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

Comparing libev/ev.pod (file contents):
Revision 1.273 by root, Tue Nov 24 14:46:59 2009 UTC vs.
Revision 1.405 by root, Thu May 3 15:07:15 2012 UTC

26 puts ("stdin ready"); 26 puts ("stdin ready");
27 // for one-shot events, one must manually stop the watcher 27 // for one-shot events, one must manually stop the watcher
28 // with its corresponding stop function. 28 // with its corresponding stop function.
29 ev_io_stop (EV_A_ w); 29 ev_io_stop (EV_A_ w);
30 30
31 // this causes all nested ev_loop's to stop iterating 31 // this causes all nested ev_run's to stop iterating
32 ev_unloop (EV_A_ EVUNLOOP_ALL); 32 ev_break (EV_A_ EVBREAK_ALL);
33 } 33 }
34 34
35 // another callback, this time for a time-out 35 // another callback, this time for a time-out
36 static void 36 static void
37 timeout_cb (EV_P_ ev_timer *w, int revents) 37 timeout_cb (EV_P_ ev_timer *w, int revents)
38 { 38 {
39 puts ("timeout"); 39 puts ("timeout");
40 // this causes the innermost ev_loop to stop iterating 40 // this causes the innermost ev_run to stop iterating
41 ev_unloop (EV_A_ EVUNLOOP_ONE); 41 ev_break (EV_A_ EVBREAK_ONE);
42 } 42 }
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);
56 // simple non-repeating 5.5 second timeout 56 // simple non-repeating 5.5 second timeout
57 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.); 57 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
58 ev_timer_start (loop, &timeout_watcher); 58 ev_timer_start (loop, &timeout_watcher);
59 59
60 // now wait for events to arrive 60 // now wait for events to arrive
61 ev_loop (loop, 0); 61 ev_run (loop, 0);
62 62
63 // unloop was called, so exit 63 // break was called, so exit
64 return 0; 64 return 0;
65 } 65 }
66 66
67=head1 ABOUT THIS DOCUMENT 67=head1 ABOUT THIS DOCUMENT
68 68
75While this document tries to be as complete as possible in documenting 75While this document tries to be as complete as possible in documenting
76libev, its usage and the rationale behind its design, it is not a tutorial 76libev, its usage and the rationale behind its design, it is not a tutorial
77on event-based programming, nor will it introduce event-based programming 77on event-based programming, nor will it introduce event-based programming
78with libev. 78with libev.
79 79
80Familarity with event based programming techniques in general is assumed 80Familiarity with event based programming techniques in general is assumed
81throughout this document. 81throughout this document.
82
83=head1 WHAT TO READ WHEN IN A HURRY
84
85This manual tries to be very detailed, but unfortunately, this also makes
86it very long. If you just want to know the basics of libev, I suggest
87reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and
88look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and
89C<ev_timer> sections in L<WATCHER TYPES>.
82 90
83=head1 ABOUT LIBEV 91=head1 ABOUT LIBEV
84 92
85Libev is an event loop: you register interest in certain events (such as a 93Libev is an event loop: you register interest in certain events (such as a
86file descriptor being readable or a timeout occurring), and it will manage 94file descriptor being readable or a timeout occurring), and it will manage
118Libev is very configurable. In this manual the default (and most common) 126Libev is very configurable. In this manual the default (and most common)
119configuration will be described, which supports multiple event loops. For 127configuration will be described, which supports multiple event loops. For
120more info about various configuration options please have a look at 128more info about various configuration options please have a look at
121B<EMBED> section in this manual. If libev was configured without support 129B<EMBED> section in this manual. If libev was configured without support
122for multiple event loops, then all functions taking an initial argument of 130for multiple event loops, then all functions taking an initial argument of
123name C<loop> (which is always of type C<ev_loop *>) will not have 131name C<loop> (which is always of type C<struct ev_loop *>) will not have
124this argument. 132this argument.
125 133
126=head2 TIME REPRESENTATION 134=head2 TIME REPRESENTATION
127 135
128Libev represents time as a single floating point number, representing 136Libev represents time as a single floating point number, representing
129the (fractional) number of seconds since the (POSIX) epoch (somewhere 137the (fractional) number of seconds since the (POSIX) epoch (in practice
130near the beginning of 1970, details are complicated, don't ask). This 138somewhere near the beginning of 1970, details are complicated, don't
131type is called C<ev_tstamp>, which is what you should use too. It usually 139ask). This type is called C<ev_tstamp>, which is what you should use
132aliases to the C<double> type in C. When you need to do any calculations 140too. It usually aliases to the C<double> type in C. When you need to do
133on it, you should treat it as some floating point value. Unlike the name 141any calculations on it, you should treat it as some floating point value.
142
134component C<stamp> might indicate, it is also used for time differences 143Unlike the name component C<stamp> might indicate, it is also used for
135throughout libev. 144time differences (e.g. delays) throughout libev.
136 145
137=head1 ERROR HANDLING 146=head1 ERROR HANDLING
138 147
139Libev knows three classes of errors: operating system errors, usage errors 148Libev knows three classes of errors: operating system errors, usage errors
140and internal errors (bugs). 149and internal errors (bugs).
164 173
165=item ev_tstamp ev_time () 174=item ev_tstamp ev_time ()
166 175
167Returns the current time as libev would use it. Please note that the 176Returns the current time as libev would use it. Please note that the
168C<ev_now> function is usually faster and also often returns the timestamp 177C<ev_now> function is usually faster and also often returns the timestamp
169you actually want to know. 178you actually want to know. Also interesting is the combination of
179C<ev_now_update> and C<ev_now>.
170 180
171=item ev_sleep (ev_tstamp interval) 181=item ev_sleep (ev_tstamp interval)
172 182
173Sleep for the given interval: The current thread will be blocked until 183Sleep for the given interval: The current thread will be blocked
174either it is interrupted or the given time interval has passed. Basically 184until either it is interrupted or the given time interval has
185passed (approximately - it might return a bit earlier even if not
186interrupted). Returns immediately if C<< interval <= 0 >>.
187
175this is a sub-second-resolution C<sleep ()>. 188Basically this is a sub-second-resolution C<sleep ()>.
189
190The range of the C<interval> is limited - libev only guarantees to work
191with sleep times of up to one day (C<< interval <= 86400 >>).
176 192
177=item int ev_version_major () 193=item int ev_version_major ()
178 194
179=item int ev_version_minor () 195=item int ev_version_minor ()
180 196
191as this indicates an incompatible change. Minor versions are usually 207as this indicates an incompatible change. Minor versions are usually
192compatible to older versions, so a larger minor version alone is usually 208compatible to older versions, so a larger minor version alone is usually
193not a problem. 209not a problem.
194 210
195Example: Make sure we haven't accidentally been linked against the wrong 211Example: Make sure we haven't accidentally been linked against the wrong
196version. 212version (note, however, that this will not detect other ABI mismatches,
213such as LFS or reentrancy).
197 214
198 assert (("libev version mismatch", 215 assert (("libev version mismatch",
199 ev_version_major () == EV_VERSION_MAJOR 216 ev_version_major () == EV_VERSION_MAJOR
200 && ev_version_minor () >= EV_VERSION_MINOR)); 217 && ev_version_minor () >= EV_VERSION_MINOR));
201 218
212 assert (("sorry, no epoll, no sex", 229 assert (("sorry, no epoll, no sex",
213 ev_supported_backends () & EVBACKEND_EPOLL)); 230 ev_supported_backends () & EVBACKEND_EPOLL));
214 231
215=item unsigned int ev_recommended_backends () 232=item unsigned int ev_recommended_backends ()
216 233
217Return the set of all backends compiled into this binary of libev and also 234Return the set of all backends compiled into this binary of libev and
218recommended for this platform. This set is often smaller than the one 235also recommended for this platform, meaning it will work for most file
236descriptor types. This set is often smaller than the one returned by
219returned by C<ev_supported_backends>, as for example kqueue is broken on 237C<ev_supported_backends>, as for example kqueue is broken on most BSDs
220most BSDs and will not be auto-detected unless you explicitly request it 238and will not be auto-detected unless you explicitly request it (assuming
221(assuming you know what you are doing). This is the set of backends that 239you know what you are doing). This is the set of backends that libev will
222libev will probe for if you specify no backends explicitly. 240probe for if you specify no backends explicitly.
223 241
224=item unsigned int ev_embeddable_backends () 242=item unsigned int ev_embeddable_backends ()
225 243
226Returns the set of backends that are embeddable in other event loops. This 244Returns the set of backends that are embeddable in other event loops. This
227is the theoretical, all-platform, value. To find which backends 245value is platform-specific but can include backends not available on the
228might be supported on the current system, you would need to look at 246current system. To find which embeddable backends might be supported on
229C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for 247the current system, you would need to look at C<ev_embeddable_backends ()
230recommended ones. 248& ev_supported_backends ()>, likewise for recommended ones.
231 249
232See the description of C<ev_embed> watchers for more info. 250See the description of C<ev_embed> watchers for more info.
233 251
234=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT] 252=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
235 253
236Sets the allocation function to use (the prototype is similar - the 254Sets the allocation function to use (the prototype is similar - the
237semantics are identical to the C<realloc> C89/SuS/POSIX function). It is 255semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
238used to allocate and free memory (no surprises here). If it returns zero 256used to allocate and free memory (no surprises here). If it returns zero
239when memory needs to be allocated (C<size != 0>), the library might abort 257when memory needs to be allocated (C<size != 0>), the library might abort
265 } 283 }
266 284
267 ... 285 ...
268 ev_set_allocator (persistent_realloc); 286 ev_set_allocator (persistent_realloc);
269 287
270=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT] 288=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
271 289
272Set the callback function to call on a retryable system call error (such 290Set the callback function to call on a retryable system call error (such
273as failed select, poll, epoll_wait). The message is a printable string 291as failed select, poll, epoll_wait). The message is a printable string
274indicating the system call or subsystem causing the problem. If this 292indicating the system call or subsystem causing the problem. If this
275callback is set, then libev will expect it to remedy the situation, no 293callback is set, then libev will expect it to remedy the situation, no
287 } 305 }
288 306
289 ... 307 ...
290 ev_set_syserr_cb (fatal_error); 308 ev_set_syserr_cb (fatal_error);
291 309
310=item ev_feed_signal (int signum)
311
312This function can be used to "simulate" a signal receive. It is completely
313safe to call this function at any time, from any context, including signal
314handlers or random threads.
315
316Its main use is to customise signal handling in your process, especially
317in the presence of threads. For example, you could block signals
318by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
319creating any loops), and in one thread, use C<sigwait> or any other
320mechanism to wait for signals, then "deliver" them to libev by calling
321C<ev_feed_signal>.
322
292=back 323=back
293 324
294=head1 FUNCTIONS CONTROLLING THE EVENT LOOP 325=head1 FUNCTIONS CONTROLLING EVENT LOOPS
295 326
296An event loop is described by a C<struct ev_loop *> (the C<struct> 327An event loop is described by a C<struct ev_loop *> (the C<struct> is
297is I<not> optional in this case, as there is also an C<ev_loop> 328I<not> optional in this case unless libev 3 compatibility is disabled, as
298I<function>). 329libev 3 had an C<ev_loop> function colliding with the struct name).
299 330
300The library knows two types of such loops, the I<default> loop, which 331The library knows two types of such loops, the I<default> loop, which
301supports signals and child events, and dynamically created loops which do 332supports child process events, and dynamically created event loops which
302not. 333do not.
303 334
304=over 4 335=over 4
305 336
306=item struct ev_loop *ev_default_loop (unsigned int flags) 337=item struct ev_loop *ev_default_loop (unsigned int flags)
307 338
308This will initialise the default event loop if it hasn't been initialised 339This returns the "default" event loop object, which is what you should
309yet and return it. If the default loop could not be initialised, returns 340normally use when you just need "the event loop". Event loop objects and
310false. If it already was initialised it simply returns it (and ignores the 341the C<flags> parameter are described in more detail in the entry for
311flags. If that is troubling you, check C<ev_backend ()> afterwards). 342C<ev_loop_new>.
343
344If the default loop is already initialised then this function simply
345returns it (and ignores the flags. If that is troubling you, check
346C<ev_backend ()> afterwards). Otherwise it will create it with the given
347flags, which should almost always be C<0>, unless the caller is also the
348one calling C<ev_run> or otherwise qualifies as "the main program".
312 349
313If you don't know what event loop to use, use the one returned from this 350If you don't know what event loop to use, use the one returned from this
314function. 351function (or via the C<EV_DEFAULT> macro).
315 352
316Note that this function is I<not> thread-safe, so if you want to use it 353Note that this function is I<not> thread-safe, so if you want to use it
317from multiple threads, you have to lock (note also that this is unlikely, 354from multiple threads, you have to employ some kind of mutex (note also
318as loops cannot be shared easily between threads anyway). 355that this case is unlikely, as loops cannot be shared easily between
356threads anyway).
319 357
320The default loop is the only loop that can handle C<ev_signal> and 358The default loop is the only loop that can handle C<ev_child> watchers,
321C<ev_child> watchers, and to do this, it always registers a handler 359and to do this, it always registers a handler for C<SIGCHLD>. If this is
322for C<SIGCHLD>. If this is a problem for your application you can either 360a problem for your application you can either create a dynamic loop with
323create a dynamic loop with C<ev_loop_new> that doesn't do that, or you 361C<ev_loop_new> which doesn't do that, or you can simply overwrite the
324can simply overwrite the C<SIGCHLD> signal handler I<after> calling 362C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
325C<ev_default_init>. 363
364Example: This is the most typical usage.
365
366 if (!ev_default_loop (0))
367 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
368
369Example: Restrict libev to the select and poll backends, and do not allow
370environment settings to be taken into account:
371
372 ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
373
374=item struct ev_loop *ev_loop_new (unsigned int flags)
375
376This will create and initialise a new event loop object. If the loop
377could not be initialised, returns false.
378
379This function is thread-safe, and one common way to use libev with
380threads is indeed to create one loop per thread, and using the default
381loop in the "main" or "initial" thread.
326 382
327The flags argument can be used to specify special behaviour or specific 383The flags argument can be used to specify special behaviour or specific
328backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>). 384backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
329 385
330The following flags are supported: 386The following flags are supported:
345useful to try out specific backends to test their performance, or to work 401useful to try out specific backends to test their performance, or to work
346around bugs. 402around bugs.
347 403
348=item C<EVFLAG_FORKCHECK> 404=item C<EVFLAG_FORKCHECK>
349 405
350Instead of calling C<ev_default_fork> or C<ev_loop_fork> manually after 406Instead of calling C<ev_loop_fork> manually after a fork, you can also
351a fork, you can also make libev check for a fork in each iteration by 407make libev check for a fork in each iteration by enabling this flag.
352enabling this flag.
353 408
354This works by calling C<getpid ()> on every iteration of the loop, 409This works by calling C<getpid ()> on every iteration of the loop,
355and thus this might slow down your event loop if you do a lot of loop 410and thus this might slow down your event loop if you do a lot of loop
356iterations and little real work, but is usually not noticeable (on my 411iterations and little real work, but is usually not noticeable (on my
357GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence 412GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
366environment variable. 421environment variable.
367 422
368=item C<EVFLAG_NOINOTIFY> 423=item C<EVFLAG_NOINOTIFY>
369 424
370When this flag is specified, then libev will not attempt to use the 425When this flag is specified, then libev will not attempt to use the
371I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and 426I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
372testing, this flag can be useful to conserve inotify file descriptors, as 427testing, this flag can be useful to conserve inotify file descriptors, as
373otherwise each loop using C<ev_stat> watchers consumes one inotify handle. 428otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
374 429
375=item C<EVFLAG_NOSIGFD> 430=item C<EVFLAG_SIGNALFD>
376 431
377When this flag is specified, then libev will not attempt to use the 432When this flag is specified, then libev will attempt to use the
378I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This is 433I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
379probably only useful to work around any bugs in libev. Consequently, this 434delivers signals synchronously, which makes it both faster and might make
380flag might go away once the signalfd functionality is considered stable, 435it possible to get the queued signal data. It can also simplify signal
381so it's useful mostly in environment variables and not in program code. 436handling with threads, as long as you properly block signals in your
437threads that are not interested in handling them.
438
439Signalfd will not be used by default as this changes your signal mask, and
440there are a lot of shoddy libraries and programs (glib's threadpool for
441example) that can't properly initialise their signal masks.
442
443=item C<EVFLAG_NOSIGMASK>
444
445When this flag is specified, then libev will avoid to modify the signal
446mask. Specifically, this means you have to make sure signals are unblocked
447when you want to receive them.
448
449This behaviour is useful when you want to do your own signal handling, or
450want to handle signals only in specific threads and want to avoid libev
451unblocking the signals.
452
453It's also required by POSIX in a threaded program, as libev calls
454C<sigprocmask>, whose behaviour is officially unspecified.
455
456This flag's behaviour will become the default in future versions of libev.
382 457
383=item C<EVBACKEND_SELECT> (value 1, portable select backend) 458=item C<EVBACKEND_SELECT> (value 1, portable select backend)
384 459
385This is your standard select(2) backend. Not I<completely> standard, as 460This is your standard select(2) backend. Not I<completely> standard, as
386libev tries to roll its own fd_set with no limits on the number of fds, 461libev tries to roll its own fd_set with no limits on the number of fds,
414=item C<EVBACKEND_EPOLL> (value 4, Linux) 489=item C<EVBACKEND_EPOLL> (value 4, Linux)
415 490
416Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9 491Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
417kernels). 492kernels).
418 493
419For few fds, this backend is a bit little slower than poll and select, 494For few fds, this backend is a bit little slower than poll and select, but
420but it scales phenomenally better. While poll and select usually scale 495it scales phenomenally better. While poll and select usually scale like
421like O(total_fds) where n is the total number of fds (or the highest fd), 496O(total_fds) where total_fds is the total number of fds (or the highest
422epoll scales either O(1) or O(active_fds). 497fd), epoll scales either O(1) or O(active_fds).
423 498
424The epoll mechanism deserves honorable mention as the most misdesigned 499The epoll mechanism deserves honorable mention as the most misdesigned
425of the more advanced event mechanisms: mere annoyances include silently 500of the more advanced event mechanisms: mere annoyances include silently
426dropping file descriptors, requiring a system call per change per file 501dropping file descriptors, requiring a system call per change per file
427descriptor (and unnecessary guessing of parameters), problems with dup and 502descriptor (and unnecessary guessing of parameters), problems with dup,
503returning before the timeout value, resulting in additional iterations
504(and only giving 5ms accuracy while select on the same platform gives
428so on. The biggest issue is fork races, however - if a program forks then 5050.1ms) and so on. The biggest issue is fork races, however - if a program
429I<both> parent and child process have to recreate the epoll set, which can 506forks then I<both> parent and child process have to recreate the epoll
430take considerable time (one syscall per file descriptor) and is of course 507set, which can take considerable time (one syscall per file descriptor)
431hard to detect. 508and is of course hard to detect.
432 509
433Epoll is also notoriously buggy - embedding epoll fds I<should> work, but 510Epoll is also notoriously buggy - embedding epoll fds I<should> work,
434of course I<doesn't>, and epoll just loves to report events for totally 511but of course I<doesn't>, and epoll just loves to report events for
435I<different> file descriptors (even already closed ones, so one cannot 512totally I<different> file descriptors (even already closed ones, so
436even remove them from the set) than registered in the set (especially 513one cannot even remove them from the set) than registered in the set
437on SMP systems). Libev tries to counter these spurious notifications by 514(especially on SMP systems). Libev tries to counter these spurious
438employing an additional generation counter and comparing that against the 515notifications by employing an additional generation counter and comparing
439events to filter out spurious ones, recreating the set when required. 516that against the events to filter out spurious ones, recreating the set
517when required. Epoll also erroneously rounds down timeouts, but gives you
518no way to know when and by how much, so sometimes you have to busy-wait
519because epoll returns immediately despite a nonzero timeout. And last
520not least, it also refuses to work with some file descriptors which work
521perfectly fine with C<select> (files, many character devices...).
522
523Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
524cobbled together in a hurry, no thought to design or interaction with
525others. Oh, the pain, will it ever stop...
440 526
441While stopping, setting and starting an I/O watcher in the same iteration 527While stopping, setting and starting an I/O watcher in the same iteration
442will result in some caching, there is still a system call per such 528will result in some caching, there is still a system call per such
443incident (because the same I<file descriptor> could point to a different 529incident (because the same I<file descriptor> could point to a different
444I<file description> now), so its best to avoid that. Also, C<dup ()>'ed 530I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
481 567
482It scales in the same way as the epoll backend, but the interface to the 568It scales in the same way as the epoll backend, but the interface to the
483kernel is more efficient (which says nothing about its actual speed, of 569kernel is more efficient (which says nothing about its actual speed, of
484course). While stopping, setting and starting an I/O watcher does never 570course). While stopping, setting and starting an I/O watcher does never
485cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to 571cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
486two event changes per incident. Support for C<fork ()> is very bad (but 572two event changes per incident. Support for C<fork ()> is very bad (you
487sane, unlike epoll) and it drops fds silently in similarly hard-to-detect 573might have to leak fd's on fork, but it's more sane than epoll) and it
488cases 574drops fds silently in similarly hard-to-detect cases
489 575
490This backend usually performs well under most conditions. 576This backend usually performs well under most conditions.
491 577
492While nominally embeddable in other event loops, this doesn't work 578While nominally embeddable in other event loops, this doesn't work
493everywhere, so you might need to test for this. And since it is broken 579everywhere, so you might need to test for this. And since it is broken
510=item C<EVBACKEND_PORT> (value 32, Solaris 10) 596=item C<EVBACKEND_PORT> (value 32, Solaris 10)
511 597
512This uses the Solaris 10 event port mechanism. As with everything on Solaris, 598This uses the Solaris 10 event port mechanism. As with everything on Solaris,
513it's really slow, but it still scales very well (O(active_fds)). 599it's really slow, but it still scales very well (O(active_fds)).
514 600
515Please note that Solaris event ports can deliver a lot of spurious
516notifications, so you need to use non-blocking I/O or other means to avoid
517blocking when no data (or space) is available.
518
519While this backend scales well, it requires one system call per active 601While this backend scales well, it requires one system call per active
520file descriptor per loop iteration. For small and medium numbers of file 602file descriptor per loop iteration. For small and medium numbers of file
521descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend 603descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
522might perform better. 604might perform better.
