Re: [PATCH V5 8/9] null_blk: serialize configfs attribute stores with device setup

On 7/13/26 9:18 PM, Zizhi Wo wrote:
>   	ret = nullb_device_##TYPE##_attr_store(&new_value, page, count);\
>   	if (ret < 0)							\
>   		return ret;						\
> +	ret = 0;							\
> +	mutex_lock(&lock);						\
>   	if (apply_fn)							\
>   		ret = apply_fn(dev, new_value);				\
>   	else if (test_bit(NULLB_DEV_FL_CONFIGURED, &dev->flags)) 	\
>   		ret = -EBUSY;						\
> +	if (!ret)							\
> +		dev->NAME = new_value;					\
> +	mutex_unlock(&lock);						\
>   	if (ret < 0)							\
>   		return ret;						\
> -	dev->NAME = new_value;						\
>   	return count;							\
>   }									\

If guard()() would be used instead of explicit mutex_lock() and
mutex_unlock() calls, the if (ret) test wouldn't have to be duplicated,
isn't it?

Additionally, the above changes modify the behavior if apply_fn()
returns a strictly positive value (> 0).

Bart.