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

Comparing libev/ev.pod (file contents):
Revision 1.136 by root, Thu Mar 13 13:06:16 2008 UTC vs.
Revision 1.160 by root, Thu May 22 03:06:58 2008 UTC

…		…
64		64
65	=head1 DESCRIPTION	65	=head1 DESCRIPTION
66		66
67	The newest version of this document is also available as an html-formatted	67	The newest version of this document is also available as an html-formatted
68	web page you might find easier to navigate when reading it for the first	68	web page you might find easier to navigate when reading it for the first
69	time: L<http://cvs.schmorp.de/libev/ev.html>.	69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
70		70
71	Libev is an event loop: you register interest in certain events (such as a	71	Libev is an event loop: you register interest in certain events (such as a
72	file descriptor being readable or a timeout occurring), and it will manage	72	file descriptor being readable or a timeout occurring), and it will manage
73	these event sources and provide your program with events.	73	these event sources and provide your program with events.
74		74
…		…
117	to the C<double> type in C, and when you need to do any calculations on	117	to the C<double> type in C, and when you need to do any calculations on
118	it, you should treat it as some floatingpoint value. Unlike the name	118	it, you should treat it as some floatingpoint value. Unlike the name
119	component C<stamp> might indicate, it is also used for time differences	119	component C<stamp> might indicate, it is also used for time differences
120	throughout libev.	120	throughout libev.
121		121
		122	=head1 ERROR HANDLING
		123
		124	Libev knows three classes of errors: operating system errors, usage errors
		125	and internal errors (bugs).
		126
		127	When libev catches an operating system error it cannot handle (for example
		128	a syscall indicating a condition libev cannot fix), it calls the callback
		129	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		130	abort. The default is to print a diagnostic message and to call C<abort
		131	()>.
		132
		133	When libev detects a usage error such as a negative timer interval, then
		134	it will print a diagnostic message and abort (via the C<assert> mechanism,
		135	so C<NDEBUG> will disable this checking): these are programming errors in
		136	the libev caller and need to be fixed there.
		137
		138	Libev also has a few internal error-checking C<assert>ions, and also has
		139	extensive consistency checking code. These do not trigger under normal
		140	circumstances, as they indicate either a bug in libev or worse.
		141
		142
122	=head1 GLOBAL FUNCTIONS	143	=head1 GLOBAL FUNCTIONS
123		144
124	These functions can be called anytime, even before initialising the	145	These functions can be called anytime, even before initialising the
125	library in any way.	146	library in any way.
126		147
…		…
196	See the description of C<ev_embed> watchers for more info.	217	See the description of C<ev_embed> watchers for more info.
197		218
198	=item ev_set_allocator (void (cb)(void *ptr, long size))	219	=item ev_set_allocator (void (cb)(void *ptr, long size))
199		220
200	Sets the allocation function to use (the prototype is similar - the	221	Sets the allocation function to use (the prototype is similar - the
201	semantics is identical - to the realloc C function). It is used to	222	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
202	allocate and free memory (no surprises here). If it returns zero when	223	used to allocate and free memory (no surprises here). If it returns zero
203	memory needs to be allocated, the library might abort or take some	224	when memory needs to be allocated (C<size != 0>), the library might abort
204	potentially destructive action. The default is your system realloc	225	or take some potentially destructive action.
205	function.	226
		227	Since some systems (at least OpenBSD and Darwin) fail to implement
		228	correct C<realloc> semantics, libev will use a wrapper around the system
		229	C<realloc> and C<free> functions by default.
206		230
207	You could override this function in high-availability programs to, say,	231	You could override this function in high-availability programs to, say,
208	free some memory if it cannot allocate memory, to use a special allocator,	232	free some memory if it cannot allocate memory, to use a special allocator,
209	or even to sleep a while and retry until some memory is available.	233	or even to sleep a while and retry until some memory is available.
210		234
211	Example: Replace the libev allocator with one that waits a bit and then	235	Example: Replace the libev allocator with one that waits a bit and then
212	retries).	236	retries (example requires a standards-compliant C<realloc>).
213		237
214	static void *	238	static void *
215	persistent_realloc (void *ptr, size_t size)	239	persistent_realloc (void *ptr, size_t size)
216	{	240	{
217	for (;;)	241	for (;;)
…		…
256		280
257	An event loop is described by a C<struct ev_loop *>. The library knows two	281	An event loop is described by a C<struct ev_loop *>. The library knows two
258	types of such loops, the I<default> loop, which supports signals and child	282	types of such loops, the I<default> loop, which supports signals and child
259	events, and dynamically created loops which do not.	283	events, and dynamically created loops which do not.
260		284
261	If you use threads, a common model is to run the default event loop
262	in your main thread (or in a separate thread) and for each thread you
263	create, you also create another event loop. Libev itself does no locking
264	whatsoever, so if you mix calls to the same event loop in different
265	threads, make sure you lock (this is usually a bad idea, though, even if
266	done correctly, because it's hideous and inefficient).
267
268	=over 4	285	=over 4
269		286
270	=item struct ev_loop *ev_default_loop (unsigned int flags)	287	=item struct ev_loop *ev_default_loop (unsigned int flags)
271		288
272	This will initialise the default event loop if it hasn't been initialised	289	This will initialise the default event loop if it hasn't been initialised
…		…
274	false. If it already was initialised it simply returns it (and ignores the	291	false. If it already was initialised it simply returns it (and ignores the
275	flags. If that is troubling you, check C<ev_backend ()> afterwards).	292	flags. If that is troubling you, check C<ev_backend ()> afterwards).
276		293
277	If you don't know what event loop to use, use the one returned from this	294	If you don't know what event loop to use, use the one returned from this
278	function.	295	function.
		296
		297	Note that this function is I<not> thread-safe, so if you want to use it
		298	from multiple threads, you have to lock (note also that this is unlikely,
		299	as loops cannot bes hared easily between threads anyway).
279		300
280	The default loop is the only loop that can handle C<ev_signal> and	301	The default loop is the only loop that can handle C<ev_signal> and
281	C<ev_child> watchers, and to do this, it always registers a handler	302	C<ev_child> watchers, and to do this, it always registers a handler
282	for C<SIGCHLD>. If this is a problem for your app you can either	303	for C<SIGCHLD>. If this is a problem for your app you can either
283	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
336	To get good performance out of this backend you need a high amount of	357	To get good performance out of this backend you need a high amount of
337	parallelity (most of the file descriptors should be busy). If you are	358	parallelity (most of the file descriptors should be busy). If you are
338	writing a server, you should C<accept ()> in a loop to accept as many	359	writing a server, you should C<accept ()> in a loop to accept as many
339	connections as possible during one iteration. You might also want to have	360	connections as possible during one iteration. You might also want to have
340	a look at C<ev_set_io_collect_interval ()> to increase the amount of	361	a look at C<ev_set_io_collect_interval ()> to increase the amount of
341	readyness notifications you get per iteration.	362	readiness notifications you get per iteration.
