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

Comparing libev/ev.pod (file contents):
Revision 1.318 by root, Fri Oct 22 09:40:22 2010 UTC vs.
Revision 1.322 by root, Sun Oct 24 17:58:41 2010 UTC

43 43
44 int 44 int
45 main (void) 45 main (void)
46 { 46 {
47 // use the default event loop unless you have special needs 47 // use the default event loop unless you have special needs
48 struct ev_loop *loop = ev_default_loop (0); 48 struct ev_loop *loop = EV_DEFAULT;
49 49
50 // initialise an io watcher, then start it 50 // initialise an io watcher, then start it
51 // this one will watch for stdin to become readable 51 // this one will watch for stdin to become readable
52 ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ); 52 ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
53 ev_io_start (loop, &stdin_watcher); 53 ev_io_start (loop, &stdin_watcher);
165 165
166=item ev_tstamp ev_time () 166=item ev_tstamp ev_time ()
167 167
168Returns the current time as libev would use it. Please note that the 168Returns the current time as libev would use it. Please note that the
169C<ev_now> function is usually faster and also often returns the timestamp 169C<ev_now> function is usually faster and also often returns the timestamp
170you actually want to know. Also interetsing is the combination of 170you actually want to know. Also interesting is the combination of
171C<ev_update_now> and C<ev_now>. 171C<ev_update_now> and C<ev_now>.
172 172
173=item ev_sleep (ev_tstamp interval) 173=item ev_sleep (ev_tstamp interval)
174 174
175Sleep for the given interval: The current thread will be blocked until 175Sleep for the given interval: The current thread will be blocked until
193as this indicates an incompatible change. Minor versions are usually 193as this indicates an incompatible change. Minor versions are usually
194compatible to older versions, so a larger minor version alone is usually 194compatible to older versions, so a larger minor version alone is usually
195not a problem. 195not a problem.
196 196
197Example: Make sure we haven't accidentally been linked against the wrong 197Example: Make sure we haven't accidentally been linked against the wrong
198version (note, however, that this will not detect ABI mismatches :). 198version (note, however, that this will not detect other ABI mismatches,
199such as LFS or reentrancy).
199 200
200 assert (("libev version mismatch", 201 assert (("libev version mismatch",
201 ev_version_major () == EV_VERSION_MAJOR 202 ev_version_major () == EV_VERSION_MAJOR
202 && ev_version_minor () >= EV_VERSION_MINOR)); 203 && ev_version_minor () >= EV_VERSION_MINOR));
203 204
225probe for if you specify no backends explicitly. 226probe for if you specify no backends explicitly.
226 227
227=item unsigned int ev_embeddable_backends () 228=item unsigned int ev_embeddable_backends ()
228 229
229Returns the set of backends that are embeddable in other event loops. This 230Returns the set of backends that are embeddable in other event loops. This
230is the theoretical, all-platform, value. To find which backends 231value is platform-specific but can include backends not available on the
231might be supported on the current system, you would need to look at 232current system. To find which embeddable backends might be supported on
232C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for 233the current system, you would need to look at C<ev_embeddable_backends ()
233recommended ones. 234& ev_supported_backends ()>, likewise for recommended ones.
234 235
235See the description of C<ev_embed> watchers for more info. 236See the description of C<ev_embed> watchers for more info.
236 237
237=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT] 238=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
238 239
292 ... 293 ...
293 ev_set_syserr_cb (fatal_error); 294 ev_set_syserr_cb (fatal_error);
294 295
295=back 296=back
296 297
297=head1 FUNCTIONS CONTROLLING THE EVENT LOOP 298=head1 FUNCTIONS CONTROLLING EVENT LOOPS
298 299
299An event loop is described by a C<struct ev_loop *> (the C<struct> is 300An event loop is described by a C<struct ev_loop *> (the C<struct> is
300I<not> optional in this case unless libev 3 compatibility is disabled, as 301I<not> optional in this case unless libev 3 compatibility is disabled, as
301libev 3 had an C<ev_loop> function colliding with the struct name). 302libev 3 had an C<ev_loop> function colliding with the struct name).
302 303
306 307
307=over 4 308=over 4
308 309
309=item struct ev_loop *ev_default_loop (unsigned int flags) 310=item struct ev_loop *ev_default_loop (unsigned int flags)
310 311
311This will initialise the default event loop if it hasn't been initialised 312This returns the "default" event loop object, which is what you should
312yet and return it. If the default loop could not be initialised, returns 313normally use when you just need "the event loop". Event loop objects and
313false. If it already was initialised it simply returns it (and ignores the 314the C<flags> parameter are described in more detail in the entry for
314flags. If that is troubling you, check C<ev_backend ()> afterwards). 315C<ev_loop_new>.
316
317If the default loop is already initialised then this function simply
318returns it (and ignores the flags. If that is troubling you, check
319C<ev_backend ()> afterwards). Otherwise it will create it with the given
320flags, which should almost always be C<0>, unless the caller is also the
321one calling C<ev_run> or otherwise qualifies as "the main program".
