[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 1.159 by root, Thu May 22 02:44:57 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
…		…
196	See the description of C<ev_embed> watchers for more info.	196	See the description of C<ev_embed> watchers for more info.
197		197
198	=item ev_set_allocator (void (cb)(void *ptr, long size))	198	=item ev_set_allocator (void (cb)(void *ptr, long size))
199		199
200	Sets the allocation function to use (the prototype is similar - the	200	Sets the allocation function to use (the prototype is similar - the
201	semantics is identical - to the realloc C function). It is used to	201	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	202	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	203	when memory needs to be allocated (C<size != 0>), the library might abort
204	potentially destructive action. The default is your system realloc	204	or take some potentially destructive action.
205	function.	205
		206	Since some systems (at least OpenBSD and Darwin) fail to implement
		207	correct C<realloc> semantics, libev will use a wrapper around the system
		208	C<realloc> and C<free> functions by default.
206		209
207	You could override this function in high-availability programs to, say,	210	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,	211	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.	212	or even to sleep a while and retry until some memory is available.
210		213
211	Example: Replace the libev allocator with one that waits a bit and then	214	Example: Replace the libev allocator with one that waits a bit and then
212	retries).	215	retries (example requires a standards-compliant C<realloc>).
213		216
214	static void *	217	static void *
215	persistent_realloc (void *ptr, size_t size)	218	persistent_realloc (void *ptr, size_t size)
216	{	219	{
217	for (;;)	220	for (;;)
…		…
256		259
257	An event loop is described by a C<struct ev_loop *>. The library knows two	260	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	261	types of such loops, the I<default> loop, which supports signals and child
259	events, and dynamically created loops which do not.	262	events, and dynamically created loops which do not.
260		263
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	264	=over 4
269		265
270	=item struct ev_loop *ev_default_loop (unsigned int flags)	266	=item struct ev_loop *ev_default_loop (unsigned int flags)
271		267
272	This will initialise the default event loop if it hasn't been initialised	268	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	270	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).	271	flags. If that is troubling you, check C<ev_backend ()> afterwards).
276		272
277	If you don't know what event loop to use, use the one returned from this	273	If you don't know what event loop to use, use the one returned from this
278	function.	274	function.
		275
		276	Note that this function is I<not> thread-safe, so if you want to use it
		277	from multiple threads, you have to lock (note also that this is unlikely,
		278	as loops cannot bes hared easily between threads anyway).
279		279
280	The default loop is the only loop that can handle C<ev_signal> and	280	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	281	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	282	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	283	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	336	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	337	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	338	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	339	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	340	a look at C<ev_set_io_collect_interval ()> to increase the amount of
341	readyness notifications you get per iteration.	341	readiness notifications you get per iteration.
342		342
343	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	343	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
344		344
345	And this is your standard poll(2) backend. It's more complicated	345	And this is your standard poll(2) backend. It's more complicated
346	than select, but handles sparse fds better and has no artificial	346	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,	354	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	355	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),	356	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	357	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	358	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	359	cases and requiring a syscall per fd change, no fork support and bad
360	support for dup.	360	support for dup.
361		361
362	While stopping, setting and starting an I/O watcher in the same iteration	362	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	363	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	364	(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	425	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	426	file descriptor per loop iteration. For small and medium numbers of file
427	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	427	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
428	might perform better.	428	might perform better.
429		429
430	On the positive side, ignoring the spurious readyness notifications, this	430	On the positive side, ignoring the spurious readiness notifications, this
431	backend actually performed to specification in all tests and is fully	431	backend actually performed to specification in all tests and is fully
432	embeddable, which is a rare feat among the OS-specific backends.	432	embeddable, which is a rare feat among the OS-specific backends.
433		433
434	=item C<EVBACKEND_ALL>	434	=item C<EVBACKEND_ALL>
435		435
…		…
465		465
466	Similar to C<ev_default_loop>, but always creates a new event loop that is	466	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	467	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	468	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).	469	undefined behaviour (or a failed assertion if assertions are enabled).
		470
		471	Note that this function I<is> thread-safe, and the recommended way to use
		472	libev with threads is indeed to create one loop per thread, and using the
		473	default loop in the "main" or "initial" thread.
470		474
471	Example: Try to create a event loop that uses epoll and nothing else.	475	Example: Try to create a event loop that uses epoll and nothing else.
472		476
473	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);	477	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);
474	if (!epoller)	478	if (!epoller)
…		…
685	interval to a value near C<0.1> or so, which is often enough for	689	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	690	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>,	691	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.	692	as this approsaches the timing granularity of most systems.
689		693
		694	=item ev_loop_verify (loop)
		695
		696	This function only does something when C<EV_VERIFY> support has been
		697	compiled in. It tries to go through all internal structures and checks
		698	them for validity. If anything is found to be inconsistent, it will print
		699	an error message to standard error and call C<abort ()>.
