Re: [PATCH v2 0/3] vmalloc enhancements

[Matthew Wilcox]

On Tue, Feb 12, 2019 at 10:36:12PM +0000, Roman Gushchin wrote:
> On Tue, Feb 12, 2019 at 12:34:09PM -0800, Andrew Morton wrote:
> > On Tue, 12 Feb 2019 13:47:24 -0500 Johannes Weiner <hannes@xxxxxxxxxxx> wrote:
> > > I don't understand what prompted this change to percpu counters.
> 
> I *think*, I see some performance difference, but it's barely measurable
> in my setup. Also as I remember, Matthew was asking why not percpu here.
> So if everybody prefers a global atomic, I'm fine with either.

I was asking why you were using an accessor instead of a direct reference
to the atomic_long_t.