1.\" Copyright (c) 2010 Alexander Motin <mav@FreeBSD.org> 2.\" All rights reserved. 3.\" 4.\" Redistribution and use in source and binary forms, with or without 5.\" modification, are permitted provided that the following conditions 6.\" are met: 7.\" 1. Redistributions of source code must retain the above copyright 8.\" notice, this list of conditions and the following disclaimer. 9.\" 2. Redistributions in binary form must reproduce the above copyright 10.\" notice, this list of conditions and the following disclaimer in the 11.\" documentation and/or other materials provided with the distribution. 12.\" 13.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 14.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 15.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 16.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 17.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 18.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 19.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 20.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 21.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 22.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 23.\" SUCH DAMAGE. 24.\" 25.\" $FreeBSD$ 26.\" 27.Dd March 13, 2012 28.Dt EVENTTIMERS 4 29.Os 30.Sh NAME 31.Nm eventtimers 32.Nd kernel event timers subsystem 33.Sh SYNOPSIS 34Kernel uses several types of time-related devices, such as: real time clocks, 35time counters and event timers. 36Real time clocks responsible for tracking real world time, mostly when system 37is down. 38Time counters are responsible for generation of monotonically increasing 39timestamps for precise uptime tracking purposes, when system is running. 40Event timers are responsible for generating interrupts at specified time or 41periodically, to run different time-based events. 42This page is about the last. 43.Sh DESCRIPTION 44Kernel uses time-based events for many different purposes: scheduling, 45statistics, time keeping, profiling and many other things, based on 46.Xr callout 9 47mechanism. 48These purposes now grouped into three main callbacks: 49.Bl -tag -width ".Fn hardclock" 50.It Fn hardclock 51.Xr callout 9 52and timekeeping events entry. 53Called with frequency defined by 54.Va hz 55variable, 56usually 1000Hz. 57.It Fn statclock 58statistics and scheduler events entry. 59Called with frequency about 128Hz. 60.It Fn profclock 61profiler events entry. 62When enabled, called with frequency about 8KHz. 63.El 64.Pp 65Different platforms provide different kinds of timer hardware. 66The goal of the event timers subsystem is to provide unified way to control 67that hardware, and to use it, supplying kernel with all required time-based 68events. 69.Pp 70Each driver implementing event timers, registers them at the subsystem. 71It is possible to see the list of present event timers, like this, via 72.Va kern.eventtimer 73sysctl: 74.Bd -literal 75kern.eventtimer.choice: HPET(550) LAPIC(400) i8254(100) RTC(0) 76kern.eventtimer.et.LAPIC.flags: 15 77kern.eventtimer.et.LAPIC.frequency: 0 78kern.eventtimer.et.LAPIC.quality: 400 79kern.eventtimer.et.i8254.flags: 1 80kern.eventtimer.et.i8254.frequency: 1193182 81kern.eventtimer.et.i8254.quality: 100 82kern.eventtimer.et.RTC.flags: 17 83kern.eventtimer.et.RTC.frequency: 32768 84kern.eventtimer.et.RTC.quality: 0 85kern.eventtimer.et.HPET.flags: 7 86kern.eventtimer.et.HPET.frequency: 14318180 87kern.eventtimer.et.HPET.quality: 550 88.Ed 89.Pp 90where: 91.Bl -inset 92.It Va kern.eventtimer.et. Ns Ar X Ns Va .flags 93is a 94bitmask, defining event timer capabilities: 95.Bl -tag -offset indent -width indent -compact 96.It 1 97periodic mode supported, 98.It 2 99one-shot mode supported, 100.It 4 101timer is per-CPU, 102.It 8 103timer may stop when CPU goes to sleep state, 104.It 16 105timer supports only power-of-2 divisors. 106.El 107.It Va kern.eventtimer.et. Ns Ar X Ns Va .frequency 108is a 109timer base frequency, 110.It Va kern.eventtimer.et. Ns Ar X Ns Va .quality 111is an 112integral value, defining how good is this timer, comparing to others. 113.El 114.Pp 115Timers management code of the kernel chooses one timer from that list. 116Current choice can be read and affected via 117.Va kern.eventtimer.timer 118tunable/sysctl. 119Several other tunables/sysctls are affecting how exactly this timer is used: 120.Bl -inset 121.It Va kern.eventtimer.periodic 122allows to choose periodic and one-shot operation mode. 123In periodic mode, periodic interrupts from timer hardware are taken as the 124only source of time for time events. 125One-shot mode instead uses currently selected time counter to precisely 126schedule all needed events and programs event timer to generate interrupt 127exactly in specified time. 128Default value depends of chosen timer capabilities, but one-shot mode is 129preferred, until other is forced by user or hardware. 130.It Va kern.eventtimer.singlemul 131in periodic mode specifies how much times higher timer frequency should be, 132to not strictly alias 133.Fn hardclock 134and 135.Fn statclock 136events. 137Default values are 1381, 2 or 4, depending on configured HZ value. 139.It Va kern.eventtimer.idletick 140makes each CPU to receive every timer interrupt independently of whether they 141busy or not. 142By default this options is disabled. 143If chosen timer is per-CPU 144and runs in periodic mode, this option has no effect - all interrupts are 145always generating. 146.El 147.Sh SEE ALSO 148.Xr apic 4 , 149.Xr atrtc 4 , 150.Xr attimer 4 , 151.Xr hpet 4 , 152.Xr timecounters 4 , 153.Xr eventtimers 9 154