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

Comparing libev/ev.pod (file contents):
Revision 1.282 by root, Wed Mar 10 08:19:39 2010 UTC vs.
Revision 1.308 by root, Thu Oct 21 02:46:59 2010 UTC

…		…
75	While this document tries to be as complete as possible in documenting	75	While this document tries to be as complete as possible in documenting
76	libev, its usage and the rationale behind its design, it is not a tutorial	76	libev, its usage and the rationale behind its design, it is not a tutorial
77	on event-based programming, nor will it introduce event-based programming	77	on event-based programming, nor will it introduce event-based programming
78	with libev.	78	with libev.
79		79
80	Familarity with event based programming techniques in general is assumed	80	Familiarity with event based programming techniques in general is assumed
81	throughout this document.	81	throughout this document.
82		82
83	=head1 ABOUT LIBEV	83	=head1 ABOUT LIBEV
84		84
85	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
…		…
124	this argument.	124	this argument.
125		125
126	=head2 TIME REPRESENTATION	126	=head2 TIME REPRESENTATION
127		127
128	Libev represents time as a single floating point number, representing	128	Libev represents time as a single floating point number, representing
129	the (fractional) number of seconds since the (POSIX) epoch (somewhere	129	the (fractional) number of seconds since the (POSIX) epoch (in practise
130	near the beginning of 1970, details are complicated, don't ask). This	130	somewhere near the beginning of 1970, details are complicated, don't
131	type is called C<ev_tstamp>, which is what you should use too. It usually	131	ask). This type is called C<ev_tstamp>, which is what you should use
132	aliases to the C<double> type in C. When you need to do any calculations	132	too. It usually aliases to the C<double> type in C. When you need to do
133	on it, you should treat it as some floating point value. Unlike the name	133	any calculations on it, you should treat it as some floating point value.
		134
134	component C<stamp> might indicate, it is also used for time differences	135	Unlike the name component C<stamp> might indicate, it is also used for
135	throughout libev.	136	time differences (e.g. delays) throughout libev.
136		137
137	=head1 ERROR HANDLING	138	=head1 ERROR HANDLING
138		139
139	Libev knows three classes of errors: operating system errors, usage errors	140	Libev knows three classes of errors: operating system errors, usage errors
140	and internal errors (bugs).	141	and internal errors (bugs).
…		…
191	as this indicates an incompatible change. Minor versions are usually	192	as this indicates an incompatible change. Minor versions are usually
192	compatible to older versions, so a larger minor version alone is usually	193	compatible to older versions, so a larger minor version alone is usually
193	not a problem.	194	not a problem.
194		195
195	Example: Make sure we haven't accidentally been linked against the wrong	196	Example: Make sure we haven't accidentally been linked against the wrong
196	version.	197	version (note, however, that this will not detect ABI mismatches :).
197		198
198	assert (("libev version mismatch",	199	assert (("libev version mismatch",
199	ev_version_major () == EV_VERSION_MAJOR	200	ev_version_major () == EV_VERSION_MAJOR
200	&& ev_version_minor () >= EV_VERSION_MINOR));	201	&& ev_version_minor () >= EV_VERSION_MINOR));
201		202
…		…
345	useful to try out specific backends to test their performance, or to work	346	useful to try out specific backends to test their performance, or to work
346	around bugs.	347	around bugs.
347		348
348	=item C<EVFLAG_FORKCHECK>	349	=item C<EVFLAG_FORKCHECK>
349		350
350	Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after	351	Instead of calling C<ev_loop_fork> manually after a fork, you can also
351	a fork, you can also make libev check for a fork in each iteration by	352	make libev check for a fork in each iteration by enabling this flag.
352	enabling this flag.
353		353
354	This works by calling C<getpid ()> on every iteration of the loop,	354	This works by calling C<getpid ()> on every iteration of the loop,
355	and thus this might slow down your event loop if you do a lot of loop	355	and thus this might slow down your event loop if you do a lot of loop
356	iterations and little real work, but is usually not noticeable (on my	356	iterations and little real work, but is usually not noticeable (on my
357	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	357	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
…		…
439	of course I<doesn't>, and epoll just loves to report events for totally	439	of course I<doesn't>, and epoll just loves to report events for totally
440	I<different> file descriptors (even already closed ones, so one cannot	440	I<different> file descriptors (even already closed ones, so one cannot
441	even remove them from the set) than registered in the set (especially	441	even remove them from the set) than registered in the set (especially
442	on SMP systems). Libev tries to counter these spurious notifications by	442	on SMP systems). Libev tries to counter these spurious notifications by
443	employing an additional generation counter and comparing that against the	443	employing an additional generation counter and comparing that against the
444	events to filter out spurious ones, recreating the set when required.	444	events to filter out spurious ones, recreating the set when required. Last
		445	not least, it also refuses to work with some file descriptors which work
		446	perfectly fine with C<select> (files, many character devices...).
445		447
446	While stopping, setting and starting an I/O watcher in the same iteration	448	While stopping, setting and starting an I/O watcher in the same iteration
447	will result in some caching, there is still a system call per such	449	will result in some caching, there is still a system call per such
448	incident (because the same I<file descriptor> could point to a different	450	incident (because the same I<file descriptor> could point to a different
449	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	451	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
567	ev_default_loop (ev_recommended_backends () \| EVBACKEND_KQUEUE);	569	ev_default_loop (ev_recommended_backends () \| EVBACKEND_KQUEUE);
568		570
569	=item struct ev_loop *ev_loop_new (unsigned int flags)	571	=item struct ev_loop *ev_loop_new (unsigned int flags)
570		572
571	Similar to C<ev_default_loop>, but always creates a new event loop that is	573	Similar to C<ev_default_loop>, but always creates a new event loop that is
572	always distinct from the default loop. Unlike the default loop, it cannot	574	always distinct from the default loop.
573	handle signal and child watchers, and attempts to do so will be greeted by
574	undefined behaviour (or a failed assertion if assertions are enabled).
575		575
576	Note that this function I<is> thread-safe, and the recommended way to use	576	Note that this function I<is> thread-safe, and one common way to use
577	libev with threads is indeed to create one loop per thread, and using the	577	libev with threads is indeed to create one loop per thread, and using the
578	default loop in the "main" or "initial" thread.	578	default loop in the "main" or "initial" thread.
579		579
580	Example: Try to create a event loop that uses epoll and nothing else.	580	Example: Try to create a event loop that uses epoll and nothing else.
581		581
…		…
583	if (!epoller)	583	if (!epoller)
584	fatal ("no epoll found here, maybe it hides under your chair");	584	fatal ("no epoll found here, maybe it hides under your chair");
585		585
586	=item ev_default_destroy ()	586	=item ev_default_destroy ()
587		587
588	Destroys the default loop again (frees all memory and kernel state	588	Destroys the default loop (frees all memory and kernel state etc.). None
589	etc.). None of the active event watchers will be stopped in the normal	589	of the active event watchers will be stopped in the normal sense, so
590	sense, so e.g. C<ev_is_active> might still return true. It is your	590	e.g. C<ev_is_active> might still return true. It is your responsibility to
591	responsibility to either stop all watchers cleanly yourself I<before>	591	either stop all watchers cleanly yourself I<before> calling this function,
592	calling this function, or cope with the fact afterwards (which is usually	592	or cope with the fact afterwards (which is usually the easiest thing, you
593	the easiest thing, you can just ignore the watchers and/or C<free ()> them	593	can just ignore the watchers and/or C<free ()> them for example).
