lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240324121418.570857a9@jic23-huawei>
Date: Sun, 24 Mar 2024 12:14:18 +0000
From: Jonathan Cameron <jic23@...nel.org>
To: Vasileios Amoiridis <vassilisamir@...il.com>
Cc: lars@...afoo.de, andriy.shevchenko@...ux.intel.com,
 ang.iglesiasg@...il.com, mazziesaccount@...il.com, ak@...klinger.de,
 petre.rodan@...dimension.ro, phil@...pberrypi.com, 579lpy@...il.com,
 linus.walleij@...aro.org, semen.protsenko@...aro.org,
 linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 6/6] iio: pressure: Add triggered buffer support for
 BMP280 driver

On Tue, 19 Mar 2024 01:29:25 +0100
Vasileios Amoiridis <vassilisamir@...il.com> wrote:

> BMP2xx, BMP3xx, and BMP5xx use consecutive buffers for their
> temperature, pressure and humidity readings. This facilitates
> the use of burst reads in order to acquire data much faster
> and in a different way from the one used in oneshot captures.
> 
> BMP085 and BMP180 use a completely different measurement
> process that is well defined and is used in their buffer_handler().
> 
> Suggested-by: Angel Iglesias <ang.iglesiasg@...il.com>
> Signed-off-by: Vasileios Amoiridis <vassilisamir@...il.com>

Various comments inline,

Thanks,
Jonathan

> ---
>  drivers/iio/pressure/Kconfig       |   2 +
>  drivers/iio/pressure/bmp280-core.c | 231 +++++++++++++++++++++++++++--
>  drivers/iio/pressure/bmp280-spi.c  |   8 +-
>  drivers/iio/pressure/bmp280.h      |  14 +-
>  4 files changed, 241 insertions(+), 14 deletions(-)
> 
> diff --git a/drivers/iio/pressure/Kconfig b/drivers/iio/pressure/Kconfig
> index 79adfd059c3a..5145b94b4679 100644
> --- a/drivers/iio/pressure/Kconfig
> +++ b/drivers/iio/pressure/Kconfig
> @@ -31,6 +31,8 @@ config BMP280
>  	select REGMAP
>  	select BMP280_I2C if (I2C)
>  	select BMP280_SPI if (SPI_MASTER)
> +	select IIO_BUFFER
> +	select IIO_TRIGGERED_BUFFER
>  	help
>  	  Say yes here to build support for Bosch Sensortec BMP180, BMP280, BMP380
>  	  and BMP580 pressure and temperature sensors. Also supports the BME280 with
> diff --git a/drivers/iio/pressure/bmp280-core.c b/drivers/iio/pressure/bmp280-core.c
> index ddfcd23f29a0..ffcd17807cf5 100644
> --- a/drivers/iio/pressure/bmp280-core.c
> +++ b/drivers/iio/pressure/bmp280-core.c
> @@ -40,7 +40,10 @@
>  #include <linux/regmap.h>
>  #include <linux/regulator/consumer.h>
>  
> +#include <linux/iio/buffer.h>
>  #include <linux/iio/iio.h>
> +#include <linux/iio/trigger_consumer.h>
> +#include <linux/iio/triggered_buffer.h>
>  
>  #include <asm/unaligned.h>
>  
> @@ -454,7 +457,7 @@ static int bmp280_read_temp(struct bmp280_data *data, s32 *comp_temp)
>  	int ret;
>  
>  	ret = regmap_bulk_read(data->regmap, BMP280_REG_TEMP_MSB,
> -			       data->buf, sizeof(data->buf));
> +			       data->buf, BMP280_NUM_TEMP_BYTES);
>  	if (ret < 0) {
>  		dev_err(data->dev, "failed to read temperature\n");
>  		return ret;
> @@ -484,7 +487,7 @@ static int bmp280_read_press(struct bmp280_data *data, u32 *comp_press)
>  		return ret;
>  
>  	ret = regmap_bulk_read(data->regmap, BMP280_REG_PRESS_MSB,
> -			       data->buf, sizeof(data->buf));
> +			       data->buf, BMP280_NUM_PRESS_BYTES);
>  	if (ret < 0) {
>  		dev_err(data->dev, "failed to read pressure\n");
>  		return ret;
> @@ -513,13 +516,13 @@ static int bmp280_read_humid(struct bmp280_data *data, u32 *comp_humidity)
>  		return ret;
>  
>  	ret = regmap_bulk_read(data->regmap, BMP280_REG_HUMIDITY_MSB,
> -			       &data->be16, sizeof(data->be16));
> +			       &data->be16, BME280_NUM_HUMIDITY_BYTES);
>  	if (ret < 0) {
>  		dev_err(data->dev, "failed to read humidity\n");
>  		return ret;
>  	}
>  
> -	adc_humidity = be16_to_cpu(data->be16);
> +	adc_humidity = get_unaligned_be16(&data->be16);