342		363
343	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	364	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
344		365
345	And this is your standard poll(2) backend. It's more complicated	366	And this is your standard poll(2) backend. It's more complicated
346	than select, but handles sparse fds better and has no artificial	367	than select, but handles sparse fds better and has no artificial
…		…
354	For few fds, this backend is a bit little slower than poll and select,	375	For few fds, this backend is a bit little slower than poll and select,
355	but it scales phenomenally better. While poll and select usually scale	376	but it scales phenomenally better. While poll and select usually scale
356	like O(total_fds) where n is the total number of fds (or the highest fd),	377	like O(total_fds) where n is the total number of fds (or the highest fd),
357	epoll scales either O(1) or O(active_fds). The epoll design has a number	378	epoll scales either O(1) or O(active_fds). The epoll design has a number
358	of shortcomings, such as silently dropping events in some hard-to-detect	379	of shortcomings, such as silently dropping events in some hard-to-detect
359	cases and rewiring a syscall per fd change, no fork support and bad	380	cases and requiring a syscall per fd change, no fork support and bad
360	support for dup.	381	support for dup.
361		382
362	While stopping, setting and starting an I/O watcher in the same iteration	383	While stopping, setting and starting an I/O watcher in the same iteration
363	will result in some caching, there is still a syscall per such incident	384	will result in some caching, there is still a syscall per such incident
364	(because the fd could point to a different file description now), so its	385	(because the fd could point to a different file description now), so its
…		…
425	While this backend scales well, it requires one system call per active	446	While this backend scales well, it requires one system call per active
426	file descriptor per loop iteration. For small and medium numbers of file	447	file descriptor per loop iteration. For small and medium numbers of file
427	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	448	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
428	might perform better.	449	might perform better.
429		450
430	On the positive side, ignoring the spurious readyness notifications, this	451	On the positive side, ignoring the spurious readiness notifications, this
431	backend actually performed to specification in all tests and is fully	452	backend actually performed to specification in all tests and is fully
432	embeddable, which is a rare feat among the OS-specific backends.	453	embeddable, which is a rare feat among the OS-specific backends.
433		454
434	=item C<EVBACKEND_ALL>	455	=item C<EVBACKEND_ALL>
435		456
…		…
465		486
466	Similar to C<ev_default_loop>, but always creates a new event loop that is	487	Similar to C<ev_default_loop>, but always creates a new event loop that is
467	always distinct from the default loop. Unlike the default loop, it cannot	488	always distinct from the default loop. Unlike the default loop, it cannot
468	handle signal and child watchers, and attempts to do so will be greeted by	489	handle signal and child watchers, and attempts to do so will be greeted by
469	undefined behaviour (or a failed assertion if assertions are enabled).	490	undefined behaviour (or a failed assertion if assertions are enabled).
		491
		492	Note that this function I<is> thread-safe, and the recommended way to use
		493	libev with threads is indeed to create one loop per thread, and using the
		494	default loop in the "main" or "initial" thread.
470		495
471	Example: Try to create a event loop that uses epoll and nothing else.	496	Example: Try to create a event loop that uses epoll and nothing else.
472		497
473	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);	498	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);
474	if (!epoller)	499	if (!epoller)
…		…
685	interval to a value near C<0.1> or so, which is often enough for	710	interval to a value near C<0.1> or so, which is often enough for
686	interactive servers (of course not for games), likewise for timeouts. It	711	interactive servers (of course not for games), likewise for timeouts. It
687	usually doesn't make much sense to set it to a lower value than C<0.01>,	712	usually doesn't make much sense to set it to a lower value than C<0.01>,
688	as this approsaches the timing granularity of most systems.	713	as this approsaches the timing granularity of most systems.
689		714
		715	=item ev_loop_verify (loop)
		716
		717	This function only does something when C<EV_VERIFY> support has been
		718	compiled in. It tries to go through all internal structures and checks
		719	them for validity. If anything is found to be inconsistent, it will print
		720	an error message to standard error and call C<abort ()>.
		721
		722	This can be used to catch bugs inside libev itself: under normal
		723	circumstances, this function will never abort as of course libev keeps its
		724	data structures consistent.
		725
690	=back	726	=back
691		727
692		728
693	=head1 ANATOMY OF A WATCHER	729	=head1 ANATOMY OF A WATCHER
694		730
…		…
1028	If you must do this, then force the use of a known-to-be-good backend	1064	If you must do this, then force the use of a known-to-be-good backend
1029	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1065	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
1030	C<EVBACKEND_POLL>).	1066	C<EVBACKEND_POLL>).
1031		1067
1032	Another thing you have to watch out for is that it is quite easy to	1068	Another thing you have to watch out for is that it is quite easy to
1033	receive "spurious" readyness notifications, that is your callback might	1069	receive "spurious" readiness notifications, that is your callback might
1034	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1070	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1035	because there is no data. Not only are some backends known to create a	1071	because there is no data. Not only are some backends known to create a
1036	lot of those (for example solaris ports), it is very easy to get into	1072	lot of those (for example solaris ports), it is very easy to get into
1037	this situation even with a relatively standard program structure. Thus	1073	this situation even with a relatively standard program structure. Thus
1038	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1074	it is best to always use non-blocking I/O: An extra C<read>(2) returning
…		…
1085	To support fork in your programs, you either have to call	1121	To support fork in your programs, you either have to call
1086	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1122	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
1087	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1123	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1088	C<EVBACKEND_POLL>.	1124	C<EVBACKEND_POLL>.
1089		1125
		1126	=head3 The special problem of SIGPIPE
		1127
		1128	While not really specific to libev, it is easy to forget about SIGPIPE:
		1129	when reading from a pipe whose other end has been closed, your program
		1130	gets send a SIGPIPE, which, by default, aborts your program. For most
		1131	programs this is sensible behaviour, for daemons, this is usually
		1132	undesirable.
		1133
		1134	So when you encounter spurious, unexplained daemon exits, make sure you
		1135	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1136	somewhere, as that would have given you a big clue).
		1137
1090		1138
1091	=head3 Watcher-Specific Functions	1139	=head3 Watcher-Specific Functions
1092		1140
1093	=over 4	1141	=over 4
1094		1142
…		…
1135		1183
1136	Timer watchers are simple relative timers that generate an event after a	1184	Timer watchers are simple relative timers that generate an event after a
1137	given time, and optionally repeating in regular intervals after that.	1185	given time, and optionally repeating in regular intervals after that.
1138		1186
1139	The timers are based on real time, that is, if you register an event that	1187	The timers are based on real time, that is, if you register an event that
1140	times out after an hour and you reset your system clock to last years	1188	times out after an hour and you reset your system clock to january last
1141	time, it will still time out after (roughly) and hour. "Roughly" because	1189	year, it will still time out after (roughly) and hour. "Roughly" because
1142	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1190	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1143	monotonic clock option helps a lot here).	1191	monotonic clock option helps a lot here).
1144		1192
1145	The relative timeouts are calculated relative to the C<ev_now ()>	1193	The relative timeouts are calculated relative to the C<ev_now ()>
1146	time. This is usually the right thing as this timestamp refers to the time	1194	time. This is usually the right thing as this timestamp refers to the time
…		…
1148	you suspect event processing to be delayed and you I<need> to base the timeout	1196	you suspect event processing to be delayed and you I<need> to base the timeout
1149	on the current time, use something like this to adjust for this:	1197	on the current time, use something like this to adjust for this:
1150		1198
1151	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1199	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1152		1200
1153	The callback is guarenteed to be invoked only when its timeout has passed,	1201	The callback is guarenteed to be invoked only after its timeout has passed,
1154	but if multiple timers become ready during the same loop iteration then	1202	but if multiple timers become ready during the same loop iteration then
1155	order of execution is undefined.	1203	order of execution is undefined.
1156		1204
1157	=head3 Watcher-Specific Functions and Data Members	1205	=head3 Watcher-Specific Functions and Data Members
1158		1206
…		…
1160		1208
1161	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1209	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1162		1210
1163	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1211	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1164		1212
1165	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1213	Configure the timer to trigger after C<after> seconds. If C<repeat>
1166	C<0.>, then it will automatically be stopped. If it is positive, then the	1214	is C<0.>, then it will automatically be stopped once the timeout is
1167	timer will automatically be configured to trigger again C<repeat> seconds	1215	reached. If it is positive, then the timer will automatically be
1168	later, again, and again, until stopped manually.	1216	configured to trigger again C<repeat> seconds later, again, and again,
		1217	until stopped manually.
1169		1218
1170	The timer itself will do a best-effort at avoiding drift, that is, if you	1219	The timer itself will do a best-effort at avoiding drift, that is, if
1171	configure a timer to trigger every 10 seconds, then it will trigger at	1220	you configure a timer to trigger every 10 seconds, then it will normally
1172	exactly 10 second intervals. If, however, your program cannot keep up with	1221	trigger at exactly 10 second intervals. If, however, your program cannot
1173	the timer (because it takes longer than those 10 seconds to do stuff) the	1222	keep up with the timer (because it takes longer than those 10 seconds to
1174	timer will not fire more than once per event loop iteration.	1223	do stuff) the timer will not fire more than once per event loop iteration.
1175		1224
1176	=item ev_timer_again (loop, ev_timer *)	1225	=item ev_timer_again (loop, ev_timer *)
1177		1226
1178	This will act as if the timer timed out and restart it again if it is	1227	This will act as if the timer timed out and restart it again if it is
1179	repeating. The exact semantics are:	1228	repeating. The exact semantics are:
…		…
1256	Periodic watchers are also timers of a kind, but they are very versatile	1305	Periodic watchers are also timers of a kind, but they are very versatile
1257	(and unfortunately a bit complex).	1306	(and unfortunately a bit complex).
1258		1307
1259	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1308	Unlike C<ev_timer>'s, they are not based on real time (or relative time)
1260	but on wallclock time (absolute time). You can tell a periodic watcher	1309	but on wallclock time (absolute time). You can tell a periodic watcher
1261	to trigger "at" some specific point in time. For example, if you tell a	1310	to trigger after some specific point in time. For example, if you tell a
1262	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	1311	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()
1263	+ 10.>) and then reset your system clock to the last year, then it will	1312	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1313	clock to january of the previous year, then it will take more than year
1264	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	1314	to trigger the event (unlike an C<ev_timer>, which would still trigger
1265	roughly 10 seconds later).	1315	roughly 10 seconds later as it uses a relative timeout).
1266		1316
1267	They can also be used to implement vastly more complex timers, such as	1317	C<ev_periodic>s can also be used to implement vastly more complex timers,
1268	triggering an event on each midnight, local time or other, complicated,	1318	such as triggering an event on each "midnight, local time", or other
1269	rules.	1319	complicated, rules.
1270		1320
1271	As with timers, the callback is guarenteed to be invoked only when the	1321	As with timers, the callback is guarenteed to be invoked only when the
1272	time (C<at>) has been passed, but if multiple periodic timers become ready	1322	time (C<at>) has passed, but if multiple periodic timers become ready
1273	during the same loop iteration then order of execution is undefined.	1323	during the same loop iteration then order of execution is undefined.
1274		1324
1275	=head3 Watcher-Specific Functions and Data Members	1325	=head3 Watcher-Specific Functions and Data Members
1276		1326
1277	=over 4	1327	=over 4
…		…
1285		1335
1286	=over 4	1336	=over 4
1287		1337
1288	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1338	=item * absolute timer (at = time, interval = reschedule_cb = 0)
1289		1339
1290	In this configuration the watcher triggers an event at the wallclock time	1340	In this configuration the watcher triggers an event after the wallclock
1291	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	1341	time C<at> has passed and doesn't repeat. It will not adjust when a time
1292	that is, if it is to be run at January 1st 2011 then it will run when the	1342	jump occurs, that is, if it is to be run at January 1st 2011 then it will
1293	system time reaches or surpasses this time.	1343	run when the system time reaches or surpasses this time.
1294		1344
1295	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1345	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
1296		1346
1297	In this mode the watcher will always be scheduled to time out at the next	1347	In this mode the watcher will always be scheduled to time out at the next
1298	C<at + N * interval> time (for some integer N, which can also be negative)	1348	C<at + N * interval> time (for some integer N, which can also be negative)
1299	and then repeat, regardless of any time jumps.	1349	and then repeat, regardless of any time jumps.
1300		1350
1301	This can be used to create timers that do not drift with respect to system	1351	This can be used to create timers that do not drift with respect to system
1302	time:	1352	time, for example, here is a C<ev_periodic> that triggers each hour, on
		1353	the hour:
1303		1354
1304	ev_periodic_set (&periodic, 0., 3600., 0);	1355	ev_periodic_set (&periodic, 0., 3600., 0);
1305		1356
1306	This doesn't mean there will always be 3600 seconds in between triggers,	1357	This doesn't mean there will always be 3600 seconds in between triggers,
1307	but only that the the callback will be called when the system time shows a	1358	but only that the the callback will be called when the system time shows a
…		…
1312	C<ev_periodic> will try to run the callback in this mode at the next possible	1363	C<ev_periodic> will try to run the callback in this mode at the next possible
1313	time where C<time = at (mod interval)>, regardless of any time jumps.	1364	time where C<time = at (mod interval)>, regardless of any time jumps.
1314		1365
1315	For numerical stability it is preferable that the C<at> value is near	1366	For numerical stability it is preferable that the C<at> value is near
1316	C<ev_now ()> (the current time), but there is no range requirement for	1367	C<ev_now ()> (the current time), but there is no range requirement for
1317	this value.	1368	this value, and in fact is often specified as zero.
		1369
		1370	Note also that there is an upper limit to how often a timer can fire (cpu
		1371	speed for example), so if C<interval> is very small then timing stability
		1372	will of course detoriate. Libev itself tries to be exact to be about one
		1373	millisecond (if the OS supports it and the machine is fast enough).
1318		1374
1319	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1375	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1320		1376
1321	In this mode the values for C<interval> and C<at> are both being	1377	In this mode the values for C<interval> and C<at> are both being
1322	ignored. Instead, each time the periodic watcher gets scheduled, the	1378	ignored. Instead, each time the periodic watcher gets scheduled, the
1323	reschedule callback will be called with the watcher as first, and the	1379	reschedule callback will be called with the watcher as first, and the
1324	current time as second argument.	1380	current time as second argument.