594	for example).
595		594
596	Note that certain global state, such as signal state (and installed signal	595	Note that certain global state, such as signal state (and installed signal
597	handlers), will not be freed by this function, and related watchers (such	596	handlers), will not be freed by this function, and related watchers (such
598	as signal and child watchers) would need to be stopped manually.	597	as signal and child watchers) would need to be stopped manually.
599		598
…		…
614	name, you can call it anytime, but it makes most sense after forking, in	613	name, you can call it anytime, but it makes most sense after forking, in
615	the child process (or both child and parent, but that again makes little	614	the child process (or both child and parent, but that again makes little
616	sense). You I<must> call it in the child before using any of the libev	615	sense). You I<must> call it in the child before using any of the libev
617	functions, and it will only take effect at the next C<ev_loop> iteration.	616	functions, and it will only take effect at the next C<ev_loop> iteration.
618		617
		618	Again, you I<have> to call it on I<any> loop that you want to re-use after
		619	a fork, I<even if you do not plan to use the loop in the parent>. This is
		620	because some kernel interfaces cough I<kqueue> cough do funny things
		621	during fork.
		622
619	On the other hand, you only need to call this function in the child	623	On the other hand, you only need to call this function in the child
620	process if and only if you want to use the event library in the child. If	624	process if and only if you want to use the event loop in the child. If you
621	you just fork+exec, you don't have to call it at all.	625	just fork+exec or create a new loop in the child, you don't have to call
		626	it at all.
622		627
623	The function itself is quite fast and it's usually not a problem to call	628	The function itself is quite fast and it's usually not a problem to call
624	it just in case after a fork. To make this easy, the function will fit in	629	it just in case after a fork. To make this easy, the function will fit in
625	quite nicely into a call to C<pthread_atfork>:	630	quite nicely into a call to C<pthread_atfork>:
626		631
…		…
628		633
629	=item ev_loop_fork (loop)	634	=item ev_loop_fork (loop)
630		635
631	Like C<ev_default_fork>, but acts on an event loop created by	636	Like C<ev_default_fork>, but acts on an event loop created by
632	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	637	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
633	after fork that you want to re-use in the child, and how you do this is	638	after fork that you want to re-use in the child, and how you keep track of
634	entirely your own problem.	639	them is entirely your own problem.
635		640
636	=item int ev_is_default_loop (loop)	641	=item int ev_is_default_loop (loop)
637		642
638	Returns true when the given loop is, in fact, the default loop, and false	643	Returns true when the given loop is, in fact, the default loop, and false
639	otherwise.	644	otherwise.
640		645
641	=item unsigned int ev_loop_count (loop)	646	=item unsigned int ev_iteration (loop)
642		647
643	Returns the count of loop iterations for the loop, which is identical to	648	Returns the current iteration count for the loop, which is identical to
644	the number of times libev did poll for new events. It starts at C<0> and	649	the number of times libev did poll for new events. It starts at C<0> and
645	happily wraps around with enough iterations.	650	happily wraps around with enough iterations.
646		651
647	This value can sometimes be useful as a generation counter of sorts (it	652	This value can sometimes be useful as a generation counter of sorts (it
648	"ticks" the number of loop iterations), as it roughly corresponds with	653	"ticks" the number of loop iterations), as it roughly corresponds with
649	C<ev_prepare> and C<ev_check> calls.	654	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		655	prepare and check phases.
650		656
651	=item unsigned int ev_loop_depth (loop)	657	=item unsigned int ev_depth (loop)
652		658
653	Returns the number of times C<ev_loop> was entered minus the number of	659	Returns the number of times C<ev_loop> was entered minus the number of
654	times C<ev_loop> was exited, in other words, the recursion depth.	660	times C<ev_loop> was exited, in other words, the recursion depth.
655		661
656	Outside C<ev_loop>, this number is zero. In a callback, this number is	662	Outside C<ev_loop>, this number is zero. In a callback, this number is
657	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),	663	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
658	in which case it is higher.	664	in which case it is higher.
659		665
660	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread	666	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
661	etc.), doesn't count as exit.	667	etc.), doesn't count as "exit" - consider this as a hint to avoid such
		668	ungentleman behaviour unless it's really convenient.
662		669
663	=item unsigned int ev_backend (loop)	670	=item unsigned int ev_backend (loop)
664		671
665	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	672	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
666	use.	673	use.
…		…
700	C<ev_resume> directly afterwards to resume timer processing.	707	C<ev_resume> directly afterwards to resume timer processing.
701		708
702	Effectively, all C<ev_timer> watchers will be delayed by the time spend	709	Effectively, all C<ev_timer> watchers will be delayed by the time spend
703	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers	710	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
704	will be rescheduled (that is, they will lose any events that would have	711	will be rescheduled (that is, they will lose any events that would have
705	occured while suspended).	712	occurred while suspended).
706		713
707	After calling C<ev_suspend> you B<must not> call I<any> function on the	714	After calling C<ev_suspend> you B<must not> call I<any> function on the
708	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>	715	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
709	without a previous call to C<ev_suspend>.	716	without a previous call to C<ev_suspend>.
710		717
…		…
787	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	794	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
788	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	795	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
789		796
790	This "unloop state" will be cleared when entering C<ev_loop> again.	797	This "unloop state" will be cleared when entering C<ev_loop> again.
791		798
792	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.	799	It is safe to call C<ev_unloop> from outside any C<ev_loop> calls.
793		800
794	=item ev_ref (loop)	801	=item ev_ref (loop)
795		802
796	=item ev_unref (loop)	803	=item ev_unref (loop)
797		804
…		…
867	usually doesn't make much sense to set it to a lower value than C<0.01>,	874	usually doesn't make much sense to set it to a lower value than C<0.01>,
868	as this approaches the timing granularity of most systems. Note that if	875	as this approaches the timing granularity of most systems. Note that if
869	you do transactions with the outside world and you can't increase the	876	you do transactions with the outside world and you can't increase the
870	parallelity, then this setting will limit your transaction rate (if you	877	parallelity, then this setting will limit your transaction rate (if you
871	need to poll once per transaction and the I/O collect interval is 0.01,	878	need to poll once per transaction and the I/O collect interval is 0.01,
872	then you can't do more than 100 transations per second).	879	then you can't do more than 100 transactions per second).
873		880
874	Setting the I<timeout collect interval> can improve the opportunity for	881	Setting the I<timeout collect interval> can improve the opportunity for
875	saving power, as the program will "bundle" timer callback invocations that	882	saving power, as the program will "bundle" timer callback invocations that
876	are "near" in time together, by delaying some, thus reducing the number of	883	are "near" in time together, by delaying some, thus reducing the number of
877	times the process sleeps and wakes up again. Another useful technique to	884	times the process sleeps and wakes up again. Another useful technique to
…		…
1032	=item C<EV_WRITE>	1039	=item C<EV_WRITE>
1033		1040
1034	The file descriptor in the C<ev_io> watcher has become readable and/or	1041	The file descriptor in the C<ev_io> watcher has become readable and/or
1035	writable.	1042	writable.
