xref: /freebsd/share/man/man9/mi_switch.9 (revision 9336e0699bda8a301cd2bfa37106b6ec5e32012e)
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 November 24, 1996
40.Dt MI_SWITCH 9
41.Os
42.Sh NAME
43.Nm mi_switch ,
44.Nm cpu_switch ,
45.Nm cpu_throw
46.Nd switch to another thread context
47.Sh SYNOPSIS
48.In sys/param.h
49.In sys/proc.h
50.Ft void
51.Fn mi_switch "void"
52.Ft void
53.Fn cpu_switch "void"
54.Ft void
55.Fn cpu_throw "void"
56.Sh DESCRIPTION
57The
58.Fn mi_switch
59function implements the machine independent prelude to a thread context
60switch.
61It is called from only a few distinguished places in the kernel
62code as a result of the principle of non-preemptable kernel mode execution.
63The various major uses of
64.Nm
65can be enumerated as follows:
66.Bl -enum -offset indent
67.It
68From within a function such as
69.Xr cv_wait 9 ,
70.Xr mtx_lock ,
71or
72.Xr tsleep 9
73when the current thread
74voluntarily relinquishes the CPU to wait for some resource or lock to become
75available.
76.It
77After handling a trap
78(e.g.\& a system call, device interrupt)
79when the kernel prepares a return to user-mode execution.
80This case is
81typically handled by machine dependent trap-handling code after detection
82of a change in the signal disposition of the current process, or when a
83higher priority thread might be available to run.
84The latter event is
85communicated by the machine independent scheduling routines by calling
86the machine defined
87.Fn need_resched .
88.It
89In the signal handling code
90(see
91.Xr issignal 9 )
92if a signal is delivered that causes a process to stop.
93.It
94When a thread dies in
95.Xr thread_exit 9
96and control of the processor can be passed to the next runnable thread.
97.It
98In
99.Xr thread_suspend_check 9
100where a thread needs to stop execution due to the suspension state of
101the process as a whole.
102.El
103.Pp
104.Fn mi_switch
105records the amount of time the current thread has been running in the
106process structures and checks this value against the CPU time limits
107allocated to the process
108(see
109.Xr getrlimit 2 ) .
110Exceeding the soft limit results in a
111.Dv SIGXCPU
112signal to be posted to the process, while exceeding the hard limit will
113cause a
114.Dv SIGKILL .
115.Pp
116If the thread is still in the
117.Dv TDS_RUNNING
118state,
119.Fn mi_switch
120will put it back onto the run queue, assuming that
121it will want to run again soon.
122If it is in one of the other
123states and KSE threading is enabled, the associated
124.Em KSE
125will be made available to any higher priority threads from the same
126group, to allow them to be scheduled next.
127.Pp
128After these administrative tasks are done,
129.Fn mi_switch
130hands over control to the machine dependent routine
131.Fn cpu_switch ,
132which will perform the actual thread context switch.
133.Pp
134.Fn cpu_switch
135first saves the context of the current thread.
136Next, it calls
137.Fn choosethread
138to determine which thread to run next.
139Finally, it reads in the saved context of the new thread and starts to
140execute the new thread.
141.Pp
142.Fn cpu_throw
143is similar to
144.Fn cpu_switch
145except that it does not save the context of the old thread.
146This function is useful when the kernel does not have an old thread
147context to save, such as when CPUs other than the boot CPU perform their
148first task switch, or when the kernel does not care about the state of the
149old thread, such as in
150.Fn thread_exit
151when the kernel terminates the current thread and switches into a new
152thread.
153.Pp
154To protect the
155.Xr runqueue 9 ,
156all of these functions must be called with the
157.Va sched_lock
158mutex held.
159.Sh SEE ALSO
160.Xr cv_wait 9 ,
161.Xr issignal 9 ,
162.Xr mutex 9 ,
163.Xr runqueue 9 ,
164.Xr tsleep 9 ,
165.Xr wakeup 9
166