Re: [PATCH] x86/power/64: Fix page-table setup for temporary text mapping

[Joerg Roedel]

Adding Dave Hansen to Cc, forgot that in the initial post.

On Thu, Apr 19, 2018 at 08:26:00PM +0200, Joerg Roedel wrote:
> From: Joerg Roedel <jroedel@xxxxxxx>
> 
> When we have a 4-level page-table there is no p4d, so we map
> the pud in the pgd. The old code before commit fb43d6cb91ef
> already did that.
> 
> With the change from above commit we end up with an invalid
> page-table which causes undefined behavior. In one report it
> caused triple faults.
> 
> Fix it by changing the p4d back to pud so that we have a
> correct page-table.
> 
> Reported-by: Borislav Petkov <bp@xxxxxxx>
> Fixes: fb43d6cb91ef ('x86/mm: Do not auto-massage page protections')
> Tested-by: Michal Kubecek <mkubecek@xxxxxxx>
> Tested-by: Borislav Petkov <bp@xxxxxxx>
> Signed-off-by: Joerg Roedel <jroedel@xxxxxxx>
> ---
>  arch/x86/power/hibernate_64.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/power/hibernate_64.c b/arch/x86/power/hibernate_64.c
> index 48b14b534897..ccf4a49bb065 100644
> --- a/arch/x86/power/hibernate_64.c
> +++ b/arch/x86/power/hibernate_64.c
> @@ -98,7 +98,7 @@ static int set_up_temporary_text_mapping(pgd_t *pgd)
>  		set_pgd(pgd + pgd_index(restore_jump_address), new_pgd);
>  	} else {
>  		/* No p4d for 4-level paging: point the pgd to the pud page table */
> -		pgd_t new_pgd = __pgd(__pa(p4d) | pgprot_val(pgtable_prot));
> +		pgd_t new_pgd = __pgd(__pa(pud) | pgprot_val(pgtable_prot));
>  		set_pgd(pgd + pgd_index(restore_jump_address), new_pgd);
>  	}
>  
> -- 
> 2.13.6