Re: [PATCH 2/2] mm/mm_init.c: do not calculate memmap and reserved pages for empty zone

[Mike Rapoport]

On Thu, Jun 01, 2023 at 06:35:36AM +0000, Haifeng Xu wrote:
> For empty zone, it's unnecessary to account for pages used by memmap.
> Also, the pages reserved in dma zone can be ignored.
> 
> Signed-off-by: Haifeng Xu <haifeng.xu@xxxxxxxxxx>
> ---
>  mm/mm_init.c | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/mm/mm_init.c b/mm/mm_init.c
> index 232efac9a929..b82718d596c3 100644
> --- a/mm/mm_init.c
> +++ b/mm/mm_init.c
> @@ -1543,6 +1543,13 @@ static void __init free_area_init_core(struct pglist_data *pgdat)
>  		size = zone->spanned_pages;
>  		freesize = zone->present_pages;
>  
> +		if (!size) {
> +			nr_kernel_pages += 0;
> +			nr_all_pages += 0;

Please drop these, they are nops anyway.

> +			zone_init_internals(zone, j, nid, 0);
> +			continue;
> +		}
> +
>  		/*
>  		 * Adjust freesize so that it accounts for how much memory
>  		 * is used by this zone for memmap. This affects the watermark
> @@ -1579,10 +1586,6 @@ static void __init free_area_init_core(struct pglist_data *pgdat)
>  		 * And all highmem pages will be managed by the buddy system.
>  		 */
>  		zone_init_internals(zone, j, nid, freesize);
> -
> -		if (!size)
> -			continue;
> -
>  		setup_usemap(zone);
>  		init_currently_empty_zone(zone, zone->zone_start_pfn, size);
>  	}
> -- 
> 2.25.1
> 

-- 
Sincerely yours,
Mike.