Re: [RFC PATCH 23/23] x86/process: Reset hardware history in context switch

[Ricardo Neri]

On Mon, Oct 03, 2022 at 12:15:32AM +0200, Borislav Petkov wrote:
> On Sun, Oct 02, 2022 at 03:12:38PM -0700, Ricardo Neri wrote:
> > Sure I can do this, Boris. I guess this implies that I also need to
> > add the DISABLE_MASK bits.
> 
> Are you adding a CONFIG_ item which can control the DISABLE_MASK bit
> too?

I am not. I could use CONFIG_INTEL_HFI_THERMAL, as using HRESET only makes
sense when Intel Thread Director (Peter argued in separate email against
having a config option specific for Intel Thread Director).

> 
> > Othewise, IIUC, cpu_feature_enabled() falls back to static_cpu_has().
> 
> And?

Since I did not implement a DISABLE_MASK bit nor a CONFIG_ option for
HRESET, I thought that static_cpu_has() was sufficient.

I am not against using cpu_feature_enabled(), I just want to confirm that
I also need to implement the DISABLE_MASK bit and the CONFIG_ option.

Thanks and BR,
Ricardo