ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/libev/ev.pod
(Generate patch)

Comparing libev/ev.pod (file contents):
Revision 1.247 by root, Wed Jul 8 02:46:05 2009 UTC vs.
Revision 1.253 by root, Tue Jul 14 18:33:48 2009 UTC

… …
856more often than 100 times per second: 856more often than 100 times per second:
857 857
858 ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1); 858 ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
859 ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01); 859 ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
860 860
861=item ev_invoke_pending (loop)
862
863This call will simply invoke all pending watchers while resetting their
864pending state. Normally, C<ev_loop> does this automatically when required,
865but when overriding the invoke callback this call comes handy.
866
867=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
868
869This overrides the invoke pending functionality of the loop: Instead of
870invoking all pending watchers when there are any, C<ev_loop> will call
871this callback instead. This is useful, for example, when you want to
872invoke the actual watchers inside another context (another thread etc.).
873
874If you want to reset the callback, use C<ev_invoke_pending> as new
875callback.
876
877=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
878
879Sometimes you want to share the same loop between multiple threads. This
880can be done relatively simply by putting mutex_lock/unlock calls around
881each call to a libev function.
882
883However, C<ev_loop> can run an indefinite time, so it is not feasible to
884wait for it to return. One way around this is to wake up the loop via
885C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
886and I<acquire> callbacks on the loop.
887
888When set, then C<release> will be called just before the thread is
889suspended waiting for new events, and C<acquire> is called just
890afterwards.
891
892Ideally, C<release> will just call your mutex_unlock function, and
893C<acquire> will just call the mutex_lock function again.
894
895=item ev_set_userdata (loop, void *data)
896
897=item ev_userdata (loop)
898
899Set and retrieve a single C<void *> associated with a loop. When
900C<ev_set_userdata> has never been called, then C<ev_userdata> returns
901C<0.>
902
903These two functions can be used to associate arbitrary data with a loop,
904and are intended solely for the C<invoke_pending_cb>, C<release> and
905C<acquire> callbacks described above, but of course can be (ab-)used for
906any other purpose as well.
907
861=item ev_loop_verify (loop) 908=item ev_loop_verify (loop)
862 909
863This function only does something when C<EV_VERIFY> support has been 910This function only does something when C<EV_VERIFY> support has been
864compiled in, which is the default for non-minimal builds. It tries to go 911compiled in, which is the default for non-minimal builds. It tries to go
865through all internal structures and checks them for validity. If anything 912through all internal structures and checks them for validity. If anything
… …
1492 1539
1493The callback is guaranteed to be invoked only I<after> its timeout has 1540The callback is guaranteed to be invoked only I<after> its timeout has
1494passed (not I<at>, so on systems with very low-resolution clocks this 1541passed (not I<at>, so on systems with very low-resolution clocks this
1495might introduce a small delay). If multiple timers become ready during the 1542might introduce a small delay). If multiple timers become ready during the
1496same loop iteration then the ones with earlier time-out values are invoked 1543same loop iteration then the ones with earlier time-out values are invoked
1497before ones with later time-out values (but this is no longer true when a 1544before ones of the same priority with later time-out values (but this is
1498callback calls C<ev_loop> recursively). 1545no longer true when a callback calls C<ev_loop> recursively).
1499 1546
1500=head3 Be smart about timeouts 1547=head3 Be smart about timeouts
1501 1548
1502Many real-world problems involve some kind of timeout, usually for error 1549Many real-world problems involve some kind of timeout, usually for error
1503recovery. A typical example is an HTTP request - if the other side hangs, 1550recovery. A typical example is an HTTP request - if the other side hangs,
… …
2031but forking and registering a watcher a few event loop iterations later or 2078but forking and registering a watcher a few event loop iterations later or
2032in the next callback invocation is not. 2079in the next callback invocation is not.
2033 2080
2034Only the default event loop is capable of handling signals, and therefore 2081Only the default event loop is capable of handling signals, and therefore
2035you can only register child watchers in the default event loop. 2082you can only register child watchers in the default event loop.
2083
2084Due to some design glitches inside libev, child watchers will always be
2085handled at maximum priority (their priority is set to C<EV_MAXPRI> by
2086libev)
2036 2087
2037=head3 Process Interaction 2088=head3 Process Interaction
2038 2089
2039Libev grabs C<SIGCHLD> as soon as the default event loop is 2090Libev grabs C<SIGCHLD> as soon as the default event loop is
2040initialised. This is necessary to guarantee proper behaviour even if 2091initialised. This is necessary to guarantee proper behaviour even if
… …
3667defined to be C<0>, then they are not. 3718defined to be C<0>, then they are not.
3668 3719
3669=item EV_MINIMAL 3720=item EV_MINIMAL
3670 3721
3671If you need to shave off some kilobytes of code at the expense of some 3722If you need to shave off some kilobytes of code at the expense of some
3672speed, define this symbol to C<1>. Currently this is used to override some 3723speed (but with the full API), define this symbol to C<1>. Currently this
3673inlining decisions, saves roughly 30% code size on amd64. It also selects a 3724is used to override some inlining decisions, saves roughly 30% code size
3674much smaller 2-heap for timer management over the default 4-heap. 3725on amd64. It also selects a much smaller 2-heap for timer management over
3726the default 4-heap.
3727
3728You can save even more by disabling watcher types you do not need
3729and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>
3730(C<-DNDEBUG>) will usually reduce code size a lot.
3731
3732Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to
3733provide a bare-bones event library. See C<ev.h> for details on what parts
3734of the API are still available, and do not complain if this subset changes
3735over time.
3675 3736
3676=item EV_PID_HASHSIZE 3737=item EV_PID_HASHSIZE
3677 3738
3678C<ev_child> watchers use a small hash table to distribute workload by 3739C<ev_child> watchers use a small hash table to distribute workload by
3679pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more 3740pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
… …
3865default loop and triggering an C<ev_async> watcher from the default loop 3926default loop and triggering an C<ev_async> watcher from the default loop
3866watcher callback into the event loop interested in the signal. 3927watcher callback into the event loop interested in the signal.
3867 3928
3868=back 3929=back
3869 3930
3931=head4 THREAD LOCKING EXAMPLE
3932
3870=head3 COROUTINES 3933=head3 COROUTINES
3871 3934
3872Libev is very accommodating to coroutines ("cooperative threads"): 3935Libev is very accommodating to coroutines ("cooperative threads"):
3873libev fully supports nesting calls to its functions from different 3936libev fully supports nesting calls to its functions from different
3874coroutines (e.g. you can call C<ev_loop> on the same loop from two 3937coroutines (e.g. you can call C<ev_loop> on the same loop from two

Diff Legend

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