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.Dd Nov 24, 1996 38.Dt MI_SWITCH 9 39.Os FreeBSD 40.Sh NAME 41.Nm mi_switch , 42.Nm cpu_switch 43.Nd switch to another process context 44.Sh SYNOPSIS 45.Ft void 46.Fn mi_switch "void" 47.Ft void 48.Fn cpu_switch "void" 49.Sh DESCRIPTION 50The 51.Fn mi_switch 52function implements the machine independent prelude to a process context 53switch. It is called from only a few distinguished places in the kernel 54code as a result of the principle of non-preemtable kernel mode execution. 55The three major uses of 56.Nm 57can be enumerated as follows: 58.Bl -enum -offset indent 59.It 60from within 61.Xr sleep 9 , and 62.Xr tsleep 9 63when the current process 64voluntarily relinquishes the CPU to wait for some resource to become 65available. 66.It 67after handling a trap 68.Pq e.g. a system call, device interrupt 69when the kernel prepares a return to user-mode execution. This case is 70typically handled by machine dependent trap-handling code after detection 71of a change in the signal disposition of the current process, or when a 72higher priority process might be available to run. The latter event is 73communicated by the machine independent scheduling routines by calling 74the machine defined 75.Fn need_resched "void" . 76.It 77in the signal handling code 78.Pq see Xr issignal 9 79if a signal is delivered that causes a process to stop. 80.El 81.Pp 82.Fn mi_switch 83records the amount of time the current process has been running in the 84process structure and checks this value against the CPU time limits 85allocated to the process 86.Pq see Xr getrlimit 2 . 87Exceeding the soft limit results in a 88.Dv SIGXCPU 89signal to be posted to the process, while exceeding the hard limit will 90cause a 91.Dv SIGKILL . 92After these administrative tasks are done, 93.Fn mi_switch 94hands over control to the machine dependent routine 95.Fn cpu_switch "void" , 96which will perform the actual process context switch. 97.Pp 98.Fn cpu_switch 99will make a choice amongst the processes which are ready to run from a 100priority queue data-structure. The priority queue consists of an array 101.Va qs[NQS] 102of queue header structures each of which identifies a list of runnable 103processes of equal priority 104.Pq see Fa <sys/proc.h> . 105A single word 106.Va whichqs 107containing a bit mask identifying non-empty queues assists in selecting 108a process quickly. 109.Fn cpu_switch 110must remove the first process from the list on the queue 111with the highest priority 112.Po lower indices in Va sq 113indicate higher priority 114.Pc , 115and assign the address of its process structure to the global variable 116.Dv curproc . 117If no processes are available on the run queues, 118.Fn cpu_switch 119shall go into an 120.Dq idle 121loop. The idle loop must allow interrupts to be taken that will eventually 122cause processes to appear again on the run queues. The variable 123.Va curproc 124should be 125.Dv NULL 126while 127.Fn cpu_switch 128waits for this to happen. 129.Pp 130Note that 131.Fn mi_switch 132and thus 133.Fn cpu_switch 134should be called at splhigh(). 135.Pp 136.Sh SEE ALSO 137.Xr issignal 9 , 138.Xr spl 9 , 139.Xr tsleep 9 , 140.Xr wakeup 9 141.Pp 142