Re: [PATCH v10 09/50] x86/traps: Define RMP violation #PF error code

[Michael Roth]

On Mon, Oct 16, 2023 at 07:14:07AM -0700, Dave Hansen wrote:
> On 10/16/23 06:27, Michael Roth wrote:
> > Bit 31 in the page fault-error bit will be set when processor encounters
> > an RMP violation. While at it, use the BIT() macro.
> 
> Any idea where the BIT() use went?  I remember seeing it in earlier
> versions.

Yah... this patch used to convert all the previous definitions over to
using BIT() as part of introducing the new RMP bit. I'm not sure what
happened, but a likely possibility is I hit a merge conflict at some
point due to upstream commit fd5439e0c9, which introduced this change:

  X86_PF_SHSTK    =               1 << 6,

and my brain probably defaulted to using the existing pattern to
resolve it. I'll get this fixed up.

-Mike