Re: [PATCH v3 3/3] iio: adc: ad7192: Add fast settling support

[Jonathan Cameron]

On Mon, 25 Sep 2023 00:51:48 +0300
alisadariana@xxxxxxxxx wrote:

> From: Alisa-Dariana Roman <alisa.roman@xxxxxxxxxx>
> 
> Add fast settling mode support for AD7193.
> 
> Add fast_settling_average_factor attribute. Expose to user the current
> value of the fast settling average factor. User can change the value by
> writing a new one.
> 
> Add fast_settling_average_factor_available attribute. Expose to user
> possible values for the fast settling average factor.
> 
> Signed-off-by: Alisa-Dariana Roman <alisa.roman@xxxxxxxxxx>

Hi Alisa-Dariana,

Some questions inline.

Jonathan

> ---
>  .../ABI/testing/sysfs-bus-iio-adc-ad7192      |  18 +++
>  drivers/iio/adc/ad7192.c                      | 128 ++++++++++++++++--
>  2 files changed, 134 insertions(+), 12 deletions(-)
> 
> diff --git a/Documentation/ABI/testing/sysfs-bus-iio-adc-ad7192 b/Documentation/ABI/testing/sysfs-bus-iio-adc-ad7192
> index f8315202c8f0..780c6841b0c3 100644
> --- a/Documentation/ABI/testing/sysfs-bus-iio-adc-ad7192
> +++ b/Documentation/ABI/testing/sysfs-bus-iio-adc-ad7192
> @@ -19,6 +19,24 @@ Description:
>  		the bridge can be disconnected (when it is not being used
>  		using the bridge_switch_en attribute.
>  
> +What:		/sys/bus/iio/devices/iio:deviceX/fast_settling_average_factor
> +KernelVersion:
> +Contact:	linux-iio@xxxxxxxxxxxxxxx
> +Description:
> +		This attribute, if available, is used to activate or deactivate
> +		fast settling mode and set the value of the average factor to
> +		1, 2, 8 or 16. If the average factor is set to 1, the fast
> +		settling mode is disabled. The data from the sinc filter is
> +		averaged by chosen value. The averaging reduces the output data
> +		rate for a given FS word, however, the RMS noise improves.

So I got distracted in earlier patches, and didn't read your description closely
so was assuming this was some weird filter control.

Despite it being described as being closely related to the sinc filter, is this
just an oversampling ratio control?  If it is, then you can use the standard
ABI for this.  Making sure to reflect changes in this on the sampling frequency
as the apparent sampling frequency will reduce.


> +
> +What:		/sys/bus/iio/devices/iio:deviceX/fast_settling_average_factor_available
> +KernelVersion:
> +Contact:	linux-iio@xxxxxxxxxxxxxxx
> +Description:
> +		Reading returns a list with the possible values for the fast
> +		settling average factor: 1, 2, 8, 16.
> +
>  What:		/sys/bus/iio/devices/iio:deviceX/in_voltagex_sys_calibration
>  KernelVersion:
>  Contact:	linux-iio@xxxxxxxxxxxxxxx
> diff --git a/drivers/iio/adc/ad7192.c b/drivers/iio/adc/ad7192.c
> index 0f9d33002d35..3b7de23b024e 100644
> --- a/drivers/iio/adc/ad7192.c
> +++ b/drivers/iio/adc/ad7192.c
> @@ -60,6 +60,8 @@
>  #define AD7192_MODE_SEL_MASK	GENMASK(23, 21) /* Operation Mode Select Mask */
>  #define AD7192_MODE_STA_MASK	BIT(20) /* Status Register transmission Mask */
>  #define AD7192_MODE_CLKSRC_MASK	GENMASK(19, 18) /* Clock Source Select Mask */
> +#define AD7192_MODE_AVG_MASK	GENMASK(17, 16)
> +		  /* Fast Settling Filter Average Select Mask (AD7193 only) */
>  #define AD7192_MODE_SINC3	BIT(15) /* SINC3 Filter Select */
>  #define AD7192_MODE_ENPAR	BIT(13) /* Parity Enable */
>  #define AD7192_MODE_CLKDIV	BIT(12) /* Clock divide by 2 (AD7190/2 only)*/
> @@ -182,6 +184,7 @@ struct ad7192_state {
>  	u32				mode;
>  	u32				conf;
>  	u32				scale_avail[8][2];
> +	u8				avg_avail[4];
>  	u8				gpocon;
>  	u8				clock_sel;
>  	struct mutex			lock;	/* protect sensor state */
> @@ -459,6 +462,13 @@ static int ad7192_setup(struct iio_dev *indio_dev, struct device_node *np)
>  		st->scale_avail[i][0] = scale_uv;
>  	}
>  
> +	if (st->chip_info->chip_id == CHIPID_AD7193) {
> +		st->avg_avail[0] = 1;
> +		st->avg_avail[1] = 2;
> +		st->avg_avail[2] = 8;
> +		st->avg_avail[3] = 16;
> +	}
> +
>  	return 0;
>  }
>  
> @@ -483,6 +493,18 @@ static ssize_t ad7192_show_bridge_switch(struct device *dev,
>  			  FIELD_GET(AD7192_GPOCON_BPDSW, st->gpocon));
>  }
>  
> +static ssize_t ad7192_show_average_factor(struct device *dev,
> +					  struct device_attribute *attr,
> +					  char *buf)
> +{
> +	struct iio_dev *indio_dev = dev_to_iio_dev(dev);
> +	struct ad7192_state *st = iio_priv(indio_dev);
> +	u8 avg_factor_index;
> +
> +	avg_factor_index = FIELD_GET(AD7192_MODE_AVG_MASK, st->mode);
> +	return sysfs_emit(buf, "%d\n", st->avg_avail[avg_factor_index]);
> +}
> +
>  static ssize_t ad7192_set(struct device *dev,
>  			  struct device_attribute *attr,
>  			  const char *buf,
> @@ -491,12 +513,10 @@ static ssize_t ad7192_set(struct device *dev,
>  	struct iio_dev *indio_dev = dev_to_iio_dev(dev);
>  	struct ad7192_state *st = iio_priv(indio_dev);
>  	struct iio_dev_attr *this_attr = to_iio_dev_attr(attr);
> +	bool val, ret_einval;
> +	u8 val_avg_factor;