523 605
524On the positive side, with the exception of the spurious readiness 606On the positive side, this backend actually performed fully to
525notifications, this backend actually performed fully to specification
526in all tests and is fully embeddable, which is a rare feat among the 607specification in all tests and is fully embeddable, which is a rare feat
527OS-specific backends (I vastly prefer correctness over speed hacks). 608among the OS-specific backends (I vastly prefer correctness over speed
609hacks).
610
611On the negative side, the interface is I<bizarre> - so bizarre that
612even sun itself gets it wrong in their code examples: The event polling
613function sometimes returns events to the caller even though an error
614occurred, but with no indication whether it has done so or not (yes, it's
615even documented that way) - deadly for edge-triggered interfaces where you
616absolutely have to know whether an event occurred or not because you have
617to re-arm the watcher.
618
619Fortunately libev seems to be able to work around these idiocies.
528 620
529This backend maps C<EV_READ> and C<EV_WRITE> in the same way as 621This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
530C<EVBACKEND_POLL>. 622C<EVBACKEND_POLL>.
531 623
532=item C<EVBACKEND_ALL> 624=item C<EVBACKEND_ALL>
533 625
534Try all backends (even potentially broken ones that wouldn't be tried 626Try all backends (even potentially broken ones that wouldn't be tried
535with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as 627with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
536C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>. 628C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
537 629
538It is definitely not recommended to use this flag. 630It is definitely not recommended to use this flag, use whatever
631C<ev_recommended_backends ()> returns, or simply do not specify a backend
632at all.
633
634=item C<EVBACKEND_MASK>
635
636Not a backend at all, but a mask to select all backend bits from a
637C<flags> value, in case you want to mask out any backends from a flags
638value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
539 639
540=back 640=back
541 641
542If one or more of the backend flags are or'ed into the flags value, 642If one or more of the backend flags are or'ed into the flags value,
543then only these backends will be tried (in the reverse order as listed 643then only these backends will be tried (in the reverse order as listed
544here). If none are specified, all backends in C<ev_recommended_backends 644here). If none are specified, all backends in C<ev_recommended_backends
545()> will be tried. 645()> will be tried.
546 646
547Example: This is the most typical usage.
548
549 if (!ev_default_loop (0))
550 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
551
552Example: Restrict libev to the select and poll backends, and do not allow
553environment settings to be taken into account:
554
555 ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
556
557Example: Use whatever libev has to offer, but make sure that kqueue is
558used if available (warning, breaks stuff, best use only with your own
559private event loop and only if you know the OS supports your types of
560fds):
561
562 ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
563
564=item struct ev_loop *ev_loop_new (unsigned int flags)
565
566Similar to C<ev_default_loop>, but always creates a new event loop that is
567always distinct from the default loop. Unlike the default loop, it cannot
568handle signal and child watchers, and attempts to do so will be greeted by
569undefined behaviour (or a failed assertion if assertions are enabled).
570
571Note that this function I<is> thread-safe, and the recommended way to use
572libev with threads is indeed to create one loop per thread, and using the
573default loop in the "main" or "initial" thread.
574
575Example: Try to create a event loop that uses epoll and nothing else. 647Example: Try to create a event loop that uses epoll and nothing else.
576 648
577 struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV); 649 struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
578 if (!epoller) 650 if (!epoller)
579 fatal ("no epoll found here, maybe it hides under your chair"); 651 fatal ("no epoll found here, maybe it hides under your chair");
580 652
653Example: Use whatever libev has to offer, but make sure that kqueue is
654used if available.
655
656 struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
657
581=item ev_default_destroy () 658=item ev_loop_destroy (loop)
582 659
583Destroys the default loop again (frees all memory and kernel state 660Destroys an event loop object (frees all memory and kernel state
584etc.). None of the active event watchers will be stopped in the normal 661etc.). None of the active event watchers will be stopped in the normal
585sense, so e.g. C<ev_is_active> might still return true. It is your 662sense, so e.g. C<ev_is_active> might still return true. It is your
586responsibility to either stop all watchers cleanly yourself I<before> 663responsibility to either stop all watchers cleanly yourself I<before>
587calling this function, or cope with the fact afterwards (which is usually 664calling this function, or cope with the fact afterwards (which is usually
588the easiest thing, you can just ignore the watchers and/or C<free ()> them 665the easiest thing, you can just ignore the watchers and/or C<free ()> them
590 667
591Note that certain global state, such as signal state (and installed signal 668Note that certain global state, such as signal state (and installed signal
592handlers), will not be freed by this function, and related watchers (such 669handlers), will not be freed by this function, and related watchers (such
593as signal and child watchers) would need to be stopped manually. 670as signal and child watchers) would need to be stopped manually.
594 671
595In general it is not advisable to call this function except in the 672This function is normally used on loop objects allocated by
596rare occasion where you really need to free e.g. the signal handling 673C<ev_loop_new>, but it can also be used on the default loop returned by
674C<ev_default_loop>, in which case it is not thread-safe.
675
676Note that it is not advisable to call this function on the default loop
677except in the rare occasion where you really need to free its resources.
597pipe fds. If you need dynamically allocated loops it is better to use 678If you need dynamically allocated loops it is better to use C<ev_loop_new>
598C<ev_loop_new> and C<ev_loop_destroy>. 679and C<ev_loop_destroy>.
599 680
600=item ev_loop_destroy (loop) 681=item ev_loop_fork (loop)
601 682
602Like C<ev_default_destroy>, but destroys an event loop created by an
603earlier call to C<ev_loop_new>.
604
605=item ev_default_fork ()
606
607This function sets a flag that causes subsequent C<ev_loop> iterations 683This function sets a flag that causes subsequent C<ev_run> iterations to
608to reinitialise the kernel state for backends that have one. Despite the 684reinitialise the kernel state for backends that have one. Despite the
609name, you can call it anytime, but it makes most sense after forking, in 685name, you can call it anytime, but it makes most sense after forking, in
610the child process (or both child and parent, but that again makes little 686the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
611sense). You I<must> call it in the child before using any of the libev 687child before resuming or calling C<ev_run>.
612functions, and it will only take effect at the next C<ev_loop> iteration. 688
689Again, you I<have> to call it on I<any> loop that you want to re-use after
690a fork, I<even if you do not plan to use the loop in the parent>. This is
691because some kernel interfaces *cough* I<kqueue> *cough* do funny things
692during fork.
613 693
614On the other hand, you only need to call this function in the child 694On the other hand, you only need to call this function in the child
615process if and only if you want to use the event library in the child. If 695process if and only if you want to use the event loop in the child. If
616you just fork+exec, you don't have to call it at all. 696you just fork+exec or create a new loop in the child, you don't have to
697call it at all (in fact, C<epoll> is so badly broken that it makes a
698difference, but libev will usually detect this case on its own and do a
699costly reset of the backend).
617 700
618The function itself is quite fast and it's usually not a problem to call 701The function itself is quite fast and it's usually not a problem to call
619it just in case after a fork. To make this easy, the function will fit in 702it just in case after a fork.
620quite nicely into a call to C<pthread_atfork>:
621 703
704Example: Automate calling C<ev_loop_fork> on the default loop when
705using pthreads.
706
707 static void
708 post_fork_child (void)
709 {
710 ev_loop_fork (EV_DEFAULT);
711 }
712
713 ...
622 pthread_atfork (0, 0, ev_default_fork); 714 pthread_atfork (0, 0, post_fork_child);
623
624=item ev_loop_fork (loop)
625
626Like C<ev_default_fork>, but acts on an event loop created by
627C<ev_loop_new>. Yes, you have to call this on every allocated event loop
628after fork that you want to re-use in the child, and how you do this is
629entirely your own problem.
630 715
631=item int ev_is_default_loop (loop) 716=item int ev_is_default_loop (loop)
632 717
633Returns true when the given loop is, in fact, the default loop, and false 718Returns true when the given loop is, in fact, the default loop, and false
634otherwise. 719otherwise.
635 720
636=item unsigned int ev_loop_count (loop) 721=item unsigned int ev_iteration (loop)
637 722
638Returns the count of loop iterations for the loop, which is identical to 723Returns the current iteration count for the event loop, which is identical
639the number of times libev did poll for new events. It starts at C<0> and 724to the number of times libev did poll for new events. It starts at C<0>
640happily wraps around with enough iterations. 725and happily wraps around with enough iterations.
641 726
642This value can sometimes be useful as a generation counter of sorts (it 727This value can sometimes be useful as a generation counter of sorts (it
643"ticks" the number of loop iterations), as it roughly corresponds with 728"ticks" the number of loop iterations), as it roughly corresponds with
644C<ev_prepare> and C<ev_check> calls. 729C<ev_prepare> and C<ev_check> calls - and is incremented between the
730prepare and check phases.
645 731
646=item unsigned int ev_loop_depth (loop) 732=item unsigned int ev_depth (loop)
647 733
648Returns the number of times C<ev_loop> was entered minus the number of 734Returns the number of times C<ev_run> was entered minus the number of
649times C<ev_loop> was exited, in other words, the recursion depth. 735times C<ev_run> was exited normally, in other words, the recursion depth.
650 736
651Outside C<ev_loop>, this number is zero. In a callback, this number is 737Outside C<ev_run>, this number is zero. In a callback, this number is
652C<1>, unless C<ev_loop> was invoked recursively (or from another thread), 738C<1>, unless C<ev_run> was invoked recursively (or from another thread),
653in which case it is higher. 739in which case it is higher.
654 740
655Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread 741Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
656etc.), doesn't count as exit. 742throwing an exception etc.), doesn't count as "exit" - consider this
743as a hint to avoid such ungentleman-like behaviour unless it's really
744convenient, in which case it is fully supported.
657 745
658=item unsigned int ev_backend (loop) 746=item unsigned int ev_backend (loop)
659 747
660Returns one of the C<EVBACKEND_*> flags indicating the event backend in 748Returns one of the C<EVBACKEND_*> flags indicating the event backend in
661use. 749use.
670 758
671=item ev_now_update (loop) 759=item ev_now_update (loop)
672 760
673Establishes the current time by querying the kernel, updating the time 761Establishes the current time by querying the kernel, updating the time
674returned by C<ev_now ()> in the progress. This is a costly operation and 762returned by C<ev_now ()> in the progress. This is a costly operation and
675is usually done automatically within C<ev_loop ()>. 763is usually done automatically within C<ev_run ()>.
676 764
677This function is rarely useful, but when some event callback runs for a 765This function is rarely useful, but when some event callback runs for a
678very long time without entering the event loop, updating libev's idea of 766very long time without entering the event loop, updating libev's idea of
679the current time is a good idea. 767the current time is a good idea.
680 768
682 770
683=item ev_suspend (loop) 771=item ev_suspend (loop)
684 772
685=item ev_resume (loop) 773=item ev_resume (loop)
686 774
687These two functions suspend and resume a loop, for use when the loop is 775These two functions suspend and resume an event loop, for use when the
688not used for a while and timeouts should not be processed. 776loop is not used for a while and timeouts should not be processed.
689 777
690A typical use case would be an interactive program such as a game: When 778A typical use case would be an interactive program such as a game: When
691the user presses C<^Z> to suspend the game and resumes it an hour later it 779the user presses C<^Z> to suspend the game and resumes it an hour later it
692would be best to handle timeouts as if no time had actually passed while 780would be best to handle timeouts as if no time had actually passed while
693the program was suspended. This can be achieved by calling C<ev_suspend> 781the program was suspended. This can be achieved by calling C<ev_suspend>
695C<ev_resume> directly afterwards to resume timer processing. 783C<ev_resume> directly afterwards to resume timer processing.
696 784
697Effectively, all C<ev_timer> watchers will be delayed by the time spend 785Effectively, all C<ev_timer> watchers will be delayed by the time spend
698between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers 786between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
699will be rescheduled (that is, they will lose any events that would have 787will be rescheduled (that is, they will lose any events that would have
700occured while suspended). 788occurred while suspended).
701 789
702After calling C<ev_suspend> you B<must not> call I<any> function on the 790After calling C<ev_suspend> you B<must not> call I<any> function on the
703given loop other than C<ev_resume>, and you B<must not> call C<ev_resume> 791given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
704without a previous call to C<ev_suspend>. 792without a previous call to C<ev_suspend>.
705 793
706Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the 794Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
707event loop time (see C<ev_now_update>). 795event loop time (see C<ev_now_update>).
708 796
709=item ev_loop (loop, int flags) 797=item bool ev_run (loop, int flags)
710 798
711Finally, this is it, the event handler. This function usually is called 799Finally, this is it, the event handler. This function usually is called
712after you have initialised all your watchers and you want to start 800after you have initialised all your watchers and you want to start
713handling events. 801handling events. It will ask the operating system for any new events, call
802the watcher callbacks, and then repeat the whole process indefinitely: This
803is why event loops are called I<loops>.
714 804
715If the flags argument is specified as C<0>, it will not return until 805If the flags argument is specified as C<0>, it will keep handling events
716either no event watchers are active anymore or C<ev_unloop> was called. 806until either no event watchers are active anymore or C<ev_break> was
807called.
717 808
809The return value is false if there are no more active watchers (which
810usually means "all jobs done" or "deadlock"), and true in all other cases
811(which usually means " you should call C<ev_run> again").
812
718Please note that an explicit C<ev_unloop> is usually better than 813Please note that an explicit C<ev_break> is usually better than
719relying on all watchers to be stopped when deciding when a program has 814relying on all watchers to be stopped when deciding when a program has
720finished (especially in interactive programs), but having a program 815finished (especially in interactive programs), but having a program
721that automatically loops as long as it has to and no longer by virtue 816that automatically loops as long as it has to and no longer by virtue
722of relying on its watchers stopping correctly, that is truly a thing of 817of relying on its watchers stopping correctly, that is truly a thing of
723beauty. 818beauty.
724 819
820This function is I<mostly> exception-safe - you can break out of a
821C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
822exception and so on. This does not decrement the C<ev_depth> value, nor
823will it clear any outstanding C<EVBREAK_ONE> breaks.
824
725A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle 825A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
726those events and any already outstanding ones, but will not block your 826those events and any already outstanding ones, but will not wait and
727process in case there are no events and will return after one iteration of 827block your process in case there are no events and will return after one
728the loop. 828iteration of the loop. This is sometimes useful to poll and handle new
829events while doing lengthy calculations, to keep the program responsive.
729 830
730A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if 831A flags value of C<EVRUN_ONCE> will look for new events (waiting if
731necessary) and will handle those and any already outstanding ones. It 832necessary) and will handle those and any already outstanding ones. It
732will block your process until at least one new event arrives (which could 833will block your process until at least one new event arrives (which could
733be an event internal to libev itself, so there is no guarantee that a 834be an event internal to libev itself, so there is no guarantee that a
734user-registered callback will be called), and will return after one 835user-registered callback will be called), and will return after one
735iteration of the loop. 836iteration of the loop.
736 837
737This is useful if you are waiting for some external event in conjunction 838This is useful if you are waiting for some external event in conjunction
738with something not expressible using other libev watchers (i.e. "roll your 839with something not expressible using other libev watchers (i.e. "roll your
739own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is 840own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
740usually a better approach for this kind of thing. 841usually a better approach for this kind of thing.
741 842
742Here are the gory details of what C<ev_loop> does: 843Here are the gory details of what C<ev_run> does (this is for your
844understanding, not a guarantee that things will work exactly like this in
845future versions):
743 846
847 - Increment loop depth.
848 - Reset the ev_break status.
744 - Before the first iteration, call any pending watchers. 849 - Before the first iteration, call any pending watchers.
850 LOOP:
745 * If EVFLAG_FORKCHECK was used, check for a fork. 851 - If EVFLAG_FORKCHECK was used, check for a fork.
746 - If a fork was detected (by any means), queue and call all fork watchers. 852 - If a fork was detected (by any means), queue and call all fork watchers.
747 - Queue and call all prepare watchers. 853 - Queue and call all prepare watchers.
854 - If ev_break was called, goto FINISH.
748 - If we have been forked, detach and recreate the kernel state 855 - If we have been forked, detach and recreate the kernel state
749 as to not disturb the other process. 856 as to not disturb the other process.
750 - Update the kernel state with all outstanding changes. 857 - Update the kernel state with all outstanding changes.
751 - Update the "event loop time" (ev_now ()). 858 - Update the "event loop time" (ev_now ()).
752 - Calculate for how long to sleep or block, if at all 859 - Calculate for how long to sleep or block, if at all
753 (active idle watchers, EVLOOP_NONBLOCK or not having 860 (active idle watchers, EVRUN_NOWAIT or not having
754 any active watchers at all will result in not sleeping). 861 any active watchers at all will result in not sleeping).
755 - Sleep if the I/O and timer collect interval say so. 862 - Sleep if the I/O and timer collect interval say so.
863 - Increment loop iteration counter.
756 - Block the process, waiting for any events. 864 - Block the process, waiting for any events.
757 - Queue all outstanding I/O (fd) events. 865 - Queue all outstanding I/O (fd) events.
758 - Update the "event loop time" (ev_now ()), and do time jump adjustments. 866 - Update the "event loop time" (ev_now ()), and do time jump adjustments.
759 - Queue all expired timers. 867 - Queue all expired timers.
760 - Queue all expired periodics. 868 - Queue all expired periodics.
761 - Unless any events are pending now, queue all idle watchers. 869 - Queue all idle watchers with priority higher than that of pending events.
762 - Queue all check watchers. 870 - Queue all check watchers.
763 - Call all queued watchers in reverse order (i.e. check watchers first). 871 - Call all queued watchers in reverse order (i.e. check watchers first).
764 Signals and child watchers are implemented as I/O watchers, and will 872 Signals and child watchers are implemented as I/O watchers, and will
765 be handled here by queueing them when their watcher gets executed. 873 be handled here by queueing them when their watcher gets executed.
766 - If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK 874 - If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
767 were used, or there are no active watchers, return, otherwise 875 were used, or there are no active watchers, goto FINISH, otherwise
768 continue with step *. 876 continue with step LOOP.
877 FINISH:
878 - Reset the ev_break status iff it was EVBREAK_ONE.
879 - Decrement the loop depth.
880 - Return.
769 881
770Example: Queue some jobs and then loop until no events are outstanding 882Example: Queue some jobs and then loop until no events are outstanding
771anymore. 883anymore.
772 884
773 ... queue jobs here, make sure they register event watchers as long 885 ... queue jobs here, make sure they register event watchers as long
774 ... as they still have work to do (even an idle watcher will do..) 886 ... as they still have work to do (even an idle watcher will do..)
775 ev_loop (my_loop, 0); 887 ev_run (my_loop, 0);
776 ... jobs done or somebody called unloop. yeah! 888 ... jobs done or somebody called break. yeah!
777 889
778=item ev_unloop (loop, how) 890=item ev_break (loop, how)
779 891
780Can be used to make a call to C<ev_loop> return early (but only after it 892Can be used to make a call to C<ev_run> return early (but only after it
781has processed all outstanding events). The C<how> argument must be either 893has processed all outstanding events). The C<how> argument must be either
782C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or 894C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
783C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return. 895C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
784 896
785This "unloop state" will be cleared when entering C<ev_loop> again. 897This "break state" will be cleared on the next call to C<ev_run>.
786 898
787It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls. 899It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
900which case it will have no effect.
788 901
789=item ev_ref (loop) 902=item ev_ref (loop)
790 903
791=item ev_unref (loop) 904=item ev_unref (loop)
792 905
793Ref/unref can be used to add or remove a reference count on the event 906Ref/unref can be used to add or remove a reference count on the event
794loop: Every watcher keeps one reference, and as long as the reference 907loop: Every watcher keeps one reference, and as long as the reference
795count is nonzero, C<ev_loop> will not return on its own. 908count is nonzero, C<ev_run> will not return on its own.
796 909
797If you have a watcher you never unregister that should not keep C<ev_loop> 910This is useful when you have a watcher that you never intend to
798from returning, call ev_unref() after starting, and ev_ref() before 911unregister, but that nevertheless should not keep C<ev_run> from
912returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
799stopping it. 913before stopping it.
800 914
801As an example, libev itself uses this for its internal signal pipe: It 915As an example, libev itself uses this for its internal signal pipe: It
802is not visible to the libev user and should not keep C<ev_loop> from 916is not visible to the libev user and should not keep C<ev_run> from
803exiting if no event watchers registered by it are active. It is also an 917exiting if no event watchers registered by it are active. It is also an
804excellent way to do this for generic recurring timers or from within 918excellent way to do this for generic recurring timers or from within
805third-party libraries. Just remember to I<unref after start> and I<ref 919third-party libraries. Just remember to I<unref after start> and I<ref
806before stop> (but only if the watcher wasn't active before, or was active 920before stop> (but only if the watcher wasn't active before, or was active
807before, respectively. Note also that libev might stop watchers itself 921before, respectively. Note also that libev might stop watchers itself
808(e.g. non-repeating timers) in which case you have to C<ev_ref> 922(e.g. non-repeating timers) in which case you have to C<ev_ref>
809in the callback). 923in the callback).
810 924
811Example: Create a signal watcher, but keep it from keeping C<ev_loop> 925Example: Create a signal watcher, but keep it from keeping C<ev_run>
812running when nothing else is active. 926running when nothing else is active.
813 927
814 ev_signal exitsig; 928 ev_signal exitsig;
815 ev_signal_init (&exitsig, sig_cb, SIGINT); 929 ev_signal_init (&exitsig, sig_cb, SIGINT);
816 ev_signal_start (loop, &exitsig); 930 ev_signal_start (loop, &exitsig);
817 evf_unref (loop); 931 ev_unref (loop);
818 932
819Example: For some weird reason, unregister the above signal handler again. 933Example: For some weird reason, unregister the above signal handler again.
820 934
821 ev_ref (loop); 935 ev_ref (loop);
822 ev_signal_stop (loop, &exitsig); 936 ev_signal_stop (loop, &exitsig);
842overhead for the actual polling but can deliver many events at once. 956overhead for the actual polling but can deliver many events at once.
