Re: [PATCH v2] mm/hugetlb: avoid hardcoding while checking if cma is enable

[Roman Gushchin]

On Tue, Jul 07, 2020 at 03:11:56PM +1200, Barry Song wrote:
> hugetlb_cma[0] can be NULL due to various reasons, for example, node0 has
> no memory. so NULL hugetlb_cma[0] doesn't necessarily mean cma is not
> enabled. gigantic pages might have been reserved on other nodes.
> 
> Fixes: cf11e85fc08c ("mm: hugetlb: optionally allocate gigantic hugepages 
> using cma")
> Cc: Roman Gushchin <g...@fb.com>
> Cc: Mike Kravetz <mike.krav...@oracle.com>
> Cc: Jonathan Cameron <jonathan.came...@huawei.com>
> Signed-off-by: Barry Song <song.bao....@hisilicon.com>
> ---
>  -v2: add hugetlb_cma_enabled() helper to improve readability according to 
> Roman
> 
>  mm/hugetlb.c | 16 +++++++++++++++-
>  1 file changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index 57ece74e3aae..d5e98ed86bb9 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -2546,6 +2546,20 @@ static void __init gather_bootmem_prealloc(void)
>       }
>  }
>  
> +bool __init hugetlb_cma_enabled(void)
> +{
> +     if (IS_ENABLED(CONFIG_CMA)) {
> +             int node;
> +
> +             for_each_online_node(node) {
> +                     if (hugetlb_cma[node])
> +                             return true;
> +             }
> +     }
> +
> +     return false;
> +}
> +

Can you, please, change it to a more canonical

#ifdef CONFIG_CMA
bool __init hugetlb_cma_enabled(void)
{
        int node;

        for_each_online_node(node)
                if (hugetlb_cma[node])
                        return true;

        return false;
}
#else
bool __init hugetlb_cma_enabled(void)
{
        return false;
}
#endif

or maybe just 

bool __init hugetlb_cma_enabled(void)
{
#ifdef CONFIG_CMA
        int node;

        for_each_online_node(node)
                if (hugetlb_cma[node])
                        return true;
#endif
        return false;
}

?

Please, feel free to add
Acked-by: Roman Gushchin <g...@fb.com> after that.

Thank you!