Re: [PATCH v2 09/10] sched/psi: per-cgroup PSI stats disable/re-enable interface

[Johannes Weiner]

On Wed, Aug 10, 2022 at 09:30:59AM +0800, Chengming Zhou wrote:
> On 2022/8/10 08:39, Chengming Zhou wrote:
> > On 2022/8/10 01:48, Tejun Heo wrote:
> >> Hello,
> >>
> >> On Mon, Aug 08, 2022 at 07:03:40PM +0800, Chengming Zhou wrote:
> >>> So this patch introduce a per-cgroup PSI stats disable/re-enable
> >>> interface "cgroup.psi", which is a read-write single value file that
> >>> allowed values are "0" and "1", the defaults is "1" so per-cgroup
> >>> PSI stats is enabled by default.
> >>
> >> Given that the knobs are named {cpu|memory|io}.pressure, I wonder whether
> >> "cgroup.psi" is the best name. Also, it doesn't convey that it's the
> >> enable/disable knob. I think it needs a better name.
> > 
> > Yes, "cgroup.psi" is not good. What abort "pressure.enable" or "cgroup.psi_enable"?
> 
> Doesn't look good either, what do you think of "cgroup.pressure.enable"?

How about just cgroup.pressure? Too ambiguous?

cgroup.pressure.enable sounds good to me too. Or, because it's
default-enabled and that likely won't change, cgroup.pressure.disable.