Re: [PATCH] mm/damon: align MMU notifier range when clearing PMD young bit

[SJ Park]

On Fri, 23 Oct 2026 13:01:31 +0800 Aveline Noir <jm5905938@xxxxxxxxx> wrote:

> Hi SJ,
> 
> On Thu, 22 Oct 2026 17:20:17 -0700 SJ Park <sj@xxxxxxxxxx> wrote:
> > I Cc-ed them.  Please Cc all
> > recipients who get_maitnainer.pl is recommending, from the next time.
> 
> Sorry for that. Maybe I will use b4 next time :)
> 
> > However, I think this patch explains its context very well, and therefore
> > this link is not needed.
> 
> Sure , That make sense.
> 
> > Were you also able to confirm the expected wrong behavior (missing part
> > of the sampled huge page and age secondary mapping), too?
> 
> Yes. A test with nested x86-64 KVM confirmed both effects. With
> 4 KiB EPT mappings, a midpoint PMD sample left the first 1 MiB young and
> aged the next THP's first 1 MiB. With 2 MiB EPT mappings, both whole THPs
> were aged. After the fix, only the sampled THP was aged in both cases.

Thank you for confirming!


Thanks,
SJ

[...]