843 957
844By setting a higher I<io collect interval> you allow libev to spend more 958By setting a higher I<io collect interval> you allow libev to spend more
845time collecting I/O events, so you can handle more events per iteration, 959time collecting I/O events, so you can handle more events per iteration,
846at the cost of increasing latency. Timeouts (both C<ev_periodic> and 960at the cost of increasing latency. Timeouts (both C<ev_periodic> and
847C<ev_timer>) will be not affected. Setting this to a non-null value will 961C<ev_timer>) will not be affected. Setting this to a non-null value will
848introduce an additional C<ev_sleep ()> call into most loop iterations. The 962introduce an additional C<ev_sleep ()> call into most loop iterations. The
849sleep time ensures that libev will not poll for I/O events more often then 963sleep time ensures that libev will not poll for I/O events more often then
850once per this interval, on average. 964once per this interval, on average (as long as the host time resolution is
965good enough).
851 966
852Likewise, by setting a higher I<timeout collect interval> you allow libev 967Likewise, by setting a higher I<timeout collect interval> you allow libev
853to spend more time collecting timeouts, at the expense of increased 968to spend more time collecting timeouts, at the expense of increased
854latency/jitter/inexactness (the watcher callback will be called 969latency/jitter/inexactness (the watcher callback will be called
855later). C<ev_io> watchers will not be affected. Setting this to a non-null 970later). C<ev_io> watchers will not be affected. Setting this to a non-null
861usually doesn't make much sense to set it to a lower value than C<0.01>, 976usually doesn't make much sense to set it to a lower value than C<0.01>,
862as this approaches the timing granularity of most systems. Note that if 977as this approaches the timing granularity of most systems. Note that if
863you do transactions with the outside world and you can't increase the 978you do transactions with the outside world and you can't increase the
864parallelity, then this setting will limit your transaction rate (if you 979parallelity, then this setting will limit your transaction rate (if you
865need to poll once per transaction and the I/O collect interval is 0.01, 980need to poll once per transaction and the I/O collect interval is 0.01,
866then you can't do more than 100 transations per second). 981then you can't do more than 100 transactions per second).
867 982
868Setting the I<timeout collect interval> can improve the opportunity for 983Setting the I<timeout collect interval> can improve the opportunity for
869saving power, as the program will "bundle" timer callback invocations that 984saving power, as the program will "bundle" timer callback invocations that
870are "near" in time together, by delaying some, thus reducing the number of 985are "near" in time together, by delaying some, thus reducing the number of
871times the process sleeps and wakes up again. Another useful technique to 986times the process sleeps and wakes up again. Another useful technique to
879 ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01); 994 ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
880 995
881=item ev_invoke_pending (loop) 996=item ev_invoke_pending (loop)
882 997
883This call will simply invoke all pending watchers while resetting their 998This call will simply invoke all pending watchers while resetting their
884pending state. Normally, C<ev_loop> does this automatically when required, 999pending state. Normally, C<ev_run> does this automatically when required,
885but when overriding the invoke callback this call comes handy. 1000but when overriding the invoke callback this call comes handy. This
1001function can be invoked from a watcher - this can be useful for example
1002when you want to do some lengthy calculation and want to pass further
1003event handling to another thread (you still have to make sure only one
1004thread executes within C<ev_invoke_pending> or C<ev_run> of course).
886 1005
887=item int ev_pending_count (loop) 1006=item int ev_pending_count (loop)
888 1007
889Returns the number of pending watchers - zero indicates that no watchers 1008Returns the number of pending watchers - zero indicates that no watchers
890are pending. 1009are pending.
891 1010
892=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P)) 1011=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
893 1012
894This overrides the invoke pending functionality of the loop: Instead of 1013This overrides the invoke pending functionality of the loop: Instead of
895invoking all pending watchers when there are any, C<ev_loop> will call 1014invoking all pending watchers when there are any, C<ev_run> will call
896this callback instead. This is useful, for example, when you want to 1015this callback instead. This is useful, for example, when you want to
897invoke the actual watchers inside another context (another thread etc.). 1016invoke the actual watchers inside another context (another thread etc.).
898 1017
899If you want to reset the callback, use C<ev_invoke_pending> as new 1018If you want to reset the callback, use C<ev_invoke_pending> as new
900callback. 1019callback.
901 1020
902=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P)) 1021=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
903 1022
904Sometimes you want to share the same loop between multiple threads. This 1023Sometimes you want to share the same loop between multiple threads. This
905can be done relatively simply by putting mutex_lock/unlock calls around 1024can be done relatively simply by putting mutex_lock/unlock calls around
906each call to a libev function. 1025each call to a libev function.
907 1026
908However, C<ev_loop> can run an indefinite time, so it is not feasible to 1027However, C<ev_run> can run an indefinite time, so it is not feasible
909wait for it to return. One way around this is to wake up the loop via 1028to wait for it to return. One way around this is to wake up the event
910C<ev_unloop> and C<av_async_send>, another way is to set these I<release> 1029loop via C<ev_break> and C<ev_async_send>, another way is to set these
911and I<acquire> callbacks on the loop. 1030I<release> and I<acquire> callbacks on the loop.
912 1031
913When set, then C<release> will be called just before the thread is 1032When set, then C<release> will be called just before the thread is
914suspended waiting for new events, and C<acquire> is called just 1033suspended waiting for new events, and C<acquire> is called just
915afterwards. 1034afterwards.
916 1035
919 1038
920While event loop modifications are allowed between invocations of 1039While event loop modifications are allowed between invocations of
921C<release> and C<acquire> (that's their only purpose after all), no 1040C<release> and C<acquire> (that's their only purpose after all), no
922modifications done will affect the event loop, i.e. adding watchers will 1041modifications done will affect the event loop, i.e. adding watchers will
923have no effect on the set of file descriptors being watched, or the time 1042have no effect on the set of file descriptors being watched, or the time
924waited. USe an C<ev_async> watcher to wake up C<ev_loop> when you want it 1043waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
925to take note of any changes you made. 1044to take note of any changes you made.
926 1045
927In theory, threads executing C<ev_loop> will be async-cancel safe between 1046In theory, threads executing C<ev_run> will be async-cancel safe between
928invocations of C<release> and C<acquire>. 1047invocations of C<release> and C<acquire>.
929 1048
930See also the locking example in the C<THREADS> section later in this 1049See also the locking example in the C<THREADS> section later in this
931document. 1050document.
932 1051
933=item ev_set_userdata (loop, void *data) 1052=item ev_set_userdata (loop, void *data)
934 1053
935=item ev_userdata (loop) 1054=item void *ev_userdata (loop)
936 1055
937Set and retrieve a single C<void *> associated with a loop. When 1056Set and retrieve a single C<void *> associated with a loop. When
938C<ev_set_userdata> has never been called, then C<ev_userdata> returns 1057C<ev_set_userdata> has never been called, then C<ev_userdata> returns
939C<0.> 1058C<0>.
940 1059
941These two functions can be used to associate arbitrary data with a loop, 1060These two functions can be used to associate arbitrary data with a loop,
942and are intended solely for the C<invoke_pending_cb>, C<release> and 1061and are intended solely for the C<invoke_pending_cb>, C<release> and
943C<acquire> callbacks described above, but of course can be (ab-)used for 1062C<acquire> callbacks described above, but of course can be (ab-)used for
944any other purpose as well. 1063any other purpose as well.
945 1064
946=item ev_loop_verify (loop) 1065=item ev_verify (loop)
947 1066
948This function only does something when C<EV_VERIFY> support has been 1067This function only does something when C<EV_VERIFY> support has been
949compiled in, which is the default for non-minimal builds. It tries to go 1068compiled in, which is the default for non-minimal builds. It tries to go
950through all internal structures and checks them for validity. If anything 1069through all internal structures and checks them for validity. If anything
951is found to be inconsistent, it will print an error message to standard 1070is found to be inconsistent, it will print an error message to standard
962 1081
963In the following description, uppercase C<TYPE> in names stands for the 1082In the following description, uppercase C<TYPE> in names stands for the
964watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer 1083watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
965watchers and C<ev_io_start> for I/O watchers. 1084watchers and C<ev_io_start> for I/O watchers.
966 1085
967A watcher is a structure that you create and register to record your 1086A watcher is an opaque structure that you allocate and register to record
968interest in some event. For instance, if you want to wait for STDIN to 1087your interest in some event. To make a concrete example, imagine you want
969become readable, you would create an C<ev_io> watcher for that: 1088to wait for STDIN to become readable, you would create an C<ev_io> watcher
1089for that:
970 1090
971 static void my_cb (struct ev_loop *loop, ev_io *w, int revents) 1091 static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
972 { 1092 {
973 ev_io_stop (w); 1093 ev_io_stop (w);
974 ev_unloop (loop, EVUNLOOP_ALL); 1094 ev_break (loop, EVBREAK_ALL);
975 } 1095 }
976 1096
977 struct ev_loop *loop = ev_default_loop (0); 1097 struct ev_loop *loop = ev_default_loop (0);
978 1098
979 ev_io stdin_watcher; 1099 ev_io stdin_watcher;
980 1100
981 ev_init (&stdin_watcher, my_cb); 1101 ev_init (&stdin_watcher, my_cb);
982 ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ); 1102 ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
983 ev_io_start (loop, &stdin_watcher); 1103 ev_io_start (loop, &stdin_watcher);
984 1104
985 ev_loop (loop, 0); 1105 ev_run (loop, 0);
986 1106
987As you can see, you are responsible for allocating the memory for your 1107As you can see, you are responsible for allocating the memory for your
988watcher structures (and it is I<usually> a bad idea to do this on the 1108watcher structures (and it is I<usually> a bad idea to do this on the
989stack). 1109stack).
990 1110
991Each watcher has an associated watcher structure (called C<struct ev_TYPE> 1111Each watcher has an associated watcher structure (called C<struct ev_TYPE>
992or simply C<ev_TYPE>, as typedefs are provided for all watcher structs). 1112or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
993 1113
994Each watcher structure must be initialised by a call to C<ev_init 1114Each watcher structure must be initialised by a call to C<ev_init (watcher
995(watcher *, callback)>, which expects a callback to be provided. This 1115*, callback)>, which expects a callback to be provided. This callback is
996callback gets invoked each time the event occurs (or, in the case of I/O 1116invoked each time the event occurs (or, in the case of I/O watchers, each
997watchers, each time the event loop detects that the file descriptor given 1117time the event loop detects that the file descriptor given is readable
998is readable and/or writable). 1118and/or writable).
999 1119
1000Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >> 1120Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
1001macro to configure it, with arguments specific to the watcher type. There 1121macro to configure it, with arguments specific to the watcher type. There
1002is also a macro to combine initialisation and setting in one call: C<< 1122is also a macro to combine initialisation and setting in one call: C<<
1003ev_TYPE_init (watcher *, callback, ...) >>. 1123ev_TYPE_init (watcher *, callback, ...) >>.
1026=item C<EV_WRITE> 1146=item C<EV_WRITE>
1027 1147
1028The file descriptor in the C<ev_io> watcher has become readable and/or 1148The file descriptor in the C<ev_io> watcher has become readable and/or
1029writable. 1149writable.
1030 1150
1031=item C<EV_TIMEOUT> 1151=item C<EV_TIMER>
1032 1152
1033The C<ev_timer> watcher has timed out. 1153The C<ev_timer> watcher has timed out.
1034 1154
1035=item C<EV_PERIODIC> 1155=item C<EV_PERIODIC>
1036 1156
1054 1174
1055=item C<EV_PREPARE> 1175=item C<EV_PREPARE>
1056 1176
1057=item C<EV_CHECK> 1177=item C<EV_CHECK>
1058 1178
1059All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts 1179All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1060to gather new events, and all C<ev_check> watchers are invoked just after 1180gather new events, and all C<ev_check> watchers are queued (not invoked)
1061C<ev_loop> has gathered them, but before it invokes any callbacks for any 1181just after C<ev_run> has gathered them, but before it queues any callbacks
1182for any received events. That means C<ev_prepare> watchers are the last
1183watchers invoked before the event loop sleeps or polls for new events, and
1184C<ev_check> watchers will be invoked before any other watchers of the same
1185or lower priority within an event loop iteration.
1186
1062received events. Callbacks of both watcher types can start and stop as 1187Callbacks of both watcher types can start and stop as many watchers as
1063many watchers as they want, and all of them will be taken into account 1188they want, and all of them will be taken into account (for example, a
1064(for example, a C<ev_prepare> watcher might start an idle watcher to keep 1189C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1065C<ev_loop> from blocking). 1190blocking).
1066 1191
1067=item C<EV_EMBED> 1192=item C<EV_EMBED>
1068 1193
1069The embedded event loop specified in the C<ev_embed> watcher needs attention. 1194The embedded event loop specified in the C<ev_embed> watcher needs attention.
1070 1195
1071=item C<EV_FORK> 1196=item C<EV_FORK>
1072 1197
1073The event loop has been resumed in the child process after fork (see 1198The event loop has been resumed in the child process after fork (see
1074C<ev_fork>). 1199C<ev_fork>).
1200
1201=item C<EV_CLEANUP>
1202
1203The event loop is about to be destroyed (see C<ev_cleanup>).
1075 1204
1076=item C<EV_ASYNC> 1205=item C<EV_ASYNC>
1077 1206
1078The given async watcher has been asynchronously notified (see C<ev_async>). 1207The given async watcher has been asynchronously notified (see C<ev_async>).
1079 1208
1126 1255
1127 ev_io w; 1256 ev_io w;
1128 ev_init (&w, my_cb); 1257 ev_init (&w, my_cb);
1129 ev_io_set (&w, STDIN_FILENO, EV_READ); 1258 ev_io_set (&w, STDIN_FILENO, EV_READ);
1130 1259
1131=item C<ev_TYPE_set> (ev_TYPE *, [args]) 1260=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
1132 1261
1133This macro initialises the type-specific parts of a watcher. You need to 1262This macro initialises the type-specific parts of a watcher. You need to
1134call C<ev_init> at least once before you call this macro, but you can 1263call C<ev_init> at least once before you call this macro, but you can
1135call C<ev_TYPE_set> any number of times. You must not, however, call this 1264call C<ev_TYPE_set> any number of times. You must not, however, call this
1136macro on a watcher that is active (it can be pending, however, which is a 1265macro on a watcher that is active (it can be pending, however, which is a
1149 1278
1150Example: Initialise and set an C<ev_io> watcher in one step. 1279Example: Initialise and set an C<ev_io> watcher in one step.
1151 1280
1152 ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ); 1281 ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
1153 1282
1154=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher) 1283=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
1155 1284
1156Starts (activates) the given watcher. Only active watchers will receive 1285Starts (activates) the given watcher. Only active watchers will receive
1157events. If the watcher is already active nothing will happen. 1286events. If the watcher is already active nothing will happen.
1158 1287
1159Example: Start the C<ev_io> watcher that is being abused as example in this 1288Example: Start the C<ev_io> watcher that is being abused as example in this
1160whole section. 1289whole section.
1161 1290
1162 ev_io_start (EV_DEFAULT_UC, &w); 1291 ev_io_start (EV_DEFAULT_UC, &w);
1163 1292
1164=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher) 1293=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
1165 1294
1166Stops the given watcher if active, and clears the pending status (whether 1295Stops the given watcher if active, and clears the pending status (whether
1167the watcher was active or not). 1296the watcher was active or not).
1168 1297
1169It is possible that stopped watchers are pending - for example, 1298It is possible that stopped watchers are pending - for example,
1194=item ev_cb_set (ev_TYPE *watcher, callback) 1323=item ev_cb_set (ev_TYPE *watcher, callback)
1195 1324
1196Change the callback. You can change the callback at virtually any time 1325Change the callback. You can change the callback at virtually any time
1197(modulo threads). 1326(modulo threads).
1198 1327
1199=item ev_set_priority (ev_TYPE *watcher, priority) 1328=item ev_set_priority (ev_TYPE *watcher, int priority)
1200 1329
1201=item int ev_priority (ev_TYPE *watcher) 1330=item int ev_priority (ev_TYPE *watcher)
1202 1331
1203Set and query the priority of the watcher. The priority is a small 1332Set and query the priority of the watcher. The priority is a small
1204integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI> 1333integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1236watcher isn't pending it does nothing and returns C<0>. 1365watcher isn't pending it does nothing and returns C<0>.
1237 1366
1238Sometimes it can be useful to "poll" a watcher instead of waiting for its 1367Sometimes it can be useful to "poll" a watcher instead of waiting for its
1239callback to be invoked, which can be accomplished with this function. 1368callback to be invoked, which can be accomplished with this function.
1240 1369
1241=item ev_feed_event (struct ev_loop *, watcher *, int revents) 1370=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
1242 1371
1243Feeds the given event set into the event loop, as if the specified event 1372Feeds the given event set into the event loop, as if the specified event
1244had happened for the specified watcher (which must be a pointer to an 1373had happened for the specified watcher (which must be a pointer to an
1245initialised but not necessarily started event watcher). Obviously you must 1374initialised but not necessarily started event watcher). Obviously you must
1246not free the watcher as long as it has pending events. 1375not free the watcher as long as it has pending events.
1252See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related 1381See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1253functions that do not need a watcher. 1382functions that do not need a watcher.
1254 1383
1255=back 1384=back
1256 1385
1386See also the L<ASSOCIATING CUSTOM DATA WITH A WATCHER> and L<BUILDING YOUR
1387OWN COMPOSITE WATCHERS> idioms.
1257 1388
1258=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER 1389=head2 WATCHER STATES
1259 1390
1260Each watcher has, by default, a member C<void *data> that you can change 1391There are various watcher states mentioned throughout this manual -
1261and read at any time: libev will completely ignore it. This can be used 1392active, pending and so on. In this section these states and the rules to
1262to associate arbitrary data with your watcher. If you need more data and 1393transition between them will be described in more detail - and while these
1263don't want to allocate memory and store a pointer to it in that data 1394rules might look complicated, they usually do "the right thing".
1264member, you can also "subclass" the watcher type and provide your own
1265data:
1266 1395
1267 struct my_io 1396=over 4
1268 {
1269 ev_io io;
1270 int otherfd;
1271 void *somedata;
1272 struct whatever *mostinteresting;
1273 };
1274 1397
1275 ... 1398=item initialiased
1276 struct my_io w;
1277 ev_io_init (&w.io, my_cb, fd, EV_READ);
1278 1399
1279And since your callback will be called with a pointer to the watcher, you 1400Before a watcher can be registered with the event loop it has to be
1280can cast it back to your own type: 1401initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
1402C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1281 1403
1282 static void my_cb (struct ev_loop *loop, ev_io *w_, int revents) 1404In this state it is simply some block of memory that is suitable for
1283 { 1405use in an event loop. It can be moved around, freed, reused etc. at
1284 struct my_io *w = (struct my_io *)w_; 1406will - as long as you either keep the memory contents intact, or call
1285 ... 1407C<ev_TYPE_init> again.
1286 }
1287 1408
1288More interesting and less C-conformant ways of casting your callback type 1409=item started/running/active
1289instead have been omitted.
1290 1410
1291Another common scenario is to use some data structure with multiple 1411Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1292embedded watchers: 1412property of the event loop, and is actively waiting for events. While in
1413this state it cannot be accessed (except in a few documented ways), moved,
1414freed or anything else - the only legal thing is to keep a pointer to it,
1415and call libev functions on it that are documented to work on active watchers.
1293 1416
1294 struct my_biggy 1417=item pending
1295 {
1296 int some_data;
1297 ev_timer t1;
1298 ev_timer t2;
1299 }
1300 1418
1301In this case getting the pointer to C<my_biggy> is a bit more 1419If a watcher is active and libev determines that an event it is interested
1302complicated: Either you store the address of your C<my_biggy> struct 1420in has occurred (such as a timer expiring), it will become pending. It will
1303in the C<data> member of the watcher (for woozies), or you need to use 1421stay in this pending state until either it is stopped or its callback is
1304some pointer arithmetic using C<offsetof> inside your watchers (for real 1422about to be invoked, so it is not normally pending inside the watcher
1305programmers): 1423callback.
1306 1424
1307 #include <stddef.h> 1425The watcher might or might not be active while it is pending (for example,
1426an expired non-repeating timer can be pending but no longer active). If it
1427is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
1428but it is still property of the event loop at this time, so cannot be
1429moved, freed or reused. And if it is active the rules described in the
1430previous item still apply.
1308 1431
1309 static void 1432It is also possible to feed an event on a watcher that is not active (e.g.
1310 t1_cb (EV_P_ ev_timer *w, int revents) 1433via C<ev_feed_event>), in which case it becomes pending without being
1311 { 1434active.
1312 struct my_biggy big = (struct my_biggy *)
1313 (((char *)w) - offsetof (struct my_biggy, t1));
1314 }
1315 1435
1316 static void 1436=item stopped
1317 t2_cb (EV_P_ ev_timer *w, int revents) 1437
1318 { 1438A watcher can be stopped implicitly by libev (in which case it might still
1319 struct my_biggy big = (struct my_biggy *) 1439be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
1320 (((char *)w) - offsetof (struct my_biggy, t2)); 1440latter will clear any pending state the watcher might be in, regardless
1321 } 1441of whether it was active or not, so stopping a watcher explicitly before
1442freeing it is often a good idea.
1443
1444While stopped (and not pending) the watcher is essentially in the
1445initialised state, that is, it can be reused, moved, modified in any way
1446you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
1447it again).
1448
1449=back
1322 1450
1323=head2 WATCHER PRIORITY MODELS 1451=head2 WATCHER PRIORITY MODELS
1324 1452
1325Many event loops support I<watcher priorities>, which are usually small 1453Many event loops support I<watcher priorities>, which are usually small
1326integers that influence the ordering of event callback invocation 1454integers that influence the ordering of event callback invocation
1369 1497
1370For example, to emulate how many other event libraries handle priorities, 1498For example, to emulate how many other event libraries handle priorities,
1371you can associate an C<ev_idle> watcher to each such watcher, and in 1499you can associate an C<ev_idle> watcher to each such watcher, and in
1372the normal watcher callback, you just start the idle watcher. The real 1500the normal watcher callback, you just start the idle watcher. The real
1373processing is done in the idle watcher callback. This causes libev to 1501processing is done in the idle watcher callback. This causes libev to
1374continously poll and process kernel event data for the watcher, but when 1502continuously poll and process kernel event data for the watcher, but when
1375the lock-out case is known to be rare (which in turn is rare :), this is 1503the lock-out case is known to be rare (which in turn is rare :), this is
1376workable. 1504workable.
