Message-ID: <9f8b6789-ca87-425c-9645-c4b97dcb2424@linaro.org>
Date: Tue, 29 Oct 2024 09:09:09 +0100
From: Neil Armstrong <neil.armstrong@...aro.org>
To: Andy Shevchenko <andy.shevchenko@...il.com>
Cc: Rob Herring <robh@...nel.org>, Krzysztof Kozlowski <krzk+dt@...nel.org>,
 Conor Dooley <conor+dt@...nel.org>, Jonathan Cameron <jic23@...nel.org>,
 Lars-Peter Clausen <lars@...afoo.de>, devicetree@...r.kernel.org,
 linux-kernel@...r.kernel.org, linux-iio@...r.kernel.org
Subject: Re: [PATCH v2 3/3] iio: magnetometer: add Allegro MicroSystems
 ALS31300 3-D Linear Hall Effect driver

Hi,

On 26/10/2024 23:17, Andy Shevchenko wrote:
> Mon, Oct 21, 2024 at 02:38:55PM +0200, Neil Armstrong kirjoitti:
>> The Allegro MicroSystems ALS31300 is a 3-D Linear Hall Effect Sensor
>> mainly used for 3D head-on motion sensing applications.
>>
>> The device is configured over I2C, and as part of the Sensor data the
>> temperature core is also provided.
>>
>> While the device provides an IRQ gpio, it depends on a configuration
>> programmed into the internal EEPROM, thus only the default mode is
>> supported and buffered input via trigger is also supported to allow
>> streaming values with the same sensing timestamp.
>>
>> The device can be configured with different sensitivities in factory,
>> but the sensitivity value used to calculate value into the Gauss
>> unit is not available from registers, thus the sensitivity is provided
>> by the compatible/device-id string which is based on the part number
>> as described in the datasheet page 2.
> 
> ...
> 
> Some headers are missing...
> 
>> +#include <linux/bitfield.h>
>> +#include <linux/bits.h>
>> +#include <linux/delay.h>
>> +#include <linux/module.h>
>> +#include <linux/i2c.h>
>> +#include <linux/regmap.h>
> 
> + pm.h
> 
>> +#include <linux/pm_runtime.h>
> 
>> +#include <linux/regulator/consumer.h>
> 
> + types.h

Ack

> 
> ...
> 
>> +#define ALS31300_DATA_X_GET(b)		\
>> +		sign_extend32(FIELD_GET(ALS31300_VOL_MSB_X_AXIS, b[0]) << 4 | \
>> +			      FIELD_GET(ALS31300_VOL_LSB_X_AXIS, b[1]), 11)
>> +#define ALS31300_DATA_Y_GET(b)		\
>> +		sign_extend32(FIELD_GET(ALS31300_VOL_MSB_Y_AXIS, b[0]) << 4 | \
>> +			      FIELD_GET(ALS31300_VOL_LSB_Y_AXIS, b[1]), 11)
>> +#define ALS31300_DATA_Z_GET(b)		\
>> +		sign_extend32(FIELD_GET(ALS31300_VOL_MSB_Z_AXIS, b[0]) << 4 | \
>> +			      FIELD_GET(ALS31300_VOL_LSB_Z_AXIS, b[1]), 11)
>> +#define ALS31300_TEMPERATURE_GET(b)	\
>> +		(FIELD_GET(ALS31300_VOL_MSB_TEMPERATURE, b[0]) << 6 | \
>> +		 FIELD_GET(ALS31300_VOL_LSB_TEMPERATURE, b[1]))
> 
> These and in the code seems like you make a home grown endianes conversion.
> I suppose all of them have to be __beXX with the respective masks that cover
> all the bits.

This has nothing do to with endianess conversion, the bits of the 4 channels
are split on 2 registers, look at the defines.

> 
>> +static int als31300_get_measure(struct als31300_data *data,
>> +				u16 *t, s16 *x, s16 *y, s16 *z)
>> +{
>> +	unsigned int count = 0;
>> +	u32 buf[2];
> 
> Shouldn't this be __be64 buf?

