Fatal trap 18

[Andriy Gapon]

on 03/11/2010 21:45 Dan Allen said the following:
> 
> On 3 Nov 2010, at 2:18 AM, Sergey Kandaurov wrote:
> 
>> +       if (cpu_logical == 0) +               cpu_logical = 1;        /*
>> XXX max_logical? */ cpu_cores /= cpu_logical; hyperthreading_cpus =
>> cpu_logical;
> 
> My machine contains an Intel Core Duo, not a Core 2 Duo.  However, both
> machines have 2 logical CPUs.  Perhaps the bug is not handling a Core Duo
> properly.
> 
> I just sync'd with csup's date command as Jeremy suggested.  It is building
> right now.  We'll see if a 2010.10.28.06.00.00 GMT system has the problem.  I
> doubt it does.  I think this Intel logical CPU code business is probably the
> culprit.

I am quite sure of that.

> It is hard to believe that I am the only one with a Core Duo machine finding
> this problem!

Yeah, true.
But there could be something unusual about your setup.
I mean these lines from your dmesg:

> WARNING: Non-uniform processors.
> WARNING: Using suboptimal topology.

What I propose is that we try to find a resolution, because I suppose you can't
stay at the last working revision forever.

P.S. so will you be trying the patch I proposed?

-- 
Andriy Gapon