Re: [PATCHv3 2/4] zsmalloc: fine-grained inuse ratio based fullness grouping

[Andrew Morton]

On Fri, 3 Mar 2023 08:58:55 -0800 Yosry Ahmed <yosryahmed@xxxxxxxxxx> wrote:

> > > +               for (fg = ZS_INUSE_RATIO_70; fg <= ZS_INUSE_RATIO_99; fg++)
> > > +                       class_almost_full = zs_stat_get(class, fg);
> > > +               for (fg = ZS_INUSE_RATIO_10; fg < ZS_INUSE_RATIO_70; fg++)
> > > +                       class_almost_empty = zs_stat_get(class, fg);
> >
> > Aren't these supposed to be += ?
> 
> Looks like they are removed by a following patch anyway.

I'll fix that up anyway, to avoid a bisection hole.  And in case [patch 4/4]
gets dropped from the series or is later reverted or whatever.