1325		1381
1326	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1382	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,
1327	ever, or make any event loop modifications>. If you need to stop it,	1383	ever, or make ANY event loop modifications whatsoever>.
1328	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by
1329	starting an C<ev_prepare> watcher, which is legal).
1330		1384
		1385	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1386	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1387	only event loop modification you are allowed to do).
		1388
1331	Its prototype is C<ev_tstamp (reschedule_cb)(struct ev_periodic w,	1389	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic
1332	ev_tstamp now)>, e.g.:	1390	*w, ev_tstamp now)>, e.g.:
1333		1391
1334	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1392	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
1335	{	1393	{
1336	return now + 60.;	1394	return now + 60.;
1337	}	1395	}
…		…
1339	It must return the next time to trigger, based on the passed time value	1397	It must return the next time to trigger, based on the passed time value
1340	(that is, the lowest time value larger than to the second argument). It	1398	(that is, the lowest time value larger than to the second argument). It
1341	will usually be called just before the callback will be triggered, but	1399	will usually be called just before the callback will be triggered, but
1342	might be called at other times, too.	1400	might be called at other times, too.
1343		1401
1344	NOTE: I<< This callback must always return a time that is later than the	1402	NOTE: I<< This callback must always return a time that is higher than or
1345	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	1403	equal to the passed C<now> value >>.
1346		1404
1347	This can be used to create very complex timers, such as a timer that	1405	This can be used to create very complex timers, such as a timer that
1348	triggers on each midnight, local time. To do this, you would calculate the	1406	triggers on "next midnight, local time". To do this, you would calculate the
1349	next midnight after C<now> and return the timestamp value for this. How	1407	next midnight after C<now> and return the timestamp value for this. How
1350	you do this is, again, up to you (but it is not trivial, which is the main	1408	you do this is, again, up to you (but it is not trivial, which is the main
1351	reason I omitted it as an example).	1409	reason I omitted it as an example).
1352		1410
1353	=back	1411	=back
…		…
1357	Simply stops and restarts the periodic watcher again. This is only useful	1415	Simply stops and restarts the periodic watcher again. This is only useful
1358	when you changed some parameters or the reschedule callback would return	1416	when you changed some parameters or the reschedule callback would return
1359	a different time than the last time it was called (e.g. in a crond like	1417	a different time than the last time it was called (e.g. in a crond like
1360	program when the crontabs have changed).	1418	program when the crontabs have changed).
1361		1419
		1420	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1421
		1422	When active, returns the absolute time that the watcher is supposed to
		1423	trigger next.
		1424
1362	=item ev_tstamp offset [read-write]	1425	=item ev_tstamp offset [read-write]
1363		1426
1364	When repeating, this contains the offset value, otherwise this is the	1427	When repeating, this contains the offset value, otherwise this is the
1365	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	1428	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
1366		1429
…		…
1376	=item ev_tstamp (reschedule_cb)(struct ev_periodic w, ev_tstamp now) [read-write]	1439	=item ev_tstamp (reschedule_cb)(struct ev_periodic w, ev_tstamp now) [read-write]
1377		1440
1378	The current reschedule callback, or C<0>, if this functionality is	1441	The current reschedule callback, or C<0>, if this functionality is
1379	switched off. Can be changed any time, but changes only take effect when	1442	switched off. Can be changed any time, but changes only take effect when
1380	the periodic timer fires or C<ev_periodic_again> is being called.	1443	the periodic timer fires or C<ev_periodic_again> is being called.
1381
1382	=item ev_tstamp at [read-only]
1383
1384	When active, contains the absolute time that the watcher is supposed to
1385	trigger next.
1386		1444
1387	=back	1445	=back
1388		1446
1389	=head3 Examples	1447	=head3 Examples
1390		1448
…		…
1594	as even with OS-supported change notifications, this can be	1652	as even with OS-supported change notifications, this can be
1595	resource-intensive.	1653	resource-intensive.
1596		1654
1597	At the time of this writing, only the Linux inotify interface is	1655	At the time of this writing, only the Linux inotify interface is
1598	implemented (implementing kqueue support is left as an exercise for the	1656	implemented (implementing kqueue support is left as an exercise for the
		1657	reader, note, however, that the author sees no way of implementing ev_stat
1599	reader). Inotify will be used to give hints only and should not change the	1658	semantics with kqueue). Inotify will be used to give hints only and should
1600	semantics of C<ev_stat> watchers, which means that libev sometimes needs	1659	not change the semantics of C<ev_stat> watchers, which means that libev
1601	to fall back to regular polling again even with inotify, but changes are	1660	sometimes needs to fall back to regular polling again even with inotify,
1602	usually detected immediately, and if the file exists there will be no	1661	but changes are usually detected immediately, and if the file exists there
1603	polling.	1662	will be no polling.
		1663
		1664	=head3 ABI Issues (Largefile Support)
		1665
		1666	Libev by default (unless the user overrides this) uses the default
		1667	compilation environment, which means that on systems with optionally
		1668	disabled large file support, you get the 32 bit version of the stat
		1669	structure. When using the library from programs that change the ABI to
		1670	use 64 bit file offsets the programs will fail. In that case you have to
		1671	compile libev with the same flags to get binary compatibility. This is
		1672	obviously the case with any flags that change the ABI, but the problem is
		1673	most noticably with ev_stat and largefile support.
1604		1674
1605	=head3 Inotify	1675	=head3 Inotify
1606		1676
1607	When C<inotify (7)> support has been compiled into libev (generally only	1677	When C<inotify (7)> support has been compiled into libev (generally only
1608	available on Linux) and present at runtime, it will be used to speed up	1678	available on Linux) and present at runtime, it will be used to speed up
1609	change detection where possible. The inotify descriptor will be created lazily	1679	change detection where possible. The inotify descriptor will be created lazily
1610	when the first C<ev_stat> watcher is being started.	1680	when the first C<ev_stat> watcher is being started.
1611		1681
1612	Inotify presense does not change the semantics of C<ev_stat> watchers	1682	Inotify presence does not change the semantics of C<ev_stat> watchers
1613	except that changes might be detected earlier, and in some cases, to avoid	1683	except that changes might be detected earlier, and in some cases, to avoid
1614	making regular C<stat> calls. Even in the presense of inotify support	1684	making regular C<stat> calls. Even in the presence of inotify support
1615	there are many cases where libev has to resort to regular C<stat> polling.	1685	there are many cases where libev has to resort to regular C<stat> polling.
1616		1686
1617	(There is no support for kqueue, as apparently it cannot be used to	1687	(There is no support for kqueue, as apparently it cannot be used to
1618	implement this functionality, due to the requirement of having a file	1688	implement this functionality, due to the requirement of having a file
1619	descriptor open on the object at all times).	1689	descriptor open on the object at all times).
…		…
1622		1692
1623	The C<stat ()> syscall only supports full-second resolution portably, and	1693	The C<stat ()> syscall only supports full-second resolution portably, and
1624	even on systems where the resolution is higher, many filesystems still	1694	even on systems where the resolution is higher, many filesystems still
1625	only support whole seconds.	1695	only support whole seconds.
