Re: [PATCH] mm: hugepage: make hstate_is_gigantic() inline

[Michal Hocko]

On Mon 22-05-17 14:46:57, Arnd Bergmann wrote:
> Ignoring the argument of both alloc_huge_page_node() and hstate_is_gigantic()
> causes a harmless warning:
> 
> mm/memory-failure.c: In function 'new_page':
> mm/memory-failure.c:1493:18: error: unused variable 'hstate' [-Werror=unused-variable]
> 
> Turning one into an inline function makes the compiler see that the variable is
> not meant to be unused.
> 
> Fixes: d81a077951f8 ("mm/madvise: enable (soft|hard) offline of HugeTLB pages at PGD level")
> Fixes: 39b80148ffd7 ("mm/follow_page_mask: add support for hugepage directory entry")

Both these sha1 are unstable because they are in Andrew's mmotm tree and
will change in future next trees. I would just drop them.

> Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>

the patch itself looks OK

> ---
>  include/linux/hugetlb.h | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
> index 76b9ddcd8c3c..c92a1f0c7240 100644
> --- a/include/linux/hugetlb.h
> +++ b/include/linux/hugetlb.h
> @@ -532,7 +532,11 @@ struct hstate {};
>  #define vma_mmu_pagesize(v) PAGE_SIZE
>  #define huge_page_order(h) 0
>  #define huge_page_shift(h) PAGE_SHIFT
> -#define hstate_is_gigantic(h) 0
> +static inline bool hstate_is_gigantic(struct hstate *h)
> +{
> +	return false;
> +}
> +
>  static inline unsigned int pages_per_huge_page(struct hstate *h)
>  {
>  	return 1;
> -- 
> 2.9.0
> 

-- 
Michal Hocko
SUSE Labs