Re: [PATCH] thermal/core: Fix lockdep_assert() warning

[Marek Szyprowski]

On 12.08.2022 15:12, Daniel Lezcano wrote:
> The function thermal_zone_device_is_enabled() must be called with the
> thermal zone lock held. In the resume path, it is called without.
>
> As the thermal_zone_device_is_enabled() is also checked in
> thermal_zone_device_update(), do the check in resume() function is
> pointless, except for saving an extra initialization which does not
> hurt if it is done in all the cases.
>
> Signed-off-by: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>

This fixes the warning I've reported. Feel free to add:

Reported-by: Marek Szyprowski <m.szyprowski@xxxxxxxxxxx>

Tested-by: Marek Szyprowski <m.szyprowski@xxxxxxxxxxx>

> ---
>   drivers/thermal/thermal_core.c | 3 ---
>   1 file changed, 3 deletions(-)
>
> diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
> index 50814009339d..dc8ff6a84df1 100644
> --- a/drivers/thermal/thermal_core.c
> +++ b/drivers/thermal/thermal_core.c
> @@ -1438,9 +1438,6 @@ static int thermal_pm_notify(struct notifier_block *nb,
>   	case PM_POST_SUSPEND:
>   		atomic_set(&in_suspend, 0);
>   		list_for_each_entry(tz, &thermal_tz_list, node) {
> -			if (!thermal_zone_device_is_enabled(tz))
> -				continue;
> -
>   			thermal_zone_device_init(tz);
>   			thermal_zone_device_update(tz,
>   						   THERMAL_EVENT_UNSPECIFIED);

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland