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

Comparing libev/ev.pod (file contents):
Revision 1.64 by root, Sat Dec 1 15:32:53 2007 UTC vs.
Revision 1.71 by root, Fri Dec 7 20:13:09 2007 UTC

…		…
47		47
48	return 0;	48	return 0;
49	}	49	}
50		50
51	=head1 DESCRIPTION	51	=head1 DESCRIPTION
		52
		53	The newest version of this document is also available as a html-formatted
		54	web page you might find easier to navigate when reading it for the first
		55	time: L<http://cvs.schmorp.de/libev/ev.html>.
52		56
53	Libev is an event loop: you register interest in certain events (such as a	57	Libev is an event loop: you register interest in certain events (such as a
54	file descriptor being readable or a timeout occuring), and it will manage	58	file descriptor being readable or a timeout occuring), and it will manage
55	these event sources and provide your program with events.	59	these event sources and provide your program with events.
56		60
…		…
274	a fork, you can also make libev check for a fork in each iteration by	278	a fork, you can also make libev check for a fork in each iteration by
275	enabling this flag.	279	enabling this flag.
276		280
277	This works by calling C<getpid ()> on every iteration of the loop,	281	This works by calling C<getpid ()> on every iteration of the loop,
278	and thus this might slow down your event loop if you do a lot of loop	282	and thus this might slow down your event loop if you do a lot of loop
279	iterations and little real work, but is usually not noticable (on my	283	iterations and little real work, but is usually not noticeable (on my
280	Linux system for example, C<getpid> is actually a simple 5-insn sequence	284	Linux system for example, C<getpid> is actually a simple 5-insn sequence
281	without a syscall and thus I<very> fast, but my Linux system also has	285	without a syscall and thus I<very> fast, but my Linux system also has
282	C<pthread_atfork> which is even faster).	286	C<pthread_atfork> which is even faster).
283		287
284	The big advantage of this flag is that you can forget about fork (and	288	The big advantage of this flag is that you can forget about fork (and
…		…
429	=item ev_loop_fork (loop)	433	=item ev_loop_fork (loop)
430		434
431	Like C<ev_default_fork>, but acts on an event loop created by	435	Like C<ev_default_fork>, but acts on an event loop created by
432	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	436	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
433	after fork, and how you do this is entirely your own problem.	437	after fork, and how you do this is entirely your own problem.
		438
		439	=item unsigned int ev_loop_count (loop)
		440
		441	Returns the count of loop iterations for the loop, which is identical to
		442	the number of times libev did poll for new events. It starts at C<0> and
		443	happily wraps around with enough iterations.
		444
		445	This value can sometimes be useful as a generation counter of sorts (it
		446	"ticks" the number of loop iterations), as it roughly corresponds with
		447	C<ev_prepare> and C<ev_check> calls.
434		448
435	=item unsigned int ev_backend (loop)	449	=item unsigned int ev_backend (loop)
436		450
437	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	451	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
438	use.	452	use.
…		…
734	=item ev_cb_set (ev_TYPE *watcher, callback)	748	=item ev_cb_set (ev_TYPE *watcher, callback)
735		749
736	Change the callback. You can change the callback at virtually any time	750	Change the callback. You can change the callback at virtually any time
737	(modulo threads).	751	(modulo threads).
738		752
		753	=item ev_set_priority (ev_TYPE *watcher, priority)
		754
		755	=item int ev_priority (ev_TYPE *watcher)
		756
		757	Set and query the priority of the watcher. The priority is a small
		758	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		759	(default: C<-2>). Pending watchers with higher priority will be invoked
		760	before watchers with lower priority, but priority will not keep watchers
		761	from being executed (except for C<ev_idle> watchers).
		762
		763	This means that priorities are I<only> used for ordering callback
		764	invocation after new events have been received. This is useful, for
		765	example, to reduce latency after idling, or more often, to bind two
		766	watchers on the same event and make sure one is called first.
