Re: [PATCH v4 3/4] mm: mprotect: convert to folio_needs_prot_numa()

[Lorenzo Stoakes <lorenzo.stoakes@oracle.com>]

On Tue, Oct 21, 2025 at 05:01:30PM +0200, David Hildenbrand wrote:
> On 21.10.25 16:56, Lorenzo Stoakes wrote:
> > On Tue, Oct 21, 2025 at 08:54:08PM +0800, Kefeng Wang wrote:
> > >
> > >
> > > On 2025/10/21 17:45, Lorenzo Stoakes wrote:
> > > > On Tue, Oct 21, 2025 at 10:25:08AM +0100, Lorenzo Stoakes wrote:
> > > > > On Tue, Oct 21, 2025 at 11:13:32AM +0200, David Hildenbrand wrote:
> > > > > > On 21.10.25 10:41, Kefeng Wang wrote:
> > > > > > > Maybe folio_needs_protnone_mapping()?
> > > > > >
> > > > > > I think prot_numa should be in the name one way or the other, because we
> > > > > > also have ordinary PROT_NONE unrealted to NUMA faults.
> > > > >
> > > > > folio_can_map_prot_numa() semes to me to be the best choice.
> > > > >
> > > >
> > > > I don't want to hold up the series so if you guys feel strongly, leave it
> > > > as-is. But this reads better to me.
> > >
> > > Let's keep things as-is for now.
> >
> > It'd be nice to have an argument as to why (I did say if you guys had
> > strong opinions...), but never mind.
> >
>
> folio_can_map_prot_numa() works for me, so we can just use that and call it
> a day :)

Thanks!

Kefeng - Since we need a respin anyway, and if you don't have any specific,
strong, objection to this, can we just change the function name to this please?

Cheers, Lorenzo