memtest running on alix1c!

[ron minnich]

it was a minicom/terminal size issue. A big minicom caused memtest to
get a divide by zero. It's running!

It has confusion:
L1 Cache: Unknown           | Test #6  [Moving inversions, 32 bit pattern]
L2 Cache: Unknown           | Testing:  120K -  256M  255M

but hey ...

ron

[Carl-Daniel Hailfinger]

Wohooo! Great! Congratulations!

On 02.02.2008 23:11, ron minnich wrote:
> it was a minicom/terminal size issue. A big minicom caused memtest to
> get a divide by zero. It's running!

Haha, a memtest bug. I guess they'd like to hear about it.
> It has confusion:
> L1 Cache: Unknown           | Test #6  [Moving inversions, 32 bit pattern]
> L2 Cache: Unknown           | Testing:  120K -  256M  255M

Good enough for now. We might want to find out why it has no idea about
cache size, though.

Regards,
Carl-Daniel

[Tom Sylla]

> > It has confusion:
> > L1 Cache: Unknown           | Test #6  [Moving inversions, 32 bit pattern]
> > L2 Cache: Unknown           | Testing:  120K -  256M  255M

See if you can make sure this is not another memtest bug. For a long 
time, memtest did not know that Geode LX CPUID meant to read the cache 
information like an AMD processor. I think the issue was fixed in more 
recent memtest versions. If memtest is ok, make sure the cache 
information MSRs for CPUID are being programmed correctly.