1036		1043
1037	=item C<EV_TIMEOUT>	1044	=item C<EV_TIMER>
1038		1045
1039	The C<ev_timer> watcher has timed out.	1046	The C<ev_timer> watcher has timed out.
1040		1047
1041	=item C<EV_PERIODIC>	1048	=item C<EV_PERIODIC>
1042		1049
…		…
1375		1382
1376	For example, to emulate how many other event libraries handle priorities,	1383	For example, to emulate how many other event libraries handle priorities,
1377	you can associate an C<ev_idle> watcher to each such watcher, and in	1384	you can associate an C<ev_idle> watcher to each such watcher, and in
1378	the normal watcher callback, you just start the idle watcher. The real	1385	the normal watcher callback, you just start the idle watcher. The real
1379	processing is done in the idle watcher callback. This causes libev to	1386	processing is done in the idle watcher callback. This causes libev to
1380	continously poll and process kernel event data for the watcher, but when	1387	continuously poll and process kernel event data for the watcher, but when
1381	the lock-out case is known to be rare (which in turn is rare :), this is	1388	the lock-out case is known to be rare (which in turn is rare :), this is
1382	workable.	1389	workable.
1383		1390
1384	Usually, however, the lock-out model implemented that way will perform	1391	Usually, however, the lock-out model implemented that way will perform
1385	miserably under the type of load it was designed to handle. In that case,	1392	miserably under the type of load it was designed to handle. In that case,
…		…
1399	{	1406	{
1400	// stop the I/O watcher, we received the event, but	1407	// stop the I/O watcher, we received the event, but
1401	// are not yet ready to handle it.	1408	// are not yet ready to handle it.
1402	ev_io_stop (EV_A_ w);	1409	ev_io_stop (EV_A_ w);
1403		1410
1404	// start the idle watcher to ahndle the actual event.	1411	// start the idle watcher to handle the actual event.
1405	// it will not be executed as long as other watchers	1412	// it will not be executed as long as other watchers
1406	// with the default priority are receiving events.	1413	// with the default priority are receiving events.
1407	ev_idle_start (EV_A_ &idle);	1414	ev_idle_start (EV_A_ &idle);
1408	}	1415	}
1409		1416
…		…
1463		1470
1464	If you cannot use non-blocking mode, then force the use of a	1471	If you cannot use non-blocking mode, then force the use of a
1465	known-to-be-good backend (at the time of this writing, this includes only	1472	known-to-be-good backend (at the time of this writing, this includes only
1466	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file	1473	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1467	descriptors for which non-blocking operation makes no sense (such as	1474	descriptors for which non-blocking operation makes no sense (such as
1468	files) - libev doesn't guarentee any specific behaviour in that case.	1475	files) - libev doesn't guarantee any specific behaviour in that case.
1469		1476
1470	Another thing you have to watch out for is that it is quite easy to	1477	Another thing you have to watch out for is that it is quite easy to
1471	receive "spurious" readiness notifications, that is your callback might	1478	receive "spurious" readiness notifications, that is your callback might
1472	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1479	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1473	because there is no data. Not only are some backends known to create a	1480	because there is no data. Not only are some backends known to create a
…		…
1538		1545
1539	So when you encounter spurious, unexplained daemon exits, make sure you	1546	So when you encounter spurious, unexplained daemon exits, make sure you
1540	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1547	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1541	somewhere, as that would have given you a big clue).	1548	somewhere, as that would have given you a big clue).
1542		1549
		1550	=head3 The special problem of accept()ing when you can't
		1551
		1552	Many implementations of the POSIX C<accept> function (for example,
		1553	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1554	connection from the pending queue in all error cases.
		1555
		1556	For example, larger servers often run out of file descriptors (because
		1557	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1558	rejecting the connection, leading to libev signalling readiness on
		1559	the next iteration again (the connection still exists after all), and
		1560	typically causing the program to loop at 100% CPU usage.
		1561
		1562	Unfortunately, the set of errors that cause this issue differs between
		1563	operating systems, there is usually little the app can do to remedy the
		1564	situation, and no known thread-safe method of removing the connection to
		1565	cope with overload is known (to me).
		1566
		1567	One of the easiest ways to handle this situation is to just ignore it
		1568	- when the program encounters an overload, it will just loop until the
		1569	situation is over. While this is a form of busy waiting, no OS offers an
		1570	event-based way to handle this situation, so it's the best one can do.
		1571
		1572	A better way to handle the situation is to log any errors other than
		1573	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1574	messages, and continue as usual, which at least gives the user an idea of
		1575	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1576	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1577	usage.
		1578
		1579	If your program is single-threaded, then you could also keep a dummy file
		1580	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1581	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1582	close that fd, and create a new dummy fd. This will gracefully refuse
		1583	clients under typical overload conditions.
		1584
		1585	The last way to handle it is to simply log the error and C<exit>, as
		1586	is often done with C<malloc> failures, but this results in an easy
		1587	opportunity for a DoS attack.
1543		1588
1544	=head3 Watcher-Specific Functions	1589	=head3 Watcher-Specific Functions
1545		1590
1546	=over 4	1591	=over 4
1547		1592
…		…
1694	ev_tstamp timeout = last_activity + 60.;	1739	ev_tstamp timeout = last_activity + 60.;
1695		1740
1696	// if last_activity + 60. is older than now, we did time out	1741	// if last_activity + 60. is older than now, we did time out
1697	if (timeout < now)	1742	if (timeout < now)
1698	{	1743	{
1699	// timeout occured, take action	1744	// timeout occurred, take action
1700	}	1745	}
1701	else	1746	else
1702	{	1747	{
1703	// callback was invoked, but there was some activity, re-arm	1748	// callback was invoked, but there was some activity, re-arm
1704	// the watcher to fire in last_activity + 60, which is	1749	// the watcher to fire in last_activity + 60, which is
…		…
1726	to the current time (meaning we just have some activity :), then call the	1771	to the current time (meaning we just have some activity :), then call the
1727	callback, which will "do the right thing" and start the timer:	1772	callback, which will "do the right thing" and start the timer:
1728		1773
1729	ev_init (timer, callback);	1774	ev_init (timer, callback);
1730	last_activity = ev_now (loop);	1775	last_activity = ev_now (loop);
1731	callback (loop, timer, EV_TIMEOUT);	1776	callback (loop, timer, EV_TIMER);
1732		1777
1733	And when there is some activity, simply store the current time in	1778	And when there is some activity, simply store the current time in
1734	C<last_activity>, no libev calls at all:	1779	C<last_activity>, no libev calls at all:
1735		1780
1736	last_actiivty = ev_now (loop);	1781	last_activity = ev_now (loop);
1737		1782
1738	This technique is slightly more complex, but in most cases where the	1783	This technique is slightly more complex, but in most cases where the
1739	time-out is unlikely to be triggered, much more efficient.	1784	time-out is unlikely to be triggered, much more efficient.
1740		1785
1741	Changing the timeout is trivial as well (if it isn't hard-coded in the	1786	Changing the timeout is trivial as well (if it isn't hard-coded in the
…		…
2080	Example: Call a callback every hour, or, more precisely, whenever the	2125	Example: Call a callback every hour, or, more precisely, whenever the
2081	system time is divisible by 3600. The callback invocation times have	2126	system time is divisible by 3600. The callback invocation times have
2082	potentially a lot of jitter, but good long-term stability.	2127	potentially a lot of jitter, but good long-term stability.
