Re: [PATCH] leds: make sure led->trigger is valid before calling trigger->activate

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

On Thu, 19 Jun 2008 15:48:47 +0400
Dmitry Baryshkov <dbaryshkov@gmail.com> wrote:

> Signed-off-by: Dmitry Baryshkov <dbaryshkov@gmail.com>
> ---
>  drivers/leds/led-triggers.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/leds/led-triggers.c b/drivers/leds/led-triggers.c
> index 0f242b3..f910eaf 100644
> --- a/drivers/leds/led-triggers.c
> +++ b/drivers/leds/led-triggers.c
> @@ -111,16 +111,17 @@ void led_trigger_set(struct led_classdev *led_cdev, struct led_trigger *trigger)
>  			flags);
>  		if (led_cdev->trigger->deactivate)
>  			led_cdev->trigger->deactivate(led_cdev);
> +		led_cdev->trigger = NULL;
>  		led_set_brightness(led_cdev, LED_OFF);
>  	}
>  	if (trigger) {
>  		write_lock_irqsave(&trigger->leddev_list_lock, flags);
>  		list_add_tail(&led_cdev->trig_list, &trigger->led_cdevs);
>  		write_unlock_irqrestore(&trigger->leddev_list_lock, flags);
> +		led_cdev->trigger = trigger;
>  		if (trigger->activate)
>  			trigger->activate(led_cdev);
>  	}
> -	led_cdev->trigger = trigger;
>  }
>  EXPORT_SYMBOL_GPL(led_trigger_set);
>  

The changelog tells us what the patch does, but it doesn't tell us
why it does it.

This matters.  For a start, there is no way in which I (at least)
am able to determine whether this change is needed in 2.6.26 nor
in 2.6.25.x.