xref: /freebsd/share/man/man9/mi_switch.9 (revision daf1cffce2e07931f27c6c6998652e90df6ba87e)
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. It is called from only a few distinguished places in the kernel
58code as a result of the principle of non-preemtable kernel mode execution.
59The three major uses of
60.Nm
61can be enumerated as follows:
62.Bl -enum -offset indent
63.It
64from within
65.Xr sleep 9 , and
66.Xr tsleep 9
67when the current process
68voluntarily relinquishes the CPU to wait for some resource to become
69available.
70.It
71after handling a trap
72.Pq e.g. a system call, device interrupt
73when the kernel prepares a return to user-mode execution. This case is
74typically handled by machine dependent trap-handling code after detection
75of a change in the signal disposition of the current process, or when a
76higher priority process might be available to run. The latter event is
77communicated by the machine independent scheduling routines by calling
78the machine defined
79.Fn need_resched .
80.It
81in the signal handling code
82.Pq see Xr issignal 9
83if a signal is delivered that causes a process to stop.
84.El
85.Pp
86.Fn mi_switch
87records the amount of time the current process has been running in the
88process structure and checks this value against the CPU time limits
89allocated to the process
90.Pq see Xr getrlimit 2 .
91Exceeding the soft limit results in a
92.Dv SIGXCPU
93signal to be posted to the process, while exceeding the hard limit will
94cause a
95.Dv SIGKILL .
96After these administrative tasks are done,
97.Fn mi_switch
98hands over control to the machine dependent routine
99.Fn cpu_switch ,
100which will perform the actual process context switch.
101.Pp
102.Fn cpu_switch
103will make a choice amongst the processes which are ready to run from a
104priority queue data-structure. The priority queue consists of an array
105.Va qs[NQS]
106of queue header structures each of which identifies a list of runnable
107processes of equal priority
108.Pq see Fa <sys/proc.h> .
109A single word
110.Va whichqs
111containing a bit mask identifying non-empty queues assists in selecting
112a process quickly.
113.Fn cpu_switch
114must remove the first process from the list on the queue
115with the highest priority
116.Po lower indices in Va qs
117indicate higher priority
118.Pc ,
119and assign the address of its process structure to the global variable
120.Dv curproc .
121If no processes are available on the run queues,
122.Fn cpu_switch
123shall go into an
124.Dq idle
125loop. The idle loop must allow interrupts to be taken that will eventually
126cause processes to appear again on the run queues. The variable
127.Va curproc
128should be
129.Dv NULL
130while
131.Fn cpu_switch
132waits for this to happen.
133.Pp
134Note that
135.Fn mi_switch
136and thus
137.Fn cpu_switch
138should be called at splhigh().
139.Pp
140.Sh SEE ALSO
141.Xr issignal 9 ,
142.Xr spl 9 ,
143.Xr tsleep 9 ,
144.Xr wakeup 9
145.Pp
146