Re: [PATCH 07/14] mm/bootmem_info: avoid using sparse_decode_mem_map()

[Lorenzo Stoakes (Oracle)]

On Tue, Mar 17, 2026 at 05:56:45PM +0100, David Hildenbrand (Arm) wrote:
> With SPARSEMEM_VMEMMAP, we can just do a pfn_to_page(). It is not super
> clear whether the start_pfn is properly aligned ... so let's just make

Maybe worth saying aligned to the start of the section?

> sure it is.
>
> We will soon might try to remove the bootmem info completely, for now,
> just keep it working as is.

*gasp*

>
> Signed-off-by: David Hildenbrand (Arm) <david@xxxxxxxxxx>

If it boots ship it, so:

Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@xxxxxxxxxx>

> ---
>  mm/bootmem_info.c | 9 ++++-----
>  1 file changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/mm/bootmem_info.c b/mm/bootmem_info.c
> index e61e08e24924..3d7675a3ae04 100644
> --- a/mm/bootmem_info.c
> +++ b/mm/bootmem_info.c
> @@ -44,17 +44,16 @@ static void __init register_page_bootmem_info_section(unsigned long start_pfn)
>  {
>  	unsigned long mapsize, section_nr, i;
>  	struct mem_section *ms;
> -	struct page *page, *memmap;
>  	struct mem_section_usage *usage;
> +	struct page *page;
>
> +	start_pfn = SECTION_ALIGN_DOWN(start_pfn);

Yeah SPARSE_VMEMMAP should make it this trivial.

>  	section_nr = pfn_to_section_nr(start_pfn);
>  	ms = __nr_to_section(section_nr);
>
> -	memmap = sparse_decode_mem_map(ms->section_mem_map, section_nr);
> -

So this is nice...

>  	if (!preinited_vmemmap_section(ms))
> -		register_page_bootmem_memmap(section_nr, memmap,
> -				PAGES_PER_SECTION);
> +		register_page_bootmem_memmap(section_nr, pfn_to_page(start_pfn),
> +					     PAGES_PER_SECTION);
>
>  	usage = ms->usage;
>  	page = virt_to_page(usage);
> --
> 2.43.0
>

Cheers, Lorenzo