2083		2128
2084	static void	2129	static void
2085	clock_cb (struct ev_loop loop, ev_io w, int revents)	2130	clock_cb (struct ev_loop loop, ev_periodic w, int revents)
2086	{	2131	{
2087	... its now a full hour (UTC, or TAI or whatever your clock follows)	2132	... its now a full hour (UTC, or TAI or whatever your clock follows)
2088	}	2133	}
2089		2134
2090	ev_periodic hourly_tick;	2135	ev_periodic hourly_tick;
…		…
2922	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2967	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2923	handlers will be invoked, too, of course.	2968	handlers will be invoked, too, of course.
2924		2969
2925	=head3 The special problem of life after fork - how is it possible?	2970	=head3 The special problem of life after fork - how is it possible?
2926		2971
2927	Most uses of C<fork()> consist of forking, then some simple calls to ste	2972	Most uses of C<fork()> consist of forking, then some simple calls to set
2928	up/change the process environment, followed by a call to C<exec()>. This	2973	up/change the process environment, followed by a call to C<exec()>. This
2929	sequence should be handled by libev without any problems.	2974	sequence should be handled by libev without any problems.
2930		2975
2931	This changes when the application actually wants to do event handling	2976	This changes when the application actually wants to do event handling
2932	in the child, or both parent in child, in effect "continuing" after the	2977	in the child, or both parent in child, in effect "continuing" after the
…		…
2966	believe me.	3011	believe me.
2967		3012
2968	=back	3013	=back
2969		3014
2970		3015
2971	=head2 C<ev_async> - how to wake up another event loop	3016	=head2 C<ev_async> - how to wake up an event loop
2972		3017
2973	In general, you cannot use an C<ev_loop> from multiple threads or other	3018	In general, you cannot use an C<ev_loop> from multiple threads or other
2974	asynchronous sources such as signal handlers (as opposed to multiple event	3019	asynchronous sources such as signal handlers (as opposed to multiple event
2975	loops - those are of course safe to use in different threads).	3020	loops - those are of course safe to use in different threads).
2976		3021
2977	Sometimes, however, you need to wake up another event loop you do not	3022	Sometimes, however, you need to wake up an event loop you do not control,
2978	control, for example because it belongs to another thread. This is what	3023	for example because it belongs to another thread. This is what C<ev_async>
2979	C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you	3024	watchers do: as long as the C<ev_async> watcher is active, you can signal
2980	can signal it by calling C<ev_async_send>, which is thread- and signal	3025	it by calling C<ev_async_send>, which is thread- and signal safe.
2981	safe.
2982		3026
2983	This functionality is very similar to C<ev_signal> watchers, as signals,	3027	This functionality is very similar to C<ev_signal> watchers, as signals,
2984	too, are asynchronous in nature, and signals, too, will be compressed	3028	too, are asynchronous in nature, and signals, too, will be compressed
2985	(i.e. the number of callback invocations may be less than the number of	3029	(i.e. the number of callback invocations may be less than the number of
2986	C<ev_async_sent> calls).	3030	C<ev_async_sent> calls).
…		…
3141		3185
3142	If C<timeout> is less than 0, then no timeout watcher will be	3186	If C<timeout> is less than 0, then no timeout watcher will be
3143	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3187	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
3144	repeat = 0) will be started. C<0> is a valid timeout.	3188	repeat = 0) will be started. C<0> is a valid timeout.
3145		3189
3146	The callback has the type C<void (cb)(int revents, void arg)> and gets	3190	The callback has the type C<void (cb)(int revents, void arg)> and is
3147	passed an C<revents> set like normal event callbacks (a combination of	3191	passed an C<revents> set like normal event callbacks (a combination of
3148	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3192	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
3149	value passed to C<ev_once>. Note that it is possible to receive I<both>	3193	value passed to C<ev_once>. Note that it is possible to receive I<both>
3150	a timeout and an io event at the same time - you probably should give io	3194	a timeout and an io event at the same time - you probably should give io
3151	events precedence.	3195	events precedence.
3152		3196
3153	Example: wait up to ten seconds for data to appear on STDIN_FILENO.	3197	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
3154		3198
3155	static void stdin_ready (int revents, void *arg)	3199	static void stdin_ready (int revents, void *arg)
3156	{	3200	{
3157	if (revents & EV_READ)	3201	if (revents & EV_READ)
3158	/* stdin might have data for us, joy! */;	3202	/* stdin might have data for us, joy! */;
3159	else if (revents & EV_TIMEOUT)	3203	else if (revents & EV_TIMER)
3160	/* doh, nothing entered */;	3204	/* doh, nothing entered */;
3161	}	3205	}
3162		3206
3163	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3207	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3164		3208
…		…
3298	myclass obj;	3342	myclass obj;
3299	ev::io iow;	3343	ev::io iow;
3300	iow.set <myclass, &myclass::io_cb> (&obj);	3344	iow.set <myclass, &myclass::io_cb> (&obj);
3301		3345
3302	=item w->set (object *)	3346	=item w->set (object *)
3303
3304	This is an B<experimental> feature that might go away in a future version.
3305		3347
3306	This is a variation of a method callback - leaving out the method to call	3348	This is a variation of a method callback - leaving out the method to call
3307	will default the method to C<operator ()>, which makes it possible to use	3349	will default the method to C<operator ()>, which makes it possible to use
3308	functor objects without having to manually specify the C<operator ()> all	3350	functor objects without having to manually specify the C<operator ()> all
3309	the time. Incidentally, you can then also leave out the template argument	3351	the time. Incidentally, you can then also leave out the template argument
…		…
3349	Associates a different C<struct ev_loop> with this watcher. You can only	3391	Associates a different C<struct ev_loop> with this watcher. You can only
3350	do this when the watcher is inactive (and not pending either).	3392	do this when the watcher is inactive (and not pending either).
3351		3393
3352	=item w->set ([arguments])	3394	=item w->set ([arguments])
3353		3395
3354	Basically the same as C<ev_TYPE_set>, with the same arguments. Must be	3396	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
3355	called at least once. Unlike the C counterpart, an active watcher gets	3397	method or a suitable start method must be called at least once. Unlike the
3356	automatically stopped and restarted when reconfiguring it with this	3398	C counterpart, an active watcher gets automatically stopped and restarted
3357	method.	3399	when reconfiguring it with this method.
3358		3400
3359	=item w->start ()	3401	=item w->start ()
3360		3402
3361	Starts the watcher. Note that there is no C<loop> argument, as the	3403	Starts the watcher. Note that there is no C<loop> argument, as the
3362	constructor already stores the event loop.	3404	constructor already stores the event loop.
3363		3405
		3406	=item w->start ([arguments])
		3407
		3408	Instead of calling C<set> and C<start> methods separately, it is often
		3409	convenient to wrap them in one call. Uses the same type of arguments as
		3410	the configure C<set> method of the watcher.
		3411
3364	=item w->stop ()	3412	=item w->stop ()
3365		3413
3366	Stops the watcher if it is active. Again, no C<loop> argument.	3414	Stops the watcher if it is active. Again, no C<loop> argument.
3367		3415
3368	=item w->again () (C<ev::timer>, C<ev::periodic> only)	3416	=item w->again () (C<ev::timer>, C<ev::periodic> only)
…		…
3380		3428
3381	=back	3429	=back
3382		3430
3383	=back	3431	=back
3384		3432
3385	Example: Define a class with an IO and idle watcher, start one of them in	3433	Example: Define a class with two I/O and idle watchers, start the I/O
3386	the constructor.	3434	watchers in the constructor.
3387		3435
3388	class myclass	3436	class myclass
3389	{	3437	{
3390	ev::io io ; void io_cb (ev::io &w, int revents);	3438	ev::io io ; void io_cb (ev::io &w, int revents);
		3439	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);
3391	ev::idle idle; void idle_cb (ev::idle &w, int revents);	3440	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3392		3441
3393	myclass (int fd)	3442	myclass (int fd)
3394	{	3443	{
3395	io .set <myclass, &myclass::io_cb > (this);	3444	io .set <myclass, &myclass::io_cb > (this);
		3445	io2 .set <myclass, &myclass::io2_cb > (this);
3396	idle.set <myclass, &myclass::idle_cb> (this);	3446	idle.set <myclass, &myclass::idle_cb> (this);
3397		3447
3398	io.start (fd, ev::READ);	3448	io.set (fd, ev::WRITE); // configure the watcher
		3449	io.start (); // start it whenever convenient
		3450
		3451	io2.start (fd, ev::READ); // set + start in one call
3399	}	3452	}
3400	};	3453	};
3401		3454
3402		3455
3403	=head1 OTHER LANGUAGE BINDINGS	3456	=head1 OTHER LANGUAGE BINDINGS
…		…
3622	define before including (or compiling) any of its files. The default in	3675	define before including (or compiling) any of its files. The default in
3623	the absence of autoconf is documented for every option.	3676	the absence of autoconf is documented for every option.
3624		3677
3625	Symbols marked with "(h)" do not change the ABI, and can have different	3678	Symbols marked with "(h)" do not change the ABI, and can have different
3626	values when compiling libev vs. including F<ev.h>, so it is permissible	3679	values when compiling libev vs. including F<ev.h>, so it is permissible
3627	to redefine them before including F<ev.h> without breakign compatibility	3680	to redefine them before including F<ev.h> without breaking compatibility
3628	to a compiled library. All other symbols change the ABI, which means all	3681	to a compiled library. All other symbols change the ABI, which means all
3629	users of libev and the libev code itself must be compiled with compatible	3682	users of libev and the libev code itself must be compiled with compatible
3630	settings.	3683	settings.
3631		3684
3632	=over 4	3685	=over 4
…		…
3838	fine.	3891	fine.
3839		3892
3840	If your embedding application does not need any priorities, defining these	3893	If your embedding application does not need any priorities, defining these
3841	both to C<0> will save some memory and CPU.	3894	both to C<0> will save some memory and CPU.
3842		3895
3843	=item EV_PERIODIC_ENABLE	3896	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		3897	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		3898	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3844		3899
3845	If undefined or defined to be C<1>, then periodic timers are supported. If	3900	If undefined or defined to be C<1> (and the platform supports it), then
3846	defined to be C<0>, then they are not. Disabling them saves a few kB of	3901	the respective watcher type is supported. If defined to be C<0>, then it
3847	code.	3902	is not. Disabling watcher types mainly saves code size.
3848		3903
3849	=item EV_IDLE_ENABLE	3904	=item EV_FEATURES
3850
3851	If undefined or defined to be C<1>, then idle watchers are supported. If
3852	defined to be C<0>, then they are not. Disabling them saves a few kB of
3853	code.
3854
3855	=item EV_EMBED_ENABLE
3856
3857	If undefined or defined to be C<1>, then embed watchers are supported. If
3858	defined to be C<0>, then they are not. Embed watchers rely on most other
3859	watcher types, which therefore must not be disabled.
3860
3861	=item EV_STAT_ENABLE
3862
3863	If undefined or defined to be C<1>, then stat watchers are supported. If
3864	defined to be C<0>, then they are not.
3865
3866	=item EV_FORK_ENABLE
3867
3868	If undefined or defined to be C<1>, then fork watchers are supported. If
3869	defined to be C<0>, then they are not.
3870
3871	=item EV_SIGNAL_ENABLE
3872
3873	If undefined or defined to be C<1>, then signal watchers are supported. If
3874	defined to be C<0>, then they are not.
3875
3876	=item EV_ASYNC_ENABLE
3877
3878	If undefined or defined to be C<1>, then async watchers are supported. If
3879	defined to be C<0>, then they are not.
3880
3881	=item EV_CHILD_ENABLE
3882
3883	If undefined or defined to be C<1> (and C<_WIN32> is not defined), then
3884	child watchers are supported. If defined to be C<0>, then they are not.
3885
3886	=item EV_MINIMAL
3887		3905
3888	If you need to shave off some kilobytes of code at the expense of some	3906	If you need to shave off some kilobytes of code at the expense of some
3889	speed (but with the full API), define this symbol to C<1>. Currently this	3907	speed (but with the full API), you can define this symbol to request
3890	is used to override some inlining decisions, saves roughly 30% code size	3908	certain subsets of functionality. The default is to enable all features
3891	on amd64. It also selects a much smaller 2-heap for timer management over	3909	that can be enabled on the platform.
3892	the default 4-heap.
3893		3910
3894	You can save even more by disabling watcher types you do not need	3911	A typical way to use this symbol is to define it to C<0> (or to a bitset
3895	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>	3912	with some broad features you want) and then selectively re-enable
3896	(C<-DNDEBUG>) will usually reduce code size a lot. Disabling inotify,	3913	additional parts you want, for example if you want everything minimal,
3897	eventfd and signalfd will further help, and disabling backends one doesn't	3914	but multiple event loop support, async and child watchers and the poll
3898	need (e.g. poll, epoll, kqueue, ports) will help further.	3915	backend, use this:
3899		3916
3900	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to	3917	#define EV_FEATURES 0
3901	provide a bare-bones event library. See C<ev.h> for details on what parts
3902	of the API are still available, and do not complain if this subset changes
3903	over time.
3904
3905	This example set of settings reduces the compiled size of libev from 24Kb
3906	to 8Kb on my GNU/Linux amd64 system (and leaves little in - there is also
3907	an effect on the amount of memory used). With an intelligent-enough linker
3908	further unused functions might be left out as well automatically.
