… | |
… | |
131 | .\} |
131 | .\} |
132 | .rm #[ #] #H #V #F C |
132 | .rm #[ #] #H #V #F C |
133 | .\" ======================================================================== |
133 | .\" ======================================================================== |
134 | .\" |
134 | .\" |
135 | .IX Title "LIBEV 3" |
135 | .IX Title "LIBEV 3" |
136 | .TH LIBEV 3 "2019-06-23" "libev-4.25" "libev - high performance full featured event loop" |
136 | .TH LIBEV 3 "2019-07-07" "libev-4.27" "libev - high performance full featured event loop" |
137 | .\" For nroff, turn off justification. Always turn off hyphenation; it makes |
137 | .\" For nroff, turn off justification. Always turn off hyphenation; it makes |
138 | .\" way too many mistakes in technical documents. |
138 | .\" way too many mistakes in technical documents. |
139 | .if n .ad l |
139 | .if n .ad l |
140 | .nh |
140 | .nh |
141 | .SH "NAME" |
141 | .SH "NAME" |
… | |
… | |
291 | When libev detects a usage error such as a negative timer interval, then |
291 | When libev detects a usage error such as a negative timer interval, then |
292 | it will print a diagnostic message and abort (via the \f(CW\*(C`assert\*(C'\fR mechanism, |
292 | it will print a diagnostic message and abort (via the \f(CW\*(C`assert\*(C'\fR mechanism, |
293 | so \f(CW\*(C`NDEBUG\*(C'\fR will disable this checking): these are programming errors in |
293 | so \f(CW\*(C`NDEBUG\*(C'\fR will disable this checking): these are programming errors in |
294 | the libev caller and need to be fixed there. |
294 | the libev caller and need to be fixed there. |
295 | .PP |
295 | .PP |
|
|
296 | Via the \f(CW\*(C`EV_FREQUENT\*(C'\fR macro you can compile in and/or enable extensive |
|
|
297 | consistency checking code inside libev that can be used to check for |
|
|
298 | internal inconsistencies, suually caused by application bugs. |
|
|
299 | .PP |
296 | Libev also has a few internal error-checking \f(CW\*(C`assert\*(C'\fRions, and also has |
300 | Libev also has a few internal error-checking \f(CW\*(C`assert\*(C'\fRions. These do not |
297 | extensive consistency checking code. These do not trigger under normal |
|
|
298 | circumstances, as they indicate either a bug in libev or worse. |
301 | trigger under normal circumstances, as they indicate either a bug in libev |
|
|
302 | or worse. |
299 | .SH "GLOBAL FUNCTIONS" |
303 | .SH "GLOBAL FUNCTIONS" |
300 | .IX Header "GLOBAL FUNCTIONS" |
304 | .IX Header "GLOBAL FUNCTIONS" |
301 | These functions can be called anytime, even before initialising the |
305 | These functions can be called anytime, even before initialising the |
302 | library in any way. |
306 | library in any way. |
303 | .IP "ev_tstamp ev_time ()" 4 |
307 | .IP "ev_tstamp ev_time ()" 4 |
… | |
… | |
638 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR to \f(CW\*(C`POLLIN | POLLERR | POLLHUP\*(C'\fR, and |
642 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR to \f(CW\*(C`POLLIN | POLLERR | POLLHUP\*(C'\fR, and |
639 | \&\f(CW\*(C`EV_WRITE\*(C'\fR to \f(CW\*(C`POLLOUT | POLLERR | POLLHUP\*(C'\fR. |
643 | \&\f(CW\*(C`EV_WRITE\*(C'\fR to \f(CW\*(C`POLLOUT | POLLERR | POLLHUP\*(C'\fR. |
640 | .ie n .IP """EVBACKEND_EPOLL"" (value 4, Linux)" 4 |
644 | .ie n .IP """EVBACKEND_EPOLL"" (value 4, Linux)" 4 |
641 | .el .IP "\f(CWEVBACKEND_EPOLL\fR (value 4, Linux)" 4 |
645 | .el .IP "\f(CWEVBACKEND_EPOLL\fR (value 4, Linux)" 4 |
642 | .IX Item "EVBACKEND_EPOLL (value 4, Linux)" |
646 | .IX Item "EVBACKEND_EPOLL (value 4, Linux)" |
643 | Use the linux-specific \fBepoll\fR\|(7) interface (for both pre\- and post\-2.6.9 |
647 | Use the Linux-specific \fBepoll\fR\|(7) interface (for both pre\- and post\-2.6.9 |
644 | kernels). |
648 | kernels). |
645 | .Sp |
649 | .Sp |
646 | For few fds, this backend is a bit little slower than poll and select, but |
