Re: [PATCH v3 10/11] video: backlight: lm3533_bl: Implement backlight_scale property

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Unprotected concurrent modification of hardware register and `bd->props.scale` causes state desynchronization.
--

commit 8adaee77dd01ba44b6fdadefc561f4ab4fed6761
Author: Svyatoslav Ryhel <clamor95@gmail.com>

video: backlight: lm3533_bl: Implement backlight_scale property

This commit implements the backlight_scale property to describe the linear
and non-linear modes supported by the device.

> diff --git a/drivers/video/backlight/lm3533_bl.c b/drivers/video/backlight/lm3533_bl.c
> index f0d88b7bc229d..945625c54fc43 100644
> --- a/drivers/video/backlight/lm3533_bl.c
> +++ b/drivers/video/backlight/lm3533_bl.c
> @@ -161,6 +161,7 @@ static ssize_t store_linear(struct device *dev,
>  					const char *buf, size_t len)
>  {
>  	struct lm3533_bl *bl = dev_get_drvdata(dev);
> +	struct backlight_device *bd = bl->bd;
>  	int id = lm3533_bl_get_ctrlbank_id(bl);
>  	unsigned long linear;
>  	int ret;
> @@ -174,6 +175,9 @@ static ssize_t store_linear(struct device *dev,
>  	if (ret)
>  		return ret;
>  
> +	bd->props.scale = linear ? BACKLIGHT_SCALE_LINEAR :
> +				   BACKLIGHT_SCALE_NON_LINEAR;

[Severity: Medium]
Is it possible for the hardware state and the software state in
bd->props.scale to become desynchronized here?

Because sysfs writes can execute concurrently, could an interleaved
execution of store_linear() by two different threads result in a race
condition where the states diverge?

For example:
Thread 1 writes linear=1
Thread 1 updates hardware via regmap_update_bits()

Thread 2 writes linear=0
Thread 2 updates hardware via regmap_update_bits()
Thread 2 updates bd->props.scale = BACKLIGHT_SCALE_NON_LINEAR

Thread 1 updates bd->props.scale = BACKLIGHT_SCALE_LINEAR

Would wrapping this update sequence in a lock ensure the hardware and
software states remain consistent?

> +
>  	return len;
>  }
>  

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260601151831.76350-1-clamor95@gmail.com?part=10