Re: [PATCH 0/2] mm: memory-failure: fix HWPoison flag race with non-atomic page flag ops

[David Hildenbrand (Arm)]

On 6/30/26 01:29, Michael S. Tsirkin wrote:
> On Mon, Jun 29, 2026 at 03:05:18PM +0200, David Hildenbrand (Arm) wrote:
>> It would be good to coordinate here.
> 
> Case in point, do you feel this should be blocking the zeroing/page
> reporting work (which adds more non atomic RMWs), or I can work more on
> that putting this aside?

For the zeroing optimization I think we'd restart with a proper discussion on
the overall requirements + design (e.g., options to communicate information,
options to do the zeroing, possible simplifications), as there are differing
opinions on the general approach ... and my gut feeling is that this needs a
proper discussion, as it involves folio allocation and severe core page
allocator changes.

.. and as GFP_ZERO handling is a mess and we'd need a bigger discussion on how
to sort that out.

That is, I'd want to see a discussion thread before sending more code to the
list (and ending in a sashiko review + resend spiral).

Of course, you can link to code, but it will make all our life easier to explore
the design space first before getting distracted by implementation details.

-- 
Cheers,

David