650 | For few fds, this backend is a bit little slower than poll and select, but |
647 | it scales phenomenally better. While poll and select usually scale like |
651 | it scales phenomenally better. While poll and select usually scale like |
648 | O(total_fds) where total_fds is the total number of fds (or the highest |
652 | O(total_fds) where total_fds is the total number of fds (or the highest |
… | |
… | |
701 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as |
705 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as |
702 | \&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR. |
706 | \&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR. |
703 | .ie n .IP """EVBACKEND_LINUXAIO"" (value 64, Linux)" 4 |
707 | .ie n .IP """EVBACKEND_LINUXAIO"" (value 64, Linux)" 4 |
704 | .el .IP "\f(CWEVBACKEND_LINUXAIO\fR (value 64, Linux)" 4 |
708 | .el .IP "\f(CWEVBACKEND_LINUXAIO\fR (value 64, Linux)" 4 |
705 | .IX Item "EVBACKEND_LINUXAIO (value 64, Linux)" |
709 | .IX Item "EVBACKEND_LINUXAIO (value 64, Linux)" |
706 | Use the linux-specific linux aio (\fInot\fR \f(CWaio(7)\fR but \f(CWio_submit(2)\fR) event interface available in post\-4.18 kernels. |
710 | Use the Linux-specific Linux \s-1AIO\s0 (\fInot\fR \f(CWaio(7)\fR but \f(CWio_submit(2)\fR) event interface available in post\-4.18 kernels (but libev |
|
|
711 | only tries to use it in 4.19+). |
|
|
712 | .Sp |
|
|
713 | This is another Linux train wreck of an event interface. |
707 | .Sp |
714 | .Sp |
708 | If this backend works for you (as of this writing, it was very |
715 | If this backend works for you (as of this writing, it was very |
709 | experimental), it is the best event interface available on linux and might |
716 | experimental), it is the best event interface available on Linux and might |
710 | be well worth enabling it \- if it isn't available in your kernel this will |
717 | be well worth enabling it \- if it isn't available in your kernel this will |
711 | be detected and this backend will be skipped. |
718 | be detected and this backend will be skipped. |
712 | .Sp |
719 | .Sp |
713 | This backend can batch oneshot requests and supports a user-space ring |
720 | This backend can batch oneshot requests and supports a user-space ring |
714 | buffer to receive events. It also doesn't suffer from most of the design |
721 | buffer to receive events. It also doesn't suffer from most of the design |
715 | problems of epoll (such as not being able to remove event sources from |
722 | problems of epoll (such as not being able to remove event sources from |
716 | the epoll set), and generally sounds too good to be true. Because, this |
723 | the epoll set), and generally sounds too good to be true. Because, this |
717 | being the linux kernel, of course it suffers from a whole new set of |
724 | being the Linux kernel, of course it suffers from a whole new set of |
718 | limitations. |
725 | limitations, forcing you to fall back to epoll, inheriting all its design |
|
|
726 | issues. |
719 | .Sp |
727 | .Sp |
720 | For one, it is not easily embeddable (but probably could be done using |
728 | For one, it is not easily embeddable (but probably could be done using |
721 | an event fd at some extra overhead). It also is subject to a system wide |
729 | an event fd at some extra overhead). It also is subject to a system wide |
722 | limit that can be configured in \fI/proc/sys/fs/aio\-max\-nr\fR \- each loop |
730 | limit that can be configured in \fI/proc/sys/fs/aio\-max\-nr\fR. If no \s-1AIO\s0 |
723 | currently requires \f(CW61\fR of this number. If no aio requests are left, this |
|
|
724 | backend will be skipped during initialisation. |
731 | requests are left, this backend will be skipped during initialisation, and |
|
|
732 | will switch to epoll when the loop is active. |
725 | .Sp |
733 | .Sp |
