaris 10 only)	235	While stopping and starting an I/O watcher in the same iteration will
		236	result in some caching, there is still a syscall per such incident
		237	(because the fd could point to a different file description now), so its
		238	best to avoid that. Also, dup()ed file descriptors might not work very
		239	well if you register events for both fds.
		240
		241	Please note that epoll sometimes generates spurious notifications, so you
		242	need to use non-blocking I/O or other means to avoid blocking when no data
		243	(or space) is available.
		244
		245	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
		246
		247	Kqueue deserves special mention, as at the time of this writing, it
		248	was broken on all BSDs except NetBSD (usually it doesn't work with
		249	anything but sockets and pipes, except on Darwin, where of course its
		250	completely useless). For this reason its not being "autodetected"
		251	unless you explicitly specify it explicitly in the flags (i.e. using
		252	C<EVBACKEND_KQUEUE>).
		253
		254	It scales in the same way as the epoll backend, but the interface to the
		255	kernel is more efficient (which says nothing about its actual speed, of
		256	course). While starting and stopping an I/O watcher does not cause an
		257	extra syscall as with epoll, it still adds up to four event changes per
		258	incident, so its best to avoid that.
		259
		260	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
		261
		262	This is not implemented yet (and might never be).
		263
		264	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
		265
		266	This uses the Solaris 10 port mechanism. As with everything on Solaris,
		267	it's really slow, but it still scales very well (O(active_fds)).
		268
		269	Please note that solaris ports can result in a lot of spurious
		270	notifications, so you need to use non-blocking I/O or other means to avoid
		271	blocking when no data (or space) is available.
		272
		273	=item C<EVBACKEND_ALL>
		274
		275	Try all backends (even potentially broken ones that wouldn't be tried
		276	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
		277	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
		278
		279	=back
159		280
160	If one or more of these are ored into the flags value, then only these	281	If one or more of these are ored into the flags value, then only these
161	backends will be tried (in the reverse order as given here). If one are	282	backends will be tried (in the reverse order as given here). If none are
162	specified, any backend will do.	283	specified, most compiled-in backend will be tried, usually in reverse
		284	order of their flag values :)
163		285
164	=back	286	The most typical usage is like this:
		287
		288	if (!ev_default_loop (0))
		289	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		290
		291	Restrict libev to the select and poll backends, and do not allow
		292	environment settings to be taken into account:
		293
		294	ev_default_loop (EVBACKEND_POLL \| EVBACKEND_SELECT \| EVFLAG_NOENV);
		295
		296	Use whatever libev has to offer, but make sure that kqueue is used if
		297	available (warning, breaks stuff, best use only with your own private
		298	event loop and only if you know the OS supports your types of fds):
		299
		300	ev_default_loop (ev_recommended_backends () \| EVBACKEND_KQUEUE);
165		301
166	=item struct ev_loop *ev_loop_new (unsigned int flags)	302	=item struct ev_loop *ev_loop_new (unsigned int flags)
167		303
168	Similar to C<ev_default_loop>, but always creates a new event loop that is	304	Similar to C<ev_default_loop>, but always creates a new event loop that is
169	always distinct from the default loop. Unlike the default loop, it cannot	305	always distinct from the default loop. Unlike the default loop, it cannot
170	handle signal and child watchers, and attempts to do so will be greeted by	306	handle signal and child watchers, and attempts to do so will be greeted by
171	undefined behaviour (or a failed assertion if assertions are enabled).	307	undefined behaviour (or a failed assertion if assertions are enabled).
172		308
		309	Example: try to create a event loop that uses epoll and nothing else.
		310
		311	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);
		312	if (!epoller)
		313	fatal ("no epoll found here, maybe it hides under your chair");
		314
173	=item ev_default_destroy ()	315	=item ev_default_destroy ()
174		316
175	Destroys the default loop again (frees all memory and kernel state	317	Destroys the default loop again (frees all memory and kernel state
176	etc.). This stops all registered event watchers (by not touching them in	318	etc.). This stops all registered event watchers (by not touching them in
177	any way whatsoever, although you cannot rely on this :).	319	any way whatsoever, although you cannot rely on this :).