		767
		768	If you need to suppress invocation when higher priority events are pending
		769	you need to look at C<ev_idle> watchers, which provide this functionality.
		770
		771	The default priority used by watchers when no priority has been set is
		772	always C<0>, which is supposed to not be too high and not be too low :).
		773
		774	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		775	fine, as long as you do not mind that the priority value you query might
		776	or might not have been adjusted to be within valid range.
		777
739	=back	778	=back
740		779
741		780
742	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	781	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
743		782
…		…
848	it is best to always use non-blocking I/O: An extra C<read>(2) returning	887	it is best to always use non-blocking I/O: An extra C<read>(2) returning
849	C<EAGAIN> is far preferable to a program hanging until some data arrives.	888	C<EAGAIN> is far preferable to a program hanging until some data arrives.
850		889
851	If you cannot run the fd in non-blocking mode (for example you should not	890	If you cannot run the fd in non-blocking mode (for example you should not
852	play around with an Xlib connection), then you have to seperately re-test	891	play around with an Xlib connection), then you have to seperately re-test
853	wether a file descriptor is really ready with a known-to-be good interface	892	whether a file descriptor is really ready with a known-to-be good interface
854	such as poll (fortunately in our Xlib example, Xlib already does this on	893	such as poll (fortunately in our Xlib example, Xlib already does this on
855	its own, so its quite safe to use).	894	its own, so its quite safe to use).
856		895
857	=over 4	896	=over 4
858		897
…		…
1341	ev_stat_start (loop, &passwd);	1380	ev_stat_start (loop, &passwd);
1342		1381
1343		1382
1344	=head2 C<ev_idle> - when you've got nothing better to do...	1383	=head2 C<ev_idle> - when you've got nothing better to do...
1345		1384
1346	Idle watchers trigger events when there are no other events are pending	1385	Idle watchers trigger events when no other events of the same or higher
1347	(prepare, check and other idle watchers do not count). That is, as long	1386	priority are pending (prepare, check and other idle watchers do not
1348	as your process is busy handling sockets or timeouts (or even signals,	1387	count).
1349	imagine) it will not be triggered. But when your process is idle all idle	1388
1350	watchers are being called again and again, once per event loop iteration -	1389	That is, as long as your process is busy handling sockets or timeouts
		1390	(or even signals, imagine) of the same or higher priority it will not be
		1391	triggered. But when your process is idle (or only lower-priority watchers
		1392	are pending), the idle watchers are being called once per event loop
1351	until stopped, that is, or your process receives more events and becomes	1393	iteration - until stopped, that is, or your process receives more events
1352	busy.	1394	and becomes busy again with higher priority stuff.
1353		1395
1354	The most noteworthy effect is that as long as any idle watchers are	1396	The most noteworthy effect is that as long as any idle watchers are
1355	active, the process will not block when waiting for new events.	1397	active, the process will not block when waiting for new events.
1356		1398
1357	Apart from keeping your process non-blocking (which is a useful	1399	Apart from keeping your process non-blocking (which is a useful
…		…
1702		1744
1703	To use it,	1745	To use it,
1704		1746
1705	#include <ev++.h>	1747	#include <ev++.h>
1706		1748
1707	(it is not installed by default). This automatically includes F<ev.h>	1749	This automatically includes F<ev.h> and puts all of its definitions (many
1708	and puts all of its definitions (many of them macros) into the global	1750	of them macros) into the global namespace. All C++ specific things are
1709	namespace. All C++ specific things are put into the C<ev> namespace.	1751	put into the C<ev> namespace. It should support all the same embedding
		1752	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1710		1753
1711	It should support all the same embedding options as F<ev.h>, most notably	1754	Care has been taken to keep the overhead low. The only data member added
1712	C<EV_MULTIPLICITY>.	1755	to the C-style watchers is the event loop the watcher is associated with
		1756	(or no additional members at all if you disable C<EV_MULTIPLICITY> when
		1757	embedding libev).
		1758
		1759	Currently, functions and static and non-static member functions can be
		1760	used as callbacks. Other types should be easy to add as long as they only
		1761	need one additional pointer for context. If you need support for other
		1762	types of functors please contact the author (preferably after implementing
		1763	it).
1713		1764
1714	Here is a list of things available in the C<ev> namespace:	1765	Here is a list of things available in the C<ev> namespace:
1715		1766
1716	=over 4	1767	=over 4
1717		1768
…		…
1733		1784
1734	All of those classes have these methods:	1785	All of those classes have these methods:
1735		1786
1736	=over 4	1787	=over 4
1737		1788
1738	=item ev::TYPE::TYPE (object , object::method )	1789	=item ev::TYPE::TYPE ()
1739		1790
1740	=item ev::TYPE::TYPE (object , object::method , struct ev_loop *)	1791	=item ev::TYPE::TYPE (struct ev_loop *)
1741		1792
1742	=item ev::TYPE::~TYPE	1793	=item ev::TYPE::~TYPE
1743		1794
1744	The constructor takes a pointer to an object and a method pointer to	1795	The constructor (optionally) takes an event loop to associate the watcher
1745	the event handler callback to call in this class. The constructor calls	1796	with. If it is omitted, it will use C<EV_DEFAULT>.
1746	C<ev_init> for you, which means you have to call the C<set> method	1797
1747	before starting it. If you do not specify a loop then the constructor	1798	The constructor calls C<ev_init> for you, which means you have to call the
1748	automatically associates the default loop with this watcher.	1799	C<set> method before starting it.
		1800
		1801	It will not set a callback, however: You have to call the templated C<set>
		1802	method to set a callback before you can start the watcher.
		1803
		1804	(The reason why you have to use a method is a limitation in C++ which does
		1805	not allow explicit template arguments for constructors).
1749		1806
1750	The destructor automatically stops the watcher if it is active.	1807	The destructor automatically stops the watcher if it is active.
		1808
		1809	=item w->set<class, &class::method> (object *)
		1810
		1811	This method sets the callback method to call. The method has to have a
		1812	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		1813	first argument and the C<revents> as second. The object must be given as
		1814	parameter and is stored in the C<data> member of the watcher.
		1815
		1816	This method synthesizes efficient thunking code to call your method from
		1817	the C callback that libev requires. If your compiler can inline your
		1818	callback (i.e. it is visible to it at the place of the C<set> call and
		1819	your compiler is good :), then the method will be fully inlined into the
		1820	thunking function, making it as fast as a direct C callback.
		1821
		1822	Example: simple class declaration and watcher initialisation
		1823
		1824	struct myclass
		1825	{
		1826	void io_cb (ev::io &w, int revents) { }
		1827	}
		1828
		1829	myclass obj;
		1830	ev::io iow;
		1831	iow.set <myclass, &myclass::io_cb> (&obj);
		1832
		1833	=item w->set (void (function)(watcher &w, int), void data = 0)
		1834
		1835	Also sets a callback, but uses a static method or plain function as
		1836	callback. The optional C<data> argument will be stored in the watcher's
		1837	C<data> member and is free for you to use.
		1838
		1839	See the method-C<set> above for more details.
1751		1840
1752	=item w->set (struct ev_loop *)	1841	=item w->set (struct ev_loop *)
1753		1842
1754	Associates a different C<struct ev_loop> with this watcher. You can only	1843	Associates a different C<struct ev_loop> with this watcher. You can only
1755	do this when the watcher is inactive (and not pending either).	1844	do this when the watcher is inactive (and not pending either).
1756		1845
1757	=item w->set ([args])	1846	=item w->set ([args])
1758		1847
1759	Basically the same as C<ev_TYPE_set>, with the same args. Must be	1848	Basically the same as C<ev_TYPE_set>, with the same args. Must be
1760	called at least once. Unlike the C counterpart, an active watcher gets	1849	called at least once. Unlike the C counterpart, an active watcher gets
1761	automatically stopped and restarted.	1850	automatically stopped and restarted when reconfiguring it with this
		1851	method.
1762		1852
1763	=item w->start ()	1853	=item w->start ()
1764		1854
1765	Starts the watcher. Note that there is no C<loop> argument as the	1855	Starts the watcher. Note that there is no C<loop> argument, as the
1766	constructor already takes the loop.	1856	constructor already stores the event loop.
1767		1857
1768	=item w->stop ()	1858	=item w->stop ()
1769		1859
1770	Stops the watcher if it is active. Again, no C<loop> argument.	1860	Stops the watcher if it is active. Again, no C<loop> argument.
1771		1861
…		…
1796		1886
1797	myclass ();	1887	myclass ();
1798	}	1888	}
1799		1889
1800	myclass::myclass (int fd)	1890	myclass::myclass (int fd)
1801	: io (this, &myclass::io_cb),
1802	idle (this, &myclass::idle_cb)
1803	{	1891	{
		1892	io .set <myclass, &myclass::io_cb > (this);
		1893	idle.set <myclass, &myclass::idle_cb> (this);
		1894
1804	io.start (fd, ev::READ);	1895	io.start (fd, ev::READ);
1805	}	1896	}
1806		1897
1807		1898
1808	=head1 MACRO MAGIC	1899	=head1 MACRO MAGIC
1809		1900
1810	Libev can be compiled with a variety of options, the most fundemantal is	1901	Libev can be compiled with a variety of options, the most fundemantal is
1811	C<EV_MULTIPLICITY>. This option determines wether (most) functions and	1902	C<EV_MULTIPLICITY>. This option determines whether (most) functions and
1812	callbacks have an initial C<struct ev_loop *> argument.	1903	callbacks have an initial C<struct ev_loop *> argument.
1813		1904
1814	To make it easier to write programs that cope with either variant, the	1905	To make it easier to write programs that cope with either variant, the
1815	following macros are defined:	1906	following macros are defined:
1816		1907
…		…
1850	loop, if multiple loops are supported ("ev loop default").	1941	loop, if multiple loops are supported ("ev loop default").
1851		1942
1852	=back	1943	=back
1853		1944
1854	Example: Declare and initialise a check watcher, utilising the above	1945	Example: Declare and initialise a check watcher, utilising the above
1855	macros so it will work regardless of wether multiple loops are supported	1946	macros so it will work regardless of whether multiple loops are supported
1856	or not.	1947	or not.
1857		1948
1858	static void	1949	static void
1859	check_cb (EV_P_ ev_timer *w, int revents)	1950	check_cb (EV_P_ ev_timer *w, int revents)
1860	{	1951	{
…		…
2085	will have the C<struct ev_loop *> as first argument, and you can create	2176	will have the C<struct ev_loop *> as first argument, and you can create
2086	additional independent event loops. Otherwise there will be no support	2177	additional independent event loops. Otherwise there will be no support
2087	for multiple event loops and there is no first event loop pointer	2178	for multiple event loops and there is no first event loop pointer
2088	argument. Instead, all functions act on the single default loop.	2179	argument. Instead, all functions act on the single default loop.
2089		2180
		2181	=item EV_MINPRI
		2182
		2183	=item EV_MAXPRI
		2184
		2185	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		2186	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		2187	provide for more priorities by overriding those symbols (usually defined
		2188	to be C<-2> and C<2>, respectively).
		2189
		2190	When doing priority-based operations, libev usually has to linearly search
		2191	all the priorities, so having many of them (hundreds) uses a lot of space
		2192	and time, so using the defaults of five priorities (-2 .. +2) is usually
		2193	fine.