1626		1696
1627	That means that, if the time is the only thing that changes, you might	1697	That means that, if the time is the only thing that changes, you can
1628	miss updates: on the first update, C<ev_stat> detects a change and calls	1698	easily miss updates: on the first update, C<ev_stat> detects a change and
1629	your callback, which does something. When there is another update within	1699	calls your callback, which does something. When there is another update
1630	the same second, C<ev_stat> will be unable to detect it.	1700	within the same second, C<ev_stat> will be unable to detect it as the stat
		1701	data does not change.
1631		1702
1632	The solution to this is to delay acting on a change for a second (or till	1703	The solution to this is to delay acting on a change for slightly more
1633	the next second boundary), using a roughly one-second delay C<ev_timer>	1704	than a second (or till slightly after the next full second boundary), using
1634	(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>	1705	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1635	is added to work around small timing inconsistencies of some operating	1706	ev_timer_again (loop, w)>).
1636	systems.	1707
		1708	The C<.02> offset is added to work around small timing inconsistencies
		1709	of some operating systems (where the second counter of the current time
		1710	might be be delayed. One such system is the Linux kernel, where a call to
		1711	C<gettimeofday> might return a timestamp with a full second later than
		1712	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		1713	update file times then there will be a small window where the kernel uses
		1714	the previous second to update file times but libev might already execute
		1715	the timer callback).
1637		1716
1638	=head3 Watcher-Specific Functions and Data Members	1717	=head3 Watcher-Specific Functions and Data Members
1639		1718
1640	=over 4	1719	=over 4
1641		1720
…		…
1647	C<path>. The C<interval> is a hint on how quickly a change is expected to	1726	C<path>. The C<interval> is a hint on how quickly a change is expected to
1648	be detected and should normally be specified as C<0> to let libev choose	1727	be detected and should normally be specified as C<0> to let libev choose
1649	a suitable value. The memory pointed to by C<path> must point to the same	1728	a suitable value. The memory pointed to by C<path> must point to the same
1650	path for as long as the watcher is active.	1729	path for as long as the watcher is active.
1651		1730
1652	The callback will be receive C<EV_STAT> when a change was detected,	1731	The callback will receive C<EV_STAT> when a change was detected, relative
1653	relative to the attributes at the time the watcher was started (or the	1732	to the attributes at the time the watcher was started (or the last change
1654	last change was detected).	1733	was detected).
1655		1734
1656	=item ev_stat_stat (loop, ev_stat *)	1735	=item ev_stat_stat (loop, ev_stat *)
1657		1736
1658	Updates the stat buffer immediately with new values. If you change the	1737	Updates the stat buffer immediately with new values. If you change the
1659	watched path in your callback, you could call this fucntion to avoid	1738	watched path in your callback, you could call this function to avoid
1660	detecting this change (while introducing a race condition). Can also be	1739	detecting this change (while introducing a race condition if you are not
1661	useful simply to find out the new values.	1740	the only one changing the path). Can also be useful simply to find out the
		1741	new values.
1662		1742
1663	=item ev_statdata attr [read-only]	1743	=item ev_statdata attr [read-only]
1664		1744
1665	The most-recently detected attributes of the file. Although the type is of	1745	The most-recently detected attributes of the file. Although the type is
1666	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types	1746	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
1667	suitable for your system. If the C<st_nlink> member is C<0>, then there	1747	suitable for your system, but you can only rely on the POSIX-standardised
		1748	members to be present. If the C<st_nlink> member is C<0>, then there was
1668	was some error while C<stat>ing the file.	1749	some error while C<stat>ing the file.
1669		1750
1670	=item ev_statdata prev [read-only]	1751	=item ev_statdata prev [read-only]
1671		1752
1672	The previous attributes of the file. The callback gets invoked whenever	1753	The previous attributes of the file. The callback gets invoked whenever
1673	C<prev> != C<attr>.	1754	C<prev> != C<attr>, or, more precisely, one or more of these members
		1755	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		1756	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1674		1757
1675	=item ev_tstamp interval [read-only]	1758	=item ev_tstamp interval [read-only]
1676		1759
1677	The specified interval.	1760	The specified interval.
1678		1761
…		…
1732	}	1815	}
1733		1816
1734	...	1817	...
1735	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	1818	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1736	ev_stat_start (loop, &passwd);	1819	ev_stat_start (loop, &passwd);
1737	ev_timer_init (&timer, timer_cb, 0., 1.01);	1820	ev_timer_init (&timer, timer_cb, 0., 1.02);
1738		1821
1739		1822
1740	=head2 C<ev_idle> - when you've got nothing better to do...	1823	=head2 C<ev_idle> - when you've got nothing better to do...
1741		1824
1742	Idle watchers trigger events when no other events of the same or higher	1825	Idle watchers trigger events when no other events of the same or higher
…		…
1830		1913
1831	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	1914	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1832	priority, to ensure that they are being run before any other watchers	1915	priority, to ensure that they are being run before any other watchers
1833	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	1916	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
1834	too) should not activate ("feed") events into libev. While libev fully	1917	too) should not activate ("feed") events into libev. While libev fully
1835	supports this, they will be called before other C<ev_check> watchers	1918	supports this, they might get executed before other C<ev_check> watchers
1836	did their job. As C<ev_check> watchers are often used to embed other	1919	did their job. As C<ev_check> watchers are often used to embed other
1837	(non-libev) event loops those other event loops might be in an unusable	1920	(non-libev) event loops those other event loops might be in an unusable
1838	state until their C<ev_check> watcher ran (always remind yourself to	1921	state until their C<ev_check> watcher ran (always remind yourself to
1839	coexist peacefully with others).	1922	coexist peacefully with others).
1840		1923
…		…
1855	=head3 Examples	1938	=head3 Examples
1856		1939
1857	There are a number of principal ways to embed other event loops or modules	1940	There are a number of principal ways to embed other event loops or modules
1858	into libev. Here are some ideas on how to include libadns into libev	1941	into libev. Here are some ideas on how to include libadns into libev
1859	(there is a Perl module named C<EV::ADNS> that does this, which you could	1942	(there is a Perl module named C<EV::ADNS> that does this, which you could
1860	use for an actually working example. Another Perl module named C<EV::Glib>	1943	use as a working example. Another Perl module named C<EV::Glib> embeds a
1861	embeds a Glib main context into libev, and finally, C<Glib::EV> embeds EV	1944	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
1862	into the Glib event loop).	1945	Glib event loop).
1863		1946
1864	Method 1: Add IO watchers and a timeout watcher in a prepare handler,	1947	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
1865	and in a check watcher, destroy them and call into libadns. What follows	1948	and in a check watcher, destroy them and call into libadns. What follows
1866	is pseudo-code only of course. This requires you to either use a low	1949	is pseudo-code only of course. This requires you to either use a low
1867	priority for the check watcher or use C<ev_clear_pending> explicitly, as	1950	priority for the check watcher or use C<ev_clear_pending> explicitly, as
…		…
2257		2340
2258	This call incurs the overhead of a syscall only once per loop iteration,	2341	This call incurs the overhead of a syscall only once per loop iteration,
2259	so while the overhead might be noticable, it doesn't apply to repeated	2342	so while the overhead might be noticable, it doesn't apply to repeated
2260	calls to C<ev_async_send>.	2343	calls to C<ev_async_send>.
