[geeks] Upgrading/troubleshooting Athlon64 laptop (was Policy for system...)

[Phil Stracchino]

On 07/26/10 20:20, der Mouse wrote:
> Unless memtest86+ simply doesn't poke the hardware that's at fault.
> This is my current favoured theory: that something (disk interface? one
> of the many bus-to-bus bridges modern machines have?) has gone flaky,
> for reasons we can speculate about but which don't really matter.  But,
> as long as it's hardware memtest86+ doesn't use and thus doesn't touch,
> and is hardware the BIOS doesn't use for booting, this fits.
> 
> The latter eliminates a few things, but not all that much; for example,
> it might not be the IDE interface, but it could be the DMA engine for
> the IDE interface (booting often enough does PIO for everything, as
> it's not performance-critical and PIO is easier and smaller to code).

I actually  finally managed to get it to boot again from a rescue CD
with a 64-bit kernel with ACPI disabled, and got a new kernel built once
it was booted.  I tested a possible theory that it might have been the
P-states driver at fault, but removing that makes no difference.

The baffling thing is, at one point I had this thing running for more
than a day *with* ACPI and cpufreqd working.  Of course, that was before
I knew the heatsinks were clogged...  I'm hoping I didn't overheat the
CPU and permanently damage it.  (Hence my asking whether it's likely
safe to try running a faster Newark-core CPU with a slightly lower core
voltage in it.)


-- 
  Phil Stracchino, CDK#2     DoD#299792458     ICBM: 43.5607, -71.355
  alaric at caerllewys.net   alaric at metrocast.net   phil at co.ordinate.org
         Renaissance Man, Unix ronin, Perl hacker, Free Stater
                 It's not the years, it's the mileage.