…		…
186	This function reinitialises the kernel state for backends that have	328	This function reinitialises the kernel state for backends that have
187	one. Despite the name, you can call it anytime, but it makes most sense	329	one. Despite the name, you can call it anytime, but it makes most sense
188	after forking, in either the parent or child process (or both, but that	330	after forking, in either the parent or child process (or both, but that
189	again makes little sense).	331	again makes little sense).
190		332
191	You I<must> call this function after forking if and only if you want to	333	You I<must> call this function in the child process after forking if and
192	use the event library in both processes. If you just fork+exec, you don't	334	only if you want to use the event library in both processes. If you just
193	have to call it.	335	fork+exec, you don't have to call it.
194		336
195	The function itself is quite fast and it's usually not a problem to call	337	The function itself is quite fast and it's usually not a problem to call
196	it just in case after a fork. To make this easy, the function will fit in	338	it just in case after a fork. To make this easy, the function will fit in
197	quite nicely into a call to C<pthread_atfork>:	339	quite nicely into a call to C<pthread_atfork>:
198		340
199	pthread_atfork (0, 0, ev_default_fork);	341	pthread_atfork (0, 0, ev_default_fork);
200		342
		343	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use
		344	without calling this function, so if you force one of those backends you
		345	do not need to care.
		346
201	=item ev_loop_fork (loop)	347	=item ev_loop_fork (loop)
202		348
203	Like C<ev_default_fork>, but acts on an event loop created by	349	Like C<ev_default_fork>, but acts on an event loop created by
204	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	350	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
205	after fork, and how you do this is entirely your own problem.	351	after fork, and how you do this is entirely your own problem.
206		352
207	=item unsigned int ev_method (loop)	353	=item unsigned int ev_backend (loop)
208		354
209	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	355	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
210	use.	356	use.
211		357
212	=item ev_tstamp ev_now (loop)	358	=item ev_tstamp ev_now (loop)
213		359
214	Returns the current "event loop time", which is the time the event loop	360	Returns the current "event loop time", which is the time the event loop
215	got events and started processing them. This timestamp does not change	361	received events and started processing them. This timestamp does not
216	as long as callbacks are being processed, and this is also the base time	362	change as long as callbacks are being processed, and this is also the base
217	used for relative timers. You can treat it as the timestamp of the event	363	time used for relative timers. You can treat it as the timestamp of the
218	occuring (or more correctly, the mainloop finding out about it).	364	event occuring (or more correctly, libev finding out about it).
219		365
220	=item ev_loop (loop, int flags)	366	=item ev_loop (loop, int flags)
221		367
222	Finally, this is it, the event handler. This function usually is called	368	Finally, this is it, the event handler. This function usually is called
223	after you initialised all your watchers and you want to start handling	369	after you initialised all your watchers and you want to start handling
224	events.	370	events.
225		371
226	If the flags argument is specified as 0, it will not return until either	372	If the flags argument is specified as C<0>, it will not return until
227	no event watchers are active anymore or C<ev_unloop> was called.	373	either no event watchers are active anymore or C<ev_unloop> was called.
		374
		375	Please note that an explicit C<ev_unloop> is usually better than
		376	relying on all watchers to be stopped when deciding when a program has
		377	finished (especially in interactive programs), but having a program that
		378	automatically loops as long as it has to and no longer by virtue of
		379	relying on its watchers stopping correctly is a thing of beauty.
228		380
229	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	381	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
230	those events and any outstanding ones, but will not block your process in	382	those events and any outstanding ones, but will not block your process in
231	case there are no events and will return after one iteration of the loop.	383	case there are no events and will return after one iteration of the loop.
232		384
233	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	385	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
234	neccessary) and will handle those and any outstanding ones. It will block	386	neccessary) and will handle those and any outstanding ones. It will block
235	your process until at least one new event arrives, and will return after	387	your process until at least one new event arrives, and will return after
236	one iteration of the loop.	388	one iteration of the loop. This is useful if you are waiting for some
		389	external event in conjunction with something not expressible using other
		390	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is
		391	usually a better approach for this kind of thing.
237		392
238	This flags value could be used to implement alternative looping	393	Here are the gory details of what C<ev_loop> does:
239	constructs, but the C<prepare> and C<check> watchers provide a better and	394
240	more generic mechanism.	395	* If there are no active watchers (reference count is zero), return.
