Re: PROBLEM: Probabilistic segfault on AMD hardware with INVLPGB

[Dave Hansen]

On 7/7/26 07:11, Rik van Riel wrote:
>>   ./repro-invlpgb --batch --rounds 20 --jobs 32 -d 5 -w 8 -m 2 -s 512
>> -q
>>
> Now that's interesting. I ran it for 100 rounds on
> a 158 core Turin system, without hitting the issue.
> 
> I'll play around with it some more to see whether
> I can make it happen by varying the parameters,
> but now I'm also wondering whether hardware
> differences might be at play.

The first thing I'd check is if you have the _exact_ same model,
stepping and (especially) microcode version.

BTW, it seems like the 'x86 mm' maintainers are cc'd here, but not the
core x86 folks, so you missed Boris. They should probably be keeping an
eye on this too.