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

Comparing libev/ev.pod (file contents):
Revision 1.448 by root, Sun Jun 23 02:02:24 2019 UTC vs.
Revision 1.458 by root, Fri Dec 20 20:51:46 2019 UTC

… …
159When libev detects a usage error such as a negative timer interval, then 159When libev detects a usage error such as a negative timer interval, then
160it will print a diagnostic message and abort (via the C<assert> mechanism, 160it will print a diagnostic message and abort (via the C<assert> mechanism,
161so C<NDEBUG> will disable this checking): these are programming errors in 161so C<NDEBUG> will disable this checking): these are programming errors in
162the libev caller and need to be fixed there. 162the libev caller and need to be fixed there.
163 163
164Via the C<EV_FREQUENT> macro you can compile in and/or enable extensive
165consistency checking code inside libev that can be used to check for
166internal inconsistencies, suually caused by application bugs.
167
164Libev also has a few internal error-checking C<assert>ions, and also has 168Libev also has a few internal error-checking C<assert>ions. These do not
165extensive consistency checking code. These do not trigger under normal
166circumstances, as they indicate either a bug in libev or worse. 169trigger under normal circumstances, as they indicate either a bug in libev
170or worse.
167 171
168 172
169=head1 GLOBAL FUNCTIONS 173=head1 GLOBAL FUNCTIONS
170 174
171These functions can be called anytime, even before initialising the 175These functions can be called anytime, even before initialising the
… …
476unblocking the signals. 480unblocking the signals.
477 481
478It's also required by POSIX in a threaded program, as libev calls 482It's also required by POSIX in a threaded program, as libev calls
479C<sigprocmask>, whose behaviour is officially unspecified. 483C<sigprocmask>, whose behaviour is officially unspecified.
480 484
481This flag's behaviour will become the default in future versions of libev. 485=item C<EVFLAG_NOTIMERFD>
486
487When this flag is specified, the libev will avoid using a C<timerfd> to
488detect time jumps. It will still be able to detect time jumps, but takes
489longer and has a lower accuracy in doing so, but saves a file descriptor
490per loop.
491
492The current implementation only tries to use a C<timerfd> when the first
493C<ev_periodic> watcher is started and falls back on other methods if it
494cannot be created, but this behaviour might change in the future.
482 495
483=item C<EVBACKEND_SELECT> (value 1, portable select backend) 496=item C<EVBACKEND_SELECT> (value 1, portable select backend)
484 497
485This is your standard select(2) backend. Not I<completely> standard, as 498This is your standard select(2) backend. Not I<completely> standard, as
486libev tries to roll its own fd_set with no limits on the number of fds, 499libev tries to roll its own fd_set with no limits on the number of fds,
… …
511This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and 524This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
512C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>. 525C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
513 526
514=item C<EVBACKEND_EPOLL> (value 4, Linux) 527=item C<EVBACKEND_EPOLL> (value 4, Linux)
515 528
516Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9 529Use the Linux-specific epoll(7) interface (for both pre- and post-2.6.9
517kernels). 530kernels).
518 531
519For few fds, this backend is a bit little slower than poll and select, but 532For few fds, this backend is a bit little slower than poll and select, but
520it scales phenomenally better. While poll and select usually scale like 533it scales phenomenally better. While poll and select usually scale like
521O(total_fds) where total_fds is the total number of fds (or the highest 534O(total_fds) where total_fds is the total number of fds (or the highest
… …
574This backend maps C<EV_READ> and C<EV_WRITE> in the same way as 587This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
575C<EVBACKEND_POLL>. 588C<EVBACKEND_POLL>.
576 589
577=item C<EVBACKEND_LINUXAIO> (value 64, Linux) 590=item C<EVBACKEND_LINUXAIO> (value 64, Linux)
578 591
579Use the linux-specific linux aio (I<not> C<< aio(7) >> but C<< 592Use the Linux-specific Linux AIO (I<not> C<< aio(7) >> but C<<
580io_submit(2) >>) event interface available in post-4.18 kernels. 593io_submit(2) >>) event interface available in post-4.18 kernels (but libev
594only tries to use it in 4.19+).