		396	- Queue prepare watchers and then call all outstanding watchers.
		397	- If we have been forked, recreate the kernel state.
		398	- Update the kernel state with all outstanding changes.
		399	- Update the "event loop time".
		400	- Calculate for how long to block.
		401	- Block the process, waiting for any events.
		402	- Queue all outstanding I/O (fd) events.
		403	- Update the "event loop time" and do time jump handling.
		404	- Queue all outstanding timers.
		405	- Queue all outstanding periodics.
		406	- If no events are pending now, queue all idle watchers.
		407	- Queue all check watchers.
		408	- Call all queued watchers in reverse order (i.e. check watchers first).
		409	Signals and child watchers are implemented as I/O watchers, and will
		410	be handled here by queueing them when their watcher gets executed.
		411	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
		412	were used, return, otherwise continue with step *.
		413
		414	Example: queue some jobs and then loop until no events are outsanding
		415	anymore.
		416
		417	... queue jobs here, make sure they register event watchers as long
		418	... as they still have work to do (even an idle watcher will do..)
		419	ev_loop (my_loop, 0);
		420	... jobs done. yeah!
241		421
242	=item ev_unloop (loop, how)	422	=item ev_unloop (loop, how)
243		423
244	Can be used to make a call to C<ev_loop> return early (but only after it	424	Can be used to make a call to C<ev_loop> return early (but only after it
245	has processed all outstanding events). The C<how> argument must be either	425	has processed all outstanding events). The C<how> argument must be either
246	C<EVUNLOOP_ONCE>, which will make the innermost C<ev_loop> call return, or	426	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
247	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	427	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
248		428
249	=item ev_ref (loop)	429	=item ev_ref (loop)
250		430
251	=item ev_unref (loop)	431	=item ev_unref (loop)
…		…
259	visible to the libev user and should not keep C<ev_loop> from exiting if	439	visible to the libev user and should not keep C<ev_loop> from exiting if
260	no event watchers registered by it are active. It is also an excellent	440	no event watchers registered by it are active. It is also an excellent
261	way to do this for generic recurring timers or from within third-party	441	way to do this for generic recurring timers or from within third-party
262	libraries. Just remember to I<unref after start> and I<ref before stop>.	442	libraries. Just remember to I<unref after start> and I<ref before stop>.
263		443
		444	Example: create a signal watcher, but keep it from keeping C<ev_loop>
		445	running when nothing else is active.
		446
		447	struct dv_signal exitsig;
		448	ev_signal_init (&exitsig, sig_cb, SIGINT);
		449	ev_signal_start (myloop, &exitsig);
		450	evf_unref (myloop);
		451
		452	Example: for some weird reason, unregister the above signal handler again.
		453
		454	ev_ref (myloop);
		455	ev_signal_stop (myloop, &exitsig);
		456
264	=back	457	=back
265		458
266	=head1 ANATOMY OF A WATCHER	459	=head1 ANATOMY OF A WATCHER
267		460
268	A watcher is a structure that you create and register to record your	461	A watcher is a structure that you create and register to record your
…		…
302	*) >>), and you can stop watching for events at any time by calling the	495	*) >>), and you can stop watching for events at any time by calling the
303	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	496	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.
304		497
305	As long as your watcher is active (has been started but not stopped) you	498	As long as your watcher is active (has been started but not stopped) you
306	must not touch the values stored in it. Most specifically you must never	499	must not touch the values stored in it. Most specifically you must never
307	reinitialise it or call its set method.	500	reinitialise it or call its set macro.
308		501
309	You can check whether an event is active by calling the C<ev_is_active	502	You can check whether an event is active by calling the C<ev_is_active
310	(watcher *)> macro. To see whether an event is outstanding (but the	503	(watcher *)> macro. To see whether an event is outstanding (but the
311	callback for it has not been called yet) you can use the C<ev_is_pending	504	callback for it has not been called yet) you can use the C<ev_is_pending
312	(watcher *)> macro.	505	(watcher *)> macro.