There isn't really that much code shared between the new use of this
and previous ones. I'd introduce a new callback for that particular sysfs
attribute instead of changing this one.


> +	unsigned int i;
>  	int ret;
> -	bool val;
> -
> -	ret = kstrtobool(buf, &val);
> -	if (ret < 0)
> -		return ret;
>  
>  	ret = iio_device_claim_direct_mode(indio_dev);
>  	if (ret)
> @@ -504,6 +524,10 @@ static ssize_t ad7192_set(struct device *dev,
>  
>  	switch ((u32)this_attr->address) {
>  	case AD7192_REG_GPOCON:
> +		ret = kstrtobool(buf, &val);
> +		if (ret < 0)
> +			return ret;
> +
>  		if (val)
>  			st->gpocon |= AD7192_GPOCON_BPDSW;
>  		else
> @@ -512,6 +536,10 @@ static ssize_t ad7192_set(struct device *dev,
>  		ad_sd_write_reg(&st->sd, AD7192_REG_GPOCON, 1, st->gpocon);
>  		break;
>  	case AD7192_REG_CONF:
> +		ret = kstrtobool(buf, &val);
> +		if (ret < 0)
> +			return ret;
> +
>  		if (val)
>  			st->conf |= AD7192_CONF_ACX;
>  		else
> @@ -519,6 +547,27 @@ static ssize_t ad7192_set(struct device *dev,
>  
>  		ad_sd_write_reg(&st->sd, AD7192_REG_CONF, 3, st->conf);
>  		break;
> +	case AD7192_REG_MODE:
> +		ret = kstrtou8(buf, 10, &val_avg_factor);
> +		if (ret < 0)
> +			return ret;
> +
> +		ret_einval = true;
I'd rather this relected the positive condition.
So maybe
		bool found = false;
> +		for (i = 0; i < ARRAY_SIZE(st->avg_avail); i++) {
> +			if (val_avg_factor == st->avg_avail[i]) {
> +				st->mode &= ~AD7192_MODE_AVG_MASK;
> +				st->mode |= FIELD_PREP(AD7192_MODE_AVG_MASK, i);
> +				ret_einval = false;
found = true;
break;
> +			}
> +		}
> +
> +		if (ret_einval)
if (!found)
	return -EINVAL;

> +			return -EINVAL;
> +
> +		ret = ad_sd_write_reg(&st->sd, AD7192_REG_MODE, 3, st->mode);
> +		if (ret)
> +			return ret;
> +		break;
>  	default:
>  		ret = -EINVAL;
>  	}
> @@ -528,15 +577,22 @@ static ssize_t ad7192_set(struct device *dev,
>  	return ret ? ret : len;
>  }