Re: [PATCH v4 1/4] mm: reduce the rcu lock duration

[Andrew Morton]

On Tue, 7 Jun 2022 17:19:53 +0800 Miaohe Lin <linmiaohe@xxxxxxxxxx> wrote:

> > 
> > 
> > If the checks are not made to guarantee they are all checking against
> > the same mm, the code needs a comment to say that there is a tiny race.
> > The comment should say we don't care about the tiny race because
> > the worst that can happen is that a page is migrated to a different
> > numa node.  And so we don't care.
> > 
> > 
> 
> I tend to do this one. To make sure, is the below code change what you suggest?

Eric went quiet.

As I understand it, any changes arising from this discussion can be
done as a followup patch.  So I'm planning on moving this 4-patch
series into the non-rebasing mm-stable branch late next week.