Re: [PATCH 4/4] thp: rewrite freeze_page()/unfreeze_page() with generic rmap walkers

[Kirill A. Shutemov]

On Wed, Feb 03, 2016 at 02:43:16PM -0800, Andrew Morton wrote:
> On Wed, 3 Feb 2016 07:42:01 -0800 Dave Hansen <dave.hansen@xxxxxxxxx> wrote:
> 
> > On 02/03/2016 07:14 AM, Kirill A. Shutemov wrote:
> > > But the new variant is somewhat slower. Current helpers iterates over
> > > VMAs the compound page is mapped to, and then over ptes within this VMA.
> > > New helpers iterates over small page, then over VMA the small page
> > > mapped to, and only then find relevant pte.
> > 
> > The code simplification here is really attractive.  Can you quantify
> > what the slowdown is?  Is it noticeable, or would it be in the noise
> > during all the other stuff that happens under memory pressure?
> 
> yup.

It is not really clear, how to quantify this properly. Let me think more
about it.

> And the "more testing is required" is a bit worrisome.  Is this
> code really ready for getting pounded upon in -next?

By now it survived 5+ hours of fuzzing in 16 VM in parallel. I'll continue
with other tests tomorrow.

-- 
 Kirill A. Shutemov