726 | Most problematic in practise, however, is that not all file descriptors |
734 | Most problematic in practice, however, is that not all file descriptors |
727 | work with it. For example, in linux 5.1, tcp sockets, pipes, event fds, |
735 | work with it. For example, in Linux 5.1, \s-1TCP\s0 sockets, pipes, event fds, |
728 | files, \fI/dev/null\fR and a few others are supported, but ttys do not work |
736 | files, \fI/dev/null\fR and many others are supported, but ttys do not work |
729 | (probably because of a bug), so this is not (yet?) a generic event polling |
737 | properly (a known bug that the kernel developers don't care about, see |
730 | interface. |
738 | <https://lore.kernel.org/patchwork/patch/1047453/>), so this is not |
|
|
739 | (yet?) a generic event polling interface. |
731 | .Sp |
740 | .Sp |
|
|
741 | Overall, it seems the Linux developers just don't want it to have a |
|
|
742 | generic event handling mechanism other than \f(CW\*(C`select\*(C'\fR or \f(CW\*(C`poll\*(C'\fR. |
|
|
743 | .Sp |
732 | To work around this latter problem, the current version of libev uses |
744 | To work around all these problem, the current version of libev uses its |
733 | epoll as a fallback for file deescriptor types that do not work. Epoll |
745 | epoll backend as a fallback for file descriptor types that do not work. Or |
734 | is used in, kind of, slow mode that hopefully avoids most of its design |
746 | falls back completely to epoll if the kernel acts up. |
735 | problems and requires 1\-3 extra syscalls per active fd every iteration. |
|
|
736 | .Sp |
747 | .Sp |
737 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as |
748 | This backend maps \f(CW\*(C`EV_READ\*(C'\fR and \f(CW\*(C`EV_WRITE\*(C'\fR in the same way as |
738 | \&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR. |
749 | \&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR. |
739 | .ie n .IP """EVBACKEND_KQUEUE"" (value 8, most \s-1BSD\s0 clones)" 4 |
750 | .ie n .IP """EVBACKEND_KQUEUE"" (value 8, most \s-1BSD\s0 clones)" 4 |
740 | .el .IP "\f(CWEVBACKEND_KQUEUE\fR (value 8, most \s-1BSD\s0 clones)" 4 |
751 | .el .IP "\f(CWEVBACKEND_KQUEUE\fR (value 8, most \s-1BSD\s0 clones)" 4 |
741 | .IX Item "EVBACKEND_KQUEUE (value 8, most BSD clones)" |
752 | .IX Item "EVBACKEND_KQUEUE (value 8, most BSD clones)" |
742 | Kqueue deserves special mention, as at the time of this writing, it |
753 | Kqueue deserves special mention, as at the time this backend was |
743 | was broken on all BSDs except NetBSD (usually it doesn't work reliably |
754 | implemented, it was broken on all BSDs except NetBSD (usually it doesn't |
744 | with anything but sockets and pipes, except on Darwin, where of course |
755 | work reliably with anything but sockets and pipes, except on Darwin, |
745 | it's completely useless). Unlike epoll, however, whose brokenness |
756 | where of course it's completely useless). Unlike epoll, however, whose |
746 | is by design, these kqueue bugs can (and eventually will) be fixed |
757 | brokenness is by design, these kqueue bugs can be (and mostly have been) |
747 | without \s-1API\s0 changes to existing programs. For this reason it's not being |
758 | fixed without \s-1API\s0 changes to existing programs. For this reason it's not |
748 | \&\*(L"auto-detected\*(R" unless you explicitly specify it in the flags (i.e. using |
759 | being \*(L"auto-detected\*(R" on all platforms unless you explicitly specify it |
749 | \&\f(CW\*(C`EVBACKEND_KQUEUE\*(C'\fR) or libev was compiled on a known-to-be-good (\-enough) |
760 | in the flags (i.e. using \f(CW\*(C`EVBACKEND_KQUEUE\*(C'\fR) or libev was compiled on a |
750 | system like NetBSD. |
761 | known-to-be-good (\-enough) system like NetBSD. |
751 | .Sp |
762 | .Sp |
752 | You still can embed kqueue into a normal poll or select backend and use it |
