Re: [PATCH v6 3/4] iommu: debug-pagealloc: Track IOMMU pages

[David Hildenbrand (Red Hat)]

On 1/13/26 16:08, Jason Gunthorpe wrote:
> On Tue, Jan 13, 2026 at 10:49:28AM +0000, Mostafa Saleh wrote:
> > Would you have a problem if we added a new function in page_ext
> > "page_ext_from_phys()" as Jason suggested?
>
> Given the hidden complexity that David just pointed out I think this
> is essential to encapsulate it in a function.
>
> The function must be able to accept any phys_addr_t and safely returns
> NULL if page_ext cannot be used for some reason.

Right. I think pfn_to_online_page() is the appropriate check.

@Mostafa, I saw you already sent a v7, is the pfn_to_online_page() stuff 
handled in there?

--
Cheers

David