		2194
		2195	If your embedding app does not need any priorities, defining these both to
		2196	C<0> will save some memory and cpu.
		2197
2090	=item EV_PERIODIC_ENABLE	2198	=item EV_PERIODIC_ENABLE
2091		2199
2092	If undefined or defined to be C<1>, then periodic timers are supported. If	2200	If undefined or defined to be C<1>, then periodic timers are supported. If
		2201	defined to be C<0>, then they are not. Disabling them saves a few kB of
		2202	code.
		2203
		2204	=item EV_IDLE_ENABLE
		2205
		2206	If undefined or defined to be C<1>, then idle watchers are supported. If
2093	defined to be C<0>, then they are not. Disabling them saves a few kB of	2207	defined to be C<0>, then they are not. Disabling them saves a few kB of
2094	code.	2208	code.
2095		2209
2096	=item EV_EMBED_ENABLE	2210	=item EV_EMBED_ENABLE
2097		2211
…		…
2190		2304
2191	In this section the complexities of (many of) the algorithms used inside	2305	In this section the complexities of (many of) the algorithms used inside
2192	libev will be explained. For complexity discussions about backends see the	2306	libev will be explained. For complexity discussions about backends see the
2193	documentation for C<ev_default_init>.	2307	documentation for C<ev_default_init>.
2194		2308
		2309	All of the following are about amortised time: If an array needs to be
		2310	extended, libev needs to realloc and move the whole array, but this
		2311	happens asymptotically never with higher number of elements, so O(1) might
		2312	mean it might do a lengthy realloc operation in rare cases, but on average
		2313	it is much faster and asymptotically approaches constant time.
		2314
2195	=over 4	2315	=over 4
2196		2316
2197	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	2317	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2198		2318
		2319	This means that, when you have a watcher that triggers in one hour and
		2320	there are 100 watchers that would trigger before that then inserting will
		2321	have to skip those 100 watchers.
		2322
2199	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)	2323	=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)
2200		2324
		2325	That means that for changing a timer costs less than removing/adding them
		2326	as only the relative motion in the event queue has to be paid for.
		2327
2201	=item Starting io/check/prepare/idle/signal/child watchers: O(1)	2328	=item Starting io/check/prepare/idle/signal/child watchers: O(1)
2202		2329
		2330	These just add the watcher into an array or at the head of a list.
2203	=item Stopping check/prepare/idle watchers: O(1)	2331	=item Stopping check/prepare/idle watchers: O(1)
2204		2332
2205	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	2333	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
2206		2334
		2335	These watchers are stored in lists then need to be walked to find the
		2336	correct watcher to remove. The lists are usually short (you don't usually
		2337	have many watchers waiting for the same fd or signal).
		2338
2207	=item Finding the next timer per loop iteration: O(1)	2339	=item Finding the next timer per loop iteration: O(1)
2208		2340
2209	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)	2341	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2210		2342
		2343	A change means an I/O watcher gets started or stopped, which requires
		2344	libev to recalculate its status (and possibly tell the kernel).
		2345
2211	=item Activating one watcher: O(1)	2346	=item Activating one watcher: O(1)
2212		2347
		2348	=item Priority handling: O(number_of_priorities)
		2349
		2350	Priorities are implemented by allocating some space for each
		2351	priority. When doing priority-based operations, libev usually has to
		2352	linearly search all the priorities.
		2353
2213	=back	2354	=back
2214		2355
2215		2356
2216	=head1 AUTHOR	2357	=head1 AUTHOR
2217		2358

Diff Legend

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

Comparing libev/ev.pod (file contents): Revision 1.64 by root, Sat Dec 1 15:32:53 2007 UTC vs. Revision 1.71 by root, Fri Dec 7 20:13:09 2007 UTC

Diff Legend

Comparing libev/ev.pod (file contents):
Revision 1.64 by root, Sat Dec 1 15:32:53 2007 UTC vs.
Revision 1.71 by root, Fri Dec 7 20:13:09 2007 UTC