[64studio-devel] kernel-2.6.22 upsets yaird on hdb

[tim hall]

Daniel James wrote:
> Hi Tim,
> 
>> BUG: sleeping function called from invalid context kpnpbiosd(92) at 
>> kernel/rtmutex.c:636
> 
> This is RT patch related, as far as I can tell. I wonder if disabling 
> PnP in your BIOS settings might help.

The BIOS interface on this machine is low on options - It's an Anaheim 
e-machines MoBo, which some on this list may recognise as belonging to 
the last century. I have seen it suggested that the BIOS on this 
computer is not fully ACPI compliant, however it seems to be running 
quite happily like this. I suspect it may be doing it in software from 
my brief reading of the dmesg. I _can_ disable ACPI in the BIOS (PHoenix 
IIRC). Incidentally it only takes matched pairs of PC100 SDRAM, so I'm 
on the lookout for two 256M slices, which are proving not so easy to get 
hold of anymore. Just thought I'd mention it. I'm running on 128M right 
now, which is painfully slow. :(

At some point in the fairly near future, hopefully, I'll be able to 
splash out on something made this century. ;)

>> Anyway, I'm running 2.6.21 for now & I'll check out 2.6.23 as soon as 
>> I have the rest of my hardware running happily on the old kernel.
> 
> So our 2.6.21-RT package works fine on this same machine?

It would seem that way:

 From dmesg:
pnp: PnP ACPI init
pnp: PnP ACPI: found 13 devices
PnPBIOS: Disabled by ACPI PNP

No errors reported.

cheers,

tim