switching above IPL_SCHED (was Re: New panic on 5.99.3)

[Iain Hibbert <plunky%rya-online.net@localhost>]

On Thu, 11 Dec 2008, Arnaud Lacombe wrote:

> On Fri, Nov 21, 2008 at 8:22 PM, Arnaud Lacombe 
> <lacombar%gmail.com@localhost> wrote:
> > acpi0: power button pressed
> > syncing disk... 2 done
> > unmounting file systems
> > unmounting /kern (kernfs)
> > unmounting /tmp (mfs:178)
> > unmounting /var (/dev/wd0f)
> > unmounting / (/dev/wd0a)
> > acpi0: entering state 5
> > panic: cpu_switchto: switching above IPL_SCHED (8)
> > fatal breakpoint ...
> > trap type 1 code 0 eip c021127c cs 8 eflags 246 cr2 bbbc658c ilevel 8
> > Stopped in pid 0.12 (system) at 0xc021127c: popl %ebp
> > db{0}> ps/l
> > [...]
> >> 12 3 204 ca7492a0 sysmon eccv
> >

I'm seeing this often (50% of time) on NetBSD 5.0_RC2 (i386 dual core)
system when shutting down with 'shutdown -p' after all disks are umounted.
(retyped) msgbuf is:

acpi0: entering state 5
panic: cpu_switchto: switching above IPL_SCHED (8)
fatal breakpoint trap in supervisor mode
trap type 1 code 0 eip c05741bc cs8 eflags ...
Stopped in pid 942.1 (halt)
db{0} bt
breakpoint(
panic(
cpu_switchto(
sleepq_block(
cv_timedwait(
acpiec_space_handler(
AcpiEvAddressSpaceDispatch(

Acpi*(  # quite a lot of these I didn't write them all down

acpi_enter_sleep_state(
cpu_reboot(
sys_reboot(

I don't *think* I've seen it when just rebooting and I'm not quite sure if
it only happens with DIAGNOSTIC or LOCKDEBUG as I've been switching
kernels about quite a bit. (that last is definitely DIAGNOSTIC and
LOCKDEBUG enabled)

> I got this one on a non stressed 5.99.4 system again.

did you make a PR?

iain