.\" $OpenBSD: spl.9,v 1.17 2004/04/27 12:05:57 jmc 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. All advertising materials mentioning features or use of this software .\" must display the following acknowledgement: .\" This product includes software developed by Michael Long. .\" 4. 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 March 11, 1997 .Dt SPL 9 .Os .Sh NAME .Nm spl .Nd modify system interrupt priority level .Sh SYNOPSIS .Fd #include .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. Code running at or above this level may not call .Fn sleep , .Fn tsleep , or .Fn wakeup , 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 .Fn sleep , .Fn tsleep , or .Fn wakeup . .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_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.