Re: [PATCH v2 2/5] powerpc/mm: drop powerpc vmemmap_can_optimize()

[Lance Yang <lance.yang@linux.dev>]

On 2026/11/6 14:09, Muchun Song wrote:
> 
> 
>> On Nov 5, 2026, at 16:02, Lance Yang <lance.yang@linux.dev> wrote:
>>
>>
>> On Thu, Nov 05, 2026 at 07:16:48AM +0000, sashiko-bot@kernel.org wrote:
>>> Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
>>> - [Low] mm: undeclared vmemmap in section_vmemmap_optimization_disable()
>>> - [Low] mm: define vmemmap_optimizable_order() before section_vmemmap_optimizable()
>>> --
>>
>> IIUC, on LoongArch, asm/pgtable.h includes vmemmap-optimization.h before
>> defining vmemmap.
> 
> Yeah. Really nasty.
> 
>>
>> How about taking a PFN and page count instead, and leaving the conversion
>> to the callers?
> 
> However, although this can solve the compilation problem, for the caller,
> it is not very straightforward because they need to calculate nr_pages.
> section_vmemmap_optimization_disable is only supposed to be used in the
> vmemmap population path, and in these paths, the most direct parameters
> are start and end. I think this function could be moved to
> mm/sparse-vmemmap.c, and that way the compilation dependency problem
> would be solved.
> 
> What do you think?

Yeah, fair point, let's do that ;)