315 322
316If you don't know what event loop to use, use the one returned from this 323If you don't know what event loop to use, use the one returned from this
317function. 324function (or via the C<EV_DEFAULT> macro).
318 325
319Note that this function is I<not> thread-safe, so if you want to use it 326Note that this function is I<not> thread-safe, so if you want to use it
320from multiple threads, you have to lock (note also that this is unlikely, 327from multiple threads, you have to employ some kind of mutex (note also
321as loops cannot be shared easily between threads anyway). 328that this case is unlikely, as loops cannot be shared easily between
329threads anyway).
322 330
323The default loop is the only loop that can handle C<ev_signal> and 331The default loop is the only loop that can handle C<ev_child> watchers,
324C<ev_child> watchers, and to do this, it always registers a handler 332and to do this, it always registers a handler for C<SIGCHLD>. If this is
325for C<SIGCHLD>. If this is a problem for your application you can either 333a problem for your application you can either create a dynamic loop with
326create a dynamic loop with C<ev_loop_new> that doesn't do that, or you 334C<ev_loop_new> which doesn't do that, or you can simply overwrite the
327can simply overwrite the C<SIGCHLD> signal handler I<after> calling 335C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
328C<ev_default_init>. 336
337Example: This is the most typical usage.
338
339 if (!ev_default_loop (0))
340 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
341
342Example: Restrict libev to the select and poll backends, and do not allow
343environment settings to be taken into account:
344
345 ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
346
347Example: Use whatever libev has to offer, but make sure that kqueue is
348used if available (warning, breaks stuff, best use only with your own
349private event loop and only if you know the OS supports your types of
350fds):
351
352 ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
353
354=item struct ev_loop *ev_loop_new (unsigned int flags)
355
356This will create and initialise a new event loop object. If the loop
357could not be initialised, returns false.
358
359Note that this function I<is> thread-safe, and one common way to use
360libev with threads is indeed to create one loop per thread, and using the
361default loop in the "main" or "initial" thread.
329 362
330The flags argument can be used to specify special behaviour or specific 363The flags argument can be used to specify special behaviour or specific
331backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>). 364backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
332 365
333The following flags are supported: 366The following flags are supported:
551If one or more of the backend flags are or'ed into the flags value, 584If one or more of the backend flags are or'ed into the flags value,
552then only these backends will be tried (in the reverse order as listed 585then only these backends will be tried (in the reverse order as listed
553here). If none are specified, all backends in C<ev_recommended_backends 586here). If none are specified, all backends in C<ev_recommended_backends
554()> will be tried. 587()> will be tried.
555 588
556Example: This is the most typical usage.
557
558 if (!ev_default_loop (0))
559 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
560
561Example: Restrict libev to the select and poll backends, and do not allow
562environment settings to be taken into account:
563
564 ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
565
566Example: Use whatever libev has to offer, but make sure that kqueue is
567used if available (warning, breaks stuff, best use only with your own
568private event loop and only if you know the OS supports your types of
569fds):
570
571 ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
572
573=item struct ev_loop *ev_loop_new (unsigned int flags)
574
575Similar to C<ev_default_loop>, but always creates a new event loop that is
576always distinct from the default loop.
577
578Note that this function I<is> thread-safe, and one common way to use
579libev with threads is indeed to create one loop per thread, and using the
580default loop in the "main" or "initial" thread.
581
582Example: Try to create a event loop that uses epoll and nothing else. 589Example: Try to create a event loop that uses epoll and nothing else.
583 590
584 struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV); 591 struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
585 if (!epoller) 592 if (!epoller)
586 fatal ("no epoll found here, maybe it hides under your chair"); 593 fatal ("no epoll found here, maybe it hides under your chair");
587 594
588=item ev_default_destroy () 595=item ev_loop_destroy (loop)
589 596
590Destroys the default loop (frees all memory and kernel state etc.). None 597Destroys an event loop object (frees all memory and kernel state
591of the active event watchers will be stopped in the normal sense, so 598etc.). None of the active event watchers will be stopped in the normal
592e.g. C<ev_is_active> might still return true. It is your responsibility to 599sense, so e.g. C<ev_is_active> might still return true. It is your
593either stop all watchers cleanly yourself I<before> calling this function, 600responsibility to either stop all watchers cleanly yourself I<before>
594or cope with the fact afterwards (which is usually the easiest thing, you 601calling this function, or cope with the fact afterwards (which is usually
595can just ignore the watchers and/or C<free ()> them for example). 602the easiest thing, you can just ignore the watchers and/or C<free ()> them
603for example).
596 604
597Note that certain global state, such as signal state (and installed signal 605Note that certain global state, such as signal state (and installed signal
598handlers), will not be freed by this function, and related watchers (such 606handlers), will not be freed by this function, and related watchers (such
599as signal and child watchers) would need to be stopped manually. 607as signal and child watchers) would need to be stopped manually.
600 608
601In general it is not advisable to call this function except in the 609This function is normally used on loop objects allocated by
602rare occasion where you really need to free e.g. the signal handling 610C<ev_loop_new>, but it can also be used on the default loop returned by
611C<ev_default_loop>, in which case it is not thread-safe.