2261		2344
		2345	=item bool = ev_async_pending (ev_async *)
		2346
		2347	Returns a non-zero value when C<ev_async_send> has been called on the
		2348	watcher but the event has not yet been processed (or even noted) by the
		2349	event loop.
		2350
		2351	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2352	the loop iterates next and checks for the watcher to have become active,
		2353	it will reset the flag again. C<ev_async_pending> can be used to very
		2354	quickly check wether invoking the loop might be a good idea.
		2355
		2356	Not that this does I<not> check wether the watcher itself is pending, only
		2357	wether it has been requested to make this watcher pending.
		2358
2262	=back	2359	=back
2263		2360
2264		2361
2265	=head1 OTHER FUNCTIONS	2362	=head1 OTHER FUNCTIONS
2266		2363
…		…
2337		2434
2338	=item * Priorities are not currently supported. Initialising priorities	2435	=item * Priorities are not currently supported. Initialising priorities
2339	will fail and all watchers will have the same priority, even though there	2436	will fail and all watchers will have the same priority, even though there
2340	is an ev_pri field.	2437	is an ev_pri field.
2341		2438
		2439	=item * In libevent, the last base created gets the signals, in libev, the
		2440	first base created (== the default loop) gets the signals.
		2441
2342	=item * Other members are not supported.	2442	=item * Other members are not supported.
2343		2443
2344	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2444	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2345	to use the libev header file and library.	2445	to use the libev header file and library.
2346		2446
…		…
2588	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	2688	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
2589		2689
2590	Similar to the other two macros, this gives you the value of the default	2690	Similar to the other two macros, this gives you the value of the default
2591	loop, if multiple loops are supported ("ev loop default").	2691	loop, if multiple loops are supported ("ev loop default").
2592		2692
		2693	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		2694
		2695	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		2696	default loop has been initialised (C<UC> == unchecked). Their behaviour
		2697	is undefined when the default loop has not been initialised by a previous
		2698	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		2699
		2700	It is often prudent to use C<EV_DEFAULT> when initialising the first
		2701	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		2702
2593	=back	2703	=back
2594		2704
2595	Example: Declare and initialise a check watcher, utilising the above	2705	Example: Declare and initialise a check watcher, utilising the above
2596	macros so it will work regardless of whether multiple loops are supported	2706	macros so it will work regardless of whether multiple loops are supported
2597	or not.	2707	or not.
…		…
2692		2802
2693	libev.m4	2803	libev.m4
2694		2804
2695	=head2 PREPROCESSOR SYMBOLS/MACROS	2805	=head2 PREPROCESSOR SYMBOLS/MACROS
2696		2806
2697	Libev can be configured via a variety of preprocessor symbols you have to define	2807	Libev can be configured via a variety of preprocessor symbols you have to
2698	before including any of its files. The default is not to build for multiplicity	2808	define before including any of its files. The default in the absense of
2699	and only include the select backend.	2809	autoconf is noted for every option.
2700		2810
2701	=over 4	2811	=over 4
2702		2812
2703	=item EV_STANDALONE	2813	=item EV_STANDALONE
2704		2814
…		…
2730	=item EV_USE_NANOSLEEP	2840	=item EV_USE_NANOSLEEP
2731		2841
2732	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	2842	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2733	and will use it for delays. Otherwise it will use C<select ()>.	2843	and will use it for delays. Otherwise it will use C<select ()>.
2734		2844
		2845	=item EV_USE_EVENTFD
		2846
		2847	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		2848	available and will probe for kernel support at runtime. This will improve
		2849	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		2850	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		2851	2.7 or newer, otherwise disabled.
		2852
2735	=item EV_USE_SELECT	2853	=item EV_USE_SELECT
2736		2854
2737	If undefined or defined to be C<1>, libev will compile in support for the	2855	If undefined or defined to be C<1>, libev will compile in support for the
2738	C<select>(2) backend. No attempt at autodetection will be done: if no	2856	C<select>(2) backend. No attempt at autodetection will be done: if no
2739	other method takes over, select will be it. Otherwise the select backend	2857	other method takes over, select will be it. Otherwise the select backend
…		…
2775		2893
2776	=item EV_USE_EPOLL	2894	=item EV_USE_EPOLL
2777		2895
2778	If defined to be C<1>, libev will compile in support for the Linux	2896	If defined to be C<1>, libev will compile in support for the Linux
2779	C<epoll>(7) backend. Its availability will be detected at runtime,	2897	C<epoll>(7) backend. Its availability will be detected at runtime,
2780	otherwise another method will be used as fallback. This is the	2898	otherwise another method will be used as fallback. This is the preferred
2781	preferred backend for GNU/Linux systems.	2899	backend for GNU/Linux systems. If undefined, it will be enabled if the
		2900	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2782		2901
2783	=item EV_USE_KQUEUE	2902	=item EV_USE_KQUEUE
2784		2903
2785	If defined to be C<1>, libev will compile in support for the BSD style	2904	If defined to be C<1>, libev will compile in support for the BSD style
2786	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	2905	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
2805		2924
2806	=item EV_USE_INOTIFY	2925	=item EV_USE_INOTIFY
2807		2926
2808	If defined to be C<1>, libev will compile in support for the Linux inotify	2927	If defined to be C<1>, libev will compile in support for the Linux inotify
2809	interface to speed up C<ev_stat> watchers. Its actual availability will	2928	interface to speed up C<ev_stat> watchers. Its actual availability will
2810	be detected at runtime.	2929	be detected at runtime. If undefined, it will be enabled if the headers
		2930	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2811		2931
2812	=item EV_ATOMIC_T	2932	=item EV_ATOMIC_T
2813		2933
2814	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	2934	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
2815	access is atomic with respect to other threads or signal contexts. No such	2935	access is atomic with respect to other threads or signal contexts. No such
…		…
2902	defined to be C<0>, then they are not.	3022	defined to be C<0>, then they are not.
2903		3023
2904	=item EV_MINIMAL	3024	=item EV_MINIMAL
2905		3025
2906	If you need to shave off some kilobytes of code at the expense of some	3026	If you need to shave off some kilobytes of code at the expense of some
2907	speed, define this symbol to C<1>. Currently only used for gcc to override	3027	speed, define this symbol to C<1>. Currently this is used to override some
2908	some inlining decisions, saves roughly 30% codesize of amd64.	3028	inlining decisions, saves roughly 30% codesize of amd64. It also selects a
		3029	much smaller 2-heap for timer management over the default 4-heap.
2909		3030
2910	=item EV_PID_HASHSIZE	3031	=item EV_PID_HASHSIZE
2911		3032
2912	C<ev_child> watchers use a small hash table to distribute workload by	3033	C<ev_child> watchers use a small hash table to distribute workload by
2913	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3034	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
2919	C<ev_stat> watchers use a small hash table to distribute workload by	3040	C<ev_stat> watchers use a small hash table to distribute workload by
2920	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	3041	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
2921	usually more than enough. If you need to manage thousands of C<ev_stat>	3042	usually more than enough. If you need to manage thousands of C<ev_stat>
2922	watchers you might want to increase this value (I<must> be a power of	3043	watchers you might want to increase this value (I<must> be a power of
2923	two).	3044	two).