763 | You still can embed kqueue into a normal poll or select backend and use it |
753 | only for sockets (after having made sure that sockets work with kqueue on |
764 | only for sockets (after having made sure that sockets work with kqueue on |
754 | the target platform). See \f(CW\*(C`ev_embed\*(C'\fR watchers for more info. |
765 | the target platform). See \f(CW\*(C`ev_embed\*(C'\fR watchers for more info. |
755 | .Sp |
766 | .Sp |
756 | It scales in the same way as the epoll backend, but the interface to the |
767 | It scales in the same way as the epoll backend, but the interface to the |
757 | kernel is more efficient (which says nothing about its actual speed, of |
768 | kernel is more efficient (which says nothing about its actual speed, of |
758 | course). While stopping, setting and starting an I/O watcher does never |
769 | course). While stopping, setting and starting an I/O watcher does never |
759 | cause an extra system call as with \f(CW\*(C`EVBACKEND_EPOLL\*(C'\fR, it still adds up to |
770 | cause an extra system call as with \f(CW\*(C`EVBACKEND_EPOLL\*(C'\fR, it still adds up to |
760 | two event changes per incident. Support for \f(CW\*(C`fork ()\*(C'\fR is very bad (you |
771 | two event changes per incident. Support for \f(CW\*(C`fork ()\*(C'\fR is very bad (you |
761 | might have to leak fd's on fork, but it's more sane than epoll) and it |
772 | might have to leak fds on fork, but it's more sane than epoll) and it |
762 | drops fds silently in similarly hard-to-detect cases. |
773 | drops fds silently in similarly hard-to-detect cases. |
763 | .Sp |
774 | .Sp |
764 | This backend usually performs well under most conditions. |
775 | This backend usually performs well under most conditions. |
765 | .Sp |
776 | .Sp |
766 | While nominally embeddable in other event loops, this doesn't work |
777 | While nominally embeddable in other event loops, this doesn't work |
… | |
… | |
1860 | reuse the same code path. |
1871 | reuse the same code path. |
1861 | .PP |
1872 | .PP |
1862 | \fIThe special problem of fork\fR |
1873 | \fIThe special problem of fork\fR |
1863 | .IX Subsection "The special problem of fork" |
1874 | .IX Subsection "The special problem of fork" |
1864 | .PP |
1875 | .PP |
1865 | Some backends (epoll, kqueue, probably linuxaio) do not support \f(CW\*(C`fork ()\*(C'\fR |
1876 | Some backends (epoll, kqueue, linuxaio, iouring) do not support \f(CW\*(C`fork ()\*(C'\fR |
1866 | at all or exhibit useless behaviour. Libev fully supports fork, but needs |
1877 | at all or exhibit useless behaviour. Libev fully supports fork, but needs |
1867 | to be told about it in the child if you want to continue to use it in the |
1878 | to be told about it in the child if you want to continue to use it in the |
1868 | child. |
1879 | child. |
1869 | .PP |
1880 | .PP |
1870 | To support fork in your child processes, you have to call \f(CW\*(C`ev_loop_fork |
1881 | To support fork in your child processes, you have to call \f(CW\*(C`ev_loop_fork |
… | |
… | |
4613 | \& |
4624 | \& |
4614 | \& ev_select.c only when select backend is enabled |
4625 | \& ev_select.c only when select backend is enabled |
4615 | \& ev_poll.c only when poll backend is enabled |
4626 | \& ev_poll.c only when poll backend is enabled |
4616 | \& ev_epoll.c only when the epoll backend is enabled |
4627 | \& ev_epoll.c only when the epoll backend is enabled |
4617 | \& ev_linuxaio.c only when the linux aio backend is enabled |
4628 | \& ev_linuxaio.c only when the linux aio backend is enabled |
|
|
4629 | \& ev_iouring.c only when the linux io_uring backend is enabled |
4618 | \& ev_kqueue.c only when the kqueue backend is enabled |
4630 | \& ev_kqueue.c only when the kqueue backend is enabled |
4619 | \& ev_port.c only when the solaris port backend is enabled |
