On Thu, 2019-10-03 at 13:49 -0700, John Hubbard wrote:
> Yes. And to clarify, I was assuming that the changes to mm/gup.c were 
> required in order to accomplish your goals. Given that assumption, I 
> wanted the generic code to be "proper", and that's what that feedback
> is about.

You assumed right. On my counting approach it's necessary count all
'lockless pagetable walks', including the ones in generic code.

And, I think even without the counting approach, it was a good way
focus this 'lockless pagetable walk' routine in a single place.

> 
> Although the other questions about file-backed THP
> make it sound like some rethinking across the board is required now.

Yeap, I need to better understand how the file-backed THP problem
works.

Thanks,

Leonardo Brás