Linus,
I meant to add, in the pull request I just sent, that we are looking
into the known unfixed regression listed by Adrian Bunk with the
subject "powerpc64: performance monitor exception". The problem is
that I am not convinced the proposed patch is safe, and in fact I
think it introduces a potential deadlock on Cell at least, but we
haven't worked out an alternative solution yet. I hope to have a good
solution within the next couple of days.
Paul.