Re: wakeup_affine_weight() is b0rked - was Re: [PATCH 2/2] sched/fair: Scale wakeup granularity relative to nr_running

[Mike Galbraith]

On Sun, 2021-10-03 at 20:34 +1300, Barry Song wrote:
>
> I am wondering if this should be the responsibility of wake_wide()?

Those event threads we stacked so high (which are kde minions btw),
don't generally accrue _any_ wakee_flips, so when X wakes a slew of the
things, wake_wide()'s heuristic rejects the lot.

So yeah, the blame game for this issue is a target rich environment.
Shoot either of 'em (or both), and you'll hit the bad guy.

	-Mike