On Wed, Feb 22, 2017 at 04:15:07PM +0100, Andrea Arcangeli wrote: > On Tue, Feb 21, 2017 at 04:25:45PM +0300, Kirill A. Shutemov wrote: > > I think more strict vma_is_anonymous() is a good thing. > > > > But I don't see a point introducing one more helper. Let's just make the > > existing helper work better. > > That would be simpler agreed. The point of having an "unsafe" faster > version was only for code running in page fault context where the > additional check is unnecessary.
Well, I don't think that the cost of additional check is significant here. And we can bring ->vm_ops a bit closer to ->vm_flags to avoid potential cache miss. -- Kirill A. Shutemov