612
613Note that it is not advisable to call this function on the default loop
614except in the rare occasion where you really need to free it's resources.
603pipe fds. If you need dynamically allocated loops it is better to use 615If you need dynamically allocated loops it is better to use C<ev_loop_new>
604C<ev_loop_new> and C<ev_loop_destroy>. 616and C<ev_loop_destroy>.
605 617
606=item ev_loop_destroy (loop) 618=item ev_loop_fork (loop)
607 619
608Like C<ev_default_destroy>, but destroys an event loop created by an
609earlier call to C<ev_loop_new>.
610
611=item ev_default_fork ()
612
613This function sets a flag that causes subsequent C<ev_run> iterations 620This function sets a flag that causes subsequent C<ev_run> iterations to
614to reinitialise the kernel state for backends that have one. Despite the 621reinitialise the kernel state for backends that have one. Despite the
615name, you can call it anytime, but it makes most sense after forking, in 622name, you can call it anytime, but it makes most sense after forking, in
616the child process (or both child and parent, but that again makes little 623the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
617sense). You I<must> call it in the child before using any of the libev 624child before resuming or calling C<ev_run>.
618functions, and it will only take effect at the next C<ev_run> iteration.
619 625
620Again, you I<have> to call it on I<any> loop that you want to re-use after 626Again, you I<have> to call it on I<any> loop that you want to re-use after
621a fork, I<even if you do not plan to use the loop in the parent>. This is 627a fork, I<even if you do not plan to use the loop in the parent>. This is
622because some kernel interfaces *cough* I<kqueue> *cough* do funny things 628because some kernel interfaces *cough* I<kqueue> *cough* do funny things
623during fork. 629during fork.
628call it at all (in fact, C<epoll> is so badly broken that it makes a 634call it at all (in fact, C<epoll> is so badly broken that it makes a
629difference, but libev will usually detect this case on its own and do a 635difference, but libev will usually detect this case on its own and do a
630costly reset of the backend). 636costly reset of the backend).
631 637
632The function itself is quite fast and it's usually not a problem to call 638The function itself is quite fast and it's usually not a problem to call
633it just in case after a fork. To make this easy, the function will fit in 639it just in case after a fork.
634quite nicely into a call to C<pthread_atfork>:
635 640
641Example: Automate calling C<ev_loop_fork> on the default loop when
642using pthreads.
643
644 static void
645 post_fork_child (void)
646 {
647 ev_loop_fork (EV_DEFAULT);
648 }
649
650 ...
636 pthread_atfork (0, 0, ev_default_fork); 651 pthread_atfork (0, 0, post_fork_child);
637
638=item ev_loop_fork (loop)
639
640Like C<ev_default_fork>, but acts on an event loop created by
641C<ev_loop_new>. Yes, you have to call this on every allocated event loop
642after fork that you want to re-use in the child, and how you keep track of
643them is entirely your own problem.
644 652
645=item int ev_is_default_loop (loop) 653=item int ev_is_default_loop (loop)
646 654
647Returns true when the given loop is, in fact, the default loop, and false 655Returns true when the given loop is, in fact, the default loop, and false
648otherwise. 656otherwise.
3074disadvantage of having to use multiple event loops (which do not support 3082disadvantage of having to use multiple event loops (which do not support
3075signal watchers). 3083signal watchers).
3076 3084
3077When this is not possible, or you want to use the default loop for 3085When this is not possible, or you want to use the default loop for
3078other reasons, then in the process that wants to start "fresh", call 3086other reasons, then in the process that wants to start "fresh", call
3079C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying 3087C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
3080the default loop will "orphan" (not stop) all registered watchers, so you 3088Destroying the default loop will "orphan" (not stop) all registered
3081have to be careful not to execute code that modifies those watchers. Note 3089watchers, so you have to be careful not to execute code that modifies
3082also that in that case, you have to re-register any signal watchers. 3090those watchers. Note also that in that case, you have to re-register any
3091signal watchers.
3083 3092
3084=head3 Watcher-Specific Functions and Data Members 3093=head3 Watcher-Specific Functions and Data Members
3085 3094
3086=over 4 3095=over 4
3087 3096
4829compatibility, so most programs should still compile. Those might be 4838compatibility, so most programs should still compile. Those might be
4830removed in later versions of libev, so better update early than late. 4839removed in later versions of libev, so better update early than late.
4831 4840
4832=over 4 4841=over 4
4833 4842
4843=item C<ev_default_destroy> and C<ev_default_fork> have been removed
4844
4845These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
4846
4847 ev_loop_destroy (EV_DEFAULT);
4848 ev_loop_fork (EV_DEFAULT);
4849
4834=item function/symbol renames 4850=item function/symbol renames
4835 4851
4836A number of functions and symbols have been renamed: 4852A number of functions and symbols have been renamed:
4837 4853
4838 ev_loop => ev_run 4854 ev_loop => ev_run

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines