Re: [PATCH 3/7] mm: Use ptep_get() for accessing PTE entries

[Ryan Roberts]

On 13/09/2024 09:44, Anshuman Khandual wrote:
> Convert PTE accesses via ptep_get() helper that defaults as READ_ONCE() but
> also provides the platform an opportunity to override when required.
> 
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: David Hildenbrand <david@xxxxxxxxxx>
> Cc: Ryan Roberts <ryan.roberts@xxxxxxx>
> Cc: "Mike Rapoport (IBM)" <rppt@xxxxxxxxxx>
> Cc: linux-mm@xxxxxxxxx
> Cc: linux-kernel@xxxxxxxxxxxxxxx
> Signed-off-by: Anshuman Khandual <anshuman.khandual@xxxxxxx>
> ---
>  include/linux/pgtable.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
> index 2a6a3cccfc36..05e6995c1b93 100644
> --- a/include/linux/pgtable.h
> +++ b/include/linux/pgtable.h
> @@ -1060,7 +1060,7 @@ static inline int pgd_same(pgd_t pgd_a, pgd_t pgd_b)
>   */
>  #define set_pte_safe(ptep, pte) \
>  ({ \
> -	WARN_ON_ONCE(pte_present(*ptep) && !pte_same(*ptep, pte)); \
> +	WARN_ON_ONCE(pte_present(ptep_get(ptep)) && !pte_same(ptep_get(ptep), pte)); \

Suggest reading once into a temporary so that the pte can't change between the 2
gets. In practice, it's not likely to be a huge problem for this instance since
its under the PTL so can only be racing with HW update of access and dirty. But
good practice IMHO:

    pte_t __old = ptep_get(ptep); \
    WARN_ON_ONCE(pte_present(__old) && !pte_same(__old, pte)); \

Thanks,
Ryan

>  	set_pte(ptep, pte); \
>  })
>