Re: [PATCH] mm: Speed up mremap on large regions

[Michal Hocko <mhocko@kernel.org>]

On Thu 11-10-18 11:17:19, Kirill A. Shutemov wrote:
[...]
> > The thing is its quite a lot of change, I wrote a coccinelle script to do it
> > tree wide, following is the diffstat:
> >  48 files changed, 91 insertions(+), 124 deletions(-)
> > 
> > Imagine then having to add the address argument back in the future in case
> > its ever needed. Is it really worth doing it?
> 
> This is the point. It will get us chance to consider if the optimization
> is still safe.
> 
> And it shouldn't be hard: [partially] revert the commit and get the address
> back into the interface.

I agree with Kirill. This will also remove quite a lot of pointless
code and make it more clear. It is impossible to see what is the address
good for and I couldn't really trace back to commit introducing it to
guess that either. So making sure nobody does anything with it is a good
pre-requisite to make further changes on top.

The chage itself is really interesting, I still have to digest it
completely to see there are no cornercases but from a quick glance it
looks reasonable.
-- 
Michal Hocko
SUSE Labs