Re: [PATCH 06/36] cpuidle: move to use bus_get_dev_root()

[Rafael J. Wysocki]

On Mon, Mar 13, 2023 at 7:30 PM Greg Kroah-Hartman
<gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
>
> Direct access to the struct bus_type dev_root pointer is going away soon
> so replace that with a call to bus_get_dev_root() instead, which is what
> it is there for.
>
> This allows us to clean up the cpuidle_add_interface() call a bit as it
> was only called in one place, with the same argument so just put that
> into the function itself.  Note that cpuidle_remove_interface() should
> also probably be removed in the future as there are no callers of it for
> some reason.
>
> Cc: "Rafael J. Wysocki" <rafael@xxxxxxxxxx>
> Cc: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>
> Cc: linux-pm@xxxxxxxxxxxxxxx
> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> ---
> Note, this is a patch that is a prepatory cleanup as part of a larger
> series of patches that is working on resolving some old driver core
> design mistakes.  It will build and apply cleanly on top of 6.3-rc2 on
> its own, but I'd prefer if I could take it through my driver-core tree
> so that the driver core changes can be taken through there for 6.4-rc1.
>
>  drivers/cpuidle/cpuidle.c |  2 +-
>  drivers/cpuidle/cpuidle.h |  2 +-
>  drivers/cpuidle/sysfs.c   | 12 +++++++++---
>  3 files changed, 11 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/cpuidle/cpuidle.c b/drivers/cpuidle/cpuidle.c
> index 0b00f21cefe3..8e929f6602ce 100644
> --- a/drivers/cpuidle/cpuidle.c
> +++ b/drivers/cpuidle/cpuidle.c
> @@ -808,7 +808,7 @@ static int __init cpuidle_init(void)
>         if (cpuidle_disabled())
>                 return -ENODEV;
>
> -       return cpuidle_add_interface(cpu_subsys.dev_root);
> +       return cpuidle_add_interface();
>  }
>
>  module_param(off, int, 0444);
> diff --git a/drivers/cpuidle/cpuidle.h b/drivers/cpuidle/cpuidle.h
> index 9f336af17fa6..52701d9588f1 100644
> --- a/drivers/cpuidle/cpuidle.h
> +++ b/drivers/cpuidle/cpuidle.h
> @@ -30,7 +30,7 @@ extern int cpuidle_switch_governor(struct cpuidle_governor *gov);
>
>  struct device;
>
> -extern int cpuidle_add_interface(struct device *dev);
> +extern int cpuidle_add_interface(void);
>  extern void cpuidle_remove_interface(struct device *dev);
>  extern int cpuidle_add_device_sysfs(struct cpuidle_device *device);
>  extern void cpuidle_remove_device_sysfs(struct cpuidle_device *device);
> diff --git a/drivers/cpuidle/sysfs.c b/drivers/cpuidle/sysfs.c
> index 48948b171749..84e4946f1072 100644
> --- a/drivers/cpuidle/sysfs.c
> +++ b/drivers/cpuidle/sysfs.c
> @@ -119,11 +119,17 @@ static struct attribute_group cpuidle_attr_group = {
>
>  /**
>   * cpuidle_add_interface - add CPU global sysfs attributes
> - * @dev: the target device
>   */
> -int cpuidle_add_interface(struct device *dev)
> +int cpuidle_add_interface(void)
>  {
> -       return sysfs_create_group(&dev->kobj, &cpuidle_attr_group);
> +       struct device *dev_root = bus_get_dev_root(&cpu_subsys);
> +       int retval = -EINVAL;
> +
> +       if (dev_root) {
> +               retval = sysfs_create_group(&dev_root->kobj, &cpuidle_attr_group);
> +               put_device(dev_root);
> +       }
> +       return retval;

I would prefer

  if (!dev_root)
          return -EINVAL;

  retval = sysfs_create_group(&dev_root->kobj, &cpuidle_attr_group);
  put_device(dev_root);
  return retval;

assuming that successful group creation will bump up the reference
counter of dev_root.