RE: Help with machine check exception

[Roger Heflin]

> 
> I would have expected 2 CPU temps (being dual-processor).  
> Maybe Remote is the second.
> 
> With 4 copies of burnK7:
> 
> machine with problems:
> 
> CPU:      +71.25°C  (low  =   +10°C, high =   +50°C)     ALARM
> Board:    +47.25°C  (low  =   +10°C, high =   +35°C)     ALARM
> Remote:   +68.50°C  (low  =   +10°C, high =   +35°C)     ALARM
> 
> machine without:
> 
> CPU:      +61.25°C  (low  =   +10°C, high =   +50°C)     ALARM
> Board:    +47.25°C  (low  =   +10°C, high =   +35°C)     ALARM
> Remote:   +74.25°C  (low  =   +10°C, high =   +35°C)     ALARM
> 
> 
> So *maybe* cooling?
> 

That is a little on the warm side, I believe AMD's posted limit
is 70C for most of their chips, assuming the measuring point is
in the correct place for the 70C limit.

Certain cpus also seem to have more issues than others, so one cpu
out of a batch can be ok with a certain setup, and another from the
same batch will mce under similar conditions.

Did you build the machines yourself or did you buy them this way?

Machines getting MCE's that often will fail the burnin testing that
we use here.

And machines that produce those kinds of temps will also fail our
burn-in process just because that seems a bit too warm.

                           Roger

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/