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

Comparing libev/ev.pod (file contents):
Revision 1.86 by root, Tue Dec 18 01:20:33 2007 UTC vs.
Revision 1.117 by root, Wed Jan 9 04:15:39 2008 UTC

… …
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 #include <ev.h> 7 #include <ev.h>
8 8
9=head1 EXAMPLE PROGRAM 9=head2 EXAMPLE PROGRAM
10 10
11 #include <ev.h> 11 #include <ev.h>
12 12
13 ev_io stdin_watcher; 13 ev_io stdin_watcher;
14 ev_timer timeout_watcher; 14 ev_timer timeout_watcher;
… …
53The newest version of this document is also available as a html-formatted 53The newest version of this document is also available as a html-formatted
54web page you might find easier to navigate when reading it for the first 54web page you might find easier to navigate when reading it for the first
55time: L<http://cvs.schmorp.de/libev/ev.html>. 55time: L<http://cvs.schmorp.de/libev/ev.html>.
56 56
57Libev is an event loop: you register interest in certain events (such as a 57Libev is an event loop: you register interest in certain events (such as a
58file descriptor being readable or a timeout occuring), and it will manage 58file descriptor being readable or a timeout occurring), and it will manage
59these event sources and provide your program with events. 59these event sources and provide your program with events.
60 60
61To do this, it must take more or less complete control over your process 61To do this, it must take more or less complete control over your process
62(or thread) by executing the I<event loop> handler, and will then 62(or thread) by executing the I<event loop> handler, and will then
63communicate events via a callback mechanism. 63communicate events via a callback mechanism.
… …
65You register interest in certain events by registering so-called I<event 65You register interest in certain events by registering so-called I<event
66watchers>, which are relatively small C structures you initialise with the 66watchers>, which are relatively small C structures you initialise with the
67details of the event, and then hand it over to libev by I<starting> the 67details of the event, and then hand it over to libev by I<starting> the
68watcher. 68watcher.
69 69
70=head1 FEATURES 70=head2 FEATURES
71 71
72Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the 72Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
73BSD-specific C<kqueue> and the Solaris-specific event port mechanisms 73BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
74for file descriptor events (C<ev_io>), the Linux C<inotify> interface 74for file descriptor events (C<ev_io>), the Linux C<inotify> interface
75(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers 75(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers
… …
82 82
83It also is quite fast (see this 83It also is quite fast (see this
84L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent 84L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
85for example). 85for example).
86 86
87=head1 CONVENTIONS 87=head2 CONVENTIONS
88 88
89Libev is very configurable. In this manual the default configuration will 89Libev is very configurable. In this manual the default configuration will
90be described, which supports multiple event loops. For more info about 90be described, which supports multiple event loops. For more info about
91various configuration options please have a look at B<EMBED> section in 91various configuration options please have a look at B<EMBED> section in
92this manual. If libev was configured without support for multiple event 92this manual. If libev was configured without support for multiple event
93loops, then all functions taking an initial argument of name C<loop> 93loops, then all functions taking an initial argument of name C<loop>
94(which is always of type C<struct ev_loop *>) will not have this argument. 94(which is always of type C<struct ev_loop *>) will not have this argument.
95 95
96=head1 TIME REPRESENTATION 96=head2 TIME REPRESENTATION
97 97
98Libev represents time as a single floating point number, representing the 98Libev represents time as a single floating point number, representing the
99(fractional) number of seconds since the (POSIX) epoch (somewhere near 99(fractional) number of seconds since the (POSIX) epoch (somewhere near
100the beginning of 1970, details are complicated, don't ask). This type is 100the beginning of 1970, details are complicated, don't ask). This type is
101called C<ev_tstamp>, which is what you should use too. It usually aliases 101called C<ev_tstamp>, which is what you should use too. It usually aliases
… …
115 115
116Returns the current time as libev would use it. Please note that the 116Returns the current time as libev would use it. Please note that the
117C<ev_now> function is usually faster and also often returns the timestamp 117C<ev_now> function is usually faster and also often returns the timestamp
118you actually want to know. 118you actually want to know.
119 119
120=item ev_sleep (ev_tstamp interval)
121
122Sleep for the given interval: The current thread will be blocked until
123either it is interrupted or the given time interval has passed. Basically
124this is a subsecond-resolution C<sleep ()>.
125
120=item int ev_version_major () 126=item int ev_version_major ()
121 127
122=item int ev_version_minor () 128=item int ev_version_minor ()
123 129
124You can find out the major and minor ABI version numbers of the library 130You can find out the major and minor ABI version numbers of the library
… …
300=item C<EVBACKEND_SELECT> (value 1, portable select backend) 306=item C<EVBACKEND_SELECT> (value 1, portable select backend)
301 307
302This is your standard select(2) backend. Not I<completely> standard, as 308This is your standard select(2) backend. Not I<completely> standard, as
303libev tries to roll its own fd_set with no limits on the number of fds, 309libev tries to roll its own fd_set with no limits on the number of fds,
304but if that fails, expect a fairly low limit on the number of fds when 310but if that fails, expect a fairly low limit on the number of fds when
305using this backend. It doesn't scale too well (O(highest_fd)), but its usually 311using this backend. It doesn't scale too well (O(highest_fd)), but its
306the fastest backend for a low number of fds. 312usually the fastest backend for a low number of (low-numbered :) fds.
313
314To get good performance out of this backend you need a high amount of
315parallelity (most of the file descriptors should be busy). If you are
316writing a server, you should C<accept ()> in a loop to accept as many
317connections as possible during one iteration. You might also want to have
318a look at C<ev_set_io_collect_interval ()> to increase the amount of
319readyness notifications you get per iteration.
307 320
308=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows) 321=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
309 322
310And this is your standard poll(2) backend. It's more complicated than 323And this is your standard poll(2) backend. It's more complicated
311select, but handles sparse fds better and has no artificial limit on the 324than select, but handles sparse fds better and has no artificial
312number of fds you can use (except it will slow down considerably with a 325limit on the number of fds you can use (except it will slow down
313lot of inactive fds). It scales similarly to select, i.e. O(total_fds). 326considerably with a lot of inactive fds). It scales similarly to select,
327i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
328performance tips.