3909
3910	// tuning and API changes
3911	#define EV_MINIMAL 2
3912	#define EV_MULTIPLICITY 0	3918	#define EV_MULTIPLICITY 1
3913	#define EV_MINPRI 0
3914	#define EV_MAXPRI 0
3915
3916	// OS-specific backends
3917	#define EV_USE_INOTIFY 0
3918	#define EV_USE_EVENTFD 0
3919	#define EV_USE_SIGNALFD 0
3920	#define EV_USE_REALTIME 0
3921	#define EV_USE_MONOTONIC 0
3922	#define EV_USE_CLOCK_SYSCALL 0
3923
3924	// disable all backends except select
3925	#define EV_USE_POLL 0	3919	#define EV_USE_POLL 1
3926	#define EV_USE_PORT 0
3927	#define EV_USE_KQUEUE 0
3928	#define EV_USE_EPOLL 0
3929
3930	// disable all watcher types that cna be disabled
3931	#define EV_STAT_ENABLE 0
3932	#define EV_PERIODIC_ENABLE 0
3933	#define EV_IDLE_ENABLE 0
3934	#define EV_FORK_ENABLE 0
3935	#define EV_SIGNAL_ENABLE 0
3936	#define EV_CHILD_ENABLE 0	3920	#define EV_CHILD_ENABLE 1
3937	#define EV_ASYNC_ENABLE 0	3921	#define EV_ASYNC_ENABLE 1
3938	#define EV_EMBED_ENABLE 0	3922
		3923	The actual value is a bitset, it can be a combination of the following
		3924	values:
		3925
		3926	=over 4
		3927
		3928	=item C<1> - faster/larger code
		3929
		3930	Use larger code to speed up some operations.
		3931
		3932	Currently this is used to override some inlining decisions (enlarging the
		3933	code size by roughly 30% on amd64).
		3934
		3935	When optimising for size, use of compiler flags such as C<-Os> with
		3936	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		3937	assertions.
		3938
		3939	=item C<2> - faster/larger data structures
		3940
		3941	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		3942	hash table sizes and so on. This will usually further increase code size
		3943	and can additionally have an effect on the size of data structures at
		3944	runtime.
		3945
		3946	=item C<4> - full API configuration
		3947
		3948	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		3949	enables multiplicity (C<EV_MULTIPLICITY>=1).
		3950
		3951	=item C<8> - full API
		3952
		3953	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		3954	details on which parts of the API are still available without this
		3955	feature, and do not complain if this subset changes over time.
		3956
		3957	=item C<16> - enable all optional watcher types
		3958
		3959	Enables all optional watcher types. If you want to selectively enable
		3960	only some watcher types other than I/O and timers (e.g. prepare,
		3961	embed, async, child...) you can enable them manually by defining
		3962	C<EV_watchertype_ENABLE> to C<1> instead.
		3963
		3964	=item C<32> - enable all backends
		3965
		3966	This enables all backends - without this feature, you need to enable at
		3967	least one backend manually (C<EV_USE_SELECT> is a good choice).
		3968
		3969	=item C<64> - enable OS-specific "helper" APIs
		3970
		3971	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		3972	default.
		3973
		3974	=back
		3975
		3976	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		3977	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		3978	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		3979	watchers, timers and monotonic clock support.
		3980
		3981	With an intelligent-enough linker (gcc+binutils are intelligent enough
		3982	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		3983	your program might be left out as well - a binary starting a timer and an
		3984	I/O watcher then might come out at only 5Kb.
3939		3985
3940	=item EV_AVOID_STDIO	3986	=item EV_AVOID_STDIO
3941		3987
3942	If this is set to C<1> at compiletime, then libev will avoid using stdio	3988	If this is set to C<1> at compiletime, then libev will avoid using stdio
3943	functions (printf, scanf, perror etc.). This will increase the codesize	3989	functions (printf, scanf, perror etc.). This will increase the code size
3944	somewhat, but if your program doesn't otherwise depend on stdio and your	3990	somewhat, but if your program doesn't otherwise depend on stdio and your
3945	libc allows it, this avoids linking in the stdio library which is quite	3991	libc allows it, this avoids linking in the stdio library which is quite
3946	big.	3992	big.
3947		3993
3948	Note that error messages might become less precise when this option is	3994	Note that error messages might become less precise when this option is
…		…
3952		3998
3953	The highest supported signal number, +1 (or, the number of	3999	The highest supported signal number, +1 (or, the number of
3954	signals): Normally, libev tries to deduce the maximum number of signals	4000	signals): Normally, libev tries to deduce the maximum number of signals
3955	automatically, but sometimes this fails, in which case it can be	4001	automatically, but sometimes this fails, in which case it can be
3956	specified. Also, using a lower number than detected (C<32> should be	4002	specified. Also, using a lower number than detected (C<32> should be
3957	good for about any system in existance) can save some memory, as libev	4003	good for about any system in existence) can save some memory, as libev
3958	statically allocates some 12-24 bytes per signal number.	4004	statically allocates some 12-24 bytes per signal number.
3959		4005
3960	=item EV_PID_HASHSIZE	4006	=item EV_PID_HASHSIZE
3961		4007
3962	C<ev_child> watchers use a small hash table to distribute workload by	4008	C<ev_child> watchers use a small hash table to distribute workload by
3963	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	4009	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3964	than enough. If you need to manage thousands of children you might want to	4010	usually more than enough. If you need to manage thousands of children you
3965	increase this value (I<must> be a power of two).	4011	might want to increase this value (I<must> be a power of two).
3966		4012
3967	=item EV_INOTIFY_HASHSIZE	4013	=item EV_INOTIFY_HASHSIZE
3968		4014
3969	C<ev_stat> watchers use a small hash table to distribute workload by	4015	C<ev_stat> watchers use a small hash table to distribute workload by
3970	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4016	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3971	usually more than enough. If you need to manage thousands of C<ev_stat>	4017	disabled), usually more than enough. If you need to manage thousands of
3972	watchers you might want to increase this value (I<must> be a power of	4018	C<ev_stat> watchers you might want to increase this value (I<must> be a
3973	two).	4019	power of two).
3974		4020
3975	=item EV_USE_4HEAP	4021	=item EV_USE_4HEAP
3976		4022
3977	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4023	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3978	timer and periodics heaps, libev uses a 4-heap when this symbol is defined	4024	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3979	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably	4025	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3980	faster performance with many (thousands) of watchers.	4026	faster performance with many (thousands) of watchers.
3981		4027
3982	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4028	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3983	(disabled).	4029	will be C<0>.
3984		4030
3985	=item EV_HEAP_CACHE_AT	4031	=item EV_HEAP_CACHE_AT
3986		4032
3987	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4033	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3988	timer and periodics heaps, libev can cache the timestamp (I<at>) within	4034	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3989	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4035	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3990	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4036	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3991	but avoids random read accesses on heap changes. This improves performance	4037	but avoids random read accesses on heap changes. This improves performance
3992	noticeably with many (hundreds) of watchers.	4038	noticeably with many (hundreds) of watchers.
3993		4039
3994	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4040	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3995	(disabled).	4041	will be C<0>.
3996		4042
3997	=item EV_VERIFY	4043	=item EV_VERIFY
3998		4044
3999	Controls how much internal verification (see C<ev_loop_verify ()>) will	4045	Controls how much internal verification (see C<ev_loop_verify ()>) will
4000	be done: If set to C<0>, no internal verification code will be compiled	4046	be done: If set to C<0>, no internal verification code will be compiled
…		…
4002	called. If set to C<2>, then the internal verification code will be	4048	called. If set to C<2>, then the internal verification code will be
4003	called once per loop, which can slow down libev. If set to C<3>, then the	4049	called once per loop, which can slow down libev. If set to C<3>, then the
4004	verification code will be called very frequently, which will slow down	4050	verification code will be called very frequently, which will slow down
4005	libev considerably.	4051	libev considerably.