		700
		701	This can be used to catch bugs inside libev itself: under normal
		702	circumstances, this function will never abort as of course libev keeps its
		703	data structures consistent.
		704
690	=back	705	=back
691		706
692		707
693	=head1 ANATOMY OF A WATCHER	708	=head1 ANATOMY OF A WATCHER
694		709
…		…
1028	If you must do this, then force the use of a known-to-be-good backend	1043	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	1044	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
1030	C<EVBACKEND_POLL>).	1045	C<EVBACKEND_POLL>).
1031		1046
1032	Another thing you have to watch out for is that it is quite easy to	1047	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	1048	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	1049	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	1050	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	1051	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	1052	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	1053	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	1100	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,	1101	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	1102	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1088	C<EVBACKEND_POLL>.	1103	C<EVBACKEND_POLL>.
1089		1104
		1105	=head3 The special problem of SIGPIPE
		1106
		1107	While not really specific to libev, it is easy to forget about SIGPIPE:
		1108	when reading from a pipe whose other end has been closed, your program
		1109	gets send a SIGPIPE, which, by default, aborts your program. For most
		1110	programs this is sensible behaviour, for daemons, this is usually
		1111	undesirable.
		1112
		1113	So when you encounter spurious, unexplained daemon exits, make sure you
		1114	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1115	somewhere, as that would have given you a big clue).
		1116
1090		1117
1091	=head3 Watcher-Specific Functions	1118	=head3 Watcher-Specific Functions
1092		1119
1093	=over 4	1120	=over 4
1094		1121
…		…
1135		1162
1136	Timer watchers are simple relative timers that generate an event after a	1163	Timer watchers are simple relative timers that generate an event after a
1137	given time, and optionally repeating in regular intervals after that.	1164	given time, and optionally repeating in regular intervals after that.
1138		1165
1139	The timers are based on real time, that is, if you register an event that	1166	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	1167	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	1168	year, it will still time out after (roughly) and hour. "Roughly" because
1142	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1169	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1143	monotonic clock option helps a lot here).	1170	monotonic clock option helps a lot here).
1144		1171
1145	The relative timeouts are calculated relative to the C<ev_now ()>	1172	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	1173	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	1175	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:	1176	on the current time, use something like this to adjust for this:
1150		1177
1151	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1178	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1152		1179
1153	The callback is guarenteed to be invoked only when its timeout has passed,	1180	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	1181	but if multiple timers become ready during the same loop iteration then
1155	order of execution is undefined.	1182	order of execution is undefined.
1156		1183
1157	=head3 Watcher-Specific Functions and Data Members	1184	=head3 Watcher-Specific Functions and Data Members
1158		1185
…		…
1160		1187
1161	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1188	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1162		1189
1163	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	1190	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1164		1191
1165	Configure the timer to trigger after C<after> seconds. If C<repeat> is	1192	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	1193	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	1194	reached. If it is positive, then the timer will automatically be
1168	later, again, and again, until stopped manually.	1195	configured to trigger again C<repeat> seconds later, again, and again,
		1196	until stopped manually.
1169		1197
1170	The timer itself will do a best-effort at avoiding drift, that is, if you	1198	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	1199	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	1200	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	1201	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.	1202	do stuff) the timer will not fire more than once per event loop iteration.
1175		1203
1176	=item ev_timer_again (loop, ev_timer *)	1204	=item ev_timer_again (loop, ev_timer *)
1177		1205
1178	This will act as if the timer timed out and restart it again if it is	1206	This will act as if the timer timed out and restart it again if it is
1179	repeating. The exact semantics are:	1207	repeating. The exact semantics are:
…		…
1256	Periodic watchers are also timers of a kind, but they are very versatile	1284	Periodic watchers are also timers of a kind, but they are very versatile
1257	(and unfortunately a bit complex).	1285	(and unfortunately a bit complex).
1258		1286
1259	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1287	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	1288	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	1289	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 ()	1290	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	1291	+ 10.>, that is, an absolute time not a delay) and then reset your system
		1292	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	1293	to trigger the event (unlike an C<ev_timer>, which would still trigger
1265	roughly 10 seconds later).	1294	roughly 10 seconds later as it uses a relative timeout).
1266		1295
1267	They can also be used to implement vastly more complex timers, such as	1296	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,	1297	such as triggering an event on each "midnight, local time", or other
1269	rules.	1298	complicated, rules.
1270		1299
1271	As with timers, the callback is guarenteed to be invoked only when the	1300	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	1301	time (C<at>) has passed, but if multiple periodic timers become ready
1273	during the same loop iteration then order of execution is undefined.	1302	during the same loop iteration then order of execution is undefined.