		3045
		3046	=item EV_USE_4HEAP
		3047
		3048	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3049	timer and periodics heap, libev uses a 4-heap when this symbol is defined
		3050	to C<1>. The 4-heap uses more complicated (longer) code but has
		3051	noticably faster performance with many (thousands) of watchers.
		3052
		3053	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3054	(disabled).
		3055
		3056	=item EV_HEAP_CACHE_AT
		3057
		3058	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3059	timer and periodics heap, libev can cache the timestamp (I<at>) within
		3060	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3061	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3062	but avoids random read accesses on heap changes. This improves performance
		3063	noticably with with many (hundreds) of watchers.
		3064
		3065	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3066	(disabled).
		3067
		3068	=item EV_VERIFY
		3069
		3070	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3071	be done: If set to C<0>, no internal verification code will be compiled
		3072	in. If set to C<1>, then verification code will be compiled in, but not
		3073	called. If set to C<2>, then the internal verification code will be
		3074	called once per loop, which can slow down libev. If set to C<3>, then the
		3075	verification code will be called very frequently, which will slow down
		3076	libev considerably.
		3077
		3078	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3079	C<0.>
2924		3080
2925	=item EV_COMMON	3081	=item EV_COMMON
2926		3082
2927	By default, all watchers have a C<void *data> member. By redefining	3083	By default, all watchers have a C<void *data> member. By redefining
2928	this macro to a something else you can include more and other types of	3084	this macro to a something else you can include more and other types of
…		…
3002		3158
3003	#include "ev_cpp.h"	3159	#include "ev_cpp.h"
3004	#include "ev.c"	3160	#include "ev.c"
3005		3161
3006		3162
		3163	=head1 THREADS AND COROUTINES
		3164
		3165	=head2 THREADS
		3166
		3167	Libev itself is completely threadsafe, but it uses no locking. This
		3168	means that you can use as many loops as you want in parallel, as long as
		3169	only one thread ever calls into one libev function with the same loop
		3170	parameter.
		3171
		3172	Or put differently: calls with different loop parameters can be done in
		3173	parallel from multiple threads, calls with the same loop parameter must be
		3174	done serially (but can be done from different threads, as long as only one
		3175	thread ever is inside a call at any point in time, e.g. by using a mutex
		3176	per loop).
		3177
		3178	If you want to know which design is best for your problem, then I cannot
		3179	help you but by giving some generic advice:
		3180
		3181	=over 4
		3182
		3183	=item * most applications have a main thread: use the default libev loop
		3184	in that thread, or create a seperate thread running only the default loop.
		3185
		3186	This helps integrating other libraries or software modules that use libev
		3187	themselves and don't care/know about threading.
		3188
		3189	=item * one loop per thread is usually a good model.
		3190
		3191	Doing this is almost never wrong, sometimes a better-performance model
		3192	exists, but it is always a good start.
		3193
		3194	=item * other models exist, such as the leader/follower pattern, where one
		3195	loop is handed through multiple threads in a kind of round-robbin fashion.
		3196
		3197	Chosing a model is hard - look around, learn, know that usually you cna do
		3198	better than you currently do :-)
		3199
		3200	=item * often you need to talk to some other thread which blocks in the
		3201	event loop - C<ev_async> watchers can be used to wake them up from other
		3202	threads safely (or from signal contexts...).
		3203
		3204	=back
		3205
		3206	=head2 COROUTINES
		3207
		3208	Libev is much more accomodating to coroutines ("cooperative threads"):
		3209	libev fully supports nesting calls to it's functions from different
		3210	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3211	different coroutines and switch freely between both coroutines running the
		3212	loop, as long as you don't confuse yourself). The only exception is that
		3213	you must not do this from C<ev_periodic> reschedule callbacks.
		3214
		3215	Care has been invested into making sure that libev does not keep local
		3216	state inside C<ev_loop>, and other calls do not usually allow coroutine
		3217	switches.
		3218
		3219
3007	=head1 COMPLEXITIES	3220	=head1 COMPLEXITIES
3008		3221
3009	In this section the complexities of (many of) the algorithms used inside	3222	In this section the complexities of (many of) the algorithms used inside
3010	libev will be explained. For complexity discussions about backends see the	3223	libev will be explained. For complexity discussions about backends see the
3011	documentation for C<ev_default_init>.	3224	documentation for C<ev_default_init>.
…		…
3041	correct watcher to remove. The lists are usually short (you don't usually	3254	correct watcher to remove. The lists are usually short (you don't usually
3042	have many watchers waiting for the same fd or signal).	3255	have many watchers waiting for the same fd or signal).
3043		3256
3044	=item Finding the next timer in each loop iteration: O(1)	3257	=item Finding the next timer in each loop iteration: O(1)
3045		3258
3046	By virtue of using a binary heap, the next timer is always found at the	3259	By virtue of using a binary or 4-heap, the next timer is always found at a
3047	beginning of the storage array.	3260	fixed position in the storage array.
3048		3261
3049	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	3262	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3050		3263
3051	A change means an I/O watcher gets started or stopped, which requires	3264	A change means an I/O watcher gets started or stopped, which requires
3052	libev to recalculate its status (and possibly tell the kernel, depending	3265	libev to recalculate its status (and possibly tell the kernel, depending
…		…
3081	model. Libev still offers limited functionality on this platform in	3294	model. Libev still offers limited functionality on this platform in
3082	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	3295	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
3083	descriptors. This only applies when using Win32 natively, not when using	3296	descriptors. This only applies when using Win32 natively, not when using
3084	e.g. cygwin.	3297	e.g. cygwin.
3085		3298
		3299	Lifting these limitations would basically require the full
		3300	re-implementation of the I/O system. If you are into these kinds of
		3301	things, then note that glib does exactly that for you in a very portable
		3302	way (note also that glib is the slowest event library known to man).
		3303
3086	There is no supported compilation method available on windows except	3304	There is no supported compilation method available on windows except
3087	embedding it into other applications.	3305	embedding it into other applications.
3088		3306
3089	Due to the many, low, and arbitrary limits on the win32 platform and the	3307	Due to the many, low, and arbitrary limits on the win32 platform and
3090	abysmal performance of winsockets, using a large number of sockets is not	3308	the abysmal performance of winsockets, using a large number of sockets
3091	recommended (and not reasonable). If your program needs to use more than	3309	is not recommended (and not reasonable). If your program needs to use
3092	a hundred or so sockets, then likely it needs to use a totally different	3310	more than a hundred or so sockets, then likely it needs to use a totally
3093	implementation for windows, as libev offers the POSIX model, which cannot	3311	different implementation for windows, as libev offers the POSIX readiness
3094	be implemented efficiently on windows (microsoft monopoly games).	3312	notification model, which cannot be implemented efficiently on windows
		3313	(microsoft monopoly games).