1377 1505
1378Usually, however, the lock-out model implemented that way will perform 1506Usually, however, the lock-out model implemented that way will perform
1379miserably under the type of load it was designed to handle. In that case, 1507miserably under the type of load it was designed to handle. In that case,
1393 { 1521 {
1394 // stop the I/O watcher, we received the event, but 1522 // stop the I/O watcher, we received the event, but
1395 // are not yet ready to handle it. 1523 // are not yet ready to handle it.
1396 ev_io_stop (EV_A_ w); 1524 ev_io_stop (EV_A_ w);
1397 1525
1398 // start the idle watcher to ahndle the actual event. 1526 // start the idle watcher to handle the actual event.
1399 // it will not be executed as long as other watchers 1527 // it will not be executed as long as other watchers
1400 // with the default priority are receiving events. 1528 // with the default priority are receiving events.
1401 ev_idle_start (EV_A_ &idle); 1529 ev_idle_start (EV_A_ &idle);
1402 } 1530 }
1403 1531
1453In general you can register as many read and/or write event watchers per 1581In general you can register as many read and/or write event watchers per
1454fd as you want (as long as you don't confuse yourself). Setting all file 1582fd as you want (as long as you don't confuse yourself). Setting all file
1455descriptors to non-blocking mode is also usually a good idea (but not 1583descriptors to non-blocking mode is also usually a good idea (but not
1456required if you know what you are doing). 1584required if you know what you are doing).
1457 1585
1458If you cannot use non-blocking mode, then force the use of a
1459known-to-be-good backend (at the time of this writing, this includes only
1460C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1461descriptors for which non-blocking operation makes no sense (such as
1462files) - libev doesn't guarentee any specific behaviour in that case.
1463
1464Another thing you have to watch out for is that it is quite easy to 1586Another thing you have to watch out for is that it is quite easy to
1465receive "spurious" readiness notifications, that is your callback might 1587receive "spurious" readiness notifications, that is, your callback might
1466be called with C<EV_READ> but a subsequent C<read>(2) will actually block 1588be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1467because there is no data. Not only are some backends known to create a 1589because there is no data. It is very easy to get into this situation even
1468lot of those (for example Solaris ports), it is very easy to get into 1590with a relatively standard program structure. Thus it is best to always
1469this situation even with a relatively standard program structure. Thus 1591use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1470it is best to always use non-blocking I/O: An extra C<read>(2) returning
1471C<EAGAIN> is far preferable to a program hanging until some data arrives. 1592preferable to a program hanging until some data arrives.
1472 1593
1473If you cannot run the fd in non-blocking mode (for example you should 1594If you cannot run the fd in non-blocking mode (for example you should
1474not play around with an Xlib connection), then you have to separately 1595not play around with an Xlib connection), then you have to separately
1475re-test whether a file descriptor is really ready with a known-to-be good 1596re-test whether a file descriptor is really ready with a known-to-be good
1476interface such as poll (fortunately in our Xlib example, Xlib already 1597interface such as poll (fortunately in the case of Xlib, it already does
1477does this on its own, so its quite safe to use). Some people additionally 1598this on its own, so its quite safe to use). Some people additionally
1478use C<SIGALRM> and an interval timer, just to be sure you won't block 1599use C<SIGALRM> and an interval timer, just to be sure you won't block
1479indefinitely. 1600indefinitely.
1480 1601
1481But really, best use non-blocking mode. 1602But really, best use non-blocking mode.
1482 1603
1510 1631
1511There is no workaround possible except not registering events 1632There is no workaround possible except not registering events
1512for potentially C<dup ()>'ed file descriptors, or to resort to 1633for potentially C<dup ()>'ed file descriptors, or to resort to
1513C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>. 1634C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1514 1635
1636=head3 The special problem of files
1637
1638Many people try to use C<select> (or libev) on file descriptors
1639representing files, and expect it to become ready when their program
1640doesn't block on disk accesses (which can take a long time on their own).
1641
1642However, this cannot ever work in the "expected" way - you get a readiness
1643notification as soon as the kernel knows whether and how much data is
1644there, and in the case of open files, that's always the case, so you
1645always get a readiness notification instantly, and your read (or possibly
1646write) will still block on the disk I/O.
1647
1648Another way to view it is that in the case of sockets, pipes, character
1649devices and so on, there is another party (the sender) that delivers data
1650on its own, but in the case of files, there is no such thing: the disk
1651will not send data on its own, simply because it doesn't know what you
1652wish to read - you would first have to request some data.
1653
1654Since files are typically not-so-well supported by advanced notification
1655mechanism, libev tries hard to emulate POSIX behaviour with respect
1656to files, even though you should not use it. The reason for this is
1657convenience: sometimes you want to watch STDIN or STDOUT, which is
1658usually a tty, often a pipe, but also sometimes files or special devices
1659(for example, C<epoll> on Linux works with F</dev/random> but not with
1660F</dev/urandom>), and even though the file might better be served with
1661asynchronous I/O instead of with non-blocking I/O, it is still useful when
1662it "just works" instead of freezing.
1663
1664So avoid file descriptors pointing to files when you know it (e.g. use
1665libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
1666when you rarely read from a file instead of from a socket, and want to
1667reuse the same code path.
1668
1515=head3 The special problem of fork 1669=head3 The special problem of fork
1516 1670
1517Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit 1671Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1518useless behaviour. Libev fully supports fork, but needs to be told about 1672useless behaviour. Libev fully supports fork, but needs to be told about
1519it in the child. 1673it in the child if you want to continue to use it in the child.
1520 1674
1521To support fork in your programs, you either have to call 1675To support fork in your child processes, you have to call C<ev_loop_fork
1522C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child, 1676()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1523enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or 1677C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1524C<EVBACKEND_POLL>.
1525 1678
1526=head3 The special problem of SIGPIPE 1679=head3 The special problem of SIGPIPE
1527 1680
1528While not really specific to libev, it is easy to forget about C<SIGPIPE>: 1681While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1529when writing to a pipe whose other end has been closed, your program gets 1682when writing to a pipe whose other end has been closed, your program gets
1532 1685
1533So when you encounter spurious, unexplained daemon exits, make sure you 1686So when you encounter spurious, unexplained daemon exits, make sure you
1534ignore SIGPIPE (and maybe make sure you log the exit status of your daemon 1687ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1535somewhere, as that would have given you a big clue). 1688somewhere, as that would have given you a big clue).
1536 1689
1690=head3 The special problem of accept()ing when you can't
1691
1692Many implementations of the POSIX C<accept> function (for example,
1693found in post-2004 Linux) have the peculiar behaviour of not removing a
1694connection from the pending queue in all error cases.
1695
1696For example, larger servers often run out of file descriptors (because
1697of resource limits), causing C<accept> to fail with C<ENFILE> but not
1698rejecting the connection, leading to libev signalling readiness on
1699the next iteration again (the connection still exists after all), and
1700typically causing the program to loop at 100% CPU usage.
1701
1702Unfortunately, the set of errors that cause this issue differs between
1703operating systems, there is usually little the app can do to remedy the
1704situation, and no known thread-safe method of removing the connection to
1705cope with overload is known (to me).
1706
1707One of the easiest ways to handle this situation is to just ignore it
1708- when the program encounters an overload, it will just loop until the
1709situation is over. While this is a form of busy waiting, no OS offers an
1710event-based way to handle this situation, so it's the best one can do.
1711
1712A better way to handle the situation is to log any errors other than
1713C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
1714messages, and continue as usual, which at least gives the user an idea of
1715what could be wrong ("raise the ulimit!"). For extra points one could stop
1716the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
1717usage.
1718
1719If your program is single-threaded, then you could also keep a dummy file
1720descriptor for overload situations (e.g. by opening F</dev/null>), and
1721when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
1722close that fd, and create a new dummy fd. This will gracefully refuse
1723clients under typical overload conditions.
1724
1725The last way to handle it is to simply log the error and C<exit>, as
1726is often done with C<malloc> failures, but this results in an easy
1727opportunity for a DoS attack.
1537 1728
1538=head3 Watcher-Specific Functions 1729=head3 Watcher-Specific Functions
1539 1730
1540=over 4 1731=over 4
1541 1732
1573 ... 1764 ...
1574 struct ev_loop *loop = ev_default_init (0); 1765 struct ev_loop *loop = ev_default_init (0);
1575 ev_io stdin_readable; 1766 ev_io stdin_readable;
1576 ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ); 1767 ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1577 ev_io_start (loop, &stdin_readable); 1768 ev_io_start (loop, &stdin_readable);
1578 ev_loop (loop, 0); 1769 ev_run (loop, 0);
1579 1770
1580 1771
1581=head2 C<ev_timer> - relative and optionally repeating timeouts 1772=head2 C<ev_timer> - relative and optionally repeating timeouts
1582 1773
1583Timer watchers are simple relative timers that generate an event after a 1774Timer watchers are simple relative timers that generate an event after a
1589detecting time jumps is hard, and some inaccuracies are unavoidable (the 1780detecting time jumps is hard, and some inaccuracies are unavoidable (the
1590monotonic clock option helps a lot here). 1781monotonic clock option helps a lot here).
1591 1782
1592The callback is guaranteed to be invoked only I<after> its timeout has 1783The callback is guaranteed to be invoked only I<after> its timeout has
1593passed (not I<at>, so on systems with very low-resolution clocks this 1784passed (not I<at>, so on systems with very low-resolution clocks this
1594might introduce a small delay). If multiple timers become ready during the 1785might introduce a small delay, see "the special problem of being too
1786early", below). If multiple timers become ready during the same loop
1595same loop iteration then the ones with earlier time-out values are invoked 1787iteration then the ones with earlier time-out values are invoked before
1596before ones of the same priority with later time-out values (but this is 1788ones of the same priority with later time-out values (but this is no
1597no longer true when a callback calls C<ev_loop> recursively). 1789longer true when a callback calls C<ev_run> recursively).
1598 1790
1599=head3 Be smart about timeouts 1791=head3 Be smart about timeouts
1600 1792
1601Many real-world problems involve some kind of timeout, usually for error 1793Many real-world problems involve some kind of timeout, usually for error
1602recovery. A typical example is an HTTP request - if the other side hangs, 1794recovery. A typical example is an HTTP request - if the other side hangs,
1677 1869
1678In this case, it would be more efficient to leave the C<ev_timer> alone, 1870In this case, it would be more efficient to leave the C<ev_timer> alone,
1679but remember the time of last activity, and check for a real timeout only 1871but remember the time of last activity, and check for a real timeout only
1680within the callback: 1872within the callback:
1681 1873
1874 ev_tstamp timeout = 60.;
1682 ev_tstamp last_activity; // time of last activity 1875 ev_tstamp last_activity; // time of last activity
1876 ev_timer timer;
1683 1877
1684 static void 1878 static void
1685 callback (EV_P_ ev_timer *w, int revents) 1879 callback (EV_P_ ev_timer *w, int revents)
1686 { 1880 {
1687 ev_tstamp now = ev_now (EV_A); 1881 // calculate when the timeout would happen
1688 ev_tstamp timeout = last_activity + 60.; 1882 ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1689 1883
1690 // if last_activity + 60. is older than now, we did time out 1884 // if negative, it means we the timeout already occurred
1691 if (timeout < now) 1885 if (after < 0.)
1692 { 1886 {
1693 // timeout occured, take action 1887 // timeout occurred, take action
1694 } 1888 }
1695 else 1889 else
1696 { 1890 {
1697 // callback was invoked, but there was some activity, re-arm 1891 // callback was invoked, but there was some recent
1698 // the watcher to fire in last_activity + 60, which is 1892 // activity. simply restart the timer to time out
1699 // guaranteed to be in the future, so "again" is positive: 1893 // after "after" seconds, which is the earliest time
1700 w->repeat = timeout - now; 1894 // the timeout can occur.
1895 ev_timer_set (w, after, 0.);
1701 ev_timer_again (EV_A_ w); 1896 ev_timer_start (EV_A_ w);
1702 } 1897 }
1703 } 1898 }
1704 1899
1705To summarise the callback: first calculate the real timeout (defined 1900To summarise the callback: first calculate in how many seconds the
1706as "60 seconds after the last activity"), then check if that time has 1901timeout will occur (by calculating the absolute time when it would occur,
1707been reached, which means something I<did>, in fact, time out. Otherwise 1902C<last_activity + timeout>, and subtracting the current time, C<ev_now
1708the callback was invoked too early (C<timeout> is in the future), so 1903(EV_A)> from that).
1709re-schedule the timer to fire at that future time, to see if maybe we have
1710a timeout then.
1711 1904
1712Note how C<ev_timer_again> is used, taking advantage of the 1905If this value is negative, then we are already past the timeout, i.e. we
1713C<ev_timer_again> optimisation when the timer is already running. 1906timed out, and need to do whatever is needed in this case.
1907
1908Otherwise, we now the earliest time at which the timeout would trigger,
1909and simply start the timer with this timeout value.
1910
1911In other words, each time the callback is invoked it will check whether
1912the timeout occurred. If not, it will simply reschedule itself to check
1913again at the earliest time it could time out. Rinse. Repeat.
1714 1914
1715This scheme causes more callback invocations (about one every 60 seconds 1915This scheme causes more callback invocations (about one every 60 seconds
1716minus half the average time between activity), but virtually no calls to 1916minus half the average time between activity), but virtually no calls to
1717libev to change the timeout. 1917libev to change the timeout.
1718 1918
1719To start the timer, simply initialise the watcher and set C<last_activity> 1919To start the machinery, simply initialise the watcher and set
1720to the current time (meaning we just have some activity :), then call the 1920C<last_activity> to the current time (meaning there was some activity just
1721callback, which will "do the right thing" and start the timer: 1921now), then call the callback, which will "do the right thing" and start
1922the timer:
1722 1923
1924 last_activity = ev_now (EV_A);
1723 ev_init (timer, callback); 1925 ev_init (&timer, callback);
1724 last_activity = ev_now (loop); 1926 callback (EV_A_ &timer, 0);
1725 callback (loop, timer, EV_TIMEOUT);
1726 1927
1727And when there is some activity, simply store the current time in 1928When there is some activity, simply store the current time in
1728C<last_activity>, no libev calls at all: 1929C<last_activity>, no libev calls at all:
1729 1930
1931 if (activity detected)
1730 last_actiivty = ev_now (loop); 1932 last_activity = ev_now (EV_A);
1933
1934When your timeout value changes, then the timeout can be changed by simply
1935providing a new value, stopping the timer and calling the callback, which
1936will again do the right thing (for example, time out immediately :).
1937
1938 timeout = new_value;
1939 ev_timer_stop (EV_A_ &timer);
1940 callback (EV_A_ &timer, 0);
1731 1941
1732This technique is slightly more complex, but in most cases where the 1942This technique is slightly more complex, but in most cases where the
1733time-out is unlikely to be triggered, much more efficient. 1943time-out is unlikely to be triggered, much more efficient.
1734
1735Changing the timeout is trivial as well (if it isn't hard-coded in the
1736callback :) - just change the timeout and invoke the callback, which will
1737fix things for you.
1738 1944
1739=item 4. Wee, just use a double-linked list for your timeouts. 1945=item 4. Wee, just use a double-linked list for your timeouts.
1740 1946
1741If there is not one request, but many thousands (millions...), all 1947If there is not one request, but many thousands (millions...), all
1742employing some kind of timeout with the same timeout value, then one can 1948employing some kind of timeout with the same timeout value, then one can
1769Method #1 is almost always a bad idea, and buys you nothing. Method #4 is 1975Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1770rather complicated, but extremely efficient, something that really pays 1976rather complicated, but extremely efficient, something that really pays
1771off after the first million or so of active timers, i.e. it's usually 1977off after the first million or so of active timers, i.e. it's usually
1772overkill :) 1978overkill :)
1773 1979
1980=head3 The special problem of being too early
1981
1982If you ask a timer to call your callback after three seconds, then
1983you expect it to be invoked after three seconds - but of course, this
1984cannot be guaranteed to infinite precision. Less obviously, it cannot be
1985guaranteed to any precision by libev - imagine somebody suspending the
1986process with a STOP signal for a few hours for example.
1987
1988So, libev tries to invoke your callback as soon as possible I<after> the
1989delay has occurred, but cannot guarantee this.
1990
1991A less obvious failure mode is calling your callback too early: many event
1992loops compare timestamps with a "elapsed delay >= requested delay", but
1993this can cause your callback to be invoked much earlier than you would
1994expect.
1995
1996To see why, imagine a system with a clock that only offers full second
1997resolution (think windows if you can't come up with a broken enough OS
1998yourself). If you schedule a one-second timer at the time 500.9, then the
1999event loop will schedule your timeout to elapse at a system time of 500
2000(500.9 truncated to the resolution) + 1, or 501.
2001
2002If an event library looks at the timeout 0.1s later, it will see "501 >=
2003501" and invoke the callback 0.1s after it was started, even though a
2004one-second delay was requested - this is being "too early", despite best
2005intentions.
2006
2007This is the reason why libev will never invoke the callback if the elapsed
2008delay equals the requested delay, but only when the elapsed delay is
2009larger than the requested delay. In the example above, libev would only invoke
2010the callback at system time 502, or 1.1s after the timer was started.
2011
2012So, while libev cannot guarantee that your callback will be invoked
2013exactly when requested, it I<can> and I<does> guarantee that the requested
2014delay has actually elapsed, or in other words, it always errs on the "too
2015late" side of things.
2016
1774=head3 The special problem of time updates 2017=head3 The special problem of time updates
1775 2018
1776Establishing the current time is a costly operation (it usually takes at 2019Establishing the current time is a costly operation (it usually takes
1777least two system calls): EV therefore updates its idea of the current 2020at least one system call): EV therefore updates its idea of the current
1778time only before and after C<ev_loop> collects new events, which causes a 2021time only before and after C<ev_run> collects new events, which causes a
1779growing difference between C<ev_now ()> and C<ev_time ()> when handling 2022growing difference between C<ev_now ()> and C<ev_time ()> when handling
1780lots of events in one iteration. 2023lots of events in one iteration.
1781 2024
1782The relative timeouts are calculated relative to the C<ev_now ()> 2025The relative timeouts are calculated relative to the C<ev_now ()>
1783time. This is usually the right thing as this timestamp refers to the time 2026time. This is usually the right thing as this timestamp refers to the time
1788 ev_timer_set (&timer, after + ev_now () - ev_time (), 0.); 2031 ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1789 2032
1790If the event loop is suspended for a long time, you can also force an 2033If the event loop is suspended for a long time, you can also force an
1791update of the time returned by C<ev_now ()> by calling C<ev_now_update 2034update of the time returned by C<ev_now ()> by calling C<ev_now_update
1792()>. 2035()>.
2036
2037=head3 The special problem of unsynchronised clocks
2038
2039Modern systems have a variety of clocks - libev itself uses the normal
2040"wall clock" clock and, if available, the monotonic clock (to avoid time
2041jumps).
2042
2043Neither of these clocks is synchronised with each other or any other clock
2044on the system, so C<ev_time ()> might return a considerably different time
2045than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
2046a call to C<gettimeofday> might return a second count that is one higher
2047than a directly following call to C<time>.
2048
2049The moral of this is to only compare libev-related timestamps with
2050C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
2051a second or so.
2052
2053One more problem arises due to this lack of synchronisation: if libev uses
2054the system monotonic clock and you compare timestamps from C<ev_time>
2055or C<ev_now> from when you started your timer and when your callback is
2056invoked, you will find that sometimes the callback is a bit "early".
2057
2058This is because C<ev_timer>s work in real time, not wall clock time, so
2059libev makes sure your callback is not invoked before the delay happened,
2060I<measured according to the real time>, not the system clock.
2061
2062If your timeouts are based on a physical timescale (e.g. "time out this
2063connection after 100 seconds") then this shouldn't bother you as it is
2064exactly the right behaviour.
2065
2066If you want to compare wall clock/system timestamps to your timers, then
2067you need to use C<ev_periodic>s, as these are based on the wall clock
2068time, where your comparisons will always generate correct results.
1793 2069
1794=head3 The special problems of suspended animation 2070=head3 The special problems of suspended animation
1795 2071
1796When you leave the server world it is quite customary to hit machines that 2072When you leave the server world it is quite customary to hit machines that
1797can suspend/hibernate - what happens to the clocks during such a suspend? 2073can suspend/hibernate - what happens to the clocks during such a suspend?
1841keep up with the timer (because it takes longer than those 10 seconds to 2117keep up with the timer (because it takes longer than those 10 seconds to
1842do stuff) the timer will not fire more than once per event loop iteration. 2118do stuff) the timer will not fire more than once per event loop iteration.
1843 2119
1844=item ev_timer_again (loop, ev_timer *) 2120=item ev_timer_again (loop, ev_timer *)
1845 2121
1846This will act as if the timer timed out and restart it again if it is 2122This will act as if the timer timed out, and restarts it again if it is
1847repeating. The exact semantics are: 2123repeating. It basically works like calling C<ev_timer_stop>, updating the
2124timeout to the C<repeat> value and calling C<ev_timer_start>.
1848 2125
2126The exact semantics are as in the following rules, all of which will be
2127applied to the watcher:
2128
2129=over 4
2130
1849If the timer is pending, its pending status is cleared. 2131=item If the timer is pending, the pending status is always cleared.
1850 2132
1851If the timer is started but non-repeating, stop it (as if it timed out). 2133=item If the timer is started but non-repeating, stop it (as if it timed
2134out, without invoking it).
1852 2135
1853If the timer is repeating, either start it if necessary (with the 2136=item If the timer is repeating, make the C<repeat> value the new timeout
1854C<repeat> value), or reset the running timer to the C<repeat> value. 2137and start the timer, if necessary.
2138
2139=back
1855 2140
1856This sounds a bit complicated, see L<Be smart about timeouts>, above, for a 2141This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1857usage example. 2142usage example.
1858 2143
1859=item ev_timer_remaining (loop, ev_timer *) 2144=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1860 2145
1861Returns the remaining time until a timer fires. If the timer is active, 2146Returns the remaining time until a timer fires. If the timer is active,
1862then this time is relative to the current event loop time, otherwise it's 2147then this time is relative to the current event loop time, otherwise it's
1863the timeout value currently configured. 2148the timeout value currently configured.
1864 2149
1865That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns 2150That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1866C<5>. When the timer is started and one second passes, C<ev_timer_remain> 2151C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
1867will return C<4>. When the timer expires and is restarted, it will return 2152will return C<4>. When the timer expires and is restarted, it will return
1868roughly C<7> (likely slightly less as callback invocation takes some time, 2153roughly C<7> (likely slightly less as callback invocation takes some time,
1869too), and so on. 2154too), and so on.
