Re: hugetlb: preserve hugetlb pte dirty state

["Ken Chen" <kenchen@google.com>]

On 2/6/07, Andrew Morton <akpm@linux-foundation.org> wrote:
> > yeah, I wonder why I didn't do that :-P  Especially after I asked you
> > a similar question the other day.  I will redo it and retest.
>
> No rush - the code should work OK as-is and given the total catastrophe
> which that change caused in core mm, making the same change to hugepages
> should be done with some care and thought and maintainer-poking.

OK.  That's comforting.  I was thinking along the same line that
hugetlb should be in-sync with core mm code and should continue to do
the same thing like core mm in the unmap path (to transfer dirty bit
into page struct); even if we dirty a page in the fault path and that
dirty state will be on for the life of the page.  I will play with the
fault path for a while then.

- Ken

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>