Re: [PATCH mm-hotfixes v2 2/2] mm/huge_memory: separate out CONFIG_PERSISTENT_HUGE_ZERO_FOLIO logic

On Wed, Aug 12, 2026 at 10:05:55AM +0200, David Hildenbrand (Arm) wrote:
> On 7/30/26 12:55, Lorenzo Stoakes (ARM) wrote:
> > Rather than mixing the refcounted and non-refcounted
> > CONFIG_PERSISTENT_HUGE_ZERO_FOLIO logic, separate the two out cleanly
> > so it is clear what happens when this configuration option is set and what
> > happens when it is not.
> >
> > Introduce HUGE_ZERO_UNSET_PFN to abstract the ~0UL assignment, only
> > introduce the refcount, lock and shrinker if
> > !CONFIG_PERSISTENT_HUGE_ZERO_FOLIO, abstract initialisation and teardown,
> > abstract the huge zero folio allocation from refcounting.
> >
> > Also change a BUG_ON() to WARN_ON_ONCE() while we're at it.
> >
> > No functional change intended.
>
> Ugh. Why is this patch already upsteam with
>
> Fixes: 3b77e8c8cde5 ("mm/thp: make is_huge_zero_pmd() safe and quicker")
> Cc: <[email protected]>

Yeah these tags weren't part of it + it was intended as a separate thing but I
guess it got lost in translation.

I should probably have just sent it entirely separately after the fix landed so
my bad on that part.

>
> It's absolutely not a fix. It's a cleanup that should have been reviewed properly.
>
> I really wanted to review this.
>
> Making David grumpy in the morning, good job everyone.

Sorry about that :) At least it's not schlop? ;))

>
> --
> Cheers,
>
> David

--
Cheers, Lorenzo