Re: [patch V3 4/6] x86/pkru: Make PKRU=0 actually work

[Thomas Gleixner]

On Tue, Jun 08 2021 at 09:06, Dave Hansen wrote:
> On 6/8/21 7:36 AM, Thomas Gleixner wrote:
>> When user space brings PKRU into init state, then the kernel handling is
>> broken:
>
> Nit: while setting PKRU=0 _can_ trigger this issue (on AMD), the
> underlying problem is truly with the init state and not simply with any
> wrpkru(0).  IOW, I like the changelog better than the subject.
>
> Maybe something like this would be more precise:
>
> 	x86/pkru: Write hardware init value to PKRU when xstate is init

Indeed.