…		…
409	=head1 WATCHER TYPES	602	=head1 WATCHER TYPES
410		603
411	This section describes each watcher in detail, but will not repeat	604	This section describes each watcher in detail, but will not repeat
412	information given in the last section.	605	information given in the last section.
413		606
		607
414	=head2 C<ev_io> - is this file descriptor readable or writable	608	=head2 C<ev_io> - is this file descriptor readable or writable
415		609
416	I/O watchers check whether a file descriptor is readable or writable	610	I/O watchers check whether a file descriptor is readable or writable
417	in each iteration of the event loop (This behaviour is called	611	in each iteration of the event loop (This behaviour is called
418	level-triggering because you keep receiving events as long as the	612	level-triggering because you keep receiving events as long as the
419	condition persists. Remember you can stop the watcher if you don't want to	613	condition persists. Remember you can stop the watcher if you don't want to
420	act on the event and neither want to receive future events).	614	act on the event and neither want to receive future events).
421		615
422	In general you can register as many read and/or write event watchers oer	616	In general you can register as many read and/or write event watchers per
423	fd as you want (as long as you don't confuse yourself). Setting all file	617	fd as you want (as long as you don't confuse yourself). Setting all file
424	descriptors to non-blocking mode is also usually a good idea (but not	618	descriptors to non-blocking mode is also usually a good idea (but not
425	required if you know what you are doing).	619	required if you know what you are doing).
426		620
427	You have to be careful with dup'ed file descriptors, though. Some backends	621	You have to be careful with dup'ed file descriptors, though. Some backends
428	(the linux epoll backend is a notable example) cannot handle dup'ed file	622	(the linux epoll backend is a notable example) cannot handle dup'ed file
429	descriptors correctly if you register interest in two or more fds pointing	623	descriptors correctly if you register interest in two or more fds pointing
430	to the same file/socket etc. description.	624	to the same underlying file/socket etc. description (that is, they share
		625	the same underlying "file open").
431		626
432	If you must do this, then force the use of a known-to-be-good backend	627	If you must do this, then force the use of a known-to-be-good backend
433	(at the time of this writing, this includes only EVMETHOD_SELECT and	628	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
434	EVMETHOD_POLL).	629	C<EVBACKEND_POLL>).
435		630
436	=over 4	631	=over 4
437		632
438	=item ev_io_init (ev_io *, callback, int fd, int events)	633	=item ev_io_init (ev_io *, callback, int fd, int events)
439		634
…		…
441		636
442	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	637	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive
443	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ \|	638	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ \|
444	EV_WRITE> to receive the given events.	639	EV_WRITE> to receive the given events.
445		640
		641	Please note that most of the more scalable backend mechanisms (for example
		642	epoll and solaris ports) can result in spurious readyness notifications
		643	for file descriptors, so you practically need to use non-blocking I/O (and
		644	treat callback invocation as hint only), or retest separately with a safe
		645	interface before doing I/O (XLib can do this), or force the use of either
		646	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>, which don't suffer from this
		647	problem. Also note that it is quite easy to have your callback invoked
		648	when the readyness condition is no longer valid even when employing
		649	typical ways of handling events, so its a good idea to use non-blocking
		650	I/O unconditionally.
		651
446	=back	652	=back
		653
		654	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well
		655	readable, but only once. Since it is likely line-buffered, you could
		656	attempt to read a whole line in the callback:
		657
		658	static void
		659	stdin_readable_cb (struct ev_loop loop, struct ev_io w, int revents)
		660	{
		661	ev_io_stop (loop, w);
		662	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		663	}
		664
		665	...
		666	struct ev_loop *loop = ev_default_init (0);
		667	struct ev_io stdin_readable;
		668	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		669	ev_io_start (loop, &stdin_readable);
		670	ev_loop (loop, 0);
		671
447		672
448	=head2 C<ev_timer> - relative and optionally recurring timeouts	673	=head2 C<ev_timer> - relative and optionally recurring timeouts
449		674
450	Timer watchers are simple relative timers that generate an event after a	675	Timer watchers are simple relative timers that generate an event after a
451	given time, and optionally repeating in regular intervals after that.	676	given time, and optionally repeating in regular intervals after that.
