Re: [PATCH 0/2] Fix migration races in rmap_walk()

[Andrea Arcangeli <aarcange@redhat.com>]

On Mon, Apr 26, 2010 at 11:37:56PM +0100, Mel Gorman wrote:
> The other issues raised about expand_downwards will need to be re-examined to
> see if they still exist and transparent hugepage support will need further
> thinking to see if split_huge_page() can deal with these situations.

So patch 1 is for aa.git too, and patch 2 is only for mainline with
the new anon-vma changes (patch 2 not needed in current aa.git, and if
I apply it, it'll deadlock so...) right?

split_huge_page is somewhat simpler and more strict in its checking
than migrate.c in this respect, and yes patch 2 will also need to be
extended to cover split_huge_page the moment I stop backing out the
new anon-vma code (but it won't be any different, whatever works for
migrate will also work for split_huge_page later).

For now I'm much more interested in patch 1 and I'll leave patch 2 to
mainline digestion and check it later hope to find all issues fixed by
the time transparent hugepage gets merged.

About patch 1 it's very interesting because I looked at the race
against fork and migrate yesterday and I didn't see issues but I'm
going to read your patch 1 in detail now to understand what is the
problem you're fixing.

Good you posted this fast, so I can try to help ;)
Andrea