1274		1303
1275	=head3 Watcher-Specific Functions and Data Members	1304	=head3 Watcher-Specific Functions and Data Members
1276		1305
1277	=over 4	1306	=over 4
…		…
1285		1314
1286	=over 4	1315	=over 4
1287		1316
1288	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1317	=item * absolute timer (at = time, interval = reschedule_cb = 0)
1289		1318
1290	In this configuration the watcher triggers an event at the wallclock time	1319	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,	1320	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	1321	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.	1322	run when the system time reaches or surpasses this time.
1294		1323
1295	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1324	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)
1296		1325
1297	In this mode the watcher will always be scheduled to time out at the next	1326	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)	1327	C<at + N * interval> time (for some integer N, which can also be negative)
1299	and then repeat, regardless of any time jumps.	1328	and then repeat, regardless of any time jumps.
1300		1329
1301	This can be used to create timers that do not drift with respect to system	1330	This can be used to create timers that do not drift with respect to system
1302	time:	1331	time, for example, here is a C<ev_periodic> that triggers each hour, on
		1332	the hour:
1303		1333
1304	ev_periodic_set (&periodic, 0., 3600., 0);	1334	ev_periodic_set (&periodic, 0., 3600., 0);
1305		1335
1306	This doesn't mean there will always be 3600 seconds in between triggers,	1336	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	1337	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	1342	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.	1343	time where C<time = at (mod interval)>, regardless of any time jumps.
1314		1344
1315	For numerical stability it is preferable that the C<at> value is near	1345	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	1346	C<ev_now ()> (the current time), but there is no range requirement for
1317	this value.	1347	this value, and in fact is often specified as zero.
		1348
		1349	Note also that there is an upper limit to how often a timer can fire (cpu
		1350	speed for example), so if C<interval> is very small then timing stability
		1351	will of course detoriate. Libev itself tries to be exact to be about one
		1352	millisecond (if the OS supports it and the machine is fast enough).
1318		1353
1319	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1354	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)
1320		1355
1321	In this mode the values for C<interval> and C<at> are both being	1356	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	1357	ignored. Instead, each time the periodic watcher gets scheduled, the
1323	reschedule callback will be called with the watcher as first, and the	1358	reschedule callback will be called with the watcher as first, and the
1324	current time as second argument.	1359	current time as second argument.
1325		1360
1326	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1361	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,	1362	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		1363
		1364	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		1365	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		1366	only event loop modification you are allowed to do).
		1367
1331	Its prototype is C<ev_tstamp (reschedule_cb)(struct ev_periodic w,	1368	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic
1332	ev_tstamp now)>, e.g.:	1369	*w, ev_tstamp now)>, e.g.:
1333		1370
1334	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1371	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
1335	{	1372	{
1336	return now + 60.;	1373	return now + 60.;
1337	}	1374	}
…		…
1339	It must return the next time to trigger, based on the passed time value	1376	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	1377	(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	1378	will usually be called just before the callback will be triggered, but
1342	might be called at other times, too.	1379	might be called at other times, too.
1343		1380
1344	NOTE: I<< This callback must always return a time that is later than the	1381	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.	1382	equal to the passed C<now> value >>.
1346		1383
1347	This can be used to create very complex timers, such as a timer that	1384	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	1385	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	1386	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	1387	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).	1388	reason I omitted it as an example).
1352		1389
1353	=back	1390	=back
…		…
1357	Simply stops and restarts the periodic watcher again. This is only useful	1394	Simply stops and restarts the periodic watcher again. This is only useful
1358	when you changed some parameters or the reschedule callback would return	1395	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	1396	a different time than the last time it was called (e.g. in a crond like
1360	program when the crontabs have changed).	1397	program when the crontabs have changed).
1361		1398
		1399	=item ev_tstamp ev_periodic_at (ev_periodic *)
		1400
		1401	When active, returns the absolute time that the watcher is supposed to
		1402	trigger next.
		1403
1362	=item ev_tstamp offset [read-write]	1404	=item ev_tstamp offset [read-write]
1363		1405
1364	When repeating, this contains the offset value, otherwise this is the	1406	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>).	1407	absolute point in time (the C<at> value passed to C<ev_periodic_set>).
1366		1408
…		…
1376	=item ev_tstamp (reschedule_cb)(struct ev_periodic w, ev_tstamp now) [read-write]	1418	=item ev_tstamp (reschedule_cb)(struct ev_periodic w, ev_tstamp now) [read-write]
1377		1419
1378	The current reschedule callback, or C<0>, if this functionality is	1420	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	1421	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.	1422	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		1423
1387	=back	1424	=back
1388		1425
1389	=head3 Examples	1426	=head3 Examples
1390		1427
…		…
1594	as even with OS-supported change notifications, this can be	1631	as even with OS-supported change notifications, this can be
1595	resource-intensive.	1632	resource-intensive.