314 329
315=item C<EVBACKEND_EPOLL> (value 4, Linux) 330=item C<EVBACKEND_EPOLL> (value 4, Linux)
316 331
317For few fds, this backend is a bit little slower than poll and select, 332For few fds, this backend is a bit little slower than poll and select,
318but it scales phenomenally better. While poll and select usually scale like 333but it scales phenomenally better. While poll and select usually scale
319O(total_fds) where n is the total number of fds (or the highest fd), epoll scales 334like O(total_fds) where n is the total number of fds (or the highest fd),
320either O(1) or O(active_fds). 335epoll scales either O(1) or O(active_fds). The epoll design has a number
336of shortcomings, such as silently dropping events in some hard-to-detect
337cases and rewiring a syscall per fd change, no fork support and bad
338support for dup.
321 339
322While stopping and starting an I/O watcher in the same iteration will 340While stopping, setting and starting an I/O watcher in the same iteration
323result in some caching, there is still a syscall per such incident 341will result in some caching, there is still a syscall per such incident
324(because the fd could point to a different file description now), so its 342(because the fd could point to a different file description now), so its
325best to avoid that. Also, dup()ed file descriptors might not work very 343best to avoid that. Also, C<dup ()>'ed file descriptors might not work
326well if you register events for both fds. 344very well if you register events for both fds.
327 345
328Please note that epoll sometimes generates spurious notifications, so you 346Please note that epoll sometimes generates spurious notifications, so you
329need to use non-blocking I/O or other means to avoid blocking when no data 347need to use non-blocking I/O or other means to avoid blocking when no data
330(or space) is available. 348(or space) is available.
331 349
350Best performance from this backend is achieved by not unregistering all
351watchers for a file descriptor until it has been closed, if possible, i.e.
352keep at least one watcher active per fd at all times.
353
354While nominally embeddeble in other event loops, this feature is broken in
355all kernel versions tested so far.
356
332=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones) 357=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
333 358
334Kqueue deserves special mention, as at the time of this writing, it 359Kqueue deserves special mention, as at the time of this writing, it
335was broken on all BSDs except NetBSD (usually it doesn't work with 360was broken on all BSDs except NetBSD (usually it doesn't work reliably
336anything but sockets and pipes, except on Darwin, where of course its 361with anything but sockets and pipes, except on Darwin, where of course
337completely useless). For this reason its not being "autodetected" 362it's completely useless). For this reason it's not being "autodetected"
338unless you explicitly specify it explicitly in the flags (i.e. using 363unless you explicitly specify it explicitly in the flags (i.e. using
339C<EVBACKEND_KQUEUE>). 364C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
365system like NetBSD.
366
367You still can embed kqueue into a normal poll or select backend and use it
368only for sockets (after having made sure that sockets work with kqueue on
369the target platform). See C<ev_embed> watchers for more info.
340 370
341It scales in the same way as the epoll backend, but the interface to the 371It scales in the same way as the epoll backend, but the interface to the
342kernel is more efficient (which says nothing about its actual speed, of 372kernel is more efficient (which says nothing about its actual speed, of
343course). While starting and stopping an I/O watcher does not cause an 373course). While stopping, setting and starting an I/O watcher does never
344extra syscall as with epoll, it still adds up to four event changes per 374cause an extra syscall as with C<EVBACKEND_EPOLL>, it still adds up to
345incident, so its best to avoid that. 375two event changes per incident, support for C<fork ()> is very bad and it
376drops fds silently in similarly hard-to-detect cases.
377
378This backend usually performs well under most conditions.
379
380While nominally embeddable in other event loops, this doesn't work
381everywhere, so you might need to test for this. And since it is broken
382almost everywhere, you should only use it when you have a lot of sockets
383(for which it usually works), by embedding it into another event loop
384(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for
385sockets.
346 386
347=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8) 387=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
348 388
349This is not implemented yet (and might never be). 389This is not implemented yet (and might never be, unless you send me an
390implementation). According to reports, C</dev/poll> only supports sockets
391and is not embeddable, which would limit the usefulness of this backend
392immensely.
350 393
351=item C<EVBACKEND_PORT> (value 32, Solaris 10) 394=item C<EVBACKEND_PORT> (value 32, Solaris 10)
352 395
353This uses the Solaris 10 port mechanism. As with everything on Solaris, 396This uses the Solaris 10 event port mechanism. As with everything on Solaris,
354it's really slow, but it still scales very well (O(active_fds)). 397it's really slow, but it still scales very well (O(active_fds)).
355 398
356Please note that solaris ports can result in a lot of spurious 399Please note that solaris event ports can deliver a lot of spurious
357notifications, so you need to use non-blocking I/O or other means to avoid 400notifications, so you need to use non-blocking I/O or other means to avoid
358blocking when no data (or space) is available. 401blocking when no data (or space) is available.
402
403While this backend scales well, it requires one system call per active
404file descriptor per loop iteration. For small and medium numbers of file
405descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
406might perform better.
407
408On the positive side, ignoring the spurious readyness notifications, this
409backend actually performed to specification in all tests and is fully
410embeddable, which is a rare feat among the OS-specific backends.
359 411
360=item C<EVBACKEND_ALL> 412=item C<EVBACKEND_ALL>
361 413
362Try all backends (even potentially broken ones that wouldn't be tried 414Try all backends (even potentially broken ones that wouldn't be tried
363with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as 415with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
364C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>. 416C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
365 417
418It is definitely not recommended to use this flag.
419
366=back 420=back
367 421
368If one or more of these are ored into the flags value, then only these 422If one or more of these are ored into the flags value, then only these
369backends will be tried (in the reverse order as given here). If none are 423backends will be tried (in the reverse order as listed here). If none are
370specified, most compiled-in backend will be tried, usually in reverse 424specified, all backends in C<ev_recommended_backends ()> will be tried.
371order of their flag values :)
372 425
373The most typical usage is like this: 426The most typical usage is like this:
374 427
375 if (!ev_default_loop (0)) 428 if (!ev_default_loop (0))
376 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?"); 429 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
… …
404Destroys the default loop again (frees all memory and kernel state 457Destroys the default loop again (frees all memory and kernel state
405etc.). None of the active event watchers will be stopped in the normal 458etc.). None of the active event watchers will be stopped in the normal
406sense, so e.g. C<ev_is_active> might still return true. It is your 459sense, so e.g. C<ev_is_active> might still return true. It is your
407responsibility to either stop all watchers cleanly yoursef I<before> 460responsibility to either stop all watchers cleanly yoursef I<before>
408calling this function, or cope with the fact afterwards (which is usually 461calling this function, or cope with the fact afterwards (which is usually
409the easiest thing, youc na just ignore the watchers and/or C<free ()> them 462the easiest thing, you can just ignore the watchers and/or C<free ()> them
410for example). 463for example).