595
596This is another Linux train wreck of an event interface.
581 597
582If this backend works for you (as of this writing, it was very 598If this backend works for you (as of this writing, it was very
583experimental), it is the best event interface available on linux and might 599experimental), it is the best event interface available on Linux and might
584be well worth enabling it - if it isn't available in your kernel this will 600be well worth enabling it - if it isn't available in your kernel this will
585be detected and this backend will be skipped. 601be detected and this backend will be skipped.
586 602
587This backend can batch oneshot requests and supports a user-space ring 603This backend can batch oneshot requests and supports a user-space ring
588buffer to receive events. It also doesn't suffer from most of the design 604buffer to receive events. It also doesn't suffer from most of the design
589problems of epoll (such as not being able to remove event sources from 605problems of epoll (such as not being able to remove event sources from
590the epoll set), and generally sounds too good to be true. Because, this 606the epoll set), and generally sounds too good to be true. Because, this
591being the linux kernel, of course it suffers from a whole new set of 607being the Linux kernel, of course it suffers from a whole new set of
592limitations. 608limitations, forcing you to fall back to epoll, inheriting all its design
609issues.
593 610
594For one, it is not easily embeddable (but probably could be done using 611For one, it is not easily embeddable (but probably could be done using
595an event fd at some extra overhead). It also is subject to a system wide 612an event fd at some extra overhead). It also is subject to a system wide
596limit that can be configured in F</proc/sys/fs/aio-max-nr> - each loop 613limit that can be configured in F</proc/sys/fs/aio-max-nr>. If no AIO
597currently requires C<61> of this number. If no aio requests are left, this
598backend will be skipped during initialisation. 614requests are left, this backend will be skipped during initialisation, and
615will switch to epoll when the loop is active.
599 616
600Most problematic in practise, however, is that not all file descriptors 617Most problematic in practice, however, is that not all file descriptors
601work with it. For example, in linux 5.1, tcp sockets, pipes, event fds, 618work with it. For example, in Linux 5.1, TCP sockets, pipes, event fds,
602files, F</dev/null> and a few others are supported, but ttys do not work 619files, F</dev/null> and many others are supported, but ttys do not work
603(probably because of a bug), so this is not (yet?) a generic event polling 620properly (a known bug that the kernel developers don't care about, see
604interface. 621L<https://lore.kernel.org/patchwork/patch/1047453/>), so this is not
622(yet?) a generic event polling interface.
605 623
624Overall, it seems the Linux developers just don't want it to have a
625generic event handling mechanism other than C<select> or C<poll>.
626
606To work around this latter problem, the current version of libev uses 627To work around all these problem, the current version of libev uses its
607epoll as a fallback for file deescriptor types that do not work. Epoll 628epoll backend as a fallback for file descriptor types that do not work. Or
608is used in, kind of, slow mode that hopefully avoids most of its design 629falls back completely to epoll if the kernel acts up.
609problems.
610 630
611This backend maps C<EV_READ> and C<EV_WRITE> in the same way as 631This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
612C<EVBACKEND_POLL>. 632C<EVBACKEND_POLL>.
613 633
614=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones) 634=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
615 635
616Kqueue deserves special mention, as at the time of this writing, it 636Kqueue deserves special mention, as at the time this backend was
617was broken on all BSDs except NetBSD (usually it doesn't work reliably 637implemented, it was broken on all BSDs except NetBSD (usually it doesn't
618with anything but sockets and pipes, except on Darwin, where of course 638work reliably with anything but sockets and pipes, except on Darwin,
619it's completely useless). Unlike epoll, however, whose brokenness 639where of course it's completely useless). Unlike epoll, however, whose
620is by design, these kqueue bugs can (and eventually will) be fixed 640brokenness is by design, these kqueue bugs can be (and mostly have been)
621without API changes to existing programs. For this reason it's not being 641fixed without API changes to existing programs. For this reason it's not
622"auto-detected" unless you explicitly specify it in the flags (i.e. using 642being "auto-detected" on all platforms unless you explicitly specify it
623C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough) 643in the flags (i.e. using C<EVBACKEND_KQUEUE>) or libev was compiled on a
624system like NetBSD. 644known-to-be-good (-enough) system like NetBSD.
625 645
626You still can embed kqueue into a normal poll or select backend and use it 646You still can embed kqueue into a normal poll or select backend and use it
627only for sockets (after having made sure that sockets work with kqueue on 647only for sockets (after having made sure that sockets work with kqueue on
628the target platform). See C<ev_embed> watchers for more info. 648the target platform). See C<ev_embed> watchers for more info.
629 649
630It scales in the same way as the epoll backend, but the interface to the 650It scales in the same way as the epoll backend, but the interface to the
631kernel is more efficient (which says nothing about its actual speed, of 651kernel is more efficient (which says nothing about its actual speed, of
632course). While stopping, setting and starting an I/O watcher does never 652course). While stopping, setting and starting an I/O watcher does never
633cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to 653cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
634two event changes per incident. Support for C<fork ()> is very bad (you 654two event changes per incident. Support for C<fork ()> is very bad (you
635might have to leak fd's on fork, but it's more sane than epoll) and it 655might have to leak fds on fork, but it's more sane than epoll) and it
636drops fds silently in similarly hard-to-detect cases. 656drops fds silently in similarly hard-to-detect cases.
637 657
638This backend usually performs well under most conditions. 658This backend usually performs well under most conditions.
639 659
640While nominally embeddable in other event loops, this doesn't work 660While nominally embeddable in other event loops, this doesn't work
… …
1524 1544
1525Many event loops support I<watcher priorities>, which are usually small 1545Many event loops support I<watcher priorities>, which are usually small
1526integers that influence the ordering of event callback invocation 1546integers that influence the ordering of event callback invocation
1527between watchers in some way, all else being equal. 1547between watchers in some way, all else being equal.
1528 1548
1529In libev, Watcher priorities can be set using C<ev_set_priority>. See its 1549In libev, watcher priorities can be set using C<ev_set_priority>. See its
1530description for the more technical details such as the actual priority 1550description for the more technical details such as the actual priority
1531range. 1551range.
1532 1552
1533There are two common ways how these these priorities are being interpreted 1553There are two common ways how these these priorities are being interpreted
1534by event loops: 1554by event loops:
… …
1738when you rarely read from a file instead of from a socket, and want to 1758when you rarely read from a file instead of from a socket, and want to
1739reuse the same code path. 1759reuse the same code path.
1740 1760
1741=head3 The special problem of fork 1761=head3 The special problem of fork
1742 1762
1743Some backends (epoll, kqueue, probably linuxaio) do not support C<fork ()> 1763Some backends (epoll, kqueue, linuxaio, iouring) do not support C<fork ()>
1744at all or exhibit useless behaviour. Libev fully supports fork, but needs 1764at all or exhibit useless behaviour. Libev fully supports fork, but needs
1745to be told about it in the child if you want to continue to use it in the 1765to be told about it in the child if you want to continue to use it in the
1746child. 1766child.
1747 1767
1748To support fork in your child processes, you have to call C<ev_loop_fork 1768To support fork in your child processes, you have to call C<ev_loop_fork
… …
4473 4493
4474 ev_select.c only when select backend is enabled 4494 ev_select.c only when select backend is enabled
4475 ev_poll.c only when poll backend is enabled 4495 ev_poll.c only when poll backend is enabled
4476 ev_epoll.c only when the epoll backend is enabled 4496 ev_epoll.c only when the epoll backend is enabled
4477 ev_linuxaio.c only when the linux aio backend is enabled 4497 ev_linuxaio.c only when the linux aio backend is enabled
4498 ev_iouring.c only when the linux io_uring backend is enabled
4478 ev_kqueue.c only when the kqueue backend is enabled 4499 ev_kqueue.c only when the kqueue backend is enabled
4479 ev_port.c only when the solaris port backend is enabled 4500 ev_port.c only when the solaris port backend is enabled
4480 4501
4481F<ev.c> includes the backend files directly when enabled, so you only need 4502F<ev.c> includes the backend files directly when enabled, so you only need
4482to compile this single file. 4503to compile this single file.
… …
4603available and will probe for kernel support at runtime. This will improve 4624available and will probe for kernel support at runtime. This will improve
4604C<ev_signal> and C<ev_async> performance and reduce resource consumption. 4625C<ev_signal> and C<ev_async> performance and reduce resource consumption.
4605If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc 4626If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
46062.7 or newer, otherwise disabled. 46272.7 or newer, otherwise disabled.
4607 4628
4629=item EV_USE_SIGNALFD
4630
4631If defined to be C<1>, then libev will assume that C<signalfd ()> is
4632available and will probe for kernel support at runtime. This enables
4633the use of EVFLAG_SIGNALFD for faster and simpler signal handling. If
4634undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
46352.7 or newer, otherwise disabled.
4636
4637=item EV_USE_TIMERFD
4638
4639If defined to be C<1>, then libev will assume that C<timerfd ()> is
4640available and will probe for kernel support at runtime. This allows
4641libev to detect time jumps accurately. If undefined, it will be enabled
4642if the headers indicate GNU/Linux + Glibc 2.8 or newer and define
4643C<TFD_TIMER_CANCEL_ON_SET>, otherwise disabled.
4644
4645=item EV_USE_EVENTFD
4646
4647If defined to be C<1>, then libev will assume that C<eventfd ()> is
4648available and will probe for kernel support at runtime. This will improve
4649C<ev_signal> and C<ev_async> performance and reduce resource consumption.
4650If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
46512.7 or newer, otherwise disabled.
4652
4608=item EV_USE_SELECT 4653=item EV_USE_SELECT
4609 4654
4610If undefined or defined to be C<1>, libev will compile in support for the 4655If undefined or defined to be C<1>, libev will compile in support for the
4611C<select>(2) backend. No attempt at auto-detection will be done: if no 4656C<select>(2) backend. No attempt at auto-detection will be done: if no
4612other method takes over, select will be it. Otherwise the select backend 4657other method takes over, select will be it. Otherwise the select backend
… …
4675backend for GNU/Linux systems. If undefined, it will be enabled if the 4720backend for GNU/Linux systems. If undefined, it will be enabled if the
4676headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled. 4721headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
4677 4722
4678=item EV_USE_LINUXAIO 4723=item EV_USE_LINUXAIO
4679 4724
4725If defined to be C<1>, libev will compile in support for the Linux aio
4726backend (C<EV_USE_EPOLL> must also be enabled). If undefined, it will be
4727enabled on linux, otherwise disabled.
4728
4729=item EV_USE_IOURING
4730
4680If defined to be C<1>, libev will compile in support for the Linux 4731If defined to be C<1>, libev will compile in support for the Linux
4681aio backend. Due to it's currenbt limitations it has to be requested 4732io_uring backend (C<EV_USE_EPOLL> must also be enabled). Due to it's
4682explicitly. If undefined, it will be enabled on linux, otherwise 4733current limitations it has to be requested explicitly. If undefined, it
4683disabled. 4734will be enabled on linux, otherwise disabled.
4684 4735
4685=item EV_USE_KQUEUE 4736=item EV_USE_KQUEUE
4686 4737
4687If defined to be C<1>, libev will compile in support for the BSD style 4738If defined to be C<1>, libev will compile in support for the BSD style
4688C<kqueue>(2) backend. Its actual availability will be detected at runtime, 4739C<kqueue>(2) backend. Its actual availability will be detected at runtime,
… …
4966called. If set to C<2>, then the internal verification code will be 5017called. If set to C<2>, then the internal verification code will be
4967called once per loop, which can slow down libev. If set to C<3>, then the 5018called once per loop, which can slow down libev. If set to C<3>, then the
4968verification code will be called very frequently, which will slow down 5019verification code will be called very frequently, which will slow down
4969libev considerably. 5020libev considerably.
4970 5021
5022Verification errors are reported via C's C<assert> mechanism, so if you
5023disable that (e.g. by defining C<NDEBUG>) then no errors will be reported.
5024
4971The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it 5025The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
4972will be C<0>. 5026will be C<0>.
4973 5027
4974=item EV_COMMON 5028=item EV_COMMON
4975 5029

Diff Legend

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