Re: [PATCH v1 2/5] PM: sleep: Remove unnecessary mutex lock when waiting on parent

[Rafael J. Wysocki]

Sorry for the delay.

On Thu, Nov 14, 2024 at 11:09 PM Saravana Kannan <saravanak@xxxxxxxxxx> wrote:
>
> Locking is not needed to do get_device(dev->parent). We either get a NULL
> (if the parent was cleared) or the actual parent. Also, when a device is
> deleted (device_del()) and removed from the dpm_list, its completion
> variable is also complete_all()-ed. So, we don't have to worry about
> waiting indefinitely on a deleted parent device.

The device_pm_initialized(dev) check before get_device(dev->parent)
doesn't make sense without the locking and that's the whole point of
it.

> Signed-off-by: Saravana Kannan <saravanak@xxxxxxxxxx>
> ---
>  drivers/base/power/main.c | 13 ++-----------
>  1 file changed, 2 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
> index 86e51b9fefab..9b9b6088e56a 100644
> --- a/drivers/base/power/main.c
> +++ b/drivers/base/power/main.c
> @@ -284,18 +284,9 @@ static bool dpm_wait_for_superior(struct device *dev, bool async)
>          * counting the parent once more unless the device has been deleted
>          * already (in which case return right away).
>          */
> -       mutex_lock(&dpm_list_mtx);
> -
> -       if (!device_pm_initialized(dev)) {
> -               mutex_unlock(&dpm_list_mtx);
> -               return false;
> -       }
> -
>         parent = get_device(dev->parent);
> -
> -       mutex_unlock(&dpm_list_mtx);
> -
> -       dpm_wait(parent, async);
> +       if (device_pm_initialized(dev))
> +               dpm_wait(parent, async);

This is racy, so what's the point?

You can just do

parent = get_device(dev->parent);
dpm_wait(parent, async);

and please update the comment above this.

>         put_device(parent);
>
>         dpm_wait_for_suppliers(dev, async);
> --