On Tue 31-01-23 16:01:16, Suren Baghdasaryan wrote:
> Provide vm_flags_reset_once() and replace the vm_flags updates which used
> WRITE_ONCE() to prevent compiler optimizations.
> 
> Fixes: 0cce31a0aa0e ("mm: replace vma->vm_flags direct modifications with 
> modifier calls")
> Reported-by: Hyeonggon Yoo <42.hye...@gmail.com>
> Signed-off-by: Suren Baghdasaryan <sur...@google.com>

This would have been better folded into the vm_flags modification patch
because it would be more obvious change. Hugh has provided a very nice
comment in mlock_vma_pages_range but the git blame would be more visible
when the conversion is from WRITE_ONCE.

One way or the other
Acked-by: Michal Hocko <mho...@suse.com>

> ---
> Notes:
> - The patch applies cleanly over mm-unstable
> - The SHA in Fixes: line is from mm-unstable, so is... unstable
> 
>  include/linux/mm.h | 7 +++++++
>  mm/mlock.c         | 4 ++--
>  2 files changed, 9 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index 5bf0ad48faaa..23ce04f6e91e 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -648,6 +648,13 @@ static inline void vm_flags_reset(struct vm_area_struct 
> *vma,
>       vm_flags_init(vma, flags);
>  }
>  
> +static inline void vm_flags_reset_once(struct vm_area_struct *vma,
> +                                    vm_flags_t flags)
> +{
> +     mmap_assert_write_locked(vma->vm_mm);
> +     WRITE_ONCE(ACCESS_PRIVATE(vma, __vm_flags), flags);
> +}
> +
>  static inline void vm_flags_set(struct vm_area_struct *vma,
>                               vm_flags_t flags)
>  {
> diff --git a/mm/mlock.c b/mm/mlock.c
> index ed49459e343e..617469fce96d 100644
> --- a/mm/mlock.c
> +++ b/mm/mlock.c
> @@ -380,7 +380,7 @@ static void mlock_vma_pages_range(struct vm_area_struct 
> *vma,
>        */
>       if (newflags & VM_LOCKED)
>               newflags |= VM_IO;
> -     vm_flags_reset(vma, newflags);
> +     vm_flags_reset_once(vma, newflags);
>  
>       lru_add_drain();
>       walk_page_range(vma->vm_mm, start, end, &mlock_walk_ops, NULL);
> @@ -388,7 +388,7 @@ static void mlock_vma_pages_range(struct vm_area_struct 
> *vma,
>  
>       if (newflags & VM_IO) {
>               newflags &= ~VM_IO;
> -             vm_flags_reset(vma, newflags);
> +             vm_flags_reset_once(vma, newflags);
>       }
>  }
>  
> -- 
> 2.39.1.456.gfc5497dd1b-goog

-- 
Michal Hocko
SUSE Labs

Reply via email to