No this is 2 separate registers with different fields, not a single 64bit big-endian register

> 
>> +	int ret;
>> +
>> +	guard(mutex)(&data->mutex);
>> +
>> +	ret = pm_runtime_resume_and_get(data->dev);
>> +	if (ret)
>> +		return ret;
> 
>> +	/* Max update rate is 2KHz, wait up to 1ms */
>> +	while (count < 50) {
>> +		/* Read Data */
>> +		ret = regmap_bulk_read(data->map, ALS31300_VOL_MSB, buf, 2);
> 
> At bare minimum ARRAY_SIZE(), but see above, I think it should use the respective type.

I'll use ARRAY_SIZE

> 
>> +		if (ret) {
>> +			dev_err(data->dev, "read data failed, error %d\n", ret);
>> +			goto out;
>> +		}
>> +
>> +		/* Check if data is valid, happens right after getting out of sleep mode */
>> +		if (FIELD_GET(ALS31300_VOL_MSB_NEW_DATA, buf[0]))
>> +			break;
>> +
>> +		usleep_range(10, 20);
>> +		++count;
>> +	}
>> +
>> +	if (count >= 50) {
>> +		ret = -ETIMEDOUT;
>> +		goto out;
>> +	}
> 
> This one of the longest variant of implementing
> 
> 	do {
> 		...
> 	} while (!FIELD_GET(...) && --count)
> 
> But you also may consider something from iopoll.h (I don't remember if we have
> regmap_read_poll_timeout() for bulk transfers.

There's no poll for bulk read, I'll try using an IOPOLL function

> 
>> +	*t = ALS31300_TEMPERATURE_GET(buf);
>> +	*x = ALS31300_DATA_X_GET(buf);
>> +	*y = ALS31300_DATA_Y_GET(buf);
>> +	*z = ALS31300_DATA_Z_GET(buf);
>> +
>> +out:
>> +	pm_runtime_mark_last_busy(data->dev);
>> +	pm_runtime_put_autosuspend(data->dev);
> 
> Last discussion with Rafael on the topic of the above puts a question mark to
> all these cases when runtime PM idle callback has not been implemented.
> 
> Shouldn't this be simply
> 
> 	pm_runtime_put(data->dev);
> 
> ?

Probably, I just used the same schema as other similar IIO drivers

> 
>> +	return ret;
>> +}
> 
> ...
> 
>> +	case IIO_CHAN_INFO_SCALE:
>> +		switch (chan->type) {
>> +		case IIO_TEMP:
>> +			/*
>> +			 * Fractional part of:
>> +			 *         1000 * 302 * (value - 1708)
>> +			 * temp = ----------------------------
>> +			 *             4096
>> +			 * to convert temperature in millicelcius
>> +			 */
>> +			*val = 1000 * 302;
> 
> MILLI from units.h ?

Well, sure, but it's still 1000, should I also define ONE to 1 and so on ?

> 
>> +			*val2 = 4096;
>> +			return IIO_VAL_FRACTIONAL;
>> +		case IIO_MAGN:
>> +			/*
>> +			 * Devices are configured in factory
>> +			 * with different sensitivities:
>> +			 * - 500 GAUSS <-> 4 LSB/Gauss
>> +			 * - 1000 GAUSS <-> 2 LSB/Gauss
>> +			 * - 2000 GAUSS <-> 1 LSB/Gauss
>> +			 * with translates by a division of the returned
>> +			 * value to get Gauss value.
>> +			 * The sensisitivity cannot be read at runtime
>> +			 * so the value depends on the model compatible
>> +			 * or device id.
>> +			 */
>> +			*val = 1;
>> +			*val2 = data->variant_info->sensitivity;
>> +			return IIO_VAL_FRACTIONAL;
>> +		default:
>> +			return -EINVAL;
>> +		}
> 
> ...
> 
>> +	struct {
>> +		u16 temperature;
>> +		s16 channels[3];
>> +		aligned_s64 timestamp;
>> +	} __packed scan;
> 
> Why __packed?

