Re: [PATCH 2/8] cpuidle: add cpuidle_unregister_driver() error check

[Andrew Morton <akpm@linux-foundation.org>]

On Wed, 26 May 2010 22:42:25 -0400 Len Brown <lenb@kernel.org> wrote:

> From: Len Brown <len.brown@intel.com>
> 
> When cpuidle_unregister_driver() is called with a driver
> other than the one that was successfully registered, do nothing.
> 
> Previously we'd NULL-out the one that was registered.
> But that required the callers to remember what this
> routine already remembers.  With this check, the callers
> can be simplified.
> 
> Signed-off-by: Len Brown <len.brown@intel.com>
> ---
>  drivers/cpuidle/driver.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/cpuidle/driver.c b/drivers/cpuidle/driver.c
> index 2257004..30bcd44 100644
> --- a/drivers/cpuidle/driver.c
> +++ b/drivers/cpuidle/driver.c
> @@ -49,7 +49,8 @@ void cpuidle_unregister_driver(struct cpuidle_driver *drv)
>  		return;
>  
>  	spin_lock(&cpuidle_driver_lock);
> -	cpuidle_curr_driver = NULL;
> +	if (drv == cpuidle_curr_driver)
> +		cpuidle_curr_driver = NULL;
>  	spin_unlock(&cpuidle_driver_lock);
>  }

This can only happen as a result of a coding bug, yes?  If so, the
kernel should go BUG() rather than secretly concealing the problem.

Also (alternatively), the `drv' arg to this function is superfluous?