[ag-automation] Latest Stable OSADL-Kernel not SMP-safe ?

[Claus Gindhart]

Robert,

i have tested the ThinkIO-Duo BIOS against the linuxfirmwarekit R3.

In fact, it claimed a problem in the HPET entry (The HPET was defined in 2 
different locations in the ACPI-Table).
For Linux dmesg, this is absolutely o.k, but the firmwarekit says
"HPET 0xFED00000 address busy" when it detects the 2nd entry.
Also other systems, for example my development notebook, show exactly the same 
behaviour.

Anyway: We fixed this table in a way, such that now firmwarekit is satisfied 
and doesnt claim any HPET fails any more. But this had absolutely no 
influence on the problem.

Cyclictest still hangs up, if started in multi-threaded mode 


Best regards,
Claus Gindhart


On Tuesday 14 October 2008 08:39:19 am Robert Schwebel wrote:
> On Tue, Oct 14, 2008 at 08:26:54AM +0200, Claus Gindhart wrote:
> > The kernel 2.6.24.7-rt17 has an additional problem, which i didnt
> > mention in the last mail. This kernel sporadically hangs up at kernel
> > boot during the ACPI initialization (does not appear, if acpi=off
> > option is added to the command line) This problem disappeared by
> > updating to 2.6.24.7-rt21
>
> PC hardware timers are a source for all kinds of weird problems. Did you
> run
>
> 	http://www.linuxfirmwarekit.org/
>
> on your board, in order to make sure your BIOS is not broken?
>
> rsc




...



...