Probably unneeded, will drop

> 
> ...
> 
>> +static int als31300_set_operating_mode(struct als31300_data *data,
>> +				       unsigned int val)
>> +{
>> +	int ret;
>> +
>> +	ret = regmap_update_bits(data->map, ALS31300_VOL_MODE,
>> +				 ALS31300_VOL_MODE_SLEEP, val);
>> +	if (ret) {
>> +		dev_err(data->dev, "failed to set operating mode (%pe)\n", ERR_PTR(ret));
>> +		return ret;
>> +	}
>> +
>> +	/* The time it takes to exit sleep mode is equivalent to Power-On Delay Time */
>> +	if (val == ALS31300_VOL_MODE_ACTIVE_MODE)
>> +		usleep_range(600, 650);
>> +
>> +	return ret;
> 
> 	return 0;

It's the same but ok

> 
>> +}
> 
> ...
> 
>> +static int als31300_probe(struct i2c_client *i2c)
>> +{
>> +	struct device *dev = &i2c->dev;
>> +	struct als31300_data *data;
>> +	struct iio_dev *indio_dev;
>> +	int ret;
>> +
>> +	indio_dev = devm_iio_device_alloc(dev, sizeof(*data));
>> +	if (!indio_dev)
>> +		return -ENOMEM;
>> +
>> +	data = iio_priv(indio_dev);
>> +	data->dev = dev;
>> +	i2c_set_clientdata(i2c, indio_dev);
> 
>> +	mutex_init(&data->mutex);
> 
> Why not devm_mutex_init()? How does this being cleaned up?

Indeed, good catch

> 
>> +	data->variant_info = i2c_get_match_data(i2c);
>> +	if (!data->variant_info)
>> +		return -EINVAL;
>> +
>> +	data->map = devm_regmap_init_i2c(i2c, &als31300_regmap_config);
>> +	if (IS_ERR(data->map))
>> +		return dev_err_probe(dev, PTR_ERR(data->map),
>> +				     "failed to allocate register map\n");
>> +
>> +	ret = devm_regulator_get_enable(dev, "vcc");
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "failed to enable regulator\n");
>> +
>> +	ret = als31300_set_operating_mode(data, ALS31300_VOL_MODE_ACTIVE_MODE);
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "failed to power on device\n");
>> +
>> +	ret = devm_add_action_or_reset(dev, als31300_power_down, data);
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "failed to add powerdown action\n");
>> +
>> +	indio_dev->info = &als31300_info;
>> +	indio_dev->modes = INDIO_DIRECT_MODE;
>> +	indio_dev->name = i2c->name;
>> +	indio_dev->channels = als31300_channels;
>> +	indio_dev->num_channels = ARRAY_SIZE(als31300_channels);
>> +	indio_dev->available_scan_masks = als31300_scan_masks;
>> +
>> +	ret = devm_iio_triggered_buffer_setup(dev, indio_dev,
>> +					      iio_pollfunc_store_time,
>> +					      als31300_trigger_handler,
>> +					      &als31300_setup_ops);
>> +	if (ret < 0) {
>> +		dev_err(dev, "iio triggered buffer setup failed\n");
>> +		return ret;
> 
> Why not return dev_err_probe(...);

Sure

> 
>> +	}
>> +
>> +	ret = pm_runtime_set_active(dev);
>> +	if (ret < 0)
>> +		return ret;
>> +
>> +	ret = devm_pm_runtime_enable(dev);
>> +	if (ret)
>> +		return ret;
>> +
>> +	pm_runtime_get_noresume(dev);
>> +	pm_runtime_set_autosuspend_delay(dev, 200);
>> +	pm_runtime_use_autosuspend(dev);
>> +
>> +	pm_runtime_mark_last_busy(dev);
>> +	pm_runtime_put_autosuspend(dev);
>> +
>> +	ret = devm_iio_device_register(dev, indio_dev);
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "device register failed\n");
>> +
>> +	return 0;
>> +}
> 

Thanks,
Neil

Powered by blists - more mailing lists