464
465Note that certain global state, such as signal state, will not be freed by
466this function, and related watchers (such as signal and child watchers)
467would need to be stopped manually.
468
469In general it is not advisable to call this function except in the
470rare occasion where you really need to free e.g. the signal handling
471pipe fds. If you need dynamically allocated loops it is better to use
472C<ev_loop_new> and C<ev_loop_destroy>).
411 473
412=item ev_loop_destroy (loop) 474=item ev_loop_destroy (loop)
413 475
414Like C<ev_default_destroy>, but destroys an event loop created by an 476Like C<ev_default_destroy>, but destroys an event loop created by an
415earlier call to C<ev_loop_new>. 477earlier call to C<ev_loop_new>.
… …
460 522
461Returns the current "event loop time", which is the time the event loop 523Returns the current "event loop time", which is the time the event loop
462received events and started processing them. This timestamp does not 524received events and started processing them. This timestamp does not
463change as long as callbacks are being processed, and this is also the base 525change as long as callbacks are being processed, and this is also the base
464time used for relative timers. You can treat it as the timestamp of the 526time used for relative timers. You can treat it as the timestamp of the
465event occuring (or more correctly, libev finding out about it). 527event occurring (or more correctly, libev finding out about it).
466 528
467=item ev_loop (loop, int flags) 529=item ev_loop (loop, int flags)
468 530
469Finally, this is it, the event handler. This function usually is called 531Finally, this is it, the event handler. This function usually is called
470after you initialised all your watchers and you want to start handling 532after you initialised all your watchers and you want to start handling
… …
492usually a better approach for this kind of thing. 554usually a better approach for this kind of thing.
493 555
494Here are the gory details of what C<ev_loop> does: 556Here are the gory details of what C<ev_loop> does:
495 557
496 - Before the first iteration, call any pending watchers. 558 - Before the first iteration, call any pending watchers.
497 * If there are no active watchers (reference count is zero), return. 559 * If EVFLAG_FORKCHECK was used, check for a fork.
498 - Queue all prepare watchers and then call all outstanding watchers. 560 - If a fork was detected, queue and call all fork watchers.
561 - Queue and call all prepare watchers.
499 - If we have been forked, recreate the kernel state. 562 - If we have been forked, recreate the kernel state.
500 - Update the kernel state with all outstanding changes. 563 - Update the kernel state with all outstanding changes.
501 - Update the "event loop time". 564 - Update the "event loop time".
502 - Calculate for how long to block. 565 - Calculate for how long to sleep or block, if at all
566 (active idle watchers, EVLOOP_NONBLOCK or not having
567 any active watchers at all will result in not sleeping).
568 - Sleep if the I/O and timer collect interval say so.
503 - Block the process, waiting for any events. 569 - Block the process, waiting for any events.
504 - Queue all outstanding I/O (fd) events. 570 - Queue all outstanding I/O (fd) events.
505 - Update the "event loop time" and do time jump handling. 571 - Update the "event loop time" and do time jump handling.
506 - Queue all outstanding timers. 572 - Queue all outstanding timers.
507 - Queue all outstanding periodics. 573 - Queue all outstanding periodics.
508 - If no events are pending now, queue all idle watchers. 574 - If no events are pending now, queue all idle watchers.
509 - Queue all check watchers. 575 - Queue all check watchers.
510 - Call all queued watchers in reverse order (i.e. check watchers first). 576 - Call all queued watchers in reverse order (i.e. check watchers first).
511 Signals and child watchers are implemented as I/O watchers, and will 577 Signals and child watchers are implemented as I/O watchers, and will
512 be handled here by queueing them when their watcher gets executed. 578 be handled here by queueing them when their watcher gets executed.
513 - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK 579 - If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
514 were used, return, otherwise continue with step *. 580 were used, or there are no active watchers, return, otherwise
581 continue with step *.
515 582
516Example: Queue some jobs and then loop until no events are outsanding 583Example: Queue some jobs and then loop until no events are outstanding
517anymore. 584anymore.
518 585
519 ... queue jobs here, make sure they register event watchers as long 586 ... queue jobs here, make sure they register event watchers as long
520 ... as they still have work to do (even an idle watcher will do..) 587 ... as they still have work to do (even an idle watcher will do..)
521 ev_loop (my_loop, 0); 588 ev_loop (my_loop, 0);
… …
525 592
526Can be used to make a call to C<ev_loop> return early (but only after it 593Can be used to make a call to C<ev_loop> return early (but only after it
527has processed all outstanding events). The C<how> argument must be either 594has processed all outstanding events). The C<how> argument must be either
528C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or 595C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
529C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return. 596C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
597
598This "unloop state" will be cleared when entering C<ev_loop> again.
530 599
531=item ev_ref (loop) 600=item ev_ref (loop)
532 601
533=item ev_unref (loop) 602=item ev_unref (loop)
534 603
… …
539returning, ev_unref() after starting, and ev_ref() before stopping it. For 608returning, ev_unref() after starting, and ev_ref() before stopping it. For
540example, libev itself uses this for its internal signal pipe: It is not 609example, libev itself uses this for its internal signal pipe: It is not
541visible to the libev user and should not keep C<ev_loop> from exiting if 610visible to the libev user and should not keep C<ev_loop> from exiting if
542no event watchers registered by it are active. It is also an excellent 611no event watchers registered by it are active. It is also an excellent
543way to do this for generic recurring timers or from within third-party 612way to do this for generic recurring timers or from within third-party
544libraries. Just remember to I<unref after start> and I<ref before stop>. 613libraries. Just remember to I<unref after start> and I<ref before stop>
614(but only if the watcher wasn't active before, or was active before,
615respectively).
545 616
546Example: Create a signal watcher, but keep it from keeping C<ev_loop> 617Example: Create a signal watcher, but keep it from keeping C<ev_loop>
547running when nothing else is active. 618running when nothing else is active.
548 619
549 struct ev_signal exitsig; 620 struct ev_signal exitsig;
… …
553 624
554Example: For some weird reason, unregister the above signal handler again. 625Example: For some weird reason, unregister the above signal handler again.