452		677
453	The timers are based on real time, that is, if you register an event that	678	The timers are based on real time, that is, if you register an event that
454	times out after an hour and youreset your system clock to last years	679	times out after an hour and you reset your system clock to last years
455	time, it will still time out after (roughly) and hour. "Roughly" because	680	time, it will still time out after (roughly) and hour. "Roughly" because
456	detecting time jumps is hard, and soem inaccuracies are unavoidable (the	681	detecting time jumps is hard, and some inaccuracies are unavoidable (the
457	monotonic clock option helps a lot here).	682	monotonic clock option helps a lot here).
458		683
459	The relative timeouts are calculated relative to the C<ev_now ()>	684	The relative timeouts are calculated relative to the C<ev_now ()>
460	time. This is usually the right thing as this timestamp refers to the time	685	time. This is usually the right thing as this timestamp refers to the time
461	of the event triggering whatever timeout you are modifying/starting. If	686	of the event triggering whatever timeout you are modifying/starting. If
462	you suspect event processing to be delayed and you need to base the timeout	687	you suspect event processing to be delayed and you I<need> to base the timeout
463	ion the current time, use something like this to adjust for this:	688	on the current time, use something like this to adjust for this:
464		689
465	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	690	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
		691
		692	The callback is guarenteed to be invoked only when its timeout has passed,
		693	but if multiple timers become ready during the same loop iteration then
		694	order of execution is undefined.
466		695
467	=over 4	696	=over 4
468		697
469	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	698	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
470		699
…		…
476	later, again, and again, until stopped manually.	705	later, again, and again, until stopped manually.
477		706
478	The timer itself will do a best-effort at avoiding drift, that is, if you	707	The timer itself will do a best-effort at avoiding drift, that is, if you
479	configure a timer to trigger every 10 seconds, then it will trigger at	708	configure a timer to trigger every 10 seconds, then it will trigger at
480	exactly 10 second intervals. If, however, your program cannot keep up with	709	exactly 10 second intervals. If, however, your program cannot keep up with
481	the timer (ecause it takes longer than those 10 seconds to do stuff) the	710	the timer (because it takes longer than those 10 seconds to do stuff) the
482	timer will not fire more than once per event loop iteration.	711	timer will not fire more than once per event loop iteration.
483		712
484	=item ev_timer_again (loop)	713	=item ev_timer_again (loop)
485		714
486	This will act as if the timer timed out and restart it again if it is	715	This will act as if the timer timed out and restart it again if it is
…		…
500	state where you do not expect data to travel on the socket, you can stop	729	state where you do not expect data to travel on the socket, you can stop
501	the timer, and again will automatically restart it if need be.	730	the timer, and again will automatically restart it if need be.
502		731
503	=back	732	=back
504		733
		734	Example: create a timer that fires after 60 seconds.
		735
		736	static void
		737	one_minute_cb (struct ev_loop loop, struct ev_timer w, int revents)
		738	{
		739	.. one minute over, w is actually stopped right here
		740	}
		741
		742	struct ev_timer mytimer;
		743	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		744	ev_timer_start (loop, &mytimer);
		745
		746	Example: create a timeout timer that times out after 10 seconds of
		747	inactivity.
		748
		749	static void
		750	timeout_cb (struct ev_loop loop, struct ev_timer w, int revents)
		751	{
		752	.. ten seconds without any activity
		753	}
		754
		755	struct ev_timer mytimer;
		756	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		757	ev_timer_again (&mytimer); /* start timer */
		758	ev_loop (loop, 0);
		759
		760	// and in some piece of code that gets executed on any "activity":
		761	// reset the timeout to start ticking again at 10 seconds
		762	ev_timer_again (&mytimer);
		763
		764
505	=head2 C<ev_periodic> - to cron or not to cron	765	=head2 C<ev_periodic> - to cron or not to cron
506		766
507	Periodic watchers are also timers of a kind, but they are very versatile	767	Periodic watchers are also timers of a kind, but they are very versatile
508	(and unfortunately a bit complex).	768	(and unfortunately a bit complex).
509		769
…		…
517	again).	777	again).
518		778
519	They can also be used to implement vastly more complex timers, such as	779	They can also be used to implement vastly more complex timers, such as
520	triggering an event on eahc midnight, local time.	780	triggering an event on eahc midnight, local time.
