Re: [PATCH v3 3/7] iio: adc: add RZ/T2H / RZ/N2H ADC driver

[Jonathan Cameron <jic23@kernel.org>]

On Wed,  1 Oct 2025 15:23:10 +0300
Cosmin Tanislav <cosmin-gabriel.tanislav.xa@renesas.com> wrote:

> Add support for the A/D 12-Bit successive approximation converters found
> in the Renesas RZ/T2H (R9A09G077) and RZ/N2H (R9A09G087) SoCs.
> 
> RZ/T2H has two ADCs with 4 channels and one with 6.
> RZ/N2H has two ADCs with 4 channels and one with 15.
> 
> Conversions can be performed in single or continuous mode. Result of the
> conversion is stored in a 16-bit data register corresponding to each
> channel.
> 
> The conversions can be started by a software trigger, a synchronous
> trigger (from MTU or from ELC) or an asynchronous external trigger (from
> ADTRGn# pin).
> 
> Only single mode with software trigger is supported for now.
> 
> Signed-off-by: Cosmin Tanislav <cosmin-gabriel.tanislav.xa@renesas.com>
> Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Hi Cosmin

A few really minor comments.

I might have ignored these or tweaked whilst applying but looks
like you'll probably be doing a v4 anyway so I thought I'd mention them.

Thanks,

Jonathan


> diff --git a/drivers/iio/adc/rzt2h_adc.c b/drivers/iio/adc/rzt2h_adc.c
> new file mode 100644
> index 000000000000..51d1852d814d
> --- /dev/null
> +++ b/drivers/iio/adc/rzt2h_adc.c
> @@ -0,0 +1,309 @@
> +// SPDX-License-Identifier: GPL-2.0
> +
> +#include <linux/bitfield.h>
> +#include <linux/cleanup.h>
> +#include <linux/completion.h>
> +#include <linux/delay.h>
> +#include <linux/iio/adc-helpers.h>
> +#include <linux/iio/iio.h>
> +#include <linux/interrupt.h>
> +#include <linux/io.h>
> +#include <linux/iopoll.h>
> +#include <linux/mod_devicetable.h>
> +#include <linux/module.h>
> +#include <linux/platform_device.h>
> +#include <linux/pm_runtime.h>
> +#include <linux/property.h>
> +
> +#define RZT2H_NAME			"rzt2h-adc"

I'm not a huge fan of defines like this as there is no inherent reason the two
places it is used should take the same value. If you feel it's really justified
then I don't mind too much though.

> +
> +static int rzt2h_adc_read_raw(struct iio_dev *indio_dev,
> +			      struct iio_chan_spec const *chan,
> +			      int *val, int *val2, long mask)
> +{
> +	struct rzt2h_adc *adc = iio_priv(indio_dev);
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_RAW:
> +		return rzt2h_adc_read_single(adc, chan->channel, val);
> +	case IIO_CHAN_INFO_SCALE:
> +		*val = RZT2H_ADC_VREF_MV;
> +		*val2 = RZT2H_ADC_RESOLUTION;

Trivial but I'd just put the two values here and get rid of the defines.
I don't see an advantage for defines when they aren't 'magic' numbers
but are real values.

> +		return IIO_VAL_FRACTIONAL_LOG2;
> +	default:
> +		return -EINVAL;
> +	}
> +}