On Tue, Jan 16, 2018 at 10:11:14AM -0800, Dave Hansen wrote:
>
> Ugh. The ghosts of PAE have come back to haunt us.
:-) Yeah, PAE caused the most trouble for me while getting this running.
>
> Could we do:
>
> static inline bool pgd_supports_nx(unsigned long)
> {
> #ifdef CONFIG_X86_64
> return (__supported_pte_mask & _PAGE_NX);
> #else
> /* No 32-bit page tables support NX at PGD level */
> return 0;
> #endif
> }
>
> Nobody will ever spot the #ifdef the way you laid it out.
Right, thats a better way to do it. I'll change it in the next version.
Thanks,
Joerg