Bug in 6.1 acpi?

[Nate Lawson]

John Baldwin wrote:
> On Friday 22 September 2006 14:12, up at 3.am wrote:
>> On Fri, 22 Sep 2006, Brooks Davis wrote:
>>
>>> [Please don't cross post]
>>>
>>> On Fri, Sep 22, 2006 at 02:03:37PM -0400, up at 3.am wrote:
>>>> (Please respond directly, as I am not subscribed)
>>>>
>>>> I've asked about this error message before, but this has gotten serious.
>>>> Not sure if it's related, but this server keeps spontaneously having 
> what
>>>> appear to be power events every 13-40 hours or so.  No errors or panic
>>>> messages or core dumps.
>>>>
>>>> The system has dual power supplies and was rock stable running 4.X.  The
>>>> problem only started occuring after upgrading to 6.1-STABLE.  Could it 
> be
>>>> related to this DMESG? :
>>>>
>>>> acpi0: <PTLTD   RSDT> on motherboard
>>>> acpi0: Power Button (fixed)
>>>> acpi: bad write to port 0x070 (8), val 0x43
>>>> acpi: bad read from port 0x071 (8)
>>>>
>>>> The hardware is an Intel L440GX+ MB with dual 1Ghz CPUs with SMP (I 
> tried
>>>> a kernel without SMP, but it didn't help), 1GB ECC RAM, 1GB Swap, 
> Adaptec
>>>> 2100 SCSI RAID level 1.  It apppears to be lightly loaded in terms of 
> CPU
>>>> and RAM.
>>> Have you verified that you're running with the latest BIOS?  On a system
>>> that old, there's a decent chance the ACPI implementation is buggy.
>>> Have you tried running with ACPI disabled?
>> Yes, I did...still had the same problem.
> 
> Then it's not an ACPI problem. :)  The bad read/write messages above I think 
> you don't need to worry about either.  Sounds more like a hardware issue of 
> sorts.
> 

I don't know of any acpi issue that could cause spontaneous reboots.

You can try booting with acpi disabled to see if it helps:
hint.acpi.0.disabled="1"

If that helps, we can try to debug acpi more.  My guess is it won't 
change anything.  The bad read/write messages are harmless since we 
don't use the RTC.  They're more for the future if we do start using it.

-- 
Nate