On Tue, 2019-10-08 at 17:43 +0000, Christopher Lameter wrote:
> On Tue, 8 Oct 2019, Leonardo Bras wrote:
> 
> > > You are creating contention on a single exclusive cacheline. Doesnt this
> > > defeat the whole purpose of the lockless page table walk? Use mmap_sem or
> > > so should cause the same performance regression?
> > 
> > Sorry, I did not understand that question.
> > I mean, this is just a refcount and never causes a lock.
> 
> Locks also use atomic operations like a refcount increment. Both require
> the cacheline to be in exclusive state. So the impact is very similar.

Thanks for explaining. :)

So you say that the performance impact of using my approach is the same
as using locks? (supposing that lock never waits)

So, there are 'lockless pagetable walks' only for the sake of better
performance? 

I thought they existed to enable doing pagetable walks in states where
locking was not safe.

Is that right?

Thanks!
Leonardo Brás,