Re: [RFT v2 2/8] PM / Domains: Handle safely genpd_syscore_switch() call on non-genpd device

[Ulf Hansson]

On 12 June 2017 at 17:17, Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote:
> genpd_syscore_switch() had two problems:
> 1. It silently assumed that device, it is being called for, belongs to
>    generic power domain and used container_of() on its power domain
>    pointer.  Such assumption might not be true always.
>
> 2. It iterated over list of generic power domains without holding
>    gpd_list_lock mutex thus list could have been modified in the same
>    time.
>
> Usage of genpd_lookup_dev() solves both problems as it is safe a call
> for non-generic power domains and uses mutex when iterating.
>
> Reported-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx>
> Signed-off-by: Krzysztof Kozlowski <krzk@xxxxxxxxxx>
>
> ---
>
> Not tested on real hardware.

When tested you may add my:

Acked-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx>

> ---
>  drivers/base/power/domain.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c
> index 2e8d0f423507..983f086ab235 100644
> --- a/drivers/base/power/domain.c
> +++ b/drivers/base/power/domain.c
> @@ -1098,8 +1098,8 @@ static void genpd_syscore_switch(struct device *dev, bool suspend)
>  {
>         struct generic_pm_domain *genpd;
>
> -       genpd = dev_to_genpd(dev);
> -       if (!pm_genpd_present(genpd))
> +       genpd = genpd_lookup_dev(dev);
> +       if (!genpd)
>                 return;
>
>         if (suspend) {
> --
> 2.9.3
>