521		781
		782	As with timers, the callback is guarenteed to be invoked only when the
		783	time (C<at>) has been passed, but if multiple periodic timers become ready
		784	during the same loop iteration then order of execution is undefined.
		785
522	=over 4	786	=over 4
523		787
524	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	788	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)
525		789
526	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	790	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)
527		791
528	Lots of arguments, lets sort it out... There are basically three modes of	792	Lots of arguments, lets sort it out... There are basically three modes of
529	operation, and we will explain them from simplest to complex:	793	operation, and we will explain them from simplest to complex:
530
531		794
532	=over 4	795	=over 4
533		796
534	=item * absolute timer (interval = reschedule_cb = 0)	797	=item * absolute timer (interval = reschedule_cb = 0)
535		798
…		…
582	(that is, the lowest time value larger than to the second argument). It	845	(that is, the lowest time value larger than to the second argument). It
583	will usually be called just before the callback will be triggered, but	846	will usually be called just before the callback will be triggered, but
584	might be called at other times, too.	847	might be called at other times, too.
585		848
586	NOTE: I<< This callback must always return a time that is later than the	849	NOTE: I<< This callback must always return a time that is later than the
587	passed C<now> value >>. Not even C<now> itself will do, it must be larger.	850	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.
588		851
589	This can be used to create very complex timers, such as a timer that	852	This can be used to create very complex timers, such as a timer that
590	triggers on each midnight, local time. To do this, you would calculate the	853	triggers on each midnight, local time. To do this, you would calculate the
591	next midnight after C<now> and return the timestamp value for this. How you do this	854	next midnight after C<now> and return the timestamp value for this. How
592	is, again, up to you (but it is not trivial).	855	you do this is, again, up to you (but it is not trivial, which is the main
		856	reason I omitted it as an example).
593		857
594	=back	858	=back
595		859
596	=item ev_periodic_again (loop, ev_periodic *)	860	=item ev_periodic_again (loop, ev_periodic *)
597		861
…		…
599	when you changed some parameters or the reschedule callback would return	863	when you changed some parameters or the reschedule callback would return
600	a different time than the last time it was called (e.g. in a crond like	864	a different time than the last time it was called (e.g. in a crond like
601	program when the crontabs have changed).	865	program when the crontabs have changed).
602		866
603	=back	867	=back
		868
		869	Example: call a callback every hour, or, more precisely, whenever the
		870	system clock is divisible by 3600. The callback invocation times have
		871	potentially a lot of jittering, but good long-term stability.
		872
		873	static void
		874	clock_cb (struct ev_loop loop, struct ev_io w, int revents)
		875	{
		876	... its now a full hour (UTC, or TAI or whatever your clock follows)
		877	}
		878
		879	struct ev_periodic hourly_tick;
		880	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		881	ev_periodic_start (loop, &hourly_tick);
		882
		883	Example: the same as above, but use a reschedule callback to do it:
		884
		885	#include <math.h>
		886
		887	static ev_tstamp
		888	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		889	{
		890	return fmod (now, 3600.) + 3600.;
		891	}
		892
		893	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		894
		895	Example: call a callback every hour, starting now:
		896
		897	struct ev_periodic hourly_tick;
		898	ev_periodic_init (&hourly_tick, clock_cb,
		899	fmod (ev_now (loop), 3600.), 3600., 0);
		900	ev_periodic_start (loop, &hourly_tick);
		901
604		902
605	=head2 C<ev_signal> - signal me when a signal gets signalled	903	=head2 C<ev_signal> - signal me when a signal gets signalled
606		904
607	Signal watchers will trigger an event when the process receives a specific	905	Signal watchers will trigger an event when the process receives a specific
608	signal one or more times. Even though signals are very asynchronous, libev	906	signal one or more times. Even though signals are very asynchronous, libev
…		…
644	the status word (use the macros from C<sys/wait.h> and see your systems	942	the status word (use the macros from C<sys/wait.h> and see your systems
645	C<waitpid> documentation). The C<rpid> member contains the pid of the	943	C<waitpid> documentation). The C<rpid> member contains the pid of the
646	process causing the status change.	944	process causing the status change.
