Re: [patch V9 00/39] x86/entry: Rework leftovers (was part V)

[Peter Zijlstra]

On Thu, Jun 04, 2020 at 08:42:52AM -0700, Andy Lutomirski wrote:

> x86 is not an architecture.  x86 is a gauntlet through which operating
> system developers must run.

That made my day :-)

> I think we can tolerate this particular mess -- can't we just say that
> a BUS LOCK DEBUG EXCEPTION is fatal if it came from kernel mode?  So
> what if it nests inside #DB -- we can survive an oops from a corrupt
> context like that.

Yes, SLD or this new thing is unconditionally fatal when from kernel
space. As long as we can get to the OOPS with our stacks completely
wrecked, we should be good I think.

We'll just need to make this one of the very first things is checks for,
to minimize the amount of code ran before OOPSing, so at to minimize the
risk for recursive exceptions.

All signs of excellent design, I'm sure.