1870 2155
1871=item ev_tstamp repeat [read-write] 2156=item ev_tstamp repeat [read-write]
1900 } 2185 }
1901 2186
1902 ev_timer mytimer; 2187 ev_timer mytimer;
1903 ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */ 2188 ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1904 ev_timer_again (&mytimer); /* start timer */ 2189 ev_timer_again (&mytimer); /* start timer */
1905 ev_loop (loop, 0); 2190 ev_run (loop, 0);
1906 2191
1907 // and in some piece of code that gets executed on any "activity": 2192 // and in some piece of code that gets executed on any "activity":
1908 // reset the timeout to start ticking again at 10 seconds 2193 // reset the timeout to start ticking again at 10 seconds
1909 ev_timer_again (&mytimer); 2194 ev_timer_again (&mytimer);
1910 2195
1936 2221
1937As with timers, the callback is guaranteed to be invoked only when the 2222As with timers, the callback is guaranteed to be invoked only when the
1938point in time where it is supposed to trigger has passed. If multiple 2223point in time where it is supposed to trigger has passed. If multiple
1939timers become ready during the same loop iteration then the ones with 2224timers become ready during the same loop iteration then the ones with
1940earlier time-out values are invoked before ones with later time-out values 2225earlier time-out values are invoked before ones with later time-out values
1941(but this is no longer true when a callback calls C<ev_loop> recursively). 2226(but this is no longer true when a callback calls C<ev_run> recursively).
1942 2227
1943=head3 Watcher-Specific Functions and Data Members 2228=head3 Watcher-Specific Functions and Data Members
1944 2229
1945=over 4 2230=over 4
1946 2231
1981 2266
1982Another way to think about it (for the mathematically inclined) is that 2267Another way to think about it (for the mathematically inclined) is that
1983C<ev_periodic> will try to run the callback in this mode at the next possible 2268C<ev_periodic> will try to run the callback in this mode at the next possible
1984time where C<time = offset (mod interval)>, regardless of any time jumps. 2269time where C<time = offset (mod interval)>, regardless of any time jumps.
1985 2270
1986For numerical stability it is preferable that the C<offset> value is near 2271The C<interval> I<MUST> be positive, and for numerical stability, the
1987C<ev_now ()> (the current time), but there is no range requirement for 2272interval value should be higher than C<1/8192> (which is around 100
1988this value, and in fact is often specified as zero. 2273microseconds) and C<offset> should be higher than C<0> and should have
2274at most a similar magnitude as the current time (say, within a factor of
2275ten). Typical values for offset are, in fact, C<0> or something between
2276C<0> and C<interval>, which is also the recommended range.
1989 2277
1990Note also that there is an upper limit to how often a timer can fire (CPU 2278Note also that there is an upper limit to how often a timer can fire (CPU
1991speed for example), so if C<interval> is very small then timing stability 2279speed for example), so if C<interval> is very small then timing stability
1992will of course deteriorate. Libev itself tries to be exact to be about one 2280will of course deteriorate. Libev itself tries to be exact to be about one
1993millisecond (if the OS supports it and the machine is fast enough). 2281millisecond (if the OS supports it and the machine is fast enough).
2074Example: Call a callback every hour, or, more precisely, whenever the 2362Example: Call a callback every hour, or, more precisely, whenever the
2075system time is divisible by 3600. The callback invocation times have 2363system time is divisible by 3600. The callback invocation times have
2076potentially a lot of jitter, but good long-term stability. 2364potentially a lot of jitter, but good long-term stability.
2077 2365
2078 static void 2366 static void
2079 clock_cb (struct ev_loop *loop, ev_io *w, int revents) 2367 clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
2080 { 2368 {
2081 ... its now a full hour (UTC, or TAI or whatever your clock follows) 2369 ... its now a full hour (UTC, or TAI or whatever your clock follows)
2082 } 2370 }
2083 2371
2084 ev_periodic hourly_tick; 2372 ev_periodic hourly_tick;
2107 2395
2108=head2 C<ev_signal> - signal me when a signal gets signalled! 2396=head2 C<ev_signal> - signal me when a signal gets signalled!
2109 2397
2110Signal watchers will trigger an event when the process receives a specific 2398Signal watchers will trigger an event when the process receives a specific
2111signal one or more times. Even though signals are very asynchronous, libev 2399signal one or more times. Even though signals are very asynchronous, libev
2112will try it's best to deliver signals synchronously, i.e. as part of the 2400will try its best to deliver signals synchronously, i.e. as part of the
2113normal event processing, like any other event. 2401normal event processing, like any other event.
2114 2402
2115If you want signals to be delivered truly asynchronously, just use 2403If you want signals to be delivered truly asynchronously, just use
2116C<sigaction> as you would do without libev and forget about sharing 2404C<sigaction> as you would do without libev and forget about sharing
2117the signal. You can even use C<ev_async> from a signal handler to 2405the signal. You can even use C<ev_async> from a signal handler to
2131C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should 2419C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2132not be unduly interrupted. If you have a problem with system calls getting 2420not be unduly interrupted. If you have a problem with system calls getting
2133interrupted by signals you can block all signals in an C<ev_check> watcher 2421interrupted by signals you can block all signals in an C<ev_check> watcher
2134and unblock them in an C<ev_prepare> watcher. 2422and unblock them in an C<ev_prepare> watcher.
2135 2423
2136=head3 The special problem of inheritance over execve 2424=head3 The special problem of inheritance over fork/execve/pthread_create
2137 2425
2138Both the signal mask (C<sigprocmask>) and the signal disposition 2426Both the signal mask (C<sigprocmask>) and the signal disposition
2139(C<sigaction>) are unspecified after starting a signal watcher (and after 2427(C<sigaction>) are unspecified after starting a signal watcher (and after
2140stopping it again), that is, libev might or might not block the signal, 2428stopping it again), that is, libev might or might not block the signal,
2141and might or might not set or restore the installed signal handler. 2429and might or might not set or restore the installed signal handler (but
2430see C<EVFLAG_NOSIGMASK>).
2142 2431
2143While this does not matter for the signal disposition (libev never 2432While this does not matter for the signal disposition (libev never
2144sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on 2433sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2145C<execve>), this matters for the signal mask: many programs do not expect 2434C<execve>), this matters for the signal mask: many programs do not expect
2146certain signals to be blocked. 2435certain signals to be blocked.
2151 2440
2152The simplest way to ensure that the signal mask is reset in the child is 2441The simplest way to ensure that the signal mask is reset in the child is
2153to install a fork handler with C<pthread_atfork> that resets it. That will 2442to install a fork handler with C<pthread_atfork> that resets it. That will
2154catch fork calls done by libraries (such as the libc) as well. 2443catch fork calls done by libraries (such as the libc) as well.
2155 2444
2156In current versions of libev, you can also ensure that the signal mask is 2445In current versions of libev, the signal will not be blocked indefinitely
2157not blocking any signals (except temporarily, so thread users watch out) 2446unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
2158by specifying the C<EVFLAG_NOSIGFD> when creating the event loop. This 2447the window of opportunity for problems, it will not go away, as libev
2159is not guaranteed for future versions, however. 2448I<has> to modify the signal mask, at least temporarily.
2449
2450So I can't stress this enough: I<If you do not reset your signal mask when
2451you expect it to be empty, you have a race condition in your code>. This
2452is not a libev-specific thing, this is true for most event libraries.
2453
2454=head3 The special problem of threads signal handling
2455
2456POSIX threads has problematic signal handling semantics, specifically,
2457a lot of functionality (sigfd, sigwait etc.) only really works if all
2458threads in a process block signals, which is hard to achieve.
2459
2460When you want to use sigwait (or mix libev signal handling with your own
2461for the same signals), you can tackle this problem by globally blocking
2462all signals before creating any threads (or creating them with a fully set
2463sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
2464loops. Then designate one thread as "signal receiver thread" which handles
2465these signals. You can pass on any signals that libev might be interested
2466in by calling C<ev_feed_signal>.
2160 2467
2161=head3 Watcher-Specific Functions and Data Members 2468=head3 Watcher-Specific Functions and Data Members
2162 2469
2163=over 4 2470=over 4
2164 2471
2180Example: Try to exit cleanly on SIGINT. 2487Example: Try to exit cleanly on SIGINT.
2181 2488
2182 static void 2489 static void
2183 sigint_cb (struct ev_loop *loop, ev_signal *w, int revents) 2490 sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
2184 { 2491 {
2185 ev_unloop (loop, EVUNLOOP_ALL); 2492 ev_break (loop, EVBREAK_ALL);
2186 } 2493 }
2187 2494
2188 ev_signal signal_watcher; 2495 ev_signal signal_watcher;
2189 ev_signal_init (&signal_watcher, sigint_cb, SIGINT); 2496 ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
2190 ev_signal_start (loop, &signal_watcher); 2497 ev_signal_start (loop, &signal_watcher);
2576 2883
2577Prepare and check watchers are usually (but not always) used in pairs: 2884Prepare and check watchers are usually (but not always) used in pairs:
2578prepare watchers get invoked before the process blocks and check watchers 2885prepare watchers get invoked before the process blocks and check watchers
2579afterwards. 2886afterwards.
2580 2887
2581You I<must not> call C<ev_loop> or similar functions that enter 2888You I<must not> call C<ev_run> or similar functions that enter
2582the current event loop from either C<ev_prepare> or C<ev_check> 2889the current event loop from either C<ev_prepare> or C<ev_check>
2583watchers. Other loops than the current one are fine, however. The 2890watchers. Other loops than the current one are fine, however. The
2584rationale behind this is that you do not need to check for recursion in 2891rationale behind this is that you do not need to check for recursion in
2585those watchers, i.e. the sequence will always be C<ev_prepare>, blocking, 2892those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
2586C<ev_check> so if you have one watcher of each kind they will always be 2893C<ev_check> so if you have one watcher of each kind they will always be
2754 3061
2755 if (timeout >= 0) 3062 if (timeout >= 0)
2756 // create/start timer 3063 // create/start timer
2757 3064
2758 // poll 3065 // poll
2759 ev_loop (EV_A_ 0); 3066 ev_run (EV_A_ 0);
2760 3067
2761 // stop timer again 3068 // stop timer again
2762 if (timeout >= 0) 3069 if (timeout >= 0)
2763 ev_timer_stop (EV_A_ &to); 3070 ev_timer_stop (EV_A_ &to);
2764 3071
2842if you do not want that, you need to temporarily stop the embed watcher). 3149if you do not want that, you need to temporarily stop the embed watcher).
2843 3150
2844=item ev_embed_sweep (loop, ev_embed *) 3151=item ev_embed_sweep (loop, ev_embed *)
2845 3152
2846Make a single, non-blocking sweep over the embedded loop. This works 3153Make a single, non-blocking sweep over the embedded loop. This works
2847similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most 3154similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
2848appropriate way for embedded loops. 3155appropriate way for embedded loops.
2849 3156
2850=item struct ev_loop *other [read-only] 3157=item struct ev_loop *other [read-only]
2851 3158
2852The embedded event loop. 3159The embedded event loop.
2912C<ev_default_fork> cheats and calls it in the wrong process, the fork 3219C<ev_default_fork> cheats and calls it in the wrong process, the fork
2913handlers will be invoked, too, of course. 3220handlers will be invoked, too, of course.
2914 3221
2915=head3 The special problem of life after fork - how is it possible? 3222=head3 The special problem of life after fork - how is it possible?
2916 3223
2917Most uses of C<fork()> consist of forking, then some simple calls to ste 3224Most uses of C<fork()> consist of forking, then some simple calls to set
2918up/change the process environment, followed by a call to C<exec()>. This 3225up/change the process environment, followed by a call to C<exec()>. This
2919sequence should be handled by libev without any problems. 3226sequence should be handled by libev without any problems.
2920 3227
2921This changes when the application actually wants to do event handling 3228This changes when the application actually wants to do event handling
2922in the child, or both parent in child, in effect "continuing" after the 3229in the child, or both parent in child, in effect "continuing" after the
2938disadvantage of having to use multiple event loops (which do not support 3245disadvantage of having to use multiple event loops (which do not support
2939signal watchers). 3246signal watchers).
2940 3247
2941When this is not possible, or you want to use the default loop for 3248When this is not possible, or you want to use the default loop for
2942other reasons, then in the process that wants to start "fresh", call 3249other reasons, then in the process that wants to start "fresh", call
2943C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying 3250C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
2944the default loop will "orphan" (not stop) all registered watchers, so you 3251Destroying the default loop will "orphan" (not stop) all registered
2945have to be careful not to execute code that modifies those watchers. Note 3252watchers, so you have to be careful not to execute code that modifies
2946also that in that case, you have to re-register any signal watchers. 3253those watchers. Note also that in that case, you have to re-register any
3254signal watchers.
2947 3255
2948=head3 Watcher-Specific Functions and Data Members 3256=head3 Watcher-Specific Functions and Data Members
2949 3257
2950=over 4 3258=over 4
2951 3259
2952=item ev_fork_init (ev_signal *, callback) 3260=item ev_fork_init (ev_fork *, callback)
2953 3261
2954Initialises and configures the fork watcher - it has no parameters of any 3262Initialises and configures the fork watcher - it has no parameters of any
2955kind. There is a C<ev_fork_set> macro, but using it is utterly pointless, 3263kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
2956believe me. 3264really.
2957 3265
2958=back 3266=back
2959 3267
2960 3268
3269=head2 C<ev_cleanup> - even the best things end
3270
3271Cleanup watchers are called just before the event loop is being destroyed
3272by a call to C<ev_loop_destroy>.
3273
3274While there is no guarantee that the event loop gets destroyed, cleanup
3275watchers provide a convenient method to install cleanup hooks for your
3276program, worker threads and so on - you just to make sure to destroy the
3277loop when you want them to be invoked.
3278
3279Cleanup watchers are invoked in the same way as any other watcher. Unlike
3280all other watchers, they do not keep a reference to the event loop (which
3281makes a lot of sense if you think about it). Like all other watchers, you
3282can call libev functions in the callback, except C<ev_cleanup_start>.
3283
3284=head3 Watcher-Specific Functions and Data Members
3285
3286=over 4
3287
3288=item ev_cleanup_init (ev_cleanup *, callback)
3289
3290Initialises and configures the cleanup watcher - it has no parameters of
3291any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
3292pointless, I assure you.
3293
3294=back
3295
3296Example: Register an atexit handler to destroy the default loop, so any
3297cleanup functions are called.
3298
3299 static void
3300 program_exits (void)
3301 {
3302 ev_loop_destroy (EV_DEFAULT_UC);
3303 }
3304
3305 ...
3306 atexit (program_exits);
3307
3308
2961=head2 C<ev_async> - how to wake up another event loop 3309=head2 C<ev_async> - how to wake up an event loop
2962 3310
2963In general, you cannot use an C<ev_loop> from multiple threads or other 3311In general, you cannot use an C<ev_loop> from multiple threads or other
2964asynchronous sources such as signal handlers (as opposed to multiple event 3312asynchronous sources such as signal handlers (as opposed to multiple event
2965loops - those are of course safe to use in different threads). 3313loops - those are of course safe to use in different threads).
2966 3314
2967Sometimes, however, you need to wake up another event loop you do not 3315Sometimes, however, you need to wake up an event loop you do not control,
2968control, for example because it belongs to another thread. This is what 3316for example because it belongs to another thread. This is what C<ev_async>
2969C<ev_async> watchers do: as long as the C<ev_async> watcher is active, you 3317watchers do: as long as the C<ev_async> watcher is active, you can signal
2970can signal it by calling C<ev_async_send>, which is thread- and signal 3318it by calling C<ev_async_send>, which is thread- and signal safe.
2971safe.
2972 3319
2973This functionality is very similar to C<ev_signal> watchers, as signals, 3320This functionality is very similar to C<ev_signal> watchers, as signals,
2974too, are asynchronous in nature, and signals, too, will be compressed 3321too, are asynchronous in nature, and signals, too, will be compressed
2975(i.e. the number of callback invocations may be less than the number of 3322(i.e. the number of callback invocations may be less than the number of
2976C<ev_async_sent> calls). 3323C<ev_async_send> calls). In fact, you could use signal watchers as a kind
2977 3324of "global async watchers" by using a watcher on an otherwise unused
2978Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not 3325signal, and C<ev_feed_signal> to signal this watcher from another thread,
2979just the default loop. 3326even without knowing which loop owns the signal.
2980 3327
2981=head3 Queueing 3328=head3 Queueing
2982 3329
2983C<ev_async> does not support queueing of data in any way. The reason 3330C<ev_async> does not support queueing of data in any way. The reason
2984is that the author does not know of a simple (or any) algorithm for a 3331is that the author does not know of a simple (or any) algorithm for a
2985multiple-writer-single-reader queue that works in all cases and doesn't 3332multiple-writer-single-reader queue that works in all cases and doesn't
2986need elaborate support such as pthreads. 3333need elaborate support such as pthreads or unportable memory access
3334semantics.
2987 3335
2988That means that if you want to queue data, you have to provide your own 3336That means that if you want to queue data, you have to provide your own
2989queue. But at least I can tell you how to implement locking around your 3337queue. But at least I can tell you how to implement locking around your
2990queue: 3338queue:
2991 3339
3075trust me. 3423trust me.
3076 3424
3077=item ev_async_send (loop, ev_async *) 3425=item ev_async_send (loop, ev_async *)
3078 3426
3079Sends/signals/activates the given C<ev_async> watcher, that is, feeds 3427Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3080an C<EV_ASYNC> event on the watcher into the event loop. Unlike 3428an C<EV_ASYNC> event on the watcher into the event loop, and instantly
3429returns.
3430
3081C<ev_feed_event>, this call is safe to do from other threads, signal or 3431Unlike C<ev_feed_event>, this call is safe to do from other threads,
3082similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding 3432signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3083section below on what exactly this means). 3433embedding section below on what exactly this means).
3084 3434
3085Note that, as with other watchers in libev, multiple events might get 3435Note that, as with other watchers in libev, multiple events might get
3086compressed into a single callback invocation (another way to look at this 3436compressed into a single callback invocation (another way to look at
3087is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>, 3437this is that C<ev_async> watchers are level-triggered: they are set on
3088reset when the event loop detects that). 3438C<ev_async_send>, reset when the event loop detects that).
3089 3439
3090This call incurs the overhead of a system call only once per event loop 3440This call incurs the overhead of at most one extra system call per event
3091iteration, so while the overhead might be noticeable, it doesn't apply to 3441loop iteration, if the event loop is blocked, and no syscall at all if
3092repeated calls to C<ev_async_send> for the same event loop. 3442the event loop (or your program) is processing events. That means that
3443repeated calls are basically free (there is no need to avoid calls for
3444performance reasons) and that the overhead becomes smaller (typically
3445zero) under load.
3093 3446
3094=item bool = ev_async_pending (ev_async *) 3447=item bool = ev_async_pending (ev_async *)
3095 3448
3096Returns a non-zero value when C<ev_async_send> has been called on the 3449Returns a non-zero value when C<ev_async_send> has been called on the
3097watcher but the event has not yet been processed (or even noted) by the 3450watcher but the event has not yet been processed (or even noted) by the
3130 3483
3131If C<timeout> is less than 0, then no timeout watcher will be 3484If C<timeout> is less than 0, then no timeout watcher will be
3132started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and 3485started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
3133repeat = 0) will be started. C<0> is a valid timeout. 3486repeat = 0) will be started. C<0> is a valid timeout.
3134 3487
3135The callback has the type C<void (*cb)(int revents, void *arg)> and gets 3488The callback has the type C<void (*cb)(int revents, void *arg)> and is
3136passed an C<revents> set like normal event callbacks (a combination of 3489passed an C<revents> set like normal event callbacks (a combination of
3137C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg> 3490C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
3138value passed to C<ev_once>. Note that it is possible to receive I<both> 3491value passed to C<ev_once>. Note that it is possible to receive I<both>
3139a timeout and an io event at the same time - you probably should give io 3492a timeout and an io event at the same time - you probably should give io
3140events precedence. 3493events precedence.
3141 3494
3142Example: wait up to ten seconds for data to appear on STDIN_FILENO. 3495Example: wait up to ten seconds for data to appear on STDIN_FILENO.
3143 3496
3144 static void stdin_ready (int revents, void *arg) 3497 static void stdin_ready (int revents, void *arg)
3145 { 3498 {
3146 if (revents & EV_READ) 3499 if (revents & EV_READ)
3147 /* stdin might have data for us, joy! */; 3500 /* stdin might have data for us, joy! */;
3148 else if (revents & EV_TIMEOUT) 3501 else if (revents & EV_TIMER)
3149 /* doh, nothing entered */; 3502 /* doh, nothing entered */;
3150 } 3503 }
3151 3504
3152 ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0); 3505 ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3153 3506
3154=item ev_feed_fd_event (struct ev_loop *, int fd, int revents) 3507=item ev_feed_fd_event (loop, int fd, int revents)
3155 3508
3156Feed an event on the given fd, as if a file descriptor backend detected 3509Feed an event on the given fd, as if a file descriptor backend detected
3157the given events it. 3510the given events.
3158 3511
3159=item ev_feed_signal_event (struct ev_loop *loop, int signum) 3512=item ev_feed_signal_event (loop, int signum)
3160 3513
3161Feed an event as if the given signal occurred (C<loop> must be the default 3514Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3162loop!). 3515which is async-safe.
3163 3516
3164=back 3517=back
3518
3519
3520=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
3521
3522This section explains some common idioms that are not immediately
3523obvious. Note that examples are sprinkled over the whole manual, and this
3524section only contains stuff that wouldn't fit anywhere else.
3525
3526=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
3527
3528Each watcher has, by default, a C<void *data> member that you can read
3529or modify at any time: libev will completely ignore it. This can be used
3530to associate arbitrary data with your watcher. If you need more data and
3531don't want to allocate memory separately and store a pointer to it in that
3532data member, you can also "subclass" the watcher type and provide your own
3533data:
3534
3535 struct my_io
3536 {
3537 ev_io io;
3538 int otherfd;
3539 void *somedata;
3540 struct whatever *mostinteresting;
3541 };
3542
3543 ...
