Re: [RFC PATCH] sched: fix the nonsense shares when load of cfs_rq is too, small

[Vincent Guittot]

On Wed, 4 Mar 2020 at 10:52, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> On Wed, Mar 04, 2020 at 09:47:34AM +0100, Vincent Guittot wrote:
> > you will add +1 of nice prio for each device
> >
> > should we use instead
> > # define scale_load_down(w) ((w >> SCHED_FIXEDPOINT_SHIFT) ? (w >>
> > SCHED_FIXEDPOINT_SHIFT) : MIN_SHARES)
>
> That's '((w >> SHIFT) ?: MIN_SHARES)', but even that is not quite right.
>
> I think we want something like:
>
> #define scale_load_down(w) \
> ({ unsigned long ___w = (w); \
>    if (___w) \
>      ____w = max(MIN_SHARES, ___w >> SHIFT); \
>    ___w; })
>
> That is, we very much want to retain 0 I'm thinking.

yes, you're right