1596		1633
1597	At the time of this writing, only the Linux inotify interface is	1634	At the time of this writing, only the Linux inotify interface is
1598	implemented (implementing kqueue support is left as an exercise for the	1635	implemented (implementing kqueue support is left as an exercise for the
		1636	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	1637	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	1638	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	1639	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	1640	but changes are usually detected immediately, and if the file exists there
1603	polling.	1641	will be no polling.
		1642
		1643	=head3 ABI Issues (Largefile Support)
		1644
		1645	Libev by default (unless the user overrides this) uses the default
		1646	compilation environment, which means that on systems with optionally
		1647	disabled large file support, you get the 32 bit version of the stat
		1648	structure. When using the library from programs that change the ABI to
		1649	use 64 bit file offsets the programs will fail. In that case you have to
		1650	compile libev with the same flags to get binary compatibility. This is
		1651	obviously the case with any flags that change the ABI, but the problem is
		1652	most noticably with ev_stat and largefile support.
1604		1653
1605	=head3 Inotify	1654	=head3 Inotify
1606		1655
1607	When C<inotify (7)> support has been compiled into libev (generally only	1656	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	1657	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	1658	change detection where possible. The inotify descriptor will be created lazily
1610	when the first C<ev_stat> watcher is being started.	1659	when the first C<ev_stat> watcher is being started.
1611		1660
1612	Inotify presense does not change the semantics of C<ev_stat> watchers	1661	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	1662	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	1663	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.	1664	there are many cases where libev has to resort to regular C<stat> polling.
1616		1665
1617	(There is no support for kqueue, as apparently it cannot be used to	1666	(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	1667	implement this functionality, due to the requirement of having a file
1619	descriptor open on the object at all times).	1668	descriptor open on the object at all times).
…		…
1622		1671
1623	The C<stat ()> syscall only supports full-second resolution portably, and	1672	The C<stat ()> syscall only supports full-second resolution portably, and
1624	even on systems where the resolution is higher, many filesystems still	1673	even on systems where the resolution is higher, many filesystems still
1625	only support whole seconds.	1674	only support whole seconds.
1626		1675
1627	That means that, if the time is the only thing that changes, you might	1676	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	1677	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	1678	calls your callback, which does something. When there is another update
1630	the same second, C<ev_stat> will be unable to detect it.	1679	within the same second, C<ev_stat> will be unable to detect it as the stat
		1680	data does not change.
1631		1681
1632	The solution to this is to delay acting on a change for a second (or till	1682	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>	1683	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>	1684	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	1685	ev_timer_again (loop, w)>).
1636	systems.	1686
		1687	The C<.02> offset is added to work around small timing inconsistencies
		1688	of some operating systems (where the second counter of the current time
		1689	might be be delayed. One such system is the Linux kernel, where a call to
		1690	C<gettimeofday> might return a timestamp with a full second later than
		1691	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		1692	update file times then there will be a small window where the kernel uses
		1693	the previous second to update file times but libev might already execute
		1694	the timer callback).
1637		1695
1638	=head3 Watcher-Specific Functions and Data Members	1696	=head3 Watcher-Specific Functions and Data Members
1639		1697
1640	=over 4	1698	=over 4
1641		1699
…		…
1647	C<path>. The C<interval> is a hint on how quickly a change is expected to	1705	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	1706	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	1707	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.	1708	path for as long as the watcher is active.
1651		1709
1652	The callback will be receive C<EV_STAT> when a change was detected,	1710	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	1711	to the attributes at the time the watcher was started (or the last change
1654	last change was detected).	1712	was detected).
1655		1713
1656	=item ev_stat_stat (loop, ev_stat *)	1714	=item ev_stat_stat (loop, ev_stat *)
1657		1715
1658	Updates the stat buffer immediately with new values. If you change the	1716	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	1717	watched path in your callback, you could call this function to avoid
1660	detecting this change (while introducing a race condition). Can also be	1718	detecting this change (while introducing a race condition if you are not
1661	useful simply to find out the new values.	1719	the only one changing the path). Can also be useful simply to find out the
		1720	new values.
1662		1721
1663	=item ev_statdata attr [read-only]	1722	=item ev_statdata attr [read-only]
1664		1723
1665	The most-recently detected attributes of the file. Although the type is of	1724	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	1725	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	1726	suitable for your system, but you can only rely on the POSIX-standardised
		1727	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.	1728	some error while C<stat>ing the file.
1669		1729
1670	=item ev_statdata prev [read-only]	1730	=item ev_statdata prev [read-only]
1671		1731
1672	The previous attributes of the file. The callback gets invoked whenever	1732	The previous attributes of the file. The callback gets invoked whenever
1673	C<prev> != C<attr>.	1733	C<prev> != C<attr>, or, more precisely, one or more of these members
		1734	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		1735	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
1674		1736
1675	=item ev_tstamp interval [read-only]	1737	=item ev_tstamp interval [read-only]
1676		1738
1677	The specified interval.	1739	The specified interval.
1678		1740
…		…
1732	}	1794	}
1733		1795
1734	...	1796	...
1735	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);	1797	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1736	ev_stat_start (loop, &passwd);	1798	ev_stat_start (loop, &passwd);
1737	ev_timer_init (&timer, timer_cb, 0., 1.01);	1799	ev_timer_init (&timer, timer_cb, 0., 1.02);
1738		1800
1739		1801
1740	=head2 C<ev_idle> - when you've got nothing better to do...	1802	=head2 C<ev_idle> - when you've got nothing better to do...
1741		1803
1742	Idle watchers trigger events when no other events of the same or higher	1804	Idle watchers trigger events when no other events of the same or higher
…		…
1830		1892
1831	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	1893	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	1894	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,	1895	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	1896	too) should not activate ("feed") events into libev. While libev fully
1835	supports this, they will be called before other C<ev_check> watchers	1897	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	1898	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	1899	(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	1900	state until their C<ev_check> watcher ran (always remind yourself to
1839	coexist peacefully with others).	1901	coexist peacefully with others).
1840		1902
…		…
1855	=head3 Examples	1917	=head3 Examples
1856		1918
1857	There are a number of principal ways to embed other event loops or modules	1919	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	1920	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	1921	(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>	1922	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	1923	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
1862	into the Glib event loop).	1924	Glib event loop).
1863		1925
1864	Method 1: Add IO watchers and a timeout watcher in a prepare handler,	1926	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	1927	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	1928	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	1929	priority for the check watcher or use C<ev_clear_pending> explicitly, as
…		…
2257		2319
2258	This call incurs the overhead of a syscall only once per loop iteration,	2320	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	2321	so while the overhead might be noticable, it doesn't apply to repeated
2260	calls to C<ev_async_send>.	2322	calls to C<ev_async_send>.
2261		2323
		2324	=item bool = ev_async_pending (ev_async *)
		2325
		2326	Returns a non-zero value when C<ev_async_send> has been called on the
		2327	watcher but the event has not yet been processed (or even noted) by the
		2328	event loop.
		2329
		2330	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		2331	the loop iterates next and checks for the watcher to have become active,
		2332	it will reset the flag again. C<ev_async_pending> can be used to very
		2333	quickly check wether invoking the loop might be a good idea.
		2334
		2335	Not that this does I<not> check wether the watcher itself is pending, only
		2336	wether it has been requested to make this watcher pending.
		2337
2262	=back	2338	=back
2263		2339
2264		2340
2265	=head1 OTHER FUNCTIONS	2341	=head1 OTHER FUNCTIONS
2266		2342
…		…
2337		2413
2338	=item * Priorities are not currently supported. Initialising priorities	2414	=item * Priorities are not currently supported. Initialising priorities
2339	will fail and all watchers will have the same priority, even though there	2415	will fail and all watchers will have the same priority, even though there
2340	is an ev_pri field.	2416	is an ev_pri field.
2341		2417
		2418	=item * In libevent, the last base created gets the signals, in libev, the
		2419	first base created (== the default loop) gets the signals.
		2420
2342	=item * Other members are not supported.	2421	=item * Other members are not supported.
2343		2422
2344	=item * The libev emulation is I<not> ABI compatible to libevent, you need	2423	=item * The libev emulation is I<not> ABI compatible to libevent, you need
2345	to use the libev header file and library.	2424	to use the libev header file and library.
2346		2425
…		…
2588	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	2667	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
2589		2668
2590	Similar to the other two macros, this gives you the value of the default	2669	Similar to the other two macros, this gives you the value of the default
2591	loop, if multiple loops are supported ("ev loop default").	2670	loop, if multiple loops are supported ("ev loop default").
2592		2671
		2672	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		2673
		2674	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		2675	default loop has been initialised (C<UC> == unchecked). Their behaviour
		2676	is undefined when the default loop has not been initialised by a previous
		2677	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		2678
		2679	It is often prudent to use C<EV_DEFAULT> when initialising the first
		2680	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		2681
2593	=back	2682	=back
2594		2683
2595	Example: Declare and initialise a check watcher, utilising the above	2684	Example: Declare and initialise a check watcher, utilising the above
2596	macros so it will work regardless of whether multiple loops are supported	2685	macros so it will work regardless of whether multiple loops are supported
2597	or not.	2686	or not.
…		…
2692		2781
2693	libev.m4	2782	libev.m4
2694		2783
2695	=head2 PREPROCESSOR SYMBOLS/MACROS	2784	=head2 PREPROCESSOR SYMBOLS/MACROS
2696		2785
2697	Libev can be configured via a variety of preprocessor symbols you have to define	2786	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	2787	define before including any of its files. The default in the absense of
2699	and only include the select backend.	2788	autoconf is noted for every option.