3544 struct my_io w;
3545 ev_io_init (&w.io, my_cb, fd, EV_READ);
3546
3547And since your callback will be called with a pointer to the watcher, you
3548can cast it back to your own type:
3549
3550 static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
3551 {
3552 struct my_io *w = (struct my_io *)w_;
3553 ...
3554 }
3555
3556More interesting and less C-conformant ways of casting your callback
3557function type instead have been omitted.
3558
3559=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
3560
3561Another common scenario is to use some data structure with multiple
3562embedded watchers, in effect creating your own watcher that combines
3563multiple libev event sources into one "super-watcher":
3564
3565 struct my_biggy
3566 {
3567 int some_data;
3568 ev_timer t1;
3569 ev_timer t2;
3570 }
3571
3572In this case getting the pointer to C<my_biggy> is a bit more
3573complicated: Either you store the address of your C<my_biggy> struct in
3574the C<data> member of the watcher (for woozies or C++ coders), or you need
3575to use some pointer arithmetic using C<offsetof> inside your watchers (for
3576real programmers):
3577
3578 #include <stddef.h>
3579
3580 static void
3581 t1_cb (EV_P_ ev_timer *w, int revents)
3582 {
3583 struct my_biggy big = (struct my_biggy *)
3584 (((char *)w) - offsetof (struct my_biggy, t1));
3585 }
3586
3587 static void
3588 t2_cb (EV_P_ ev_timer *w, int revents)
3589 {
3590 struct my_biggy big = (struct my_biggy *)
3591 (((char *)w) - offsetof (struct my_biggy, t2));
3592 }
3593
3594=head2 AVOIDING FINISHING BEFORE RETURNING
3595
3596Often you have structures like this in event-based programs:
3597
3598 callback ()
3599 {
3600 free (request);
3601 }
3602
3603 request = start_new_request (..., callback);
3604
3605The intent is to start some "lengthy" operation. The C<request> could be
3606used to cancel the operation, or do other things with it.
3607
3608It's not uncommon to have code paths in C<start_new_request> that
3609immediately invoke the callback, for example, to report errors. Or you add
3610some caching layer that finds that it can skip the lengthy aspects of the
3611operation and simply invoke the callback with the result.
3612
3613The problem here is that this will happen I<before> C<start_new_request>
3614has returned, so C<request> is not set.
3615
3616Even if you pass the request by some safer means to the callback, you
3617might want to do something to the request after starting it, such as
3618canceling it, which probably isn't working so well when the callback has
3619already been invoked.
3620
3621A common way around all these issues is to make sure that
3622C<start_new_request> I<always> returns before the callback is invoked. If
3623C<start_new_request> immediately knows the result, it can artificially
3624delay invoking the callback by e.g. using a C<prepare> or C<idle> watcher
3625for example, or more sneakily, by reusing an existing (stopped) watcher
3626and pushing it into the pending queue:
3627
3628 ev_set_cb (watcher, callback);
3629 ev_feed_event (EV_A_ watcher, 0);
3630
3631This way, C<start_new_request> can safely return before the callback is
3632invoked, while not delaying callback invocation too much.
3633
3634=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
3635
3636Often (especially in GUI toolkits) there are places where you have
3637I<modal> interaction, which is most easily implemented by recursively
3638invoking C<ev_run>.
3639
3640This brings the problem of exiting - a callback might want to finish the
3641main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
3642a modal "Are you sure?" dialog is still waiting), or just the nested one
3643and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
3644other combination: In these cases, C<ev_break> will not work alone.
3645
3646The solution is to maintain "break this loop" variable for each C<ev_run>
3647invocation, and use a loop around C<ev_run> until the condition is
3648triggered, using C<EVRUN_ONCE>:
3649
3650 // main loop
3651 int exit_main_loop = 0;
3652
3653 while (!exit_main_loop)
3654 ev_run (EV_DEFAULT_ EVRUN_ONCE);
3655
3656 // in a modal watcher
3657 int exit_nested_loop = 0;
3658
3659 while (!exit_nested_loop)
3660 ev_run (EV_A_ EVRUN_ONCE);
3661
3662To exit from any of these loops, just set the corresponding exit variable:
3663
3664 // exit modal loop
3665 exit_nested_loop = 1;
3666
3667 // exit main program, after modal loop is finished
3668 exit_main_loop = 1;
3669
3670 // exit both
3671 exit_main_loop = exit_nested_loop = 1;
3672
3673=head2 THREAD LOCKING EXAMPLE
3674
3675Here is a fictitious example of how to run an event loop in a different
3676thread from where callbacks are being invoked and watchers are
3677created/added/removed.
3678
3679For a real-world example, see the C<EV::Loop::Async> perl module,
3680which uses exactly this technique (which is suited for many high-level
3681languages).
3682
3683The example uses a pthread mutex to protect the loop data, a condition
3684variable to wait for callback invocations, an async watcher to notify the
3685event loop thread and an unspecified mechanism to wake up the main thread.
3686
3687First, you need to associate some data with the event loop:
3688
3689 typedef struct {
3690 mutex_t lock; /* global loop lock */
3691 ev_async async_w;
3692 thread_t tid;
3693 cond_t invoke_cv;
3694 } userdata;
3695
3696 void prepare_loop (EV_P)
3697 {
3698 // for simplicity, we use a static userdata struct.
3699 static userdata u;
3700
3701 ev_async_init (&u->async_w, async_cb);
3702 ev_async_start (EV_A_ &u->async_w);
3703
3704 pthread_mutex_init (&u->lock, 0);
3705 pthread_cond_init (&u->invoke_cv, 0);
3706
3707 // now associate this with the loop
3708 ev_set_userdata (EV_A_ u);
3709 ev_set_invoke_pending_cb (EV_A_ l_invoke);
3710 ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
3711
3712 // then create the thread running ev_run
3713 pthread_create (&u->tid, 0, l_run, EV_A);
3714 }
3715
3716The callback for the C<ev_async> watcher does nothing: the watcher is used
3717solely to wake up the event loop so it takes notice of any new watchers
3718that might have been added:
3719
3720 static void
3721 async_cb (EV_P_ ev_async *w, int revents)
3722 {
3723 // just used for the side effects
3724 }
3725
3726The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
3727protecting the loop data, respectively.
3728
3729 static void
3730 l_release (EV_P)
3731 {
3732 userdata *u = ev_userdata (EV_A);
3733 pthread_mutex_unlock (&u->lock);
3734 }
3735
3736 static void
3737 l_acquire (EV_P)
3738 {
3739 userdata *u = ev_userdata (EV_A);
3740 pthread_mutex_lock (&u->lock);
3741 }
3742
3743The event loop thread first acquires the mutex, and then jumps straight
3744into C<ev_run>:
3745
3746 void *
3747 l_run (void *thr_arg)
3748 {
3749 struct ev_loop *loop = (struct ev_loop *)thr_arg;
3750
3751 l_acquire (EV_A);
3752 pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
3753 ev_run (EV_A_ 0);
3754 l_release (EV_A);
3755
3756 return 0;
3757 }
3758
3759Instead of invoking all pending watchers, the C<l_invoke> callback will
3760signal the main thread via some unspecified mechanism (signals? pipe
3761writes? C<Async::Interrupt>?) and then waits until all pending watchers
3762have been called (in a while loop because a) spurious wakeups are possible
3763and b) skipping inter-thread-communication when there are no pending
3764watchers is very beneficial):
3765
3766 static void
3767 l_invoke (EV_P)
3768 {
3769 userdata *u = ev_userdata (EV_A);
3770
3771 while (ev_pending_count (EV_A))
3772 {
3773 wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
3774 pthread_cond_wait (&u->invoke_cv, &u->lock);
3775 }
3776 }
3777
3778Now, whenever the main thread gets told to invoke pending watchers, it
3779will grab the lock, call C<ev_invoke_pending> and then signal the loop
3780thread to continue:
3781
3782 static void
3783 real_invoke_pending (EV_P)
3784 {
3785 userdata *u = ev_userdata (EV_A);
3786
3787 pthread_mutex_lock (&u->lock);
3788 ev_invoke_pending (EV_A);
3789 pthread_cond_signal (&u->invoke_cv);
3790 pthread_mutex_unlock (&u->lock);
3791 }
3792
3793Whenever you want to start/stop a watcher or do other modifications to an
3794event loop, you will now have to lock:
3795
3796 ev_timer timeout_watcher;
3797 userdata *u = ev_userdata (EV_A);
3798
3799 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
3800
3801 pthread_mutex_lock (&u->lock);
3802 ev_timer_start (EV_A_ &timeout_watcher);
3803 ev_async_send (EV_A_ &u->async_w);
3804 pthread_mutex_unlock (&u->lock);
3805
3806Note that sending the C<ev_async> watcher is required because otherwise
3807an event loop currently blocking in the kernel will have no knowledge
3808about the newly added timer. By waking up the loop it will pick up any new
3809watchers in the next event loop iteration.
3810
3811=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
3812
3813While the overhead of a callback that e.g. schedules a thread is small, it
3814is still an overhead. If you embed libev, and your main usage is with some
3815kind of threads or coroutines, you might want to customise libev so that
3816doesn't need callbacks anymore.
3817
3818Imagine you have coroutines that you can switch to using a function
3819C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
3820and that due to some magic, the currently active coroutine is stored in a
3821global called C<current_coro>. Then you can build your own "wait for libev
3822event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
3823the differing C<;> conventions):
3824
3825 #define EV_CB_DECLARE(type) struct my_coro *cb;
3826 #define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
3827
3828That means instead of having a C callback function, you store the
3829coroutine to switch to in each watcher, and instead of having libev call
3830your callback, you instead have it switch to that coroutine.
3831
3832A coroutine might now wait for an event with a function called
3833C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
3834matter when, or whether the watcher is active or not when this function is
3835called):
3836
3837 void
3838 wait_for_event (ev_watcher *w)
3839 {
3840 ev_cb_set (w) = current_coro;
3841 switch_to (libev_coro);
3842 }
3843
3844That basically suspends the coroutine inside C<wait_for_event> and
3845continues the libev coroutine, which, when appropriate, switches back to
3846this or any other coroutine.
3847
3848You can do similar tricks if you have, say, threads with an event queue -
3849instead of storing a coroutine, you store the queue object and instead of
3850switching to a coroutine, you push the watcher onto the queue and notify
3851any waiters.
3852
3853To embed libev, see L<EMBEDDING>, but in short, it's easiest to create two
3854files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
3855
3856 // my_ev.h
3857 #define EV_CB_DECLARE(type) struct my_coro *cb;
3858 #define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb);
3859 #include "../libev/ev.h"
3860
3861 // my_ev.c
3862 #define EV_H "my_ev.h"
3863 #include "../libev/ev.c"
3864
3865And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
3866F<my_ev.c> into your project. When properly specifying include paths, you
3867can even use F<ev.h> as header file name directly.
3165 3868
3166 3869
3167=head1 LIBEVENT EMULATION 3870=head1 LIBEVENT EMULATION
3168 3871
3169Libev offers a compatibility emulation layer for libevent. It cannot 3872Libev offers a compatibility emulation layer for libevent. It cannot
3170emulate the internals of libevent, so here are some usage hints: 3873emulate the internals of libevent, so here are some usage hints:
3171 3874
3172=over 4 3875=over 4
3876
3877=item * Only the libevent-1.4.1-beta API is being emulated.
3878
3879This was the newest libevent version available when libev was implemented,
3880and is still mostly unchanged in 2010.
3173 3881
3174=item * Use it by including <event.h>, as usual. 3882=item * Use it by including <event.h>, as usual.
3175 3883
3176=item * The following members are fully supported: ev_base, ev_callback, 3884=item * The following members are fully supported: ev_base, ev_callback,
3177ev_arg, ev_fd, ev_res, ev_events. 3885ev_arg, ev_fd, ev_res, ev_events.
3183=item * Priorities are not currently supported. Initialising priorities 3891=item * Priorities are not currently supported. Initialising priorities
3184will fail and all watchers will have the same priority, even though there 3892will fail and all watchers will have the same priority, even though there
3185is an ev_pri field. 3893is an ev_pri field.
3186 3894
3187=item * In libevent, the last base created gets the signals, in libev, the 3895=item * In libevent, the last base created gets the signals, in libev, the
3188first base created (== the default loop) gets the signals. 3896base that registered the signal gets the signals.
3189 3897
3190=item * Other members are not supported. 3898=item * Other members are not supported.
3191 3899
3192=item * The libev emulation is I<not> ABI compatible to libevent, you need 3900=item * The libev emulation is I<not> ABI compatible to libevent, you need
3193to use the libev header file and library. 3901to use the libev header file and library.
3194 3902
3195=back 3903=back
3196 3904
3197=head1 C++ SUPPORT 3905=head1 C++ SUPPORT
3906
3907=head2 C API
3908
3909The normal C API should work fine when used from C++: both ev.h and the
3910libev sources can be compiled as C++. Therefore, code that uses the C API
3911will work fine.
3912
3913Proper exception specifications might have to be added to callbacks passed
3914to libev: exceptions may be thrown only from watcher callbacks, all
3915other callbacks (allocator, syserr, loop acquire/release and periodioc
3916reschedule callbacks) must not throw exceptions, and might need a C<throw
3917()> specification. If you have code that needs to be compiled as both C
3918and C++ you can use the C<EV_THROW> macro for this:
3919
3920 static void
3921 fatal_error (const char *msg) EV_THROW
3922 {
3923 perror (msg);
3924 abort ();
3925 }
3926
3927 ...
3928 ev_set_syserr_cb (fatal_error);
3929
3930The only API functions that can currently throw exceptions are C<ev_run>,
3931C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
3932because it runs cleanup watchers).
3933
3934Throwing exceptions in watcher callbacks is only supported if libev itself
3935is compiled with a C++ compiler or your C and C++ environments allow
3936throwing exceptions through C libraries (most do).
3937
3938=head2 C++ API
3198 3939
3199Libev comes with some simplistic wrapper classes for C++ that mainly allow 3940Libev comes with some simplistic wrapper classes for C++ that mainly allow
3200you to use some convenience methods to start/stop watchers and also change 3941you to use some convenience methods to start/stop watchers and also change
3201the callback model to a model using method callbacks on objects. 3942the callback model to a model using method callbacks on objects.
3202 3943
3212Care has been taken to keep the overhead low. The only data member the C++ 3953Care has been taken to keep the overhead low. The only data member the C++
3213classes add (compared to plain C-style watchers) is the event loop pointer 3954classes add (compared to plain C-style watchers) is the event loop pointer
3214that the watcher is associated with (or no additional members at all if 3955that the watcher is associated with (or no additional members at all if
3215you disable C<EV_MULTIPLICITY> when embedding libev). 3956you disable C<EV_MULTIPLICITY> when embedding libev).
3216 3957
3217Currently, functions, and static and non-static member functions can be 3958Currently, functions, static and non-static member functions and classes
3218used as callbacks. Other types should be easy to add as long as they only 3959with C<operator ()> can be used as callbacks. Other types should be easy
3219need one additional pointer for context. If you need support for other 3960to add as long as they only need one additional pointer for context. If
3220types of functors please contact the author (preferably after implementing 3961you need support for other types of functors please contact the author
3221it). 3962(preferably after implementing it).
3963
3964For all this to work, your C++ compiler either has to use the same calling
3965conventions as your C compiler (for static member functions), or you have
3966to embed libev and compile libev itself as C++.
3222 3967
3223Here is a list of things available in the C<ev> namespace: 3968Here is a list of things available in the C<ev> namespace:
3224 3969
3225=over 4 3970=over 4
3226 3971
3236=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc. 3981=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
3237 3982
3238For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of 3983For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
3239the same name in the C<ev> namespace, with the exception of C<ev_signal> 3984the same name in the C<ev> namespace, with the exception of C<ev_signal>
3240which is called C<ev::sig> to avoid clashes with the C<signal> macro 3985which is called C<ev::sig> to avoid clashes with the C<signal> macro
3241defines by many implementations. 3986defined by many implementations.
3242 3987
3243All of those classes have these methods: 3988All of those classes have these methods:
3244 3989
3245=over 4 3990=over 4
3246 3991
3247=item ev::TYPE::TYPE () 3992=item ev::TYPE::TYPE ()
3248 3993
3249=item ev::TYPE::TYPE (struct ev_loop *) 3994=item ev::TYPE::TYPE (loop)
3250 3995
3251=item ev::TYPE::~TYPE 3996=item ev::TYPE::~TYPE
3252 3997
3253The constructor (optionally) takes an event loop to associate the watcher 3998The constructor (optionally) takes an event loop to associate the watcher
3254with. If it is omitted, it will use C<EV_DEFAULT>. 3999with. If it is omitted, it will use C<EV_DEFAULT>.
3287 myclass obj; 4032 myclass obj;
3288 ev::io iow; 4033 ev::io iow;
3289 iow.set <myclass, &myclass::io_cb> (&obj); 4034 iow.set <myclass, &myclass::io_cb> (&obj);
3290 4035
3291=item w->set (object *) 4036=item w->set (object *)
3292
3293This is an B<experimental> feature that might go away in a future version.
3294 4037
3295This is a variation of a method callback - leaving out the method to call 4038This is a variation of a method callback - leaving out the method to call
3296will default the method to C<operator ()>, which makes it possible to use 4039will default the method to C<operator ()>, which makes it possible to use
3297functor objects without having to manually specify the C<operator ()> all 4040functor objects without having to manually specify the C<operator ()> all
3298the time. Incidentally, you can then also leave out the template argument 4041the time. Incidentally, you can then also leave out the template argument
3331Example: Use a plain function as callback. 4074Example: Use a plain function as callback.
3332 4075
3333 static void io_cb (ev::io &w, int revents) { } 4076 static void io_cb (ev::io &w, int revents) { }
3334 iow.set <io_cb> (); 4077 iow.set <io_cb> ();
3335 4078
3336=item w->set (struct ev_loop *) 4079=item w->set (loop)
3337 4080
3338Associates a different C<struct ev_loop> with this watcher. You can only 4081Associates a different C<struct ev_loop> with this watcher. You can only
3339do this when the watcher is inactive (and not pending either). 4082do this when the watcher is inactive (and not pending either).
3340 4083
3341=item w->set ([arguments]) 4084=item w->set ([arguments])
3342 4085
3343Basically the same as C<ev_TYPE_set>, with the same arguments. Must be 4086Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
3344called at least once. Unlike the C counterpart, an active watcher gets 4087method or a suitable start method must be called at least once. Unlike the
3345automatically stopped and restarted when reconfiguring it with this 4088C counterpart, an active watcher gets automatically stopped and restarted
3346method. 4089when reconfiguring it with this method.
3347 4090
3348=item w->start () 4091=item w->start ()
3349 4092
3350Starts the watcher. Note that there is no C<loop> argument, as the 4093Starts the watcher. Note that there is no C<loop> argument, as the
3351constructor already stores the event loop. 4094constructor already stores the event loop.
3352 4095
4096=item w->start ([arguments])
4097
4098Instead of calling C<set> and C<start> methods separately, it is often
4099convenient to wrap them in one call. Uses the same type of arguments as
4100the configure C<set> method of the watcher.
4101
3353=item w->stop () 4102=item w->stop ()
3354 4103
3355Stops the watcher if it is active. Again, no C<loop> argument. 4104Stops the watcher if it is active. Again, no C<loop> argument.
3356 4105
3357=item w->again () (C<ev::timer>, C<ev::periodic> only) 4106=item w->again () (C<ev::timer>, C<ev::periodic> only)
3369 4118
3370=back 4119=back
3371 4120
3372=back 4121=back
3373 4122
3374Example: Define a class with an IO and idle watcher, start one of them in 4123Example: Define a class with two I/O and idle watchers, start the I/O
3375the constructor. 4124watchers in the constructor.
3376 4125
3377 class myclass 4126 class myclass
3378 { 4127 {
3379 ev::io io ; void io_cb (ev::io &w, int revents); 4128 ev::io io ; void io_cb (ev::io &w, int revents);
4129 ev::io io2 ; void io2_cb (ev::io &w, int revents);
3380 ev::idle idle; void idle_cb (ev::idle &w, int revents); 4130 ev::idle idle; void idle_cb (ev::idle &w, int revents);
3381 4131
3382 myclass (int fd) 4132 myclass (int fd)
3383 { 4133 {
3384 io .set <myclass, &myclass::io_cb > (this); 4134 io .set <myclass, &myclass::io_cb > (this);
4135 io2 .set <myclass, &myclass::io2_cb > (this);
3385 idle.set <myclass, &myclass::idle_cb> (this); 4136 idle.set <myclass, &myclass::idle_cb> (this);
3386 4137
3387 io.start (fd, ev::READ); 4138 io.set (fd, ev::WRITE); // configure the watcher
4139 io.start (); // start it whenever convenient
4140
4141 io2.start (fd, ev::READ); // set + start in one call
3388 } 4142 }
3389 }; 4143 };
3390 4144
3391 4145
3392=head1 OTHER LANGUAGE BINDINGS 4146=head1 OTHER LANGUAGE BINDINGS
3431L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>. 4185L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
3432 4186
3433=item D 4187=item D
3434 4188
3435Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to 4189Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
3436be found at L<http://proj.llucax.com.ar/wiki/evd>. 4190be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
3437 4191
3438=item Ocaml 4192=item Ocaml
3439 4193
3440Erkki Seppala has written Ocaml bindings for libev, to be found at 4194Erkki Seppala has written Ocaml bindings for libev, to be found at
3441L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>. 4195L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
3442 4196
3443=item Lua 4197=item Lua
3444 4198
3445Brian Maher has written a partial interface to libev 4199Brian Maher has written a partial interface to libev for lua (at the
3446for lua (only C<ev_io> and C<ev_timer>), to be found at 4200time of this writing, only C<ev_io> and C<ev_timer>), to be found at
3447L<http://github.com/brimworks/lua-ev>. 4201L<http://github.com/brimworks/lua-ev>.
3448 4202
3449=back 4203=back
3450 4204
3451 4205
3466loop argument"). The C<EV_A> form is used when this is the sole argument, 4220loop argument"). The C<EV_A> form is used when this is the sole argument,
3467C<EV_A_> is used when other arguments are following. Example: 4221C<EV_A_> is used when other arguments are following. Example:
3468 4222
3469 ev_unref (EV_A); 4223 ev_unref (EV_A);
3470 ev_timer_add (EV_A_ watcher); 4224 ev_timer_add (EV_A_ watcher);
3471 ev_loop (EV_A_ 0); 4225 ev_run (EV_A_ 0);
3472 4226
3473It assumes the variable C<loop> of type C<struct ev_loop *> is in scope, 4227It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
3474which is often provided by the following macro. 4228which is often provided by the following macro.