4631 | \& ev_port.c only when the solaris port backend is enabled |
4620 | .Ve |
4632 | .Ve |
4621 | .PP |
4633 | .PP |
4622 | \&\fIev.c\fR includes the backend files directly when enabled, so you only need |
4634 | \&\fIev.c\fR includes the backend files directly when enabled, so you only need |
… | |
… | |
4805 | otherwise another method will be used as fallback. This is the preferred |
4817 | otherwise another method will be used as fallback. This is the preferred |
4806 | backend for GNU/Linux systems. If undefined, it will be enabled if the |
4818 | backend for GNU/Linux systems. If undefined, it will be enabled if the |
4807 | headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled. |
4819 | headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled. |
4808 | .IP "\s-1EV_USE_LINUXAIO\s0" 4 |
4820 | .IP "\s-1EV_USE_LINUXAIO\s0" 4 |
4809 | .IX Item "EV_USE_LINUXAIO" |
4821 | .IX Item "EV_USE_LINUXAIO" |
|
|
4822 | If defined to be \f(CW1\fR, libev will compile in support for the Linux aio |
|
|
4823 | backend (\f(CW\*(C`EV_USE_EPOLL\*(C'\fR must also be enabled). If undefined, it will be |
|
|
4824 | enabled on linux, otherwise disabled. |
|
|
4825 | .IP "\s-1EV_USE_IOURING\s0" 4 |
|
|
4826 | .IX Item "EV_USE_IOURING" |
4810 | If defined to be \f(CW1\fR, libev will compile in support for the Linux |
4827 | If defined to be \f(CW1\fR, libev will compile in support for the Linux |
4811 | aio backend. Due to it's currenbt limitations it has to be requested |
4828 | io_uring backend (\f(CW\*(C`EV_USE_EPOLL\*(C'\fR must also be enabled). Due to it's |
4812 | explicitly. If undefined, it will be enabled on linux, otherwise |
4829 | current limitations it has to be requested explicitly. If undefined, it |
4813 | disabled. |
4830 | will be enabled on linux, otherwise disabled. |
4814 | .IP "\s-1EV_USE_KQUEUE\s0" 4 |
4831 | .IP "\s-1EV_USE_KQUEUE\s0" 4 |
4815 | .IX Item "EV_USE_KQUEUE" |
4832 | .IX Item "EV_USE_KQUEUE" |
4816 | If defined to be \f(CW1\fR, libev will compile in support for the \s-1BSD\s0 style |
4833 | If defined to be \f(CW1\fR, libev will compile in support for the \s-1BSD\s0 style |
4817 | \&\f(CW\*(C`kqueue\*(C'\fR(2) backend. Its actual availability will be detected at runtime, |
4834 | \&\f(CW\*(C`kqueue\*(C'\fR(2) backend. Its actual availability will be detected at runtime, |
4818 | otherwise another method will be used as fallback. This is the preferred |
4835 | otherwise another method will be used as fallback. This is the preferred |
… | |
… | |
5074 | in. If set to \f(CW1\fR, then verification code will be compiled in, but not |
5091 | in. If set to \f(CW1\fR, then verification code will be compiled in, but not |
5075 | called. If set to \f(CW2\fR, then the internal verification code will be |
5092 | called. If set to \f(CW2\fR, then the internal verification code will be |
5076 | called once per loop, which can slow down libev. If set to \f(CW3\fR, then the |
5093 | called once per loop, which can slow down libev. If set to \f(CW3\fR, then the |
5077 | verification code will be called very frequently, which will slow down |
5094 | verification code will be called very frequently, which will slow down |
5078 | libev considerably. |
5095 | libev considerably. |
|
|
5096 | .Sp |
|
|
5097 | Verification errors are reported via C's \f(CW\*(C`assert\*(C'\fR mechanism, so if you |
|
|
5098 | disable that (e.g. by defining \f(CW\*(C`NDEBUG\*(C'\fR) then no errors will be reported. |
5079 | .Sp |
5099 | .Sp |
5080 | The default is \f(CW1\fR, unless \f(CW\*(C`EV_FEATURES\*(C'\fR overrides it, in which case it |
5100 | The default is \f(CW1\fR, unless \f(CW\*(C`EV_FEATURES\*(C'\fR overrides it, in which case it |
5081 | will be \f(CW0\fR. |
5101 | will be \f(CW0\fR. |
5082 | .IP "\s-1EV_COMMON\s0" 4 |
5102 | .IP "\s-1EV_COMMON\s0" 4 |
5083 | .IX Item "EV_COMMON" |
5103 | .IX Item "EV_COMMON" |