4006		4052
4007	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4053	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
4008	C<0>.	4054	will be C<0>.
4009		4055
4010	=item EV_COMMON	4056	=item EV_COMMON
4011		4057
4012	By default, all watchers have a C<void *data> member. By redefining	4058	By default, all watchers have a C<void *data> member. By redefining
4013	this macro to a something else you can include more and other types of	4059	this macro to something else you can include more and other types of
4014	members. You have to define it each time you include one of the files,	4060	members. You have to define it each time you include one of the files,
4015	though, and it must be identical each time.	4061	though, and it must be identical each time.
4016		4062
4017	For example, the perl EV module uses something like this:	4063	For example, the perl EV module uses something like this:
4018		4064
…		…
4071	file.	4117	file.
4072		4118
4073	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4119	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
4074	that everybody includes and which overrides some configure choices:	4120	that everybody includes and which overrides some configure choices:
4075		4121
4076	#define EV_MINIMAL 1	4122	#define EV_FEATURES 8
4077	#define EV_USE_POLL 0	4123	#define EV_USE_SELECT 1
4078	#define EV_MULTIPLICITY 0
4079	#define EV_PERIODIC_ENABLE 0	4124	#define EV_PREPARE_ENABLE 1
		4125	#define EV_IDLE_ENABLE 1
4080	#define EV_STAT_ENABLE 0	4126	#define EV_SIGNAL_ENABLE 1
4081	#define EV_FORK_ENABLE 0	4127	#define EV_CHILD_ENABLE 1
		4128	#define EV_USE_STDEXCEPT 0
4082	#define EV_CONFIG_H <config.h>	4129	#define EV_CONFIG_H <config.h>
4083	#define EV_MINPRI 0
4084	#define EV_MAXPRI 0
4085		4130
4086	#include "ev++.h"	4131	#include "ev++.h"
4087		4132
4088	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4133	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4089		4134
…		…
4318	maintainable.	4363	maintainable.
4319		4364
4320	And of course, some compiler warnings are just plain stupid, or simply	4365	And of course, some compiler warnings are just plain stupid, or simply
4321	wrong (because they don't actually warn about the condition their message	4366	wrong (because they don't actually warn about the condition their message
4322	seems to warn about). For example, certain older gcc versions had some	4367	seems to warn about). For example, certain older gcc versions had some
4323	warnings that resulted an extreme number of false positives. These have	4368	warnings that resulted in an extreme number of false positives. These have
4324	been fixed, but some people still insist on making code warn-free with	4369	been fixed, but some people still insist on making code warn-free with
4325	such buggy versions.	4370	such buggy versions.
4326		4371
4327	While libev is written to generate as few warnings as possible,	4372	While libev is written to generate as few warnings as possible,
4328	"warn-free" code is not a goal, and it is recommended not to build libev	4373	"warn-free" code is not a goal, and it is recommended not to build libev
…		…
4364	I suggest using suppression lists.	4409	I suggest using suppression lists.
4365		4410
4366		4411
4367	=head1 PORTABILITY NOTES	4412	=head1 PORTABILITY NOTES
4368		4413
		4414	=head2 GNU/LINUX 32 BIT LIMITATIONS
		4415
		4416	GNU/Linux is the only common platform that supports 64 bit file/large file
		4417	interfaces but I<disables> them by default.
		4418
		4419	That means that libev compiled in the default environment doesn't support
		4420	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		4421
		4422	Unfortunately, many programs try to work around this GNU/Linux issue
		4423	by enabling the large file API, which makes them incompatible with the
		4424	standard libev compiled for their system.
		4425
		4426	Likewise, libev cannot enable the large file API itself as this would
		4427	suddenly make it incompatible to the default compile time environment,
		4428	i.e. all programs not using special compile switches.
		4429
		4430	=head2 OS/X AND DARWIN BUGS
		4431
		4432	The whole thing is a bug if you ask me - basically any system interface
		4433	you touch is broken, whether it is locales, poll, kqueue or even the
		4434	OpenGL drivers.
		4435
		4436	=head3 C<kqueue> is buggy
		4437
		4438	The kqueue syscall is broken in all known versions - most versions support
		4439	only sockets, many support pipes.
		4440
		4441	Libev tries to work around this by not using C<kqueue> by default on
		4442	this rotten platform, but of course you can still ask for it when creating
		4443	a loop.
		4444
		4445	=head3 C<poll> is buggy
		4446
		4447	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4448	implementation by something calling C<kqueue> internally around the 10.5.6
		4449	release, so now C<kqueue> I<and> C<poll> are broken.
		4450
		4451	Libev tries to work around this by not using C<poll> by default on
		4452	this rotten platform, but of course you can still ask for it when creating
		4453	a loop.
		4454
		4455	=head3 C<select> is buggy
		4456
		4457	All that's left is C<select>, and of course Apple found a way to fuck this
		4458	one up as well: On OS/X, C<select> actively limits the number of file
		4459	descriptors you can pass in to 1024 - your program suddenly crashes when
		4460	you use more.
		4461
		4462	There is an undocumented "workaround" for this - defining
		4463	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		4464	work on OS/X.
		4465
		4466	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		4467
		4468	=head3 C<errno> reentrancy
		4469
		4470	The default compile environment on Solaris is unfortunately so
		4471	thread-unsafe that you can't even use components/libraries compiled
		4472	without C<-D_REENTRANT> (as long as they use C<errno>), which, of course,
		4473	isn't defined by default.
		4474
		4475	If you want to use libev in threaded environments you have to make sure
		4476	it's compiled with C<_REENTRANT> defined.
		4477
		4478	=head3 Event port backend
		4479
		4480	The scalable event interface for Solaris is called "event ports". Unfortunately,
		4481	this mechanism is very buggy. If you run into high CPU usage, your program
		4482	freezes or you get a large number of spurious wakeups, make sure you have
		4483	all the relevant and latest kernel patches applied. No, I don't know which
		4484	ones, but there are multiple ones.
		4485
		4486	If you can't get it to work, you can try running the program by setting
		4487	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		4488	C<select> backends.
		4489
		4490	=head2 AIX POLL BUG
		4491
		4492	AIX unfortunately has a broken C<poll.h> header. Libev works around
		4493	this by trying to avoid the poll backend altogether (i.e. it's not even
		4494	compiled in), which normally isn't a big problem as C<select> works fine
		4495	with large bitsets, and AIX is dead anyway.
		4496
4369	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4497	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4498
		4499	=head3 General issues
4370		4500
4371	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4501	Win32 doesn't support any of the standards (e.g. POSIX) that libev
4372	requires, and its I/O model is fundamentally incompatible with the POSIX	4502	requires, and its I/O model is fundamentally incompatible with the POSIX
4373	model. Libev still offers limited functionality on this platform in	4503	model. Libev still offers limited functionality on this platform in
4374	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4504	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4375	descriptors. This only applies when using Win32 natively, not when using	4505	descriptors. This only applies when using Win32 natively, not when using
4376	e.g. cygwin.	4506	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		4507	as every compielr comes with a slightly differently broken/incompatible
		4508	environment.
4377		4509
4378	Lifting these limitations would basically require the full	4510	Lifting these limitations would basically require the full
4379	re-implementation of the I/O system. If you are into these kinds of	4511	re-implementation of the I/O system. If you are into this kind of thing,
4380	things, then note that glib does exactly that for you in a very portable	4512	then note that glib does exactly that for you in a very portable way (note
4381	way (note also that glib is the slowest event library known to man).	4513	also that glib is the slowest event library known to man).