3475 4229
3476=item C<EV_P>, C<EV_P_> 4230=item C<EV_P>, C<EV_P_>
3489suitable for use with C<EV_A>. 4243suitable for use with C<EV_A>.
3490 4244
3491=item C<EV_DEFAULT>, C<EV_DEFAULT_> 4245=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3492 4246
3493Similar to the other two macros, this gives you the value of the default 4247Similar to the other two macros, this gives you the value of the default
3494loop, if multiple loops are supported ("ev loop default"). 4248loop, if multiple loops are supported ("ev loop default"). The default loop
4249will be initialised if it isn't already initialised.
4250
4251For non-multiplicity builds, these macros do nothing, so you always have
4252to initialise the loop somewhere.
3495 4253
3496=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_> 4254=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3497 4255
3498Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the 4256Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3499default loop has been initialised (C<UC> == unchecked). Their behaviour 4257default loop has been initialised (C<UC> == unchecked). Their behaviour
3516 } 4274 }
3517 4275
3518 ev_check check; 4276 ev_check check;
3519 ev_check_init (&check, check_cb); 4277 ev_check_init (&check, check_cb);
3520 ev_check_start (EV_DEFAULT_ &check); 4278 ev_check_start (EV_DEFAULT_ &check);
3521 ev_loop (EV_DEFAULT_ 0); 4279 ev_run (EV_DEFAULT_ 0);
3522 4280
3523=head1 EMBEDDING 4281=head1 EMBEDDING
3524 4282
3525Libev can (and often is) directly embedded into host 4283Libev can (and often is) directly embedded into host
3526applications. Examples of applications that embed it include the Deliantra 4284applications. Examples of applications that embed it include the Deliantra
3606 libev.m4 4364 libev.m4
3607 4365
3608=head2 PREPROCESSOR SYMBOLS/MACROS 4366=head2 PREPROCESSOR SYMBOLS/MACROS
3609 4367
3610Libev can be configured via a variety of preprocessor symbols you have to 4368Libev can be configured via a variety of preprocessor symbols you have to
3611define before including any of its files. The default in the absence of 4369define before including (or compiling) any of its files. The default in
3612autoconf is documented for every option. 4370the absence of autoconf is documented for every option.
4371
4372Symbols marked with "(h)" do not change the ABI, and can have different
4373values when compiling libev vs. including F<ev.h>, so it is permissible
4374to redefine them before including F<ev.h> without breaking compatibility
4375to a compiled library. All other symbols change the ABI, which means all
4376users of libev and the libev code itself must be compiled with compatible
4377settings.
3613 4378
3614=over 4 4379=over 4
3615 4380
4381=item EV_COMPAT3 (h)
4382
4383Backwards compatibility is a major concern for libev. This is why this
4384release of libev comes with wrappers for the functions and symbols that
4385have been renamed between libev version 3 and 4.
4386
4387You can disable these wrappers (to test compatibility with future
4388versions) by defining C<EV_COMPAT3> to C<0> when compiling your
4389sources. This has the additional advantage that you can drop the C<struct>
4390from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
4391typedef in that case.
4392
4393In some future version, the default for C<EV_COMPAT3> will become C<0>,
4394and in some even more future version the compatibility code will be
4395removed completely.
4396
3616=item EV_STANDALONE 4397=item EV_STANDALONE (h)
3617 4398
3618Must always be C<1> if you do not use autoconf configuration, which 4399Must always be C<1> if you do not use autoconf configuration, which
3619keeps libev from including F<config.h>, and it also defines dummy 4400keeps libev from including F<config.h>, and it also defines dummy
3620implementations for some libevent functions (such as logging, which is not 4401implementations for some libevent functions (such as logging, which is not
3621supported). It will also not define any of the structs usually found in 4402supported). It will also not define any of the structs usually found in
3622F<event.h> that are not directly supported by the libev core alone. 4403F<event.h> that are not directly supported by the libev core alone.
3623 4404
3624In standalone mode, libev will still try to automatically deduce the 4405In standalone mode, libev will still try to automatically deduce the
3625configuration, but has to be more conservative. 4406configuration, but has to be more conservative.
4407
4408=item EV_USE_FLOOR
4409
4410If defined to be C<1>, libev will use the C<floor ()> function for its
4411periodic reschedule calculations, otherwise libev will fall back on a
4412portable (slower) implementation. If you enable this, you usually have to
4413link against libm or something equivalent. Enabling this when the C<floor>
4414function is not available will fail, so the safe default is to not enable
4415this.
3626 4416
3627=item EV_USE_MONOTONIC 4417=item EV_USE_MONOTONIC
3628 4418
3629If defined to be C<1>, libev will try to detect the availability of the 4419If defined to be C<1>, libev will try to detect the availability of the
3630monotonic clock option at both compile time and runtime. Otherwise no 4420monotonic clock option at both compile time and runtime. Otherwise no
3760If defined to be C<1>, libev will compile in support for the Linux inotify 4550If defined to be C<1>, libev will compile in support for the Linux inotify
3761interface to speed up C<ev_stat> watchers. Its actual availability will 4551interface to speed up C<ev_stat> watchers. Its actual availability will
3762be detected at runtime. If undefined, it will be enabled if the headers 4552be detected at runtime. If undefined, it will be enabled if the headers
3763indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled. 4553indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3764 4554
4555=item EV_NO_SMP
4556
4557If defined to be C<1>, libev will assume that memory is always coherent
4558between threads, that is, threads can be used, but threads never run on
4559different cpus (or different cpu cores). This reduces dependencies
4560and makes libev faster.
4561
4562=item EV_NO_THREADS
4563
4564If defined to be C<1>, libev will assume that it will never be called
4565from different threads, which is a stronger assumption than C<EV_NO_SMP>,
4566above. This reduces dependencies and makes libev faster.
4567
3765=item EV_ATOMIC_T 4568=item EV_ATOMIC_T
3766 4569
3767Libev requires an integer type (suitable for storing C<0> or C<1>) whose 4570Libev requires an integer type (suitable for storing C<0> or C<1>) whose
3768access is atomic with respect to other threads or signal contexts. No such 4571access is atomic and serialised with respect to other threads or signal
3769type is easily found in the C language, so you can provide your own type 4572contexts. No such type is easily found in the C language, so you can
3770that you know is safe for your purposes. It is used both for signal handler "locking" 4573provide your own type that you know is safe for your purposes. It is used
3771as well as for signal and thread safety in C<ev_async> watchers. 4574both for signal handler "locking" as well as for signal and thread safety
4575in C<ev_async> watchers.
3772 4576
3773In the absence of this define, libev will use C<sig_atomic_t volatile> 4577In the absence of this define, libev will use C<sig_atomic_t volatile>
3774(from F<signal.h>), which is usually good enough on most platforms. 4578(from F<signal.h>), which is usually good enough on most platforms,
4579although strictly speaking using a type that also implies a memory fence
4580is required.
3775 4581
3776=item EV_H 4582=item EV_H (h)
3777 4583
3778The name of the F<ev.h> header file used to include it. The default if 4584The name of the F<ev.h> header file used to include it. The default if
3779undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be 4585undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
3780used to virtually rename the F<ev.h> header file in case of conflicts. 4586used to virtually rename the F<ev.h> header file in case of conflicts.
3781 4587
3782=item EV_CONFIG_H 4588=item EV_CONFIG_H (h)
3783 4589
3784If C<EV_STANDALONE> isn't C<1>, this variable can be used to override 4590If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
3785F<ev.c>'s idea of where to find the F<config.h> file, similarly to 4591F<ev.c>'s idea of where to find the F<config.h> file, similarly to
3786C<EV_H>, above. 4592C<EV_H>, above.
3787 4593
3788=item EV_EVENT_H 4594=item EV_EVENT_H (h)
3789 4595
3790Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea 4596Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
3791of how the F<event.h> header can be found, the default is C<"event.h">. 4597of how the F<event.h> header can be found, the default is C<"event.h">.
3792 4598
3793=item EV_PROTOTYPES 4599=item EV_PROTOTYPES (h)
3794 4600
3795If defined to be C<0>, then F<ev.h> will not define any function 4601If defined to be C<0>, then F<ev.h> will not define any function
3796prototypes, but still define all the structs and other symbols. This is 4602prototypes, but still define all the structs and other symbols. This is
3797occasionally useful if you want to provide your own wrapper functions 4603occasionally useful if you want to provide your own wrapper functions
3798around libev functions. 4604around libev functions.
3803will have the C<struct ev_loop *> as first argument, and you can create 4609will have the C<struct ev_loop *> as first argument, and you can create
3804additional independent event loops. Otherwise there will be no support 4610additional independent event loops. Otherwise there will be no support
3805for multiple event loops and there is no first event loop pointer 4611for multiple event loops and there is no first event loop pointer
3806argument. Instead, all functions act on the single default loop. 4612argument. Instead, all functions act on the single default loop.
3807 4613
4614Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
4615default loop when multiplicity is switched off - you always have to
4616initialise the loop manually in this case.
4617
3808=item EV_MINPRI 4618=item EV_MINPRI
3809 4619
3810=item EV_MAXPRI 4620=item EV_MAXPRI
3811 4621
3812The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to 4622The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
3820fine. 4630fine.
3821 4631
3822If your embedding application does not need any priorities, defining these 4632If your embedding application does not need any priorities, defining these
3823both to C<0> will save some memory and CPU. 4633both to C<0> will save some memory and CPU.
3824 4634
3825=item EV_PERIODIC_ENABLE 4635=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
4636EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
4637EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3826 4638
3827If undefined or defined to be C<1>, then periodic timers are supported. If 4639If undefined or defined to be C<1> (and the platform supports it), then
3828defined to be C<0>, then they are not. Disabling them saves a few kB of 4640the respective watcher type is supported. If defined to be C<0>, then it
3829code. 4641is not. Disabling watcher types mainly saves code size.
3830 4642
3831=item EV_IDLE_ENABLE 4643=item EV_FEATURES
3832
3833If undefined or defined to be C<1>, then idle watchers are supported. If
3834defined to be C<0>, then they are not. Disabling them saves a few kB of
3835code.
3836
3837=item EV_EMBED_ENABLE
3838
3839If undefined or defined to be C<1>, then embed watchers are supported. If
3840defined to be C<0>, then they are not. Embed watchers rely on most other
3841watcher types, which therefore must not be disabled.
3842
3843=item EV_STAT_ENABLE
3844
3845If undefined or defined to be C<1>, then stat watchers are supported. If
3846defined to be C<0>, then they are not.
3847
3848=item EV_FORK_ENABLE
3849
3850If undefined or defined to be C<1>, then fork watchers are supported. If
3851defined to be C<0>, then they are not.
3852
3853=item EV_ASYNC_ENABLE
3854
3855If undefined or defined to be C<1>, then async watchers are supported. If
3856defined to be C<0>, then they are not.
3857
3858=item EV_MINIMAL
3859 4644
3860If you need to shave off some kilobytes of code at the expense of some 4645If you need to shave off some kilobytes of code at the expense of some
3861speed (but with the full API), define this symbol to C<1>. Currently this 4646speed (but with the full API), you can define this symbol to request
3862is used to override some inlining decisions, saves roughly 30% code size 4647certain subsets of functionality. The default is to enable all features
3863on amd64. It also selects a much smaller 2-heap for timer management over 4648that can be enabled on the platform.
3864the default 4-heap.
3865 4649
3866You can save even more by disabling watcher types you do not need 4650A typical way to use this symbol is to define it to C<0> (or to a bitset
3867and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert> 4651with some broad features you want) and then selectively re-enable
3868(C<-DNDEBUG>) will usually reduce code size a lot. 4652additional parts you want, for example if you want everything minimal,
4653but multiple event loop support, async and child watchers and the poll
4654backend, use this:
3869 4655
3870Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to 4656 #define EV_FEATURES 0
3871provide a bare-bones event library. See C<ev.h> for details on what parts 4657 #define EV_MULTIPLICITY 1
3872of the API are still available, and do not complain if this subset changes 4658 #define EV_USE_POLL 1
3873over time. 4659 #define EV_CHILD_ENABLE 1
4660 #define EV_ASYNC_ENABLE 1
4661
4662The actual value is a bitset, it can be a combination of the following
4663values (by default, all of these are enabled):
4664
4665=over 4
4666
4667=item C<1> - faster/larger code
4668
4669Use larger code to speed up some operations.
4670
4671Currently this is used to override some inlining decisions (enlarging the
4672code size by roughly 30% on amd64).
4673
4674When optimising for size, use of compiler flags such as C<-Os> with
4675gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
4676assertions.
4677
4678The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
4679(e.g. gcc with C<-Os>).
4680
4681=item C<2> - faster/larger data structures
4682
4683Replaces the small 2-heap for timer management by a faster 4-heap, larger
4684hash table sizes and so on. This will usually further increase code size
4685and can additionally have an effect on the size of data structures at
4686runtime.
4687
4688The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
4689(e.g. gcc with C<-Os>).
4690
4691=item C<4> - full API configuration
4692
4693This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
4694enables multiplicity (C<EV_MULTIPLICITY>=1).
4695
4696=item C<8> - full API
4697
4698This enables a lot of the "lesser used" API functions. See C<ev.h> for
4699details on which parts of the API are still available without this
4700feature, and do not complain if this subset changes over time.
4701
4702=item C<16> - enable all optional watcher types
4703
4704Enables all optional watcher types. If you want to selectively enable
4705only some watcher types other than I/O and timers (e.g. prepare,
4706embed, async, child...) you can enable them manually by defining
4707C<EV_watchertype_ENABLE> to C<1> instead.
4708
4709=item C<32> - enable all backends
4710
4711This enables all backends - without this feature, you need to enable at
4712least one backend manually (C<EV_USE_SELECT> is a good choice).
4713
4714=item C<64> - enable OS-specific "helper" APIs
4715
4716Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
4717default.
4718
4719=back
4720
4721Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
4722reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
4723code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
4724watchers, timers and monotonic clock support.
4725
4726With an intelligent-enough linker (gcc+binutils are intelligent enough
4727when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
4728your program might be left out as well - a binary starting a timer and an
4729I/O watcher then might come out at only 5Kb.
4730
4731=item EV_API_STATIC
4732
4733If this symbol is defined (by default it is not), then all identifiers
4734will have static linkage. This means that libev will not export any
4735identifiers, and you cannot link against libev anymore. This can be useful
4736when you embed libev, only want to use libev functions in a single file,
4737and do not want its identifiers to be visible.
4738
4739To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
4740wants to use libev.
4741
4742This option only works when libev is compiled with a C compiler, as C++
4743doesn't support the required declaration syntax.
4744
4745=item EV_AVOID_STDIO
4746
4747If this is set to C<1> at compiletime, then libev will avoid using stdio
4748functions (printf, scanf, perror etc.). This will increase the code size
4749somewhat, but if your program doesn't otherwise depend on stdio and your
4750libc allows it, this avoids linking in the stdio library which is quite
4751big.
4752
4753Note that error messages might become less precise when this option is
4754enabled.
3874 4755
3875=item EV_NSIG 4756=item EV_NSIG
3876 4757
3877The highest supported signal number, +1 (or, the number of 4758The highest supported signal number, +1 (or, the number of
3878signals): Normally, libev tries to deduce the maximum number of signals 4759signals): Normally, libev tries to deduce the maximum number of signals
3879automatically, but sometimes this fails, in which case it can be 4760automatically, but sometimes this fails, in which case it can be
3880specified. Also, using a lower number than detected (C<32> should be 4761specified. Also, using a lower number than detected (C<32> should be
3881good for about any system in existance) can save some memory, as libev 4762good for about any system in existence) can save some memory, as libev
3882statically allocates some 12-24 bytes per signal number. 4763statically allocates some 12-24 bytes per signal number.
3883 4764
3884=item EV_PID_HASHSIZE 4765=item EV_PID_HASHSIZE
3885 4766
3886C<ev_child> watchers use a small hash table to distribute workload by 4767C<ev_child> watchers use a small hash table to distribute workload by
3887pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more 4768pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3888than enough. If you need to manage thousands of children you might want to 4769usually more than enough. If you need to manage thousands of children you
3889increase this value (I<must> be a power of two). 4770might want to increase this value (I<must> be a power of two).
3890 4771
3891=item EV_INOTIFY_HASHSIZE 4772=item EV_INOTIFY_HASHSIZE
3892 4773
3893C<ev_stat> watchers use a small hash table to distribute workload by 4774C<ev_stat> watchers use a small hash table to distribute workload by
3894inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>), 4775inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3895usually more than enough. If you need to manage thousands of C<ev_stat> 4776disabled), usually more than enough. If you need to manage thousands of
3896watchers you might want to increase this value (I<must> be a power of 4777C<ev_stat> watchers you might want to increase this value (I<must> be a
3897two). 4778power of two).
3898 4779
3899=item EV_USE_4HEAP 4780=item EV_USE_4HEAP
3900 4781
3901Heaps are not very cache-efficient. To improve the cache-efficiency of the 4782Heaps are not very cache-efficient. To improve the cache-efficiency of the
3902timer and periodics heaps, libev uses a 4-heap when this symbol is defined 4783timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3903to C<1>. The 4-heap uses more complicated (longer) code but has noticeably 4784to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3904faster performance with many (thousands) of watchers. 4785faster performance with many (thousands) of watchers.
3905 4786
3906The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0> 4787The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3907(disabled). 4788will be C<0>.
3908 4789
3909=item EV_HEAP_CACHE_AT 4790=item EV_HEAP_CACHE_AT
3910 4791
3911Heaps are not very cache-efficient. To improve the cache-efficiency of the 4792Heaps are not very cache-efficient. To improve the cache-efficiency of the
3912timer and periodics heaps, libev can cache the timestamp (I<at>) within 4793timer and periodics heaps, libev can cache the timestamp (I<at>) within
3913the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>), 4794the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3914which uses 8-12 bytes more per watcher and a few hundred bytes more code, 4795which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3915but avoids random read accesses on heap changes. This improves performance 4796but avoids random read accesses on heap changes. This improves performance
3916noticeably with many (hundreds) of watchers. 4797noticeably with many (hundreds) of watchers.
3917 4798
3918The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0> 4799The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3919(disabled). 4800will be C<0>.
3920 4801
3921=item EV_VERIFY 4802=item EV_VERIFY
3922 4803
3923Controls how much internal verification (see C<ev_loop_verify ()>) will 4804Controls how much internal verification (see C<ev_verify ()>) will
3924be done: If set to C<0>, no internal verification code will be compiled 4805be done: If set to C<0>, no internal verification code will be compiled
3925in. If set to C<1>, then verification code will be compiled in, but not 4806in. If set to C<1>, then verification code will be compiled in, but not
3926called. If set to C<2>, then the internal verification code will be 4807called. If set to C<2>, then the internal verification code will be
3927called once per loop, which can slow down libev. If set to C<3>, then the 4808called once per loop, which can slow down libev. If set to C<3>, then the
3928verification code will be called very frequently, which will slow down 4809verification code will be called very frequently, which will slow down
3929libev considerably. 4810libev considerably.
3930 4811
3931The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be 4812The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3932C<0>. 4813will be C<0>.
3933 4814
3934=item EV_COMMON 4815=item EV_COMMON
3935 4816
3936By default, all watchers have a C<void *data> member. By redefining 4817By default, all watchers have a C<void *data> member. By redefining
3937this macro to a something else you can include more and other types of 4818this macro to something else you can include more and other types of
3938members. You have to define it each time you include one of the files, 4819members. You have to define it each time you include one of the files,
3939though, and it must be identical each time. 4820though, and it must be identical each time.
3940 4821
3941For example, the perl EV module uses something like this: 4822For example, the perl EV module uses something like this:
3942 4823
3995file. 4876file.
3996 4877
3997The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file 4878The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3998that everybody includes and which overrides some configure choices: 4879that everybody includes and which overrides some configure choices:
3999 4880
4000 #define EV_MINIMAL 1 4881 #define EV_FEATURES 8
4001 #define EV_USE_POLL 0 4882 #define EV_USE_SELECT 1
4002 #define EV_MULTIPLICITY 0
4003 #define EV_PERIODIC_ENABLE 0 4883 #define EV_PREPARE_ENABLE 1
4884 #define EV_IDLE_ENABLE 1
4004 #define EV_STAT_ENABLE 0 4885 #define EV_SIGNAL_ENABLE 1
4005 #define EV_FORK_ENABLE 0 4886 #define EV_CHILD_ENABLE 1
4887 #define EV_USE_STDEXCEPT 0
4006 #define EV_CONFIG_H <config.h> 4888 #define EV_CONFIG_H <config.h>
4007 #define EV_MINPRI 0
4008 #define EV_MAXPRI 0
4009 4889
4010 #include "ev++.h" 4890 #include "ev++.h"
4011 4891
4012And a F<ev_cpp.C> implementation file that contains libev proper and is compiled: 4892And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4013 4893
4014 #include "ev_cpp.h" 4894 #include "ev_cpp.h"
4015 #include "ev.c" 4895 #include "ev.c"
4016 4896
4017=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES 4897=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
4018 4898
4019=head2 THREADS AND COROUTINES 4899=head2 THREADS AND COROUTINES
4020 4900
4021=head3 THREADS 4901=head3 THREADS
4022 4902
4073default loop and triggering an C<ev_async> watcher from the default loop 4953default loop and triggering an C<ev_async> watcher from the default loop
4074watcher callback into the event loop interested in the signal. 4954watcher callback into the event loop interested in the signal.
4075 4955
4076=back 4956=back
4077 4957
4078=head4 THREAD LOCKING EXAMPLE 4958See also L<THREAD LOCKING EXAMPLE>.
4079
4080Here is a fictitious example of how to run an event loop in a different
4081thread than where callbacks are being invoked and watchers are
4082created/added/removed.
4083
4084For a real-world example, see the C<EV::Loop::Async> perl module,
4085which uses exactly this technique (which is suited for many high-level
4086languages).
4087
4088The example uses a pthread mutex to protect the loop data, a condition
4089variable to wait for callback invocations, an async watcher to notify the
4090event loop thread and an unspecified mechanism to wake up the main thread.