2700		2789
2701	=over 4	2790	=over 4
2702		2791
2703	=item EV_STANDALONE	2792	=item EV_STANDALONE
2704		2793
…		…
2730	=item EV_USE_NANOSLEEP	2819	=item EV_USE_NANOSLEEP
2731		2820
2732	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	2821	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 ()>.	2822	and will use it for delays. Otherwise it will use C<select ()>.
2734		2823
		2824	=item EV_USE_EVENTFD
		2825
		2826	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		2827	available and will probe for kernel support at runtime. This will improve
		2828	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		2829	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		2830	2.7 or newer, otherwise disabled.
		2831
2735	=item EV_USE_SELECT	2832	=item EV_USE_SELECT
2736		2833
2737	If undefined or defined to be C<1>, libev will compile in support for the	2834	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	2835	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	2836	other method takes over, select will be it. Otherwise the select backend
…		…
2775		2872
2776	=item EV_USE_EPOLL	2873	=item EV_USE_EPOLL
2777		2874
2778	If defined to be C<1>, libev will compile in support for the Linux	2875	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,	2876	C<epoll>(7) backend. Its availability will be detected at runtime,
2780	otherwise another method will be used as fallback. This is the	2877	otherwise another method will be used as fallback. This is the preferred
2781	preferred backend for GNU/Linux systems.	2878	backend for GNU/Linux systems. If undefined, it will be enabled if the
		2879	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2782		2880
2783	=item EV_USE_KQUEUE	2881	=item EV_USE_KQUEUE
2784		2882
2785	If defined to be C<1>, libev will compile in support for the BSD style	2883	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,	2884	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
2805		2903
2806	=item EV_USE_INOTIFY	2904	=item EV_USE_INOTIFY
2807		2905
2808	If defined to be C<1>, libev will compile in support for the Linux inotify	2906	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	2907	interface to speed up C<ev_stat> watchers. Its actual availability will
2810	be detected at runtime.	2908	be detected at runtime. If undefined, it will be enabled if the headers
		2909	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
2811		2910
2812	=item EV_ATOMIC_T	2911	=item EV_ATOMIC_T
2813		2912
2814	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	2913	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	2914	access is atomic with respect to other threads or signal contexts. No such
…		…
2902	defined to be C<0>, then they are not.	3001	defined to be C<0>, then they are not.
2903		3002
2904	=item EV_MINIMAL	3003	=item EV_MINIMAL
2905		3004
2906	If you need to shave off some kilobytes of code at the expense of some	3005	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	3006	speed, define this symbol to C<1>. Currently this is used to override some
2908	some inlining decisions, saves roughly 30% codesize of amd64.	3007	inlining decisions, saves roughly 30% codesize of amd64. It also selects a
		3008	much smaller 2-heap for timer management over the default 4-heap.
2909		3009
2910	=item EV_PID_HASHSIZE	3010	=item EV_PID_HASHSIZE
2911		3011
2912	C<ev_child> watchers use a small hash table to distribute workload by	3012	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	3013	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	3019	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>),	3020	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>	3021	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	3022	watchers you might want to increase this value (I<must> be a power of
2923	two).	3023	two).
		3024
		3025	=item EV_USE_4HEAP
		3026
		3027	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3028	timer and periodics heap, libev uses a 4-heap when this symbol is defined
		3029	to C<1>. The 4-heap uses more complicated (longer) code but has
		3030	noticably faster performance with many (thousands) of watchers.
		3031
		3032	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3033	(disabled).
		3034
		3035	=item EV_HEAP_CACHE_AT
		3036
		3037	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		3038	timer and periodics heap, libev can cache the timestamp (I<at>) within
		3039	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		3040	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		3041	but avoids random read accesses on heap changes. This improves performance
		3042	noticably with with many (hundreds) of watchers.
		3043
		3044	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
		3045	(disabled).
		3046
		3047	=item EV_VERIFY
		3048
		3049	Controls how much internal verification (see C<ev_loop_verify ()>) will
		3050	be done: If set to C<0>, no internal verification code will be compiled
		3051	in. If set to C<1>, then verification code will be compiled in, but not
		3052	called. If set to C<2>, then the internal verification code will be
		3053	called once per loop, which can slow down libev. If set to C<3>, then the
		3054	verification code will be called very frequently, which will slow down
		3055	libev considerably.