555 626
556 ev_ref (loop); 627 ev_ref (loop);
557 ev_signal_stop (loop, &exitsig); 628 ev_signal_stop (loop, &exitsig);
629
630=item ev_set_io_collect_interval (loop, ev_tstamp interval)
631
632=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
633
634These advanced functions influence the time that libev will spend waiting
635for events. Both are by default C<0>, meaning that libev will try to
636invoke timer/periodic callbacks and I/O callbacks with minimum latency.
637
638Setting these to a higher value (the C<interval> I<must> be >= C<0>)
639allows libev to delay invocation of I/O and timer/periodic callbacks to
640increase efficiency of loop iterations.
641
642The background is that sometimes your program runs just fast enough to
643handle one (or very few) event(s) per loop iteration. While this makes
644the program responsive, it also wastes a lot of CPU time to poll for new
645events, especially with backends like C<select ()> which have a high
646overhead for the actual polling but can deliver many events at once.
647
648By setting a higher I<io collect interval> you allow libev to spend more
649time collecting I/O events, so you can handle more events per iteration,
650at the cost of increasing latency. Timeouts (both C<ev_periodic> and
651C<ev_timer>) will be not affected. Setting this to a non-null value will
652introduce an additional C<ev_sleep ()> call into most loop iterations.
653
654Likewise, by setting a higher I<timeout collect interval> you allow libev
655to spend more time collecting timeouts, at the expense of increased
656latency (the watcher callback will be called later). C<ev_io> watchers
657will not be affected. Setting this to a non-null value will not introduce
658any overhead in libev.
659
660Many (busy) programs can usually benefit by setting the io collect
661interval to a value near C<0.1> or so, which is often enough for
662interactive servers (of course not for games), likewise for timeouts. It
663usually doesn't make much sense to set it to a lower value than C<0.01>,
664as this approsaches the timing granularity of most systems.
558 665
559=back 666=back
560 667
561 668
562=head1 ANATOMY OF A WATCHER 669=head1 ANATOMY OF A WATCHER
… …
888In general you can register as many read and/or write event watchers per 995In general you can register as many read and/or write event watchers per
889fd as you want (as long as you don't confuse yourself). Setting all file 996fd as you want (as long as you don't confuse yourself). Setting all file
890descriptors to non-blocking mode is also usually a good idea (but not 997descriptors to non-blocking mode is also usually a good idea (but not
891required if you know what you are doing). 998required if you know what you are doing).
892 999
893You have to be careful with dup'ed file descriptors, though. Some backends
894(the linux epoll backend is a notable example) cannot handle dup'ed file
895descriptors correctly if you register interest in two or more fds pointing
896to the same underlying file/socket/etc. description (that is, they share
897the same underlying "file open").
898
899If you must do this, then force the use of a known-to-be-good backend 1000If you must do this, then force the use of a known-to-be-good backend
900(at the time of this writing, this includes only C<EVBACKEND_SELECT> and 1001(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
901C<EVBACKEND_POLL>). 1002C<EVBACKEND_POLL>).
902 1003
903Another thing you have to watch out for is that it is quite easy to 1004Another thing you have to watch out for is that it is quite easy to
… …
915such as poll (fortunately in our Xlib example, Xlib already does this on 1016such as poll (fortunately in our Xlib example, Xlib already does this on
916its own, so its quite safe to use). 1017its own, so its quite safe to use).
917 1018
918=head3 The special problem of disappearing file descriptors 1019=head3 The special problem of disappearing file descriptors
919 1020
920Some backends (e.g kqueue, epoll) need to be told about closing a file 1021Some backends (e.g. kqueue, epoll) need to be told about closing a file
921descriptor (either by calling C<close> explicitly or by any other means, 1022descriptor (either by calling C<close> explicitly or by any other means,
922such as C<dup>). The reason is that you register interest in some file 1023such as C<dup>). The reason is that you register interest in some file
923descriptor, but when it goes away, the operating system will silently drop 1024descriptor, but when it goes away, the operating system will silently drop
924this interest. If another file descriptor with the same number then is 1025this interest. If another file descriptor with the same number then is
925registered with libev, there is no efficient way to see that this is, in 1026registered with libev, there is no efficient way to see that this is, in
… …
934 1035
935This is how one would do it normally anyway, the important point is that 1036This is how one would do it normally anyway, the important point is that
936the libev application should not optimise around libev but should leave 1037the libev application should not optimise around libev but should leave
937optimisations to libev. 1038optimisations to libev.
938 1039
1040=head3 The special problem of dup'ed file descriptors
1041
1042Some backends (e.g. epoll), cannot register events for file descriptors,
1043but only events for the underlying file descriptions. That means when you
1044have C<dup ()>'ed file descriptors or weirder constellations, and register
1045events for them, only one file descriptor might actually receive events.
1046
1047There is no workaround possible except not registering events
1048for potentially C<dup ()>'ed file descriptors, or to resort to
1049C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1050
1051=head3 The special problem of fork
1052
1053Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
1054useless behaviour. Libev fully supports fork, but needs to be told about
1055it in the child.
1056
1057To support fork in your programs, you either have to call
1058C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,
1059enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1060C<EVBACKEND_POLL>.
1061
939 1062
940=head3 Watcher-Specific Functions 1063=head3 Watcher-Specific Functions
941 1064
942=over 4 1065=over 4
943 1066
… …
956=item int events [read-only] 1079=item int events [read-only]
957 1080
958The events being watched. 1081The events being watched.
959 1082
960=back 1083=back
1084
1085=head3 Examples
961 1086
962Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well 1087Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
963readable, but only once. Since it is likely line-buffered, you could 1088readable, but only once. Since it is likely line-buffered, you could
964attempt to read a whole line in the callback. 1089attempt to read a whole line in the callback.
965 1090
… …
1063or C<ev_timer_again> is called and determines the next timeout (if any), 1188or C<ev_timer_again> is called and determines the next timeout (if any),
1064which is also when any modifications are taken into account. 1189which is also when any modifications are taken into account.
1065 1190
1066=back 1191=back
1067 1192
1193=head3 Examples
1194
1068Example: Create a timer that fires after 60 seconds. 1195Example: Create a timer that fires after 60 seconds.
1069 1196
1070 static void 1197 static void
1071 one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents) 1198 one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1072 { 1199 {
… …
1229When active, contains the absolute time that the watcher is supposed to 1356When active, contains the absolute time that the watcher is supposed to
1230trigger next. 1357trigger next.
