Re: [PATCH 2/2] iio: proximity: add support for PulsedLight LIDAR

[Jonathan Cameron <jic23@kernel.org>]

On 02/08/15 10:42, Lars-Peter Clausen wrote:
> On 08/01/2015 05:58 AM, Matt Ranostay wrote:
> [...]
>> +
>> +struct lidar_data {
>> +	struct mutex lock;
>> +	struct iio_dev *indio_dev;
>> +	struct i2c_client *client;
>> +
>> +	/* config */
>> +	int calib_bias;
>> +
>> +	u16 buffer[5]; /* 2 byte distance + 8 byte timestamp */
> 
> Needs to be in its own cacheline to avoid issues if the I2C controller is
> using DMA.
Would do if spi, but in the case of i2c I thought all bus drivers were
obliged to deal with this rather than leaving it to the client drivers?

Has this changed?

> 
> 	u16 buffer[5] ____cacheline_aligned;
> 
>> +};
> [...]
>> +static inline int lidar_read_byte(struct lidar_data *data, int reg)
> 
> I'd drop the inline. The compiler is smart enough to figure out whether it
> makes sense to inline it or not.
> 
>> +{
>> +	struct i2c_client *client = data->client;
>> +	int ret;
>> +
>> +	ret = i2c_smbus_write_byte(client, reg);
>> +	if (ret < 0) {
>> +		dev_err(&client->dev, "cannot write addr value");
>> +		return ret;
>> +	}
>> +
>> +	ret = i2c_smbus_read_byte(client);
>> +	if (ret < 0) {
>> +		dev_err(&client->dev, "cannot read data value");
>> +		return ret;
>> +	}
> 
> Instead of using a write_byte/read_byte combination rather use
> i2c_smbus_read_byte_data(). I think that will do the same, but in one atomic
> operation.
> 
>> +
>> +	return ret;
>> +}
> [...]
>> +static int lidar_read_measurement(struct lidar_data *data, u16 *reg)
>> +{
>> +	int ret;
>> +	int val;
>> +
>> +	ret = lidar_read_byte(data, LIDAR_REG_DATA_HBYTE);
>> +	if (ret < 0)
>> +		return ret;
>> +	val = ret << 8;
>> +
>> +	ret = lidar_read_byte(data, LIDAR_REG_DATA_LBYTE);
>> +	if (ret < 0)
>> +		return ret;
>> +	val |= ret;
>> +
>> +	/* correct any possible overflow or underflow */
>> +	val += data->calib_bias / 10000;
> 
> Typically calib bias is only meant for a offset correction that is done in
> hardware rather than in software. What's the usecase for doing it in
> kernelspace rather than in userspace?
> 
>> +	if (val < 0)
>> +		val = 0;
>> +
>> +	if (val > LIDAR_REG_DATA_MAX)
>> +		val = LIDAR_REG_DATA_MAX;
>> +
>> +	*reg = val;
>> +
>> +	return 0;
>> +}
> [...]
>> +static int lidar_read_raw(struct iio_dev *indio_dev,
>> +			  struct iio_chan_spec const *chan,
>> +			  int *val, int *val2, long mask)
>> +{
>> +	struct lidar_data *data = iio_priv(indio_dev);
>> +	int ret = -EINVAL;
>> +
>> +	if (iio_buffer_enabled(indio_dev) && mask == IIO_CHAN_INFO_RAW)
>> +		return -EBUSY;
>> +
>> +	mutex_lock(&data->lock);
>> +
>> +	switch (mask) {
>> +	case IIO_CHAN_INFO_RAW: {
>> +		u16 reg;
>> +
>> +		ret = lidar_get_measurement(data, &reg);
>> +		if (!ret) {
>> +			*val = reg / 100;
>> +			*val2 = (reg % 100) * 10000;
>> +			ret = IIO_VAL_INT_PLUS_MICRO;
> 
> The raw attribute is supposed to return the raw value as a IIO_VAL_INT. A
> scale attribute should be used to indicate the conversion factor to get the
> proper unit.
> 
>> +		}
>> +		break;
>> +	}
>> +	case IIO_CHAN_INFO_CALIBBIAS:
>> +		*val = 0;
>> +		*val2 = data->calib_bias;
>> +		ret = IIO_VAL_INT_PLUS_MICRO;
>> +		break;
>> +	}
>> +
>> +	mutex_unlock(&data->lock);
>> +
>> +	return ret;
>> +}
> [...]
>> +static struct iio_info lidar_info = {
> 
> const
> 
>> +	.driver_module = THIS_MODULE,
>> +	.read_raw = lidar_read_raw,
>> +	.write_raw = lidar_write_raw,
>> +};
> [...]
>> +static struct i2c_driver lidar_driver = {
>> +	.driver = {
>> +		.name	= LIDAR_DRV_NAME,
>> +		.owner	= THIS_MODULE,
> 
> You added a DT vendor prefix, but there is no of match table for the driver.
> 
>> +	},
>> +	.probe		= lidar_probe,
>> +	.remove		= lidar_remove,
>> +	.id_table	= lidar_id,
>> +};
>> +module_i2c_driver(lidar_driver);
> --
> To unsubscribe from this list: send the line "unsubscribe linux-iio" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>