3095		3314
3096	=over 4	3315	=over 4
3097		3316
3098	=item The winsocket select function	3317	=item The winsocket select function
3099		3318
3100	The winsocket C<select> function doesn't follow POSIX in that it requires	3319	The winsocket C<select> function doesn't follow POSIX in that it
3101	socket I<handles> and not socket I<file descriptors>. This makes select	3320	requires socket I<handles> and not socket I<file descriptors> (it is
3102	very inefficient, and also requires a mapping from file descriptors	3321	also extremely buggy). This makes select very inefficient, and also
3103	to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,	3322	requires a mapping from file descriptors to socket handles. See the
3104	C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor	3323	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
3105	symbols for more info.	3324	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
3106		3325
3107	The configuration for a "naked" win32 using the microsoft runtime	3326	The configuration for a "naked" win32 using the microsoft runtime
3108	libraries and raw winsocket select is:	3327	libraries and raw winsocket select is:
3109		3328
3110	#define EV_USE_SELECT 1	3329	#define EV_USE_SELECT 1
…		…
3113	Note that winsockets handling of fd sets is O(n), so you can easily get a	3332	Note that winsockets handling of fd sets is O(n), so you can easily get a
3114	complexity in the O(n²) range when using win32.	3333	complexity in the O(n²) range when using win32.
3115		3334
3116	=item Limited number of file descriptors	3335	=item Limited number of file descriptors
3117		3336
3118	Windows has numerous arbitrary (and low) limits on things. Early versions	3337	Windows has numerous arbitrary (and low) limits on things.
3119	of winsocket's select only supported waiting for a max. of C<64> handles	3338
		3339	Early versions of winsocket's select only supported waiting for a maximum
3120	(probably owning to the fact that all windows kernels can only wait for	3340	of C<64> handles (probably owning to the fact that all windows kernels
3121	C<64> things at the same time internally; microsoft recommends spawning a	3341	can only wait for C<64> things at the same time internally; microsoft
3122	chain of threads and wait for 63 handles and the previous thread in each).	3342	recommends spawning a chain of threads and wait for 63 handles and the
		3343	previous thread in each. Great).
3123		3344
3124	Newer versions support more handles, but you need to define C<FD_SETSIZE>	3345	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3125	to some high number (e.g. C<2048>) before compiling the winsocket select	3346	to some high number (e.g. C<2048>) before compiling the winsocket select
3126	call (which might be in libev or elsewhere, for example, perl does its own	3347	call (which might be in libev or elsewhere, for example, perl does its own
3127	select emulation on windows).	3348	select emulation on windows).
…		…
3139	calling select (O(n²)) will likely make this unworkable.	3360	calling select (O(n²)) will likely make this unworkable.
3140		3361
3141	=back	3362	=back
3142		3363
3143		3364
		3365	=head1 PORTABILITY REQUIREMENTS
		3366
		3367	In addition to a working ISO-C implementation, libev relies on a few
		3368	additional extensions:
		3369
		3370	=over 4
		3371
		3372	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3373
		3374	The type C<sig_atomic_t volatile> (or whatever is defined as
		3375	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different
		3376	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3377	believed to be sufficiently portable.
		3378
		3379	=item C<sigprocmask> must work in a threaded environment
		3380
		3381	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3382	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3383	pthread implementations will either allow C<sigprocmask> in the "main
		3384	thread" or will block signals process-wide, both behaviours would
		3385	be compatible with libev. Interaction between C<sigprocmask> and
		3386	C<pthread_sigmask> could complicate things, however.
		3387
		3388	The most portable way to handle signals is to block signals in all threads
		3389	except the initial one, and run the default loop in the initial thread as
		3390	well.
		3391
		3392	=item C<long> must be large enough for common memory allocation sizes
		3393
		3394	To improve portability and simplify using libev, libev uses C<long>
		3395	internally instead of C<size_t> when allocating its data structures. On
		3396	non-POSIX systems (Microsoft...) this might be unexpectedly low, but
		3397	is still at least 31 bits everywhere, which is enough for hundreds of
		3398	millions of watchers.
		3399
		3400	=item C<double> must hold a time value in seconds with enough accuracy
		3401
		3402	The type C<double> is used to represent timestamps. It is required to
		3403	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3404	enough for at least into the year 4000. This requirement is fulfilled by
		3405	implementations implementing IEEE 754 (basically all existing ones).
		3406
		3407	=back
		3408
		3409	If you know of other additional requirements drop me a note.
		3410
		3411
		3412	=head1 COMPILER WARNINGS
		3413
		3414	Depending on your compiler and compiler settings, you might get no or a
		3415	lot of warnings when compiling libev code. Some people are apparently
		3416	scared by this.
		3417
		3418	However, these are unavoidable for many reasons. For one, each compiler
		3419	has different warnings, and each user has different tastes regarding
		3420	warning options. "Warn-free" code therefore cannot be a goal except when
		3421	targetting a specific compiler and compiler-version.
		3422
		3423	Another reason is that some compiler warnings require elaborate
		3424	workarounds, or other changes to the code that make it less clear and less
		3425	maintainable.
		3426
		3427	And of course, some compiler warnings are just plain stupid, or simply
		3428	wrong (because they don't actually warn about the cindition their message
		3429	seems to warn about).
		3430
		3431	While libev is written to generate as few warnings as possible,
		3432	"warn-free" code is not a goal, and it is recommended not to build libev
		3433	with any compiler warnings enabled unless you are prepared to cope with
		3434	them (e.g. by ignoring them). Remember that warnings are just that:
		3435	warnings, not errors, or proof of bugs.
		3436
		3437
		3438	=head1 VALGRIND
		3439
		3440	Valgrind has a special section here because it is a popular tool that is
		3441	highly useful, but valgrind reports are very hard to interpret.
		3442
		3443	If you think you found a bug (memory leak, uninitialised data access etc.)
		3444	in libev, then check twice: If valgrind reports something like:
		3445
		3446	==2274== definitely lost: 0 bytes in 0 blocks.
		3447	==2274== possibly lost: 0 bytes in 0 blocks.
		3448	==2274== still reachable: 256 bytes in 1 blocks.
		3449
		3450	then there is no memory leak. Similarly, under some circumstances,
		3451	valgrind might report kernel bugs as if it were a bug in libev, or it
		3452	might be confused (it is a very good tool, but only a tool).
		3453
		3454	If you are unsure about something, feel free to contact the mailing list
		3455	with the full valgrind report and an explanation on why you think this is
		3456	a bug in libev. However, don't be annoyed when you get a brisk "this is
		3457	no bug" answer and take the chance of learning how to interpret valgrind
		3458	properly.
		3459
		3460	If you need, for some reason, empty reports from valgrind for your project
		3461	I suggest using suppression lists.
		3462
		3463
3144	=head1 AUTHOR	3464	=head1 AUTHOR
3145		3465
3146	Marc Lehmann <libev@schmorp.de>.	3466	Marc Lehmann <libev@schmorp.de>.
3147		3467

Diff Legend

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

Comparing libev/ev.pod (file contents): Revision 1.136 by root, Thu Mar 13 13:06:16 2008 UTC vs. Revision 1.160 by root, Thu May 22 03:06:58 2008 UTC

Diff Legend

Comparing libev/ev.pod (file contents):
Revision 1.136 by root, Thu Mar 13 13:06:16 2008 UTC vs.
Revision 1.160 by root, Thu May 22 03:06:58 2008 UTC