Re: [PATCH v2 1/2] powercap: dtpm_cpu: Guard em_cpu_get() against NULL return in sysfs callbacks

[Daniel Lezcano]

On 6/24/26 22:31, sivany32@xxxxxxxxx wrote:
> From: Sivan Zohar-Kotzer <sivany32@xxxxxxxxx>
>
> em_cpu_get() can return NULL when a CPU becomes impossible. Two call
> sites miss the NULL test:
>
> 1. set_pd_power_limit() — reachable from sysfs via the powercap
>     constraint power_limit_uw store path. If the CPU has become
>     impossible then pd will be NULL, return the current power limit
>     unchanged (no-op) instead of crashing.

Can you provide a script or a recipe spotting these issues ?

> 2. update_pd_power_uw() — called from CPU hotplug handlers via
>     dtpm_update_power(). While reaching this with a NULL em may be
>     technically impossible today, the check is kept for defense in
>     depth against future refactoring or subtle races.
>
> The other sysfs-reachable path, get_pd_power_uw(), already has a NULL
> guard returning 0.
>
> Fixes: 0e8f68d7f048 ("powercap/drivers/dtpm: Add CPU energy model based support")
> Signed-off-by: Sivan Zohar-Kotzer <sivany32@xxxxxxxxx>
> Co-developed-by: Elazar Leibovich <elazarl@xxxxxxxxx>
> Signed-off-by: Elazar Leibovich <elazarl@xxxxxxxxx>
> ---
>   drivers/powercap/dtpm_cpu.c | 6 ++++++
>   1 file changed, 6 insertions(+)
>
> diff --git a/drivers/powercap/dtpm_cpu.c b/drivers/powercap/dtpm_cpu.c
> index 21355db64..886cbe922 100644
> --- a/drivers/powercap/dtpm_cpu.c
> +++ b/drivers/powercap/dtpm_cpu.c
> @@ -47,6 +47,9 @@ static u64 set_pd_power_limit(struct dtpm *dtpm, u64 power_limit)
>   	u64 power;
>   	int i, nr_cpus;
>   
> +	if (!pd)
> +		return dtpm->power_limit;
> +
>   	nr_cpus = cpumask_weight_and(cpu_online_mask, to_cpumask(pd->cpus));
>   
>   	rcu_read_lock();
> @@ -125,6 +128,9 @@ static int update_pd_power_uw(struct dtpm *dtpm)
>   	struct em_perf_state *table;
>   	int nr_cpus;
>   
> +	if (!em)
> +		return -EINVAL;
> +
>   	nr_cpus = cpumask_weight_and(cpu_online_mask, to_cpumask(em->cpus));
>   
>   	rcu_read_lock();