On Mon, 2020-05-25 at 10:20 +0200, Borislav Petkov wrote: > On Fri, May 22, 2020 at 10:50:17PM +0300, Jarkko Sakkinen wrote: > > If other values except two are never going to be used it is more than a > > legit point to validate this field. > > > > It also the potential to use ~0x8086 bits to be defined later if ever > > needed lets say for some kernel specific purpose. > > Yah, let's cover our ass for the future. We have all seen the "this > won't be used" but then "we're using it" change of heart. IOW, let's > align with what the hw checks and we can always relax that in the future > but not the other way around. > I fully agree with this conclusion and that is exactly how I think about the issue too. I'll add a check for this. /Jarkko
