Re: [PATCH] x86: fix user address masking non-canonical speculation issue

[Kirill A. Shutemov]

On Thu, Oct 24, 2024 at 10:35:33AM -0700, Linus Torvalds wrote:
> On Wed, 23 Oct 2024 at 23:13, Josh Poimboeuf <jpoimboe@xxxxxxxxxx> wrote:
> >
> > I'm probably missing something but once LAM is enabled, how wouldn't
> > this allow non-canonical address speculation?
> 
> Once LAM is enabled, together with LASS, non-canonical addresses
> basically don't exit.

That's not true.

With LAM, canonically check is relaxed to bit 63 == bit 47/56.

I try to confirm internally that we don't speculate past this relaxed
canonically check. I believe we don't, but I want to double-check.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov