Oops with ACPI (in sched:566)

Oops with ACPI (in sched:566)

[Marcin Gogolewski]

I have STL2 machine, if I compile kernel (2.4.18) with ACPI I got:
[...] (I hope this is OK, I write it down from monitor ;-) )
ACPI: Core Subsystem version [20011018]
Scheduling in interrupt
kernel BUG at sched:566!
invalid operand
CPU:0
EIP:0010:[<c0117546>] Not tainted
EFLAGS: 00010286
eax: 0x16 ebx: 0x0 ecx:c02d2764 edx:00002480
esi:ffdf4000 edi: 0x0 ebp: f7df5fd4 esp:f7df5f9c

ds:0018 es: 0018 ss:0018
Process ksoftirq_CPU0(pid:3, stackpage=f7df5000)
Stack: c02918e2 00000236 00000000 00000000 00000018 c0320018 f7df4000 c0121660
00000000 f7df4000 00000246 00000000 f7df4000 f7df4000 00000000 c0121ca1
00000000 00010f00 c2113fac 00000000 c0350f40 c0105766 00000000 c0121bb0
Call Trace:[<c0121660>][<c0121ca1>][<c0105766>][<c0121bb0>]
Code:0f 0b 56 5e 8b 55 ec 8b 4a 1c 85 c9 78 3e 81 3d 04 84 32 c0
<0> Kernel panic: Aiee, killing interrupt handler!
In interrupt handler - not syncing
[...]

without ACPI there isn't any problem, with ACPI in (I think) 50% starts.

My system:
Intel STL2 with 2xPIII 1 GHz, I can enclose details (lcpci, etc.)
I can't do more tests (it's a production machine).

                                        Marcin

Re: Oops with ACPI (in sched:566)

[Michael Smith]

I haven't checked for BIOS updates for the SLT2 for a while, but in the 
code as of 6-12 months ago there is definitely a mutex deadlock.

Andrew already has all the details.

> > From: Marcin Gogolewski [mailto:marcing@ms-itti.com.pl]
> > I have STL2 machine, if I compile kernel (2.4.18) with ACPI I got:
> > [...] (I hope this is OK, I write it down from monitor ;-) )
> > ACPI: Core Subsystem version [20011018]
> > Scheduling in interrupt
> > kernel BUG at sched:566!
> 
> I've heard of problems with STL2 before, due to a poor ACPI BIOS.
> 
> If you're willing, I'd be great if you could get the ACPI pmtools
> (http://developer.intel.com/technology/iapc/acpi/downloads.htm), run
> acpidmp, and send me the output. At least then I can put the STL2 on the
> ACPI blacklist, and maybe we can figure out more about the problem, too.
> 
> Regards -- Andy

-- 
To announce that there must be no criticism of the president,
or that we are to stand by the president, right or wrong, is not
only unpatriotic and servile, but is morally treasonable to 
the American public.  - Theodore Roosevelt

RE: Oops with ACPI (in sched:566)

[Grover, Andrew]

> From: Marcin Gogolewski [mailto:marcing@ms-itti.com.pl]
> I have STL2 machine, if I compile kernel (2.4.18) with ACPI I got:
> [...] (I hope this is OK, I write it down from monitor ;-) )
> ACPI: Core Subsystem version [20011018]
> Scheduling in interrupt
> kernel BUG at sched:566!

I've heard of problems with STL2 before, due to a poor ACPI BIOS.

If you're willing, I'd be great if you could get the ACPI pmtools
(http://developer.intel.com/technology/iapc/acpi/downloads.htm), run
acpidmp, and send me the output. At least then I can put the STL2 on the
ACPI blacklist, and maybe we can figure out more about the problem, too.

Regards -- Andy