| Date | Fri, 28 Feb 2014 02:18:58 +0200 | From | "Kirill A. Shutemov" <> | Subject | Re: [PATCHv3 1/2] mm: introduce vm_ops->map_pages() |
| |
On Thu, Feb 27, 2014 at 02:34:55PM -0800, Dave Hansen wrote: > Kirill's git test suite runs did show that it _can_ hurt in some cases.
And see last use-case for how much it can hurt. :)
It shouldn't differs much for the same *number* of pages between [u]archs unless setup of the pte is significantly more expensive or page fault is faster.
But of course, I can move FAULT_AROUND_PAGES to arch/x86/ and let architecture mantainers to decide if they want the feature. ;)
-- Kirill A. Shutemov
|