Re: [PATCH v4] PM: EM: Fix potential division-by-zero error in em_compute_costs()

[Yaxiong Tian]

在 2025/4/17 13:57, Lukasz Luba 写道:
> On 4/17/25 02:07, Yaxiong Tian wrote:
> > From: Yaxiong Tian <tianyaxiong@xxxxxxxxxx>
> >
> > When the device is of a non-CPU type, table[i].performance won't be
> > initialized in the previous em_init_performance(), resulting in division
> > by zero when calculating costs in em_compute_costs().
> >
> > Since the 'cost' algorithm is only used for EAS energy efficiency
> > calculations and is currently not utilized by other device drivers, we
> > should add the _is_cpu_device(dev) check to prevent this division-by-zero
> > issue.
> >
> > Fixes: 1b600da51073 ("PM: EM: Optimize em_cpu_energy() and remove 
> > division")
> > Signed-off-by: Yaxiong Tian <tianyaxiong@xxxxxxxxxx>
> > ---
> >   kernel/power/energy_model.c | 4 ++++
> >   1 file changed, 4 insertions(+)
> >
> > diff --git a/kernel/power/energy_model.c b/kernel/power/energy_model.c
> > index d9b7e2b38c7a..41606247c277 100644
> > --- a/kernel/power/energy_model.c
> > +++ b/kernel/power/energy_model.c
> > @@ -233,6 +233,10 @@ static int em_compute_costs(struct device *dev, 
> > struct em_perf_state *table,
> >       unsigned long prev_cost = ULONG_MAX;
> >       int i, ret;
> > +    /* This is needed only for CPUs and EAS skip other devices */
> > +    if (!_is_cpu_device(dev))
> > +        return 0;
> > +
> >       /* Compute the cost of each performance state. */
> >       for (i = nr_states - 1; i >= 0; i--) {
> >           unsigned long power_res, cost;
>
>
> Please stop for a while. I have to check what happened that you
> faced the issue in the first place. I have been testing the GPU
> EMs and there was no issues...
>
> Let me debug that today.

Of course. Since I don't have actual hardware, I can only logically
deduce that this issue might exist.