1231 1358
1232=back 1359=back
1233 1360
1361=head3 Examples
1362
1234Example: Call a callback every hour, or, more precisely, whenever the 1363Example: Call a callback every hour, or, more precisely, whenever the
1235system clock is divisible by 3600. The callback invocation times have 1364system clock is divisible by 3600. The callback invocation times have
1236potentially a lot of jittering, but good long-term stability. 1365potentially a lot of jittering, but good long-term stability.
1237 1366
1238 static void 1367 static void
… …
1329 1458
1330The process exit/trace status caused by C<rpid> (see your systems 1459The process exit/trace status caused by C<rpid> (see your systems
1331C<waitpid> and C<sys/wait.h> documentation for details). 1460C<waitpid> and C<sys/wait.h> documentation for details).
1332 1461
1333=back 1462=back
1463
1464=head3 Examples
1334 1465
1335Example: Try to exit cleanly on SIGINT and SIGTERM. 1466Example: Try to exit cleanly on SIGINT and SIGTERM.
1336 1467
1337 static void 1468 static void
1338 sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents) 1469 sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
… …
1379semantics of C<ev_stat> watchers, which means that libev sometimes needs 1510semantics of C<ev_stat> watchers, which means that libev sometimes needs
1380to fall back to regular polling again even with inotify, but changes are 1511to fall back to regular polling again even with inotify, but changes are
1381usually detected immediately, and if the file exists there will be no 1512usually detected immediately, and if the file exists there will be no
1382polling. 1513polling.
1383 1514
1515=head3 Inotify
1516
1517When C<inotify (7)> support has been compiled into libev (generally only
1518available on Linux) and present at runtime, it will be used to speed up
1519change detection where possible. The inotify descriptor will be created lazily
1520when the first C<ev_stat> watcher is being started.
1521
1522Inotify presense does not change the semantics of C<ev_stat> watchers
1523except that changes might be detected earlier, and in some cases, to avoid
1524making regular C<stat> calls. Even in the presense of inotify support
1525there are many cases where libev has to resort to regular C<stat> polling.
1526
1527(There is no support for kqueue, as apparently it cannot be used to
1528implement this functionality, due to the requirement of having a file
1529descriptor open on the object at all times).
1530
1531=head3 The special problem of stat time resolution
1532
1533The C<stat ()> syscall only supports full-second resolution portably, and
1534even on systems where the resolution is higher, many filesystems still
1535only support whole seconds.
1536
1537That means that, if the time is the only thing that changes, you might
1538miss updates: on the first update, C<ev_stat> detects a change and calls
1539your callback, which does something. When there is another update within
1540the same second, C<ev_stat> will be unable to detect it.
1541
1542The solution to this is to delay acting on a change for a second (or till
1543the next second boundary), using a roughly one-second delay C<ev_timer>
1544(C<ev_timer_set (w, 0., 1.01); ev_timer_again (loop, w)>). The C<.01>
1545is added to work around small timing inconsistencies of some operating
1546systems.
1547
1384=head3 Watcher-Specific Functions and Data Members 1548=head3 Watcher-Specific Functions and Data Members
1385 1549
1386=over 4 1550=over 4
1387 1551
1388=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval) 1552=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
… …
1425=item const char *path [read-only] 1589=item const char *path [read-only]
1426 1590
1427The filesystem path that is being watched. 1591The filesystem path that is being watched.
1428 1592
1429=back 1593=back
1594
1595=head3 Examples
1430 1596
1431Example: Watch C</etc/passwd> for attribute changes. 1597Example: Watch C</etc/passwd> for attribute changes.
1432 1598
1433 static void 1599 static void
1434 passwd_cb (struct ev_loop *loop, ev_stat *w, int revents) 1600 passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
… …
1447 } 1613 }
1448 1614
1449 ... 1615 ...
1450 ev_stat passwd; 1616 ev_stat passwd;
1451 1617
1452 ev_stat_init (&passwd, passwd_cb, "/etc/passwd"); 1618 ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
1453 ev_stat_start (loop, &passwd); 1619 ev_stat_start (loop, &passwd);
1620
1621Example: Like above, but additionally use a one-second delay so we do not
1622miss updates (however, frequent updates will delay processing, too, so
1623one might do the work both on C<ev_stat> callback invocation I<and> on
1624C<ev_timer> callback invocation).
1625
1626 static ev_stat passwd;
1627 static ev_timer timer;
1628
1629 static void
1630 timer_cb (EV_P_ ev_timer *w, int revents)
1631 {
1632 ev_timer_stop (EV_A_ w);
1633
1634 /* now it's one second after the most recent passwd change */
1635 }
1636
1637 static void
1638 stat_cb (EV_P_ ev_stat *w, int revents)
1639 {
1640 /* reset the one-second timer */
1641 ev_timer_again (EV_A_ &timer);
1642 }
1643
1644 ...
1645 ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
1646 ev_stat_start (loop, &passwd);
1647 ev_timer_init (&timer, timer_cb, 0., 1.01);
1454 1648
1455 1649
1456=head2 C<ev_idle> - when you've got nothing better to do... 1650=head2 C<ev_idle> - when you've got nothing better to do...
1457 1651
1458Idle watchers trigger events when no other events of the same or higher 1652Idle watchers trigger events when no other events of the same or higher
… …
1483Initialises and configures the idle watcher - it has no parameters of any 1677Initialises and configures the idle watcher - it has no parameters of any
1484kind. There is a C<ev_idle_set> macro, but using it is utterly pointless, 1678kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1485believe me. 1679believe me.
1486 1680
1487=back 1681=back
1682
1683=head3 Examples
1488 1684
1489Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the 1685Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1490callback, free it. Also, use no error checking, as usual. 1686callback, free it. Also, use no error checking, as usual.
1491 1687
1492 static void 1688 static void
… …
1544 1740
1545It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>) 1741It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1546priority, to ensure that they are being run before any other watchers 1742priority, to ensure that they are being run before any other watchers
1547after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers, 1743after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,
1548too) should not activate ("feed") events into libev. While libev fully 1744too) should not activate ("feed") events into libev. While libev fully
1549supports this, they will be called before other C<ev_check> watchers did 1745supports this, they will be called before other C<ev_check> watchers
1550their job. As C<ev_check> watchers are often used to embed other event 1746did their job. As C<ev_check> watchers are often used to embed other
1551loops those other event loops might be in an unusable state until their 1747(non-libev) event loops those other event loops might be in an unusable
1552C<ev_check> watcher ran (always remind yourself to coexist peacefully with 1748state until their C<ev_check> watcher ran (always remind yourself to
1553others). 1749coexist peacefully with others).
