Re: [PATCH RFC v2 2/6] gpio: aggregator: refactor the forwarder part.

[Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>]

On Mon, Mar 17, 2025 at 04:38:00PM +0100, Thomas Richard wrote:
> Prepare the code to create a gpio-fwd library. This library will allow to
> create and register a gpiochip forwarder.

...

>  struct gpiochip_fwd {
> +	struct device *dev;
>  	struct gpio_chip chip;

Have you checked the code generation?
Also, is this new pointer the same as chip.parent?

>  	struct gpio_desc **descs;
>  	union {

>  };

...

> +static struct gpiochip_fwd *devm_gpiochip_fwd_alloc(struct device *dev,
> +						    unsigned int ngpios)

I would rather split as

static struct gpiochip_fwd *
devm_gpiochip_fwd_alloc(struct device *dev, unsigned int ngpios)

...

> +	fwd->descs = devm_kcalloc(dev, ngpios, sizeof(*fwd->descs),
> +				  GFP_KERNEL);

One line.

...

> +static int gpiochip_fwd_register(struct gpiochip_fwd *fwd)
> +{
> +	struct gpio_chip *chip = &fwd->chip;
> +	struct device *dev = fwd->dev;
> +	int error;
>  
>  	if (chip->can_sleep)
>  		mutex_init(&fwd->mlock);
>  	else
>  		spin_lock_init(&fwd->slock);
>  
> +	error = devm_gpiochip_add_data(dev, chip, fwd);
> +
> +	return error;

	return devm_...

> +}

...

Overall it looks and feels like this can be split to more simpler logically
isolated changes. At least I see that folding function parameters can be a
separate patch.

-- 
With Best Regards,
Andy Shevchenko