Re: [PATCH v3 14/20] sched/debug: Provide debugfs to enable/disable steal monitor

[Shrikanth Hegde]

On 6/8/26 11:58 PM, Ilya Leoshkevich wrote:
> On 5/14/26 17:21, Shrikanth Hegde wrote:
> > Add a debugfs "enable" file to enable steal time monitor.
> >
> > Computing steal time and acting on it periodically are to be opted by
> > the user. This helps to avoid any overhead when the feature
> > is disabled.
> >
> > It is disabled by default.
> >
> > Signed-off-by: Shrikanth Hegde <sshegde@xxxxxxxxxxxxx>
> > ---
> >   kernel/sched/core.c  |  1 +
> >   kernel/sched/debug.c | 29 +++++++++++++++++++++++++++++
> >   kernel/sched/sched.h |  2 ++
> >   3 files changed, 32 insertions(+)
> >
> > diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> > index 0d1995c65ce6..1533a44d1a6b 100644
> > --- a/kernel/sched/core.c
> > +++ b/kernel/sched/core.c
>
> [...]
>
> > @@ -592,6 +592,33 @@ static void debugfs_ext_server_init(void)
> >   #endif /* CONFIG_SCHED_CLASS_EXT */
> >   #ifdef CONFIG_PREFERRED_CPU
> > +__read_mostly bool sched_sm_wr_enable;
> > +
> > +static ssize_t sched_sm_en_write(struct file *filp, const char __user 
> > *ubuf,
> > +                 size_t cnt, loff_t *ppos)
> > +{
> > +    bool orig = sched_sm_wr_enable;
> > +    ssize_t result;
> > +
> > +    result = debugfs_write_file_bool(filp, ubuf, cnt, ppos);
> > +
> > +    if (sched_sm_wr_enable && !orig) {
> > +        static_branch_enable(&__sched_sm_enable);
>
> Would it make sense to clear the steal monitor state when enabling it?
> Otherwise we may immediately trigger a bogus "dec" if user did a disable
> during contention and re-enabled when everything is quiet.

Yes, it may lead to one wrong direction decision but for the case when disable
was during contention and re-enabled again during contention.

I think setting previous_decision = 0 during disable to should fix it. So previous
direction of inc/dec won't carry over the enable.