1554 1750
1555=head3 Watcher-Specific Functions and Data Members 1751=head3 Watcher-Specific Functions and Data Members
1556 1752
1557=over 4 1753=over 4
1558 1754
… …
1563Initialises and configures the prepare or check watcher - they have no 1759Initialises and configures the prepare or check watcher - they have no
1564parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set> 1760parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1565macros, but using them is utterly, utterly and completely pointless. 1761macros, but using them is utterly, utterly and completely pointless.
1566 1762
1567=back 1763=back
1764
1765=head3 Examples
1568 1766
1569There are a number of principal ways to embed other event loops or modules 1767There are a number of principal ways to embed other event loops or modules
1570into libev. Here are some ideas on how to include libadns into libev 1768into libev. Here are some ideas on how to include libadns into libev
1571(there is a Perl module named C<EV::ADNS> that does this, which you could 1769(there is a Perl module named C<EV::ADNS> that does this, which you could
1572use for an actually working example. Another Perl module named C<EV::Glib> 1770use for an actually working example. Another Perl module named C<EV::Glib>
… …
1741portable one. 1939portable one.
1742 1940
1743So when you want to use this feature you will always have to be prepared 1941So when you want to use this feature you will always have to be prepared
1744that you cannot get an embeddable loop. The recommended way to get around 1942that you cannot get an embeddable loop. The recommended way to get around
1745this is to have a separate variables for your embeddable loop, try to 1943this is to have a separate variables for your embeddable loop, try to
1746create it, and if that fails, use the normal loop for everything: 1944create it, and if that fails, use the normal loop for everything.
1945
1946=head3 Watcher-Specific Functions and Data Members
1947
1948=over 4
1949
1950=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
1951
1952=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
1953
1954Configures the watcher to embed the given loop, which must be
1955embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1956invoked automatically, otherwise it is the responsibility of the callback
1957to invoke it (it will continue to be called until the sweep has been done,
1958if you do not want thta, you need to temporarily stop the embed watcher).
1959
1960=item ev_embed_sweep (loop, ev_embed *)
1961
1962Make a single, non-blocking sweep over the embedded loop. This works
1963similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1964apropriate way for embedded loops.
1965
1966=item struct ev_loop *other [read-only]
1967
1968The embedded event loop.
1969
1970=back
1971
1972=head3 Examples
1973
1974Example: Try to get an embeddable event loop and embed it into the default
1975event loop. If that is not possible, use the default loop. The default
1976loop is stored in C<loop_hi>, while the mebeddable loop is stored in
1977C<loop_lo> (which is C<loop_hi> in the acse no embeddable loop can be
1978used).
1747 1979
1748 struct ev_loop *loop_hi = ev_default_init (0); 1980 struct ev_loop *loop_hi = ev_default_init (0);
1749 struct ev_loop *loop_lo = 0; 1981 struct ev_loop *loop_lo = 0;
1750 struct ev_embed embed; 1982 struct ev_embed embed;
1751 1983
… …
1762 ev_embed_start (loop_hi, &embed); 1994 ev_embed_start (loop_hi, &embed);
1763 } 1995 }
1764 else 1996 else
1765 loop_lo = loop_hi; 1997 loop_lo = loop_hi;
1766 1998
1767=head3 Watcher-Specific Functions and Data Members 1999Example: Check if kqueue is available but not recommended and create
2000a kqueue backend for use with sockets (which usually work with any
2001kqueue implementation). Store the kqueue/socket-only event loop in
2002C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
1768 2003
1769=over 4 2004 struct ev_loop *loop = ev_default_init (0);
2005 struct ev_loop *loop_socket = 0;
2006 struct ev_embed embed;
2007
2008 if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2009 if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2010 {
2011 ev_embed_init (&embed, 0, loop_socket);
2012 ev_embed_start (loop, &embed);
2013 }
1770 2014
1771=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop) 2015 if (!loop_socket)
2016 loop_socket = loop;
1772 2017
1773=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop) 2018 // now use loop_socket for all sockets, and loop for everything else
1774
1775Configures the watcher to embed the given loop, which must be
1776embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1777invoked automatically, otherwise it is the responsibility of the callback
1778to invoke it (it will continue to be called until the sweep has been done,
1779if you do not want thta, you need to temporarily stop the embed watcher).
1780
1781=item ev_embed_sweep (loop, ev_embed *)
1782
1783Make a single, non-blocking sweep over the embedded loop. This works
1784similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most
1785apropriate way for embedded loops.
1786
1787=item struct ev_loop *loop [read-only]
1788
1789The embedded event loop.
1790
1791=back
1792 2019
1793 2020
1794=head2 C<ev_fork> - the audacity to resume the event loop after a fork 2021=head2 C<ev_fork> - the audacity to resume the event loop after a fork
1795 2022
1796Fork watchers are called when a C<fork ()> was detected (usually because 2023Fork watchers are called when a C<fork ()> was detected (usually because
… …
2131Libev can (and often is) directly embedded into host 2358Libev can (and often is) directly embedded into host
2132applications. Examples of applications that embed it include the Deliantra 2359applications. Examples of applications that embed it include the Deliantra
2133Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe) 2360Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
2134and rxvt-unicode. 2361and rxvt-unicode.
2135 2362
2136The goal is to enable you to just copy the neecssary files into your 2363The goal is to enable you to just copy the necessary files into your
2137source directory without having to change even a single line in them, so 2364source directory without having to change even a single line in them, so
2138you can easily upgrade by simply copying (or having a checked-out copy of 2365you can easily upgrade by simply copying (or having a checked-out copy of
2139libev somewhere in your source tree). 2366libev somewhere in your source tree).
2140 2367
2141=head2 FILESETS 2368=head2 FILESETS
… …
2231 2458
2232If defined to be C<1>, libev will try to detect the availability of the 2459If defined to be C<1>, libev will try to detect the availability of the
2233monotonic clock option at both compiletime and runtime. Otherwise no use 2460monotonic clock option at both compiletime and runtime. Otherwise no use
2234of the monotonic clock option will be attempted. If you enable this, you 2461of the monotonic clock option will be attempted. If you enable this, you
2235usually have to link against librt or something similar. Enabling it when 2462usually have to link against librt or something similar. Enabling it when
2236the functionality isn't available is safe, though, althoguh you have 2463the functionality isn't available is safe, though, although you have
2237to make sure you link against any libraries where the C<clock_gettime> 2464to make sure you link against any libraries where the C<clock_gettime>
2238function is hiding in (often F<-lrt>). 2465function is hiding in (often F<-lrt>).
2239 2466
2240=item EV_USE_REALTIME 2467=item EV_USE_REALTIME
2241 2468
2242If defined to be C<1>, libev will try to detect the availability of the 2469If defined to be C<1>, libev will try to detect the availability of the
2243realtime clock option at compiletime (and assume its availability at 2470realtime clock option at compiletime (and assume its availability at
2244runtime if successful). Otherwise no use of the realtime clock option will 2471runtime if successful). Otherwise no use of the realtime clock option will
2245be attempted. This effectively replaces C<gettimeofday> by C<clock_get 2472be attempted. This effectively replaces C<gettimeofday> by C<clock_get
2246(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries 2473(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the
2247in the description of C<EV_USE_MONOTONIC>, though. 2474note about libraries in the description of C<EV_USE_MONOTONIC>, though.
2475
2476=item EV_USE_NANOSLEEP
2477
2478If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2479and will use it for delays. Otherwise it will use C<select ()>.
2248 2480
2249=item EV_USE_SELECT 2481=item EV_USE_SELECT
2250 2482
2251If undefined or defined to be C<1>, libev will compile in support for the 2483If undefined or defined to be C<1>, libev will compile in support for the
2252C<select>(2) backend. No attempt at autodetection will be done: if no 2484C<select>(2) backend. No attempt at autodetection will be done: if no
… …
2270wants osf handles on win32 (this is the case when the select to 2502wants osf handles on win32 (this is the case when the select to
2271be used is the winsock select). This means that it will call 2503be used is the winsock select). This means that it will call
2272C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise, 2504C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2273it is assumed that all these functions actually work on fds, even 2505it is assumed that all these functions actually work on fds, even
2274on win32. Should not be defined on non-win32 platforms. 2506on win32. Should not be defined on non-win32 platforms.
2507
2508=item EV_FD_TO_WIN32_HANDLE
2509
2510If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2511file descriptors to socket handles. When not defining this symbol (the
2512default), then libev will call C<_get_osfhandle>, which is usually
2513correct. In some cases, programs use their own file descriptor management,
2514in which case they can provide this function to map fds to socket handles.
2275 2515
2276=item EV_USE_POLL 2516=item EV_USE_POLL
2277 2517
2278If defined to be C<1>, libev will compile in support for the C<poll>(2) 2518If defined to be C<1>, libev will compile in support for the C<poll>(2)
2279backend. Otherwise it will be enabled on non-win32 platforms. It 2519backend. Otherwise it will be enabled on non-win32 platforms. It
… …
2316be detected at runtime. 2556be detected at runtime.
2317 2557
2318=item EV_H 2558=item EV_H
2319 2559
2320The name of the F<ev.h> header file used to include it. The default if 2560The name of the F<ev.h> header file used to include it. The default if
2321undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This 2561undefined is C<"ev.h"> in F<event.h> and F<ev.c>. This can be used to
2322can be used to virtually rename the F<ev.h> header file in case of conflicts. 2562virtually rename the F<ev.h> header file in case of conflicts.
2323 2563
2324=item EV_CONFIG_H 2564=item EV_CONFIG_H
2325 2565
2326If C<EV_STANDALONE> isn't C<1>, this variable can be used to override 2566If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2327F<ev.c>'s idea of where to find the F<config.h> file, similarly to 2567F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2328C<EV_H>, above. 2568C<EV_H>, above.
2329 2569
2330=item EV_EVENT_H 2570=item EV_EVENT_H
2331 2571
2332Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea 2572Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2333of how the F<event.h> header can be found. 2573of how the F<event.h> header can be found, the dfeault is C<"event.h">.
2334 2574
2335=item EV_PROTOTYPES 2575=item EV_PROTOTYPES
2336 2576
2337If defined to be C<0>, then F<ev.h> will not define any function 2577If defined to be C<0>, then F<ev.h> will not define any function
2338prototypes, but still define all the structs and other symbols. This is 2578prototypes, but still define all the structs and other symbols. This is
… …
2404than enough. If you need to manage thousands of children you might want to 2644than enough. If you need to manage thousands of children you might want to
2405increase this value (I<must> be a power of two). 2645increase this value (I<must> be a power of two).
2406 2646
2407=item EV_INOTIFY_HASHSIZE 2647=item EV_INOTIFY_HASHSIZE
2408 2648
2409C<ev_staz> watchers use a small hash table to distribute workload by 2649C<ev_stat> watchers use a small hash table to distribute workload by
2410inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>), 2650inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),
2411usually more than enough. If you need to manage thousands of C<ev_stat> 2651usually more than enough. If you need to manage thousands of C<ev_stat>
2412watchers you might want to increase this value (I<must> be a power of 2652watchers you might want to increase this value (I<must> be a power of
2413two). 2653two).
2414 2654
… …
2431 2671
2432=item ev_set_cb (ev, cb) 2672=item ev_set_cb (ev, cb)
2433 2673
2434Can be used to change the callback member declaration in each watcher, 2674Can be used to change the callback member declaration in each watcher,
2435and the way callbacks are invoked and set. Must expand to a struct member 2675and the way callbacks are invoked and set. Must expand to a struct member
2436definition and a statement, respectively. See the F<ev.v> header file for 2676definition and a statement, respectively. See the F<ev.h> header file for
2437their default definitions. One possible use for overriding these is to 2677their default definitions. One possible use for overriding these is to
2438avoid the C<struct ev_loop *> as first argument in all cases, or to use 2678avoid the C<struct ev_loop *> as first argument in all cases, or to use
2439method calls instead of plain function calls in C++. 2679method calls instead of plain function calls in C++.
2680
2681=head2 EXPORTED API SYMBOLS
2682
2683If you need to re-export the API (e.g. via a dll) and you need a list of
2684exported symbols, you can use the provided F<Symbol.*> files which list
2685all public symbols, one per line:
2686
2687 Symbols.ev for libev proper
2688 Symbols.event for the libevent emulation
2689
2690This can also be used to rename all public symbols to avoid clashes with
2691multiple versions of libev linked together (which is obviously bad in
2692itself, but sometimes it is inconvinient to avoid this).
2693
2694A sed command like this will create wrapper C<#define>'s that you need to
2695include before including F<ev.h>:
2696
2697 <Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
2698
2699This would create a file F<wrap.h> which essentially looks like this:
2700
2701 #define ev_backend myprefix_ev_backend
2702 #define ev_check_start myprefix_ev_check_start
2703 #define ev_check_stop myprefix_ev_check_stop
2704 ...
2440 2705
2441=head2 EXAMPLES 2706=head2 EXAMPLES
2442 2707
2443For a real-world example of a program the includes libev 2708For a real-world example of a program the includes libev
2444verbatim, you can have a look at the EV perl module 2709verbatim, you can have a look at the EV perl module
… …
2485 2750
2486=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers) 2751=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
2487 2752
2488This means that, when you have a watcher that triggers in one hour and 2753This means that, when you have a watcher that triggers in one hour and
2489there are 100 watchers that would trigger before that then inserting will 2754there are 100 watchers that would trigger before that then inserting will
2490have to skip those 100 watchers. 2755have to skip roughly seven (C<ld 100>) of these watchers.
2491 2756
2492=item Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers) 2757=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
2493 2758
2494That means that for changing a timer costs less than removing/adding them 2759That means that changing a timer costs less than removing/adding them
2495as only the relative motion in the event queue has to be paid for. 2760as only the relative motion in the event queue has to be paid for.
2496 2761
2497=item Starting io/check/prepare/idle/signal/child watchers: O(1) 2762=item Starting io/check/prepare/idle/signal/child watchers: O(1)
2498 2763
2499These just add the watcher into an array or at the head of a list. 2764These just add the watcher into an array or at the head of a list.
2765
2500=item Stopping check/prepare/idle watchers: O(1) 2766=item Stopping check/prepare/idle watchers: O(1)
2501 2767
2502=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE)) 2768=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
2503 2769
2504These watchers are stored in lists then need to be walked to find the 2770These watchers are stored in lists then need to be walked to find the
2505correct watcher to remove. The lists are usually short (you don't usually 2771correct watcher to remove. The lists are usually short (you don't usually
2506have many watchers waiting for the same fd or signal). 2772have many watchers waiting for the same fd or signal).
2507 2773
2508=item Finding the next timer per loop iteration: O(1) 2774=item Finding the next timer in each loop iteration: O(1)
2775
2776By virtue of using a binary heap, the next timer is always found at the
2777beginning of the storage array.
2509 2778
2510=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd) 2779=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
2511 2780
2512A change means an I/O watcher gets started or stopped, which requires 2781A change means an I/O watcher gets started or stopped, which requires
2513libev to recalculate its status (and possibly tell the kernel). 2782libev to recalculate its status (and possibly tell the kernel, depending
2783on backend and wether C<ev_io_set> was used).
2514 2784
2515=item Activating one watcher: O(1) 2785=item Activating one watcher (putting it into the pending state): O(1)
2516 2786
2517=item Priority handling: O(number_of_priorities) 2787=item Priority handling: O(number_of_priorities)
2518 2788
2519Priorities are implemented by allocating some space for each 2789Priorities are implemented by allocating some space for each
2520priority. When doing priority-based operations, libev usually has to 2790priority. When doing priority-based operations, libev usually has to
2521linearly search all the priorities. 2791linearly search all the priorities, but starting/stopping and activating
2792watchers becomes O(1) w.r.t. prioritiy handling.
2522 2793
2523=back 2794=back
2524 2795
2525 2796
2797=head1 Win32 platform limitations and workarounds
2798
2799Win32 doesn't support any of the standards (e.g. POSIX) that libev
2800requires, and its I/O model is fundamentally incompatible with the POSIX
2801model. Libev still offers limited functionality on this platform in
2802the form of the C<EVBACKEND_SELECT> backend, and only supports socket
2803descriptors. This only applies when using Win32 natively, not when using
2804e.g. cygwin.
2805
2806There is no supported compilation method available on windows except
2807embedding it into other applications.
2808
2809Due to the many, low, and arbitrary limits on the win32 platform and the
2810abysmal performance of winsockets, using a large number of sockets is not
2811recommended (and not reasonable). If your program needs to use more than
2812a hundred or so sockets, then likely it needs to use a totally different
2813implementation for windows, as libev offers the POSIX model, which cannot
2814be implemented efficiently on windows (microsoft monopoly games).
2815
2816=over 4
2817
2818=item The winsocket select function
2819
2820The winsocket C<select> function doesn't follow POSIX in that it requires
2821socket I<handles> and not socket I<file descriptors>. This makes select
2822very inefficient, and also requires a mapping from file descriptors
2823to socket handles. See the discussion of the C<EV_SELECT_USE_FD_SET>,
2824C<EV_SELECT_IS_WINSOCKET> and C<EV_FD_TO_WIN32_HANDLE> preprocessor
2825symbols for more info.
2826
2827The configuration for a "naked" win32 using the microsoft runtime
2828libraries and raw winsocket select is:
2829
2830 #define EV_USE_SELECT 1
2831 #define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
2832
2833Note that winsockets handling of fd sets is O(n), so you can easily get a
2834complexity in the O(n²) range when using win32.
2835
2836=item Limited number of file descriptors
2837
2838Windows has numerous arbitrary (and low) limits on things. Early versions
2839of winsocket's select only supported waiting for a max. of C<64> handles
2840(probably owning to the fact that all windows kernels can only wait for
2841C<64> things at the same time internally; microsoft recommends spawning a
2842chain of threads and wait for 63 handles and the previous thread in each).
2843
2844Newer versions support more handles, but you need to define C<FD_SETSIZE>
2845to some high number (e.g. C<2048>) before compiling the winsocket select
2846call (which might be in libev or elsewhere, for example, perl does its own
2847select emulation on windows).
2848
2849Another limit is the number of file descriptors in the microsoft runtime
2850libraries, which by default is C<64> (there must be a hidden I<64> fetish
2851or something like this inside microsoft). You can increase this by calling
2852C<_setmaxstdio>, which can increase this limit to C<2048> (another
2853arbitrary limit), but is broken in many versions of the microsoft runtime
2854libraries.
2855
2856This might get you to about C<512> or C<2048> sockets (depending on
2857windows version and/or the phase of the moon). To get more, you need to
2858wrap all I/O functions and provide your own fd management, but the cost of
2859calling select (O(n²)) will likely make this unworkable.
2860
2861=back
2862
2863
2526=head1 AUTHOR 2864=head1 AUTHOR
2527 2865
2528Marc Lehmann <libev@schmorp.de>. 2866Marc Lehmann <libev@schmorp.de>.
2529 2867

Diff Legend

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