		3056
		3057	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
		3058	C<0.>
2924		3059
2925	=item EV_COMMON	3060	=item EV_COMMON
2926		3061
2927	By default, all watchers have a C<void *data> member. By redefining	3062	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	3063	this macro to a something else you can include more and other types of
…		…
3002		3137
3003	#include "ev_cpp.h"	3138	#include "ev_cpp.h"
3004	#include "ev.c"	3139	#include "ev.c"
3005		3140
3006		3141
		3142	=head1 THREADS AND COROUTINES
		3143
		3144	=head2 THREADS
		3145
		3146	Libev itself is completely threadsafe, but it uses no locking. This
		3147	means that you can use as many loops as you want in parallel, as long as
		3148	only one thread ever calls into one libev function with the same loop
		3149	parameter.
		3150
		3151	Or put differently: calls with different loop parameters can be done in
		3152	parallel from multiple threads, calls with the same loop parameter must be
		3153	done serially (but can be done from different threads, as long as only one
		3154	thread ever is inside a call at any point in time, e.g. by using a mutex
		3155	per loop).
		3156
		3157	If you want to know which design is best for your problem, then I cannot
		3158	help you but by giving some generic advice:
		3159
		3160	=over 4
		3161
		3162	=item * most applications have a main thread: use the default libev loop
		3163	in that thread, or create a seperate thread running only the default loop.
		3164
		3165	This helps integrating other libraries or software modules that use libev
		3166	themselves and don't care/know about threading.
		3167
		3168	=item * one loop per thread is usually a good model.
		3169
		3170	Doing this is almost never wrong, sometimes a better-performance model
		3171	exists, but it is always a good start.
		3172
		3173	=item * other models exist, such as the leader/follower pattern, where one
		3174	loop is handed through multiple threads in a kind of round-robbin fashion.
		3175
		3176	Chosing a model is hard - look around, learn, know that usually you cna do
		3177	better than you currently do :-)
		3178
		3179	=item * often you need to talk to some other thread which blocks in the
		3180	event loop - C<ev_async> watchers can be used to wake them up from other
		3181	threads safely (or from signal contexts...).
		3182
		3183	=back
		3184
		3185	=head2 COROUTINES
		3186
		3187	Libev is much more accomodating to coroutines ("cooperative threads"):
		3188	libev fully supports nesting calls to it's functions from different
		3189	coroutines (e.g. you can call C<ev_loop> on the same loop from two
		3190	different coroutines and switch freely between both coroutines running the
		3191	loop, as long as you don't confuse yourself). The only exception is that
		3192	you must not do this from C<ev_periodic> reschedule callbacks.
		3193
		3194	Care has been invested into making sure that libev does not keep local
		3195	state inside C<ev_loop>, and other calls do not usually allow coroutine
		3196	switches.
		3197
		3198
3007	=head1 COMPLEXITIES	3199	=head1 COMPLEXITIES
3008		3200
3009	In this section the complexities of (many of) the algorithms used inside	3201	In this section the complexities of (many of) the algorithms used inside
3010	libev will be explained. For complexity discussions about backends see the	3202	libev will be explained. For complexity discussions about backends see the
3011	documentation for C<ev_default_init>.	3203	documentation for C<ev_default_init>.
…		…
3041	correct watcher to remove. The lists are usually short (you don't usually	3233	correct watcher to remove. The lists are usually short (you don't usually
3042	have many watchers waiting for the same fd or signal).	3234	have many watchers waiting for the same fd or signal).
3043		3235
3044	=item Finding the next timer in each loop iteration: O(1)	3236	=item Finding the next timer in each loop iteration: O(1)
3045		3237
3046	By virtue of using a binary heap, the next timer is always found at the	3238	By virtue of using a binary or 4-heap, the next timer is always found at a
3047	beginning of the storage array.	3239	fixed position in the storage array.
3048		3240
3049	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	3241	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3050		3242
3051	A change means an I/O watcher gets started or stopped, which requires	3243	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	3244	libev to recalculate its status (and possibly tell the kernel, depending
…		…
3081	model. Libev still offers limited functionality on this platform in	3273	model. Libev still offers limited functionality on this platform in
3082	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	3274	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
3083	descriptors. This only applies when using Win32 natively, not when using	3275	descriptors. This only applies when using Win32 natively, not when using
3084	e.g. cygwin.	3276	e.g. cygwin.
3085		3277
		3278	Lifting these limitations would basically require the full
		3279	re-implementation of the I/O system. If you are into these kinds of
		3280	things, then note that glib does exactly that for you in a very portable
		3281	way (note also that glib is the slowest event library known to man).
		3282
3086	There is no supported compilation method available on windows except	3283	There is no supported compilation method available on windows except
3087	embedding it into other applications.	3284	embedding it into other applications.
3088		3285
3089	Due to the many, low, and arbitrary limits on the win32 platform and the	3286	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	3287	the abysmal performance of winsockets, using a large number of sockets
3091	recommended (and not reasonable). If your program needs to use more than	3288	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	3289	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	3290	different implementation for windows, as libev offers the POSIX readiness
3094	be implemented efficiently on windows (microsoft monopoly games).	3291	notification model, which cannot be implemented efficiently on windows
		3292	(microsoft monopoly games).
