Re: [PATCH v2 05/19] staging: iio: resolver: ad2s1210: fix probe

[Jonathan Cameron <jic23@kernel.org>]

On Thu, 21 Sep 2023 09:43:46 -0500
David Lechner <dlechner@baylibre.com> wrote:

> This fixes a number of issues in the ad2s1210_probe() function:
> - Move call to ad2s1210_setup_gpios() after `st->sdev = spi;`. This
>   fixes use of this pointer before it is initialized.
> - Check return value on ad2s1210_initial().
Should mention moving it as well.
> - Move devm_iio_device_register() to the end to avoid race of
>   registering before fully initialized.
> - Remove call to spi_setup(). Note: MODE_3 was incorrect, it should be
>   MODE_1 but we can let the device tree select this.
> - Change default value for fclkin. This is an external oscillator, not
>   the SPI bus clock. (Will use device tree to get the correct value
>   in a future patch. For now, using the eval board value.)
> - Remove spi_set_drvdata().

Hmm. This is a lot of different things. I'd prefer it more split up
as a few of these are not completely trivial.

> 
> Signed-off-by: David Lechner <dlechner@baylibre.com>

I think the patch split up broke on this one...

> ---
>  drivers/staging/iio/resolver/ad2s1210.c | 30 ++++++++++++-------------
>  1 file changed, 15 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/staging/iio/resolver/ad2s1210.c b/drivers/staging/iio/resolver/ad2s1210.c
> index 0bdd5a30d45d..9c7f76114360 100644
> --- a/drivers/staging/iio/resolver/ad2s1210.c
> +++ b/drivers/staging/iio/resolver/ad2s1210.c
> @@ -3,6 +3,7 @@
>   * ad2s1210.c support for the ADI Resolver to Digital Converters: AD2S1210
>   *
>   * Copyright (c) 2010-2010 Analog Devices Inc.
> + * Copyright (C) 2023 BayLibre, SAS

Bit early to justify that, but I'm fine with it anyway as it will soon
be justified!

>   */
>  #include <linux/types.h>
>  #include <linux/mutex.h>
> @@ -657,12 +658,8 @@ static int ad2s1210_probe(struct spi_device *spi)
>  	indio_dev = devm_iio_device_alloc(&spi->dev, sizeof(*st));
>  	if (!indio_dev)
>  		return -ENOMEM;
> -	st = iio_priv(indio_dev);
> -	ret = ad2s1210_setup_gpios(st);
> -	if (ret < 0)
> -		return ret;
>  
> -	spi_set_drvdata(spi, indio_dev);

Looks fine, but unconnected to the rest of this patch.

> +	st = iio_priv(indio_dev);
>  
>  	mutex_init(&st->lock);
>  	st->sdev = spi;
> @@ -671,22 +668,25 @@ static int ad2s1210_probe(struct spi_device *spi)
>  	st->resolution = 12;
>  	st->fexcit = AD2S1210_DEF_EXCIT;
>  
> +	ret = ad2s1210_setup_clocks(st);

doesn't exist yet.

> +	if (ret < 0)
> +		return ret;
> +
> +	ret = ad2s1210_setup_gpios(st);

> +	if (ret < 0)
> +		return ret;
> +
> +	ret = ad2s1210_initial(st);
> +	if (ret < 0)
> +		return ret;
> +
>  	indio_dev->info = &ad2s1210_info;
>  	indio_dev->modes = INDIO_DIRECT_MODE;
>  	indio_dev->channels = ad2s1210_channels;
>  	indio_dev->num_channels = ARRAY_SIZE(ad2s1210_channels);
>  	indio_dev->name = spi_get_device_id(spi)->name;
>  
> -	ret = devm_iio_device_register(&spi->dev, indio_dev);
> -	if (ret)
> -		return ret;
> -
> -	st->fclkin = spi->max_speed_hz;
> -	spi->mode = SPI_MODE_3;
> -	spi_setup(spi);
> -	ad2s1210_initial(st);
> -
> -	return 0;
> +	return devm_iio_device_register(&spi->dev, indio_dev);
>  }
>  
>  static const struct of_device_id ad2s1210_of_match[] = {