4382		4514
4383	There is no supported compilation method available on windows except	4515	There is no supported compilation method available on windows except
4384	embedding it into other applications.	4516	embedding it into other applications.
4385		4517
4386	Sensible signal handling is officially unsupported by Microsoft - libev	4518	Sensible signal handling is officially unsupported by Microsoft - libev
…		…
4414	you do I<not> compile the F<ev.c> or any other embedded source files!):	4546	you do I<not> compile the F<ev.c> or any other embedded source files!):
4415		4547
4416	#include "evwrap.h"	4548	#include "evwrap.h"
4417	#include "ev.c"	4549	#include "ev.c"
4418		4550
4419	=over 4
4420
4421	=item The winsocket select function	4551	=head3 The winsocket C<select> function
4422		4552
4423	The winsocket C<select> function doesn't follow POSIX in that it	4553	The winsocket C<select> function doesn't follow POSIX in that it
4424	requires socket I<handles> and not socket I<file descriptors> (it is	4554	requires socket I<handles> and not socket I<file descriptors> (it is
4425	also extremely buggy). This makes select very inefficient, and also	4555	also extremely buggy). This makes select very inefficient, and also
4426	requires a mapping from file descriptors to socket handles (the Microsoft	4556	requires a mapping from file descriptors to socket handles (the Microsoft
…		…
4435	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */	4565	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
4436		4566
4437	Note that winsockets handling of fd sets is O(n), so you can easily get a	4567	Note that winsockets handling of fd sets is O(n), so you can easily get a
4438	complexity in the O(n²) range when using win32.	4568	complexity in the O(n²) range when using win32.
4439		4569
4440	=item Limited number of file descriptors	4570	=head3 Limited number of file descriptors
4441		4571
4442	Windows has numerous arbitrary (and low) limits on things.	4572	Windows has numerous arbitrary (and low) limits on things.
4443		4573
4444	Early versions of winsocket's select only supported waiting for a maximum	4574	Early versions of winsocket's select only supported waiting for a maximum
4445	of C<64> handles (probably owning to the fact that all windows kernels	4575	of C<64> handles (probably owning to the fact that all windows kernels
…		…
4460	runtime libraries. This might get you to about C<512> or C<2048> sockets	4590	runtime libraries. This might get you to about C<512> or C<2048> sockets
4461	(depending on windows version and/or the phase of the moon). To get more,	4591	(depending on windows version and/or the phase of the moon). To get more,
4462	you need to wrap all I/O functions and provide your own fd management, but	4592	you need to wrap all I/O functions and provide your own fd management, but
4463	the cost of calling select (O(n²)) will likely make this unworkable.	4593	the cost of calling select (O(n²)) will likely make this unworkable.
4464		4594
4465	=back
4466
4467	=head2 PORTABILITY REQUIREMENTS	4595	=head2 PORTABILITY REQUIREMENTS
4468		4596
4469	In addition to a working ISO-C implementation and of course the	4597	In addition to a working ISO-C implementation and of course the
4470	backend-specific APIs, libev relies on a few additional extensions:	4598	backend-specific APIs, libev relies on a few additional extensions:
4471		4599
…		…
4509	watchers.	4637	watchers.
4510		4638
4511	=item C<double> must hold a time value in seconds with enough accuracy	4639	=item C<double> must hold a time value in seconds with enough accuracy
4512		4640
4513	The type C<double> is used to represent timestamps. It is required to	4641	The type C<double> is used to represent timestamps. It is required to
4514	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4642	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4515	enough for at least into the year 4000. This requirement is fulfilled by	4643	good enough for at least into the year 4000 with millisecond accuracy
		4644	(the design goal for libev). This requirement is overfulfilled by
4516	implementations implementing IEEE 754, which is basically all existing	4645	implementations using IEEE 754, which is basically all existing ones. With
4517	ones. With IEEE 754 doubles, you get microsecond accuracy until at least	4646	IEEE 754 doubles, you get microsecond accuracy until at least 2200.
4518	2200.
4519		4647
4520	=back	4648	=back
4521		4649
4522	If you know of other additional requirements drop me a note.	4650	If you know of other additional requirements drop me a note.
4523		4651
…		…
4591	involves iterating over all running async watchers or all signal numbers.	4719	involves iterating over all running async watchers or all signal numbers.
4592		4720
4593	=back	4721	=back
4594		4722
4595		4723
		4724	=head1 PORTING FROM LIBEV 3.X TO 4.X
		4725
		4726	The major version 4 introduced some minor incompatible changes to the API.
		4727
		4728	At the moment, the C<ev.h> header file tries to implement superficial
		4729	compatibility, so most programs should still compile. Those might be
		4730	removed in later versions of libev, so better update early than late.
		4731
		4732	=over 4
		4733
		4734	=item C<ev_loop_count> renamed to C<ev_iteration>
		4735
		4736	=item C<ev_loop_depth> renamed to C<ev_depth>
		4737
		4738	=item C<ev_loop_verify> renamed to C<ev_verify>
		4739
		4740	Most functions working on C<struct ev_loop> objects don't have an
		4741	C<ev_loop_> prefix, so it was removed. Note that C<ev_loop_fork> is
		4742	still called C<ev_loop_fork> because it would otherwise clash with the
		4743	C<ev_fork> typedef.
		4744
		4745	=item C<EV_TIMEOUT> renamed to C<EV_TIMER> in C<revents>
		4746
		4747	This is a simple rename - all other watcher types use their name
		4748	as revents flag, and now C<ev_timer> does, too.
		4749
		4750	Both C<EV_TIMER> and C<EV_TIMEOUT> symbols were present in 3.x versions
		4751	and continue to be present for the foreseeable future, so this is mostly a
		4752	documentation change.
		4753
		4754	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		4755
		4756	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		4757	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		4758	and work, but the library code will of course be larger.
		4759
		4760	=back
		4761
		4762
4596	=head1 GLOSSARY	4763	=head1 GLOSSARY
4597		4764
4598	=over 4	4765	=over 4
4599		4766
4600	=item active	4767	=item active
…		…
4621	A change of state of some external event, such as data now being available	4788	A change of state of some external event, such as data now being available
4622	for reading on a file descriptor, time having passed or simply not having	4789	for reading on a file descriptor, time having passed or simply not having
4623	any other events happening anymore.	4790	any other events happening anymore.
4624		4791
4625	In libev, events are represented as single bits (such as C<EV_READ> or	4792	In libev, events are represented as single bits (such as C<EV_READ> or
4626	C<EV_TIMEOUT>).	4793	C<EV_TIMER>).
4627		4794
4628	=item event library	4795	=item event library
4629		4796
4630	A software package implementing an event model and loop.	4797	A software package implementing an event model and loop.
4631		4798

Diff Legend

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

Comparing libev/ev.pod (file contents): Revision 1.282 by root, Wed Mar 10 08:19:39 2010 UTC vs. Revision 1.308 by root, Thu Oct 21 02:46:59 2010 UTC

Diff Legend

Comparing libev/ev.pod (file contents):
Revision 1.282 by root, Wed Mar 10 08:19:39 2010 UTC vs.
Revision 1.308 by root, Thu Oct 21 02:46:59 2010 UTC