4091
4092First, you need to associate some data with the event loop:
4093
4094 typedef struct {
4095 mutex_t lock; /* global loop lock */
4096 ev_async async_w;
4097 thread_t tid;
4098 cond_t invoke_cv;
4099 } userdata;
4100
4101 void prepare_loop (EV_P)
4102 {
4103 // for simplicity, we use a static userdata struct.
4104 static userdata u;
4105
4106 ev_async_init (&u->async_w, async_cb);
4107 ev_async_start (EV_A_ &u->async_w);
4108
4109 pthread_mutex_init (&u->lock, 0);
4110 pthread_cond_init (&u->invoke_cv, 0);
4111
4112 // now associate this with the loop
4113 ev_set_userdata (EV_A_ u);
4114 ev_set_invoke_pending_cb (EV_A_ l_invoke);
4115 ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4116
4117 // then create the thread running ev_loop
4118 pthread_create (&u->tid, 0, l_run, EV_A);
4119 }
4120
4121The callback for the C<ev_async> watcher does nothing: the watcher is used
4122solely to wake up the event loop so it takes notice of any new watchers
4123that might have been added:
4124
4125 static void
4126 async_cb (EV_P_ ev_async *w, int revents)
4127 {
4128 // just used for the side effects
4129 }
4130
4131The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4132protecting the loop data, respectively.
4133
4134 static void
4135 l_release (EV_P)
4136 {
4137 userdata *u = ev_userdata (EV_A);
4138 pthread_mutex_unlock (&u->lock);
4139 }
4140
4141 static void
4142 l_acquire (EV_P)
4143 {
4144 userdata *u = ev_userdata (EV_A);
4145 pthread_mutex_lock (&u->lock);
4146 }
4147
4148The event loop thread first acquires the mutex, and then jumps straight
4149into C<ev_loop>:
4150
4151 void *
4152 l_run (void *thr_arg)
4153 {
4154 struct ev_loop *loop = (struct ev_loop *)thr_arg;
4155
4156 l_acquire (EV_A);
4157 pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4158 ev_loop (EV_A_ 0);
4159 l_release (EV_A);
4160
4161 return 0;
4162 }
4163
4164Instead of invoking all pending watchers, the C<l_invoke> callback will
4165signal the main thread via some unspecified mechanism (signals? pipe
4166writes? C<Async::Interrupt>?) and then waits until all pending watchers
4167have been called (in a while loop because a) spurious wakeups are possible
4168and b) skipping inter-thread-communication when there are no pending
4169watchers is very beneficial):
4170
4171 static void
4172 l_invoke (EV_P)
4173 {
4174 userdata *u = ev_userdata (EV_A);
4175
4176 while (ev_pending_count (EV_A))
4177 {
4178 wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4179 pthread_cond_wait (&u->invoke_cv, &u->lock);
4180 }
4181 }
4182
4183Now, whenever the main thread gets told to invoke pending watchers, it
4184will grab the lock, call C<ev_invoke_pending> and then signal the loop
4185thread to continue:
4186
4187 static void
4188 real_invoke_pending (EV_P)
4189 {
4190 userdata *u = ev_userdata (EV_A);
4191
4192 pthread_mutex_lock (&u->lock);
4193 ev_invoke_pending (EV_A);
4194 pthread_cond_signal (&u->invoke_cv);
4195 pthread_mutex_unlock (&u->lock);
4196 }
4197
4198Whenever you want to start/stop a watcher or do other modifications to an
4199event loop, you will now have to lock:
4200
4201 ev_timer timeout_watcher;
4202 userdata *u = ev_userdata (EV_A);
4203
4204 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4205
4206 pthread_mutex_lock (&u->lock);
4207 ev_timer_start (EV_A_ &timeout_watcher);
4208 ev_async_send (EV_A_ &u->async_w);
4209 pthread_mutex_unlock (&u->lock);
4210
4211Note that sending the C<ev_async> watcher is required because otherwise
4212an event loop currently blocking in the kernel will have no knowledge
4213about the newly added timer. By waking up the loop it will pick up any new
4214watchers in the next event loop iteration.
4215 4959
4216=head3 COROUTINES 4960=head3 COROUTINES
4217 4961
4218Libev is very accommodating to coroutines ("cooperative threads"): 4962Libev is very accommodating to coroutines ("cooperative threads"):
4219libev fully supports nesting calls to its functions from different 4963libev fully supports nesting calls to its functions from different
4220coroutines (e.g. you can call C<ev_loop> on the same loop from two 4964coroutines (e.g. you can call C<ev_run> on the same loop from two
4221different coroutines, and switch freely between both coroutines running 4965different coroutines, and switch freely between both coroutines running
4222the loop, as long as you don't confuse yourself). The only exception is 4966the loop, as long as you don't confuse yourself). The only exception is
4223that you must not do this from C<ev_periodic> reschedule callbacks. 4967that you must not do this from C<ev_periodic> reschedule callbacks.
4224 4968
4225Care has been taken to ensure that libev does not keep local state inside 4969Care has been taken to ensure that libev does not keep local state inside
4226C<ev_loop>, and other calls do not usually allow for coroutine switches as 4970C<ev_run>, and other calls do not usually allow for coroutine switches as
4227they do not call any callbacks. 4971they do not call any callbacks.
4228 4972
4229=head2 COMPILER WARNINGS 4973=head2 COMPILER WARNINGS
4230 4974
4231Depending on your compiler and compiler settings, you might get no or a 4975Depending on your compiler and compiler settings, you might get no or a
4242maintainable. 4986maintainable.
4243 4987
4244And of course, some compiler warnings are just plain stupid, or simply 4988And of course, some compiler warnings are just plain stupid, or simply
4245wrong (because they don't actually warn about the condition their message 4989wrong (because they don't actually warn about the condition their message
4246seems to warn about). For example, certain older gcc versions had some 4990seems to warn about). For example, certain older gcc versions had some
4247warnings that resulted an extreme number of false positives. These have 4991warnings that resulted in an extreme number of false positives. These have
4248been fixed, but some people still insist on making code warn-free with 4992been fixed, but some people still insist on making code warn-free with
4249such buggy versions. 4993such buggy versions.
4250 4994
4251While libev is written to generate as few warnings as possible, 4995While libev is written to generate as few warnings as possible,
4252"warn-free" code is not a goal, and it is recommended not to build libev 4996"warn-free" code is not a goal, and it is recommended not to build libev
4288I suggest using suppression lists. 5032I suggest using suppression lists.
4289 5033
4290 5034
4291=head1 PORTABILITY NOTES 5035=head1 PORTABILITY NOTES
4292 5036
5037=head2 GNU/LINUX 32 BIT LIMITATIONS
5038
5039GNU/Linux is the only common platform that supports 64 bit file/large file
5040interfaces but I<disables> them by default.
5041
5042That means that libev compiled in the default environment doesn't support
5043files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
5044
5045Unfortunately, many programs try to work around this GNU/Linux issue
5046by enabling the large file API, which makes them incompatible with the
5047standard libev compiled for their system.
5048
5049Likewise, libev cannot enable the large file API itself as this would
5050suddenly make it incompatible to the default compile time environment,
5051i.e. all programs not using special compile switches.
5052
5053=head2 OS/X AND DARWIN BUGS
5054
5055The whole thing is a bug if you ask me - basically any system interface
5056you touch is broken, whether it is locales, poll, kqueue or even the
5057OpenGL drivers.
5058
5059=head3 C<kqueue> is buggy
5060
5061The kqueue syscall is broken in all known versions - most versions support
5062only sockets, many support pipes.
5063
5064Libev tries to work around this by not using C<kqueue> by default on this
5065rotten platform, but of course you can still ask for it when creating a
5066loop - embedding a socket-only kqueue loop into a select-based one is
5067probably going to work well.
5068
5069=head3 C<poll> is buggy
5070
5071Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
5072implementation by something calling C<kqueue> internally around the 10.5.6
5073release, so now C<kqueue> I<and> C<poll> are broken.
5074
5075Libev tries to work around this by not using C<poll> by default on
5076this rotten platform, but of course you can still ask for it when creating
5077a loop.
5078
5079=head3 C<select> is buggy
5080
5081All that's left is C<select>, and of course Apple found a way to fuck this
5082one up as well: On OS/X, C<select> actively limits the number of file
5083descriptors you can pass in to 1024 - your program suddenly crashes when
5084you use more.
5085
5086There is an undocumented "workaround" for this - defining
5087C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
5088work on OS/X.
5089
5090=head2 SOLARIS PROBLEMS AND WORKAROUNDS
5091
5092=head3 C<errno> reentrancy
5093
5094The default compile environment on Solaris is unfortunately so
5095thread-unsafe that you can't even use components/libraries compiled
5096without C<-D_REENTRANT> in a threaded program, which, of course, isn't
5097defined by default. A valid, if stupid, implementation choice.
5098
5099If you want to use libev in threaded environments you have to make sure
5100it's compiled with C<_REENTRANT> defined.
5101
5102=head3 Event port backend
5103
5104The scalable event interface for Solaris is called "event
5105ports". Unfortunately, this mechanism is very buggy in all major
5106releases. If you run into high CPU usage, your program freezes or you get
5107a large number of spurious wakeups, make sure you have all the relevant
5108and latest kernel patches applied. No, I don't know which ones, but there
5109are multiple ones to apply, and afterwards, event ports actually work
5110great.
5111
5112If you can't get it to work, you can try running the program by setting
5113the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
5114C<select> backends.
5115
5116=head2 AIX POLL BUG
5117
5118AIX unfortunately has a broken C<poll.h> header. Libev works around
5119this by trying to avoid the poll backend altogether (i.e. it's not even
5120compiled in), which normally isn't a big problem as C<select> works fine
5121with large bitsets on AIX, and AIX is dead anyway.
5122
4293=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS 5123=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
5124
5125=head3 General issues
4294 5126
4295Win32 doesn't support any of the standards (e.g. POSIX) that libev 5127Win32 doesn't support any of the standards (e.g. POSIX) that libev
4296requires, and its I/O model is fundamentally incompatible with the POSIX 5128requires, and its I/O model is fundamentally incompatible with the POSIX
4297model. Libev still offers limited functionality on this platform in 5129model. Libev still offers limited functionality on this platform in
4298the form of the C<EVBACKEND_SELECT> backend, and only supports socket 5130the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4299descriptors. This only applies when using Win32 natively, not when using 5131descriptors. This only applies when using Win32 natively, not when using
4300e.g. cygwin. 5132e.g. cygwin. Actually, it only applies to the microsofts own compilers,
5133as every compiler comes with a slightly differently broken/incompatible
5134environment.
4301 5135
4302Lifting these limitations would basically require the full 5136Lifting these limitations would basically require the full
4303re-implementation of the I/O system. If you are into these kinds of 5137re-implementation of the I/O system. If you are into this kind of thing,
4304things, then note that glib does exactly that for you in a very portable 5138then note that glib does exactly that for you in a very portable way (note
4305way (note also that glib is the slowest event library known to man). 5139also that glib is the slowest event library known to man).
4306 5140
4307There is no supported compilation method available on windows except 5141There is no supported compilation method available on windows except
4308embedding it into other applications. 5142embedding it into other applications.
4309 5143
4310Sensible signal handling is officially unsupported by Microsoft - libev 5144Sensible signal handling is officially unsupported by Microsoft - libev
4338you do I<not> compile the F<ev.c> or any other embedded source files!): 5172you do I<not> compile the F<ev.c> or any other embedded source files!):
4339 5173
4340 #include "evwrap.h" 5174 #include "evwrap.h"
4341 #include "ev.c" 5175 #include "ev.c"
4342 5176
4343=over 4
4344
4345=item The winsocket select function 5177=head3 The winsocket C<select> function
4346 5178
4347The winsocket C<select> function doesn't follow POSIX in that it 5179The winsocket C<select> function doesn't follow POSIX in that it
4348requires socket I<handles> and not socket I<file descriptors> (it is 5180requires socket I<handles> and not socket I<file descriptors> (it is
4349also extremely buggy). This makes select very inefficient, and also 5181also extremely buggy). This makes select very inefficient, and also
4350requires a mapping from file descriptors to socket handles (the Microsoft 5182requires a mapping from file descriptors to socket handles (the Microsoft
4359 #define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */ 5191 #define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
4360 5192
4361Note that winsockets handling of fd sets is O(n), so you can easily get a 5193Note that winsockets handling of fd sets is O(n), so you can easily get a
4362complexity in the O(n²) range when using win32. 5194complexity in the O(n²) range when using win32.
4363 5195
4364=item Limited number of file descriptors 5196=head3 Limited number of file descriptors
4365 5197
4366Windows has numerous arbitrary (and low) limits on things. 5198Windows has numerous arbitrary (and low) limits on things.
4367 5199
4368Early versions of winsocket's select only supported waiting for a maximum 5200Early versions of winsocket's select only supported waiting for a maximum
4369of C<64> handles (probably owning to the fact that all windows kernels 5201of C<64> handles (probably owning to the fact that all windows kernels
4384runtime libraries. This might get you to about C<512> or C<2048> sockets 5216runtime libraries. This might get you to about C<512> or C<2048> sockets
4385(depending on windows version and/or the phase of the moon). To get more, 5217(depending on windows version and/or the phase of the moon). To get more,
4386you need to wrap all I/O functions and provide your own fd management, but 5218you need to wrap all I/O functions and provide your own fd management, but
4387the cost of calling select (O(n²)) will likely make this unworkable. 5219the cost of calling select (O(n²)) will likely make this unworkable.
4388 5220
4389=back
4390
4391=head2 PORTABILITY REQUIREMENTS 5221=head2 PORTABILITY REQUIREMENTS
4392 5222
4393In addition to a working ISO-C implementation and of course the 5223In addition to a working ISO-C implementation and of course the
4394backend-specific APIs, libev relies on a few additional extensions: 5224backend-specific APIs, libev relies on a few additional extensions:
4395 5225
4401Libev assumes not only that all watcher pointers have the same internal 5231Libev assumes not only that all watcher pointers have the same internal
4402structure (guaranteed by POSIX but not by ISO C for example), but it also 5232structure (guaranteed by POSIX but not by ISO C for example), but it also
4403assumes that the same (machine) code can be used to call any watcher 5233assumes that the same (machine) code can be used to call any watcher
4404callback: The watcher callbacks have different type signatures, but libev 5234callback: The watcher callbacks have different type signatures, but libev
4405calls them using an C<ev_watcher *> internally. 5235calls them using an C<ev_watcher *> internally.
5236
5237=item pointer accesses must be thread-atomic
5238
5239Accessing a pointer value must be atomic, it must both be readable and
5240writable in one piece - this is the case on all current architectures.
4406 5241
4407=item C<sig_atomic_t volatile> must be thread-atomic as well 5242=item C<sig_atomic_t volatile> must be thread-atomic as well
4408 5243
4409The type C<sig_atomic_t volatile> (or whatever is defined as 5244The type C<sig_atomic_t volatile> (or whatever is defined as
4410C<EV_ATOMIC_T>) must be atomic with respect to accesses from different 5245C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
4433watchers. 5268watchers.
4434 5269
4435=item C<double> must hold a time value in seconds with enough accuracy 5270=item C<double> must hold a time value in seconds with enough accuracy
4436 5271
4437The type C<double> is used to represent timestamps. It is required to 5272The type C<double> is used to represent timestamps. It is required to
4438have at least 51 bits of mantissa (and 9 bits of exponent), which is good 5273have at least 51 bits of mantissa (and 9 bits of exponent), which is
4439enough for at least into the year 4000. This requirement is fulfilled by 5274good enough for at least into the year 4000 with millisecond accuracy
5275(the design goal for libev). This requirement is overfulfilled by
4440implementations implementing IEEE 754, which is basically all existing 5276implementations using IEEE 754, which is basically all existing ones.
5277
4441ones. With IEEE 754 doubles, you get microsecond accuracy until at least 5278With IEEE 754 doubles, you get microsecond accuracy until at least the
44422200. 5279year 2255 (and millisecond accuracy till the year 287396 - by then, libev
5280is either obsolete or somebody patched it to use C<long double> or
5281something like that, just kidding).
4443 5282
4444=back 5283=back
4445 5284
4446If you know of other additional requirements drop me a note. 5285If you know of other additional requirements drop me a note.
4447 5286
4509=item Processing ev_async_send: O(number_of_async_watchers) 5348=item Processing ev_async_send: O(number_of_async_watchers)
4510 5349
4511=item Processing signals: O(max_signal_number) 5350=item Processing signals: O(max_signal_number)
4512 5351
4513Sending involves a system call I<iff> there were no other C<ev_async_send> 5352Sending involves a system call I<iff> there were no other C<ev_async_send>
4514calls in the current loop iteration. Checking for async and signal events 5353calls in the current loop iteration and the loop is currently
5354blocked. Checking for async and signal events involves iterating over all
4515involves iterating over all running async watchers or all signal numbers. 5355running async watchers or all signal numbers.
4516 5356
4517=back 5357=back
4518 5358
4519 5359
5360=head1 PORTING FROM LIBEV 3.X TO 4.X
5361
5362The major version 4 introduced some incompatible changes to the API.
5363
5364At the moment, the C<ev.h> header file provides compatibility definitions
5365for all changes, so most programs should still compile. The compatibility
5366layer might be removed in later versions of libev, so better update to the
5367new API early than late.
5368
5369=over 4
5370
5371=item C<EV_COMPAT3> backwards compatibility mechanism
5372
5373The backward compatibility mechanism can be controlled by
5374C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
5375section.
5376
5377=item C<ev_default_destroy> and C<ev_default_fork> have been removed
5378
5379These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
5380
5381 ev_loop_destroy (EV_DEFAULT_UC);
5382 ev_loop_fork (EV_DEFAULT);
5383
5384=item function/symbol renames
5385
5386A number of functions and symbols have been renamed:
5387
5388 ev_loop => ev_run
5389 EVLOOP_NONBLOCK => EVRUN_NOWAIT
5390 EVLOOP_ONESHOT => EVRUN_ONCE
5391
5392 ev_unloop => ev_break
5393 EVUNLOOP_CANCEL => EVBREAK_CANCEL
5394 EVUNLOOP_ONE => EVBREAK_ONE
5395 EVUNLOOP_ALL => EVBREAK_ALL
5396
5397 EV_TIMEOUT => EV_TIMER
5398
5399 ev_loop_count => ev_iteration
5400 ev_loop_depth => ev_depth
5401 ev_loop_verify => ev_verify
5402
5403Most functions working on C<struct ev_loop> objects don't have an
5404C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
5405associated constants have been renamed to not collide with the C<struct
5406ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
5407as all other watcher types. Note that C<ev_loop_fork> is still called
5408C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
5409typedef.
5410
5411=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
5412
5413The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
5414mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
5415and work, but the library code will of course be larger.
5416
5417=back
5418
5419
4520=head1 GLOSSARY 5420=head1 GLOSSARY
4521 5421
4522=over 4 5422=over 4
4523 5423
4524=item active 5424=item active
4525 5425
4526A watcher is active as long as it has been started (has been attached to 5426A watcher is active as long as it has been started and not yet stopped.
4527an event loop) but not yet stopped (disassociated from the event loop). 5427See L<WATCHER STATES> for details.
4528 5428
4529=item application 5429=item application
4530 5430
4531In this document, an application is whatever is using libev. 5431In this document, an application is whatever is using libev.
5432
5433=item backend
5434
5435The part of the code dealing with the operating system interfaces.
4532 5436
4533=item callback 5437=item callback
4534 5438
4535The address of a function that is called when some event has been 5439The address of a function that is called when some event has been
4536detected. Callbacks are being passed the event loop, the watcher that 5440detected. Callbacks are being passed the event loop, the watcher that
4537received the event, and the actual event bitset. 5441received the event, and the actual event bitset.
4538 5442
4539=item callback invocation 5443=item callback/watcher invocation
4540 5444
4541The act of calling the callback associated with a watcher. 5445The act of calling the callback associated with a watcher.
4542 5446
4543=item event 5447=item event
4544 5448
4545A change of state of some external event, such as data now being available 5449A change of state of some external event, such as data now being available
4546for reading on a file descriptor, time having passed or simply not having 5450for reading on a file descriptor, time having passed or simply not having
4547any other events happening anymore. 5451any other events happening anymore.
4548 5452
4549In libev, events are represented as single bits (such as C<EV_READ> or 5453In libev, events are represented as single bits (such as C<EV_READ> or
4550C<EV_TIMEOUT>). 5454C<EV_TIMER>).
4551 5455
4552=item event library 5456=item event library
4553 5457
4554A software package implementing an event model and loop. 5458A software package implementing an event model and loop.
4555 5459
4563The model used to describe how an event loop handles and processes 5467The model used to describe how an event loop handles and processes
4564watchers and events. 5468watchers and events.
4565 5469
4566=item pending 5470=item pending
4567 5471
4568A watcher is pending as soon as the corresponding event has been detected, 5472A watcher is pending as soon as the corresponding event has been
4569and stops being pending as soon as the watcher will be invoked or its 5473detected. See L<WATCHER STATES> for details.
4570pending status is explicitly cleared by the application.
4571
4572A watcher can be pending, but not active. Stopping a watcher also clears
4573its pending status.
4574 5474
4575=item real time 5475=item real time
4576 5476
4577The physical time that is observed. It is apparently strictly monotonic :) 5477The physical time that is observed. It is apparently strictly monotonic :)
4578 5478
4579=item wall-clock time 5479=item wall-clock time
4580 5480
4581The time and date as shown on clocks. Unlike real time, it can actually 5481The time and date as shown on clocks. Unlike real time, it can actually
4582be wrong and jump forwards and backwards, e.g. when the you adjust your 5482be wrong and jump forwards and backwards, e.g. when you adjust your
4583clock. 5483clock.
4584 5484
4585=item watcher 5485=item watcher
4586 5486
4587A data structure that describes interest in certain events. Watchers need 5487A data structure that describes interest in certain events. Watchers need
4588to be started (attached to an event loop) before they can receive events. 5488to be started (attached to an event loop) before they can receive events.
4589 5489
4590=item watcher invocation
4591
4592The act of calling the callback associated with a watcher.
4593
4594=back 5490=back
4595 5491
4596=head1 AUTHOR 5492=head1 AUTHOR
4597 5493
4598Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson. 5494Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
5495Magnusson and Emanuele Giaquinta, and minor corrections by many others.
4599 5496

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines