OpenBSD-4.6/share/man/man9/spl.9

Compare this file to the similar file:
Show the results in this format:
.\"	$OpenBSD: spl.9,v 1.23 2007/09/14 16:16:08 mk Exp $
.\"	$NetBSD: spl.9,v 1.1 1997/03/11 06:15:05 mikel Exp $
.\"
.\" Copyright (c) 1997 Michael Long.
.\" Copyright (c) 1997 Jonathan Stone.
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\"    notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\" 3. The name of the author may not be used to endorse or promote products
.\"    derived from this software without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
.\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
.\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
.\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
.\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
.\"
.Dd $Mdocdate: September 14 2007 $
.Dt SPL 9
.Os
.Sh NAME
.Nm spl
.Nd modify system interrupt priority level
.Sh SYNOPSIS
.Fd #include <machine/intr.h>
.Ft int
.Fn splhigh void
.Ft int
.Fn splserial void
.Ft int
.Fn splsched void
.Ft int
.Fn splclock void
.Ft int
.Fn splstatclock void
.Ft int
.Fn splvm void
.Ft int
.Fn spltty void
.Ft int
.Fn splsofttty void
.Ft int
.Fn splnet void
.Ft int
.Fn splbio void
.Ft int
.Fn splsoftnet void
.Ft int
.Fn splsoftclock void
.Ft int
.Fn spllowersoftclock void
.Ft int
.Fn spl0 void
.Ft void
.Fn splx "int s"
.Ft void
.Fn splassert "int s"
.Sh DESCRIPTION
These functions raise and lower the system priority level.
They are used by kernel code to block interrupts with priority less
than or equal to the named level (i.e.,
.Fn spltty
blocks interrupts of priority less than or equal to
.Dv IPL_TTY ) .
The code may then safely access variables and data structures which
are used by kernel code that runs at an equal or lower priority level.
.Pp
An
.Nm
function exists for each distinct priority level which can exist in
the system.
These macros and the corresponding priority levels are
used for various defined purposes, and may be divided into two main
types: hard and soft.
Hard interrupts are generated by hardware
devices, while soft interrupts are generated by callouts and called
from the kernel's periodic timer interrupt service routine.
.Pp
In order of highest to lowest priority, the priority-raising macros
are:
.Bl -tag -width splsoftclockXX
.It Fn splhigh
blocks all hard and soft interrupts.
It is used for code that cannot
tolerate any interrupts, like hardware context switching code and the
.Xr ddb 4
in-kernel debugger.
.It Fn splserial
blocks hard interrupts from serial interfaces.
Code running at this level may not access the tty subsystem.
.It Fn splsched
blocks interrupts that may access scheduler data structures.
Specifically the clock interrupt that invokes the
.Fn schedclock
function needs to be blocked.
On some systems this is a separate clock;
on others it is the same as the statistics clock and, on these,
.Fn splsched
must block everything that
.Fn splstatclock
does.
Code running at or above this level may not call
.Xr tsleep 9
or
.Xr wakeup 9 ,
nor may it post signals.
Note that "running" means invoked by an interrupt handler that
operates at this level or higher.
Kernel code that operates in the context of a process and has called
.Fn splhigh
for blocking purposes can use
.Xr tsleep 9
or
.Xr wakeup 9 .
.It Fn splclock
blocks the hardware clock interrupt.
It is used by
.Fn hardclock
to update kernel and process times, and must be used by any other code
that accesses time-related data.
.It Fn splstatclock
blocks the hardware statistics clock interrupt.
It is used by
.Fn statclock
to update kernel profiling and other statistics, and must be used by
any code that accesses that data.
This level is identical to
.Fn splclock
if there is no separate statistics clock.
.It Fn splvm
blocks hard interrupts from all devices that are allowed to use the
kernel
.Xr malloc 9 .
That includes all disk, network, and tty device interrupts.
.It Fn spltty
blocks hard interrupts from TTY devices.
.It Fn splsofttty
blocks soft interrupts generated by serial devices.
.It Fn splnet
blocks hard interrupts from network interfaces.
.It Fn splbio
blocks hard interrupts from disks and other mass-storage devices.
.It Fn splsoftnet
blocks soft network interrupts.
.It Fn splsoftclock
blocks soft clock interrupts.
.El
.Pp
Two macros lower the system priority level.
They are:
.Bl -tag -width spllowersoftclockXX
.It Fn spllowersoftclock
unblocks all interrupts but the soft clock interrupt.
.It Fn spl0
unblocks all interrupts.
.El
.Pp
The
.Fn splx
macro restores the system priority level to the one encoded in
.Fa s ,
which must be a value previously returned by one of the other
.Nm
macros.
.Pp
The
.Fn splassert
function checks that the system is running at a certain priority level.
The argument
.Fa s
should be one of these constants:
.Pp
.Bl -tag -width IPL_SOFTCLOCKXX -compact -offset indent
.It Dv IPL_STATCLOCK
.It Dv IPL_SCHED
.It Dv IPL_CLOCK
.It Dv IPL_VM
.It Dv IPL_BIO
.It Dv IPL_TTY
.It Dv IPL_NET
.It Dv IPL_SOFTNET
.It Dv IPL_SOFTCLOCK
.It Dv IPL_NONE
.El
.Pp
The
.Fn splassert
function is optional and is not necessarily implemented on
all architectures nor enabled in all kernel configurations.
It checks the current system priority level to see if it's
at least at the level specified in the argument
.Fa s .
If possible, it also checks if it hasn't been called from an
interrupt handler with a level higher than the one requested, which
must be an error (if some code is protected from
.Dv IPL_SOFTNET
interrupts, but accessed from an
.Dv IPL_NET
interrupt, it must be a design error in the code).
.Pp
The behavior of the
.Fn splassert
function is controlled by the kern.splassert
.Xr sysctl 8 .
Valid values for it are:
.Pp
.Bl -tag -width 1nr -compact -offset indent
.It 0
disable error checking
.It 1
print a message if an error is detected
.It 2
print a message and a stack trace if possible
.It 3
like 2 but also drop into the kernel debugger
.El
.Pp
Any other value causes a system panic on errors.