647		945
648	=back	946	=back
		947
		948	Example: try to exit cleanly on SIGINT and SIGTERM.
		949
		950	static void
		951	sigint_cb (struct ev_loop loop, struct ev_signal w, int revents)
		952	{
		953	ev_unloop (loop, EVUNLOOP_ALL);
		954	}
		955
		956	struct ev_signal signal_watcher;
		957	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		958	ev_signal_start (loop, &sigint_cb);
		959
649		960
650	=head2 C<ev_idle> - when you've got nothing better to do	961	=head2 C<ev_idle> - when you've got nothing better to do
651		962
652	Idle watchers trigger events when there are no other events are pending	963	Idle watchers trigger events when there are no other events are pending
653	(prepare, check and other idle watchers do not count). That is, as long	964	(prepare, check and other idle watchers do not count). That is, as long
…		…
673	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	984	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
674	believe me.	985	believe me.
675		986
676	=back	987	=back
677		988
		989	Example: dynamically allocate an C<ev_idle>, start it, and in the
		990	callback, free it. Alos, use no error checking, as usual.
		991
		992	static void
		993	idle_cb (struct ev_loop loop, struct ev_idle w, int revents)
		994	{
		995	free (w);
		996	// now do something you wanted to do when the program has
		997	// no longer asnything immediate to do.
		998	}
		999
		1000	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1001	ev_idle_init (idle_watcher, idle_cb);
		1002	ev_idle_start (loop, idle_cb);
		1003
		1004
678	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	1005	=head2 C<ev_prepare> and C<ev_check> - customise your event loop
679		1006
680	Prepare and check watchers are usually (but not always) used in tandem:	1007	Prepare and check watchers are usually (but not always) used in tandem:
681	Prepare watchers get invoked before the process blocks and check watchers	1008	prepare watchers get invoked before the process blocks and check watchers
682	afterwards.	1009	afterwards.
683		1010
684	Their main purpose is to integrate other event mechanisms into libev. This	1011	Their main purpose is to integrate other event mechanisms into libev. This
685	could be used, for example, to track variable changes, implement your own	1012	could be used, for example, to track variable changes, implement your own
686	watchers, integrate net-snmp or a coroutine library and lots more.	1013	watchers, integrate net-snmp or a coroutine library and lots more.
…		…
689	to be watched by the other library, registering C<ev_io> watchers for	1016	to be watched by the other library, registering C<ev_io> watchers for
690	them and starting an C<ev_timer> watcher for any timeouts (many libraries	1017	them and starting an C<ev_timer> watcher for any timeouts (many libraries
691	provide just this functionality). Then, in the check watcher you check for	1018	provide just this functionality). Then, in the check watcher you check for
692	any events that occured (by checking the pending status of all watchers	1019	any events that occured (by checking the pending status of all watchers
693	and stopping them) and call back into the library. The I/O and timer	1020	and stopping them) and call back into the library. The I/O and timer
694	callbacks will never actually be called (but must be valid neverthelles,	1021	callbacks will never actually be called (but must be valid nevertheless,
695	because you never know, you know?).	1022	because you never know, you know?).
696		1023
697	As another example, the Perl Coro module uses these hooks to integrate	1024	As another example, the Perl Coro module uses these hooks to integrate
698	coroutines into libev programs, by yielding to other active coroutines	1025	coroutines into libev programs, by yielding to other active coroutines
699	during each prepare and only letting the process block if no coroutines	1026	during each prepare and only letting the process block if no coroutines
700	are ready to run (its actually more complicated, it only runs coroutines	1027	are ready to run (it's actually more complicated: it only runs coroutines
701	with priority higher than the event loop and one lower priority once,	1028	with priority higher than or equal to the event loop and one coroutine
702	using idle watchers to keep the event loop from blocking if lower-priority	1029	of lower priority, but only once, using idle watchers to keep the event
703	coroutines exist, thus mapping low-priority coroutines to idle/background	1030	loop from blocking if lower-priority coroutines are active, thus mapping
704	tasks).	1031	low-priority coroutines to idle/background tasks).
705		1032
706	=over 4	1033	=over 4
707		1034
708	=item ev_prepare_init (ev_prepare *, callback)	1035	=item ev_prepare_init (ev_prepare *, callback)
709		1036
…		…
713	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	1040	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
714	macros, but using them is utterly, utterly and completely pointless.	1041	macros, but using them is utterly, utterly and completely pointless.
715		1042
716	=back	1043	=back
717		1044
		1045	Example: TODO.
		1046
		1047
718	=head1 OTHER FUNCTIONS	1048	=head1 OTHER FUNCTIONS
719		1049
720	There are some other functions of possible interest. Described. Here. Now.	1050	There are some other functions of possible interest. Described. Here. Now.
721		1051
722	=over 4	1052	=over 4
…		…
724	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	1054	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
725		1055
726	This function combines a simple timer and an I/O watcher, calls your	1056	This function combines a simple timer and an I/O watcher, calls your
727	callback on whichever event happens first and automatically stop both	1057	callback on whichever event happens first and automatically stop both
728	watchers. This is useful if you want to wait for a single event on an fd	1058	watchers. This is useful if you want to wait for a single event on an fd
729	or timeout without havign to allocate/configure/start/stop/free one or	1059	or timeout without having to allocate/configure/start/stop/free one or
730	more watchers yourself.	1060	more watchers yourself.
731		1061
732	If C<fd> is less than 0, then no I/O watcher will be started and events	1062	If C<fd> is less than 0, then no I/O watcher will be started and events
733	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	1063	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and
734	C<events> set will be craeted and started.	1064	C<events> set will be craeted and started.
…		…
737	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	1067	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
738	repeat = 0) will be started. While C<0> is a valid timeout, it is of	1068	repeat = 0) will be started. While C<0> is a valid timeout, it is of
739	dubious value.	1069	dubious value.
740		1070
741	The callback has the type C<void (cb)(int revents, void arg)> and gets	1071	The callback has the type C<void (cb)(int revents, void arg)> and gets
742	passed an events set like normal event callbacks (with a combination of	1072	passed an C<revents> set like normal event callbacks (a combination of
743	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	1073	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
744	value passed to C<ev_once>:	1074	value passed to C<ev_once>:
745		1075
746	static void stdin_ready (int revents, void *arg)	1076	static void stdin_ready (int revents, void *arg)
747	{	1077	{
…		…
768		1098
769	Feed an event as if the given signal occured (loop must be the default loop!).	1099	Feed an event as if the given signal occured (loop must be the default loop!).
770		1100
771	=back	1101	=back
772		1102
		1103
		1104	=head1 LIBEVENT EMULATION
		1105
		1106	Libev offers a compatibility emulation layer for libevent. It cannot
		1107	emulate the internals of libevent, so here are some usage hints:
		1108
		1109	=over 4
		1110
		1111	=item * Use it by including <event.h>, as usual.
		1112
		1113	=item * The following members are fully supported: ev_base, ev_callback,
		1114	ev_arg, ev_fd, ev_res, ev_events.
		1115
		1116	=item * Avoid using ev_flags and the EVLIST_*-macros, while it is
		1117	maintained by libev, it does not work exactly the same way as in libevent (consider
		1118	it a private API).
		1119
		1120	=item * Priorities are not currently supported. Initialising priorities
		1121	will fail and all watchers will have the same priority, even though there
		1122	is an ev_pri field.
		1123
		1124	=item * Other members are not supported.
		1125
		1126	=item * The libev emulation is I<not> ABI compatible to libevent, you need
		1127	to use the libev header file and library.
		1128
		1129	=back
		1130
		1131	=head1 C++ SUPPORT
		1132
		1133	TBD.
		1134
773	=head1 AUTHOR	1135	=head1 AUTHOR
774		1136
775	Marc Lehmann <libev@schmorp.de>.	1137	Marc Lehmann <libev@schmorp.de>.
776		1138

Diff Legend

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

Comparing libev/ev.pod (file contents): Revision 1.18 by root, Mon Nov 12 09:01:12 2007 UTC vs. Revision 1.34 by root, Fri Nov 23 16:17:12 2007 UTC

Diff Legend

Comparing libev/ev.pod (file contents):
Revision 1.18 by root, Mon Nov 12 09:01:12 2007 UTC vs.
Revision 1.34 by root, Fri Nov 23 16:17:12 2007 UTC