1.\" $NetBSD: ctxsw.9,v 1.2 1996/12/02 00:11:31 tls Exp $ 2.\" 3.\" Copyright (c) 1996 The NetBSD Foundation, Inc. 4.\" All rights reserved. 5.\" 6.\" This code is derived from software contributed to The NetBSD Foundation 7.\" by Paul Kranenburg. 8.\" 9.\" Redistribution and use in source and binary forms, with or without 10.\" modification, are permitted provided that the following conditions 11.\" are met: 12.\" 1. Redistributions of source code must retain the above copyright 13.\" notice, this list of conditions and the following disclaimer. 14.\" 2. Redistributions in binary form must reproduce the above copyright 15.\" notice, this list of conditions and the following disclaimer in the 16.\" documentation and/or other materials provided with the distribution. 17.\" 3. All advertising materials mentioning features or use of this software 18.\" must display the following acknowledgement: 19.\" This product includes software developed by the NetBSD 20.\" Foundation, Inc. and its contributors. 21.\" 4. Neither the name of The NetBSD Foundation nor the names of its 22.\" contributors may be used to endorse or promote products derived 23.\" from this software without specific prior written permission. 24.\" 25.\" THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 26.\" ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 27.\" TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 28.\" PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE 29.\" LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 30.\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 31.\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 32.\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 33.\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 34.\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 35.\" POSSIBILITY OF SUCH DAMAGE. 36.\" 37.\" $FreeBSD$ 38.\" 39.Dd Nov 24, 1996 40.Dt MI_SWITCH 9 41.Os FreeBSD 42.Sh NAME 43.Nm mi_switch , 44.Nm cpu_switch 45.Nd switch to another process context 46.Sh SYNOPSIS 47.Fd #include <sys/param.h> 48.Fd #include <sys/proc.h> 49.Ft void 50.Fn mi_switch "void" 51.Ft void 52.Fn cpu_switch "struct proc *p" 53.Sh DESCRIPTION 54The 55.Fn mi_switch 56function implements the machine independent prelude to a process context 57switch. 58It is called from only a few distinguished places in the kernel 59code as a result of the principle of non-preemtable kernel mode execution. 60The three major uses of 61.Nm 62can be enumerated as follows: 63.Bl -enum -offset indent 64.It 65from within 66.Xr sleep 9 , and 67.Xr tsleep 9 68when the current process 69voluntarily relinquishes the CPU to wait for some resource to become 70available. 71.It 72after handling a trap 73.Pq e.g. a system call, device interrupt 74when the kernel prepares a return to user-mode execution. 75This case is 76typically handled by machine dependent trap-handling code after detection 77of a change in the signal disposition of the current process, or when a 78higher priority process might be available to run. 79The latter event is 80communicated by the machine independent scheduling routines by calling 81the machine defined 82.Fn need_resched . 83.It 84in the signal handling code 85.Pq see Xr issignal 9 86if a signal is delivered that causes a process to stop. 87.El 88.Pp 89.Fn mi_switch 90records the amount of time the current process has been running in the 91process structure and checks this value against the CPU time limits 92allocated to the process 93.Pq see Xr getrlimit 2 . 94Exceeding the soft limit results in a 95.Dv SIGXCPU 96signal to be posted to the process, while exceeding the hard limit will 97cause a 98.Dv SIGKILL . 99After these administrative tasks are done, 100.Fn mi_switch 101hands over control to the machine dependent routine 102.Fn cpu_switch , 103which will perform the actual process context switch. 104.Pp 105.Fn cpu_switch 106will make a choice amongst the processes which are ready to run from a 107priority queue data-structure. 108The priority queue consists of an array 109.Va qs[NQS] 110of queue header structures each of which identifies a list of runnable 111processes of equal priority 112.Pq see Fa <sys/proc.h> . 113A single word 114.Va whichqs 115containing a bit mask identifying non-empty queues assists in selecting 116a process quickly. 117.Fn cpu_switch 118must remove the first process from the list on the queue 119with the highest priority 120.Po lower indices in Va qs 121indicate higher priority 122.Pc , 123and assign the address of its process structure to the global variable 124.Dv curproc . 125If no processes are available on the run queues, 126.Fn cpu_switch 127shall go into an 128.Dq idle 129loop. 130The idle loop must allow interrupts to be taken that will eventually 131cause processes to appear again on the run queues. 132The variable 133.Va curproc 134should be 135.Dv NULL 136while 137.Fn cpu_switch 138waits for this to happen. 139.Pp 140Note that 141.Fn mi_switch 142and thus 143.Fn cpu_switch 144should be called at splhigh(). 145.Pp 146.Sh SEE ALSO 147.Xr issignal 9 , 148.Xr spl 9 , 149.Xr tsleep 9 , 150.Xr wakeup 9 151.Pp 152