On Mon, Aug 03, 2026 at 11:39:11AM +0200, David Hildenbrand (Arm) wrote:
> On 7/30/26 08:32, Gregory Price wrote:
> > On Wed, Jul 29, 2026 at 05:48:39PM +0100, Lorenzo Stoakes (ARM) wrote:
> >> This is potentially useful debugging information and matches the existing
> >> page offset provided.
> >>
> >> Use the raw __linear_anon_page_index() function so as to always output this
> >> value regardless of whether the mapping is file-backed or not.
> >>
> >
> > Begs the question whether we should mark the value based on whether
> > it is file backed or not, but i suppose you can gleen that from the
> > other printed values (anon_mem, mapping, vm_ops, etc).
>
> You mean something anon_pgoff_t or what did you have in mind?
I suspect he means e.g. adding a (A) suffix or something? Like page offset xxx
(F) or yyy (A) I guess?
For anon_pgoff_t it'd have to be a typedef struct { pgoff_t val; } anon_pgoff_t
to be meaningful and then you'd have weirdness with assignment to
e.g. folio->index/other calculations, and especially linear_folio_index() which
returns the right pgoff_t depending on file-backed/anon :)
So I don't think it'd be helpful, and the asserts already prevent incorrect
usage.
>
> --
> Cheers,
>
> David
--
Cheers, Lorenzo