3095		3293
3096	=over 4	3294	=over 4
3097		3295
3098	=item The winsocket select function	3296	=item The winsocket select function
3099		3297
…		…
3113	Note that winsockets handling of fd sets is O(n), so you can easily get a	3311	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.	3312	complexity in the O(n²) range when using win32.
3115		3313
3116	=item Limited number of file descriptors	3314	=item Limited number of file descriptors
3117		3315
3118	Windows has numerous arbitrary (and low) limits on things. Early versions	3316	Windows has numerous arbitrary (and low) limits on things.
3119	of winsocket's select only supported waiting for a max. of C<64> handles	3317
		3318	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	3319	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	3320	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).	3321	recommends spawning a chain of threads and wait for 63 handles and the
		3322	previous thread in each. Great).
3123		3323
3124	Newer versions support more handles, but you need to define C<FD_SETSIZE>	3324	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	3325	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	3326	call (which might be in libev or elsewhere, for example, perl does its own
3127	select emulation on windows).	3327	select emulation on windows).
…		…
3139	calling select (O(n²)) will likely make this unworkable.	3339	calling select (O(n²)) will likely make this unworkable.
3140		3340
3141	=back	3341	=back
3142		3342
3143		3343
		3344	=head1 PORTABILITY REQUIREMENTS
		3345
		3346	In addition to a working ISO-C implementation, libev relies on a few
		3347	additional extensions:
		3348
		3349	=over 4
		3350
		3351	=item C<sig_atomic_t volatile> must be thread-atomic as well
		3352
		3353	The type C<sig_atomic_t volatile> (or whatever is defined as
		3354	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different
		3355	threads. This is not part of the specification for C<sig_atomic_t>, but is
		3356	believed to be sufficiently portable.
		3357
		3358	=item C<sigprocmask> must work in a threaded environment
		3359
		3360	Libev uses C<sigprocmask> to temporarily block signals. This is not
		3361	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		3362	pthread implementations will either allow C<sigprocmask> in the "main
		3363	thread" or will block signals process-wide, both behaviours would
		3364	be compatible with libev. Interaction between C<sigprocmask> and
		3365	C<pthread_sigmask> could complicate things, however.
		3366
		3367	The most portable way to handle signals is to block signals in all threads
		3368	except the initial one, and run the default loop in the initial thread as
		3369	well.
		3370
		3371	=item C<long> must be large enough for common memory allocation sizes
		3372
		3373	To improve portability and simplify using libev, libev uses C<long>
		3374	internally instead of C<size_t> when allocating its data structures. On
		3375	non-POSIX systems (Microsoft...) this might be unexpectedly low, but
		3376	is still at least 31 bits everywhere, which is enough for hundreds of
		3377	millions of watchers.
		3378
		3379	=item C<double> must hold a time value in seconds with enough accuracy
		3380
		3381	The type C<double> is used to represent timestamps. It is required to
		3382	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
		3383	enough for at least into the year 4000. This requirement is fulfilled by
		3384	implementations implementing IEEE 754 (basically all existing ones).
		3385
		3386	=back
		3387
		3388	If you know of other additional requirements drop me a note.
		3389
		3390
		3391	=head1 VALGRIND
		3392
		3393	Valgrind has a special section here because it is a popular tool that is
		3394	highly useful, but valgrind reports are very hard to interpret.
		3395
		3396	If you think you found a bug (memory leak, uninitialised data access etc.)
		3397	in libev, then check twice: If valgrind reports something like:
		3398
		3399	==2274== definitely lost: 0 bytes in 0 blocks.
		3400	==2274== possibly lost: 0 bytes in 0 blocks.
		3401	==2274== still reachable: 256 bytes in 1 blocks.
		3402
		3403	then there is no memory leak. Similarly, under some circumstances,
		3404	valgrind might report kernel bugs as if it were a bug in libev, or it
		3405	might be confused (it is a very good tool, but only a tool).
		3406
		3407	If you are unsure about something, feel free to contact the mailing list
		3408	with the full valgrind report and an explanation on why you think this is
		3409	a bug in libev. However, don't be annoyed when you get a brisk "this is
		3410	no bug" answer and take the chance of learning how to interpret valgrind
		3411	properly.
		3412
		3413	If you need, for some reason, empty reports from valgrind for your project
		3414	I suggest using suppression lists.
		3415
		3416
3144	=head1 AUTHOR	3417	=head1 AUTHOR
3145		3418
3146	Marc Lehmann <libev@schmorp.de>.	3419	Marc Lehmann <libev@schmorp.de>.
3147		3420

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 1.159 by root, Thu May 22 02:44:57 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 1.159 by root, Thu May 22 02:44:57 2008 UTC