If it's a be16, how did it become unaligned?

>  	if (adc_humidity == BMP280_HUMIDITY_SKIPPED) {
>  		/* reading was skipped */
>  		dev_err(data->dev, "reading humidity skipped\n");
> @@ -931,6 +934,73 @@ static int bmp280_chip_config(struct bmp280_data *data)
>  	return ret;
>  }
>  
> +static irqreturn_t bmp280_buffer_handler(int irq, void *p)
> +{
> +	struct iio_poll_func *pf = p;
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct bmp280_data *data = iio_priv(indio_dev);
> +	s32 adc_temp, adc_press, adc_humidity;
> +	u8 size_of_burst_read;
> +	int ret, chan_value;
> +
> +	guard(mutex)(&data->lock);
> +
> +	if (test_bit(BME280_HUMID, indio_dev->active_scan_mask))

This confuses me a little. Is it allowing reuse of this function for
multiple devices or aiming to optimise the read in the case of
the humidity channel being disabled (in which case I don't think
it works because you aren't providing that combination in avail_scan_masks.)

Add a comment to explain.

> +		size_of_burst_read = 8;
> +	else
> +		size_of_burst_read = 6;
> +
> +	/* Burst read data registers */
> +	ret = regmap_bulk_read(data->regmap, BMP280_REG_PRESS_MSB,
> +			       data->buf, 8);
> +	if (ret) {
> +		dev_err(data->dev, "failed to read sensor data\n");
> +		return ret;
> +	}
> +
> +	/* Temperature calculations */
> +	memcpy(&chan_value, &data->buf[3], 3);
> +
> +	adc_temp = FIELD_GET(BMP280_MEAS_TRIM_MASK, get_unaligned_be24(&chan_value));
> +	if (adc_temp == BMP280_TEMP_SKIPPED) {
> +		/* reading was skipped */
Similar comments on this to below (I read backwards as normally code makes
more sense that way :)
> +		dev_err(data->dev, "reading temperature skipped\n");
> +		return -EIO;
> +	}
> +	data->iio_buf.sensor_data[0] = bmp280_compensate_temp(data, adc_temp);
> +
> +	/* Pressure calculations */
> +	memcpy(&chan_value, &data->buf[0], 3);
> +
> +	adc_press = FIELD_GET(BMP280_MEAS_TRIM_MASK, get_unaligned_be24(&chan_value));
> +	if (adc_press == BMP280_PRESS_SKIPPED) {
> +		/* reading was skipped */
> +		dev_err(data->dev, "reading pressure skipped\n");
> +		return -EIO;
> +	}
> +	data->iio_buf.sensor_data[1] = bmp280_compensate_press(data, adc_press);
> +
> +	/* Humidity calculations */
> +	if (test_bit(BME280_HUMID, indio_dev->active_scan_mask)) {
> +		memcpy(&chan_value, &data->buf[6], 2);
> +
> +		adc_humidity = get_unaligned_be16(&chan_value);
> +		if (adc_humidity == BMP280_HUMIDITY_SKIPPED) {
> +			/* reading was skipped */
> +			dev_err(data->dev, "reading humidity skipped\n");
> +			return -EIO;
> +		}
> +	data->iio_buf.sensor_data[2] = bmp280_compensate_humidity(data, adc_humidity);
Alignment of code looks wrong.

> +	}
> +
> +	iio_push_to_buffers_with_timestamp(indio_dev, &data->iio_buf,
> +					   iio_get_time_ns(indio_dev));
> +
> +	iio_trigger_notify_done(indio_dev->trig);
> +
> +	return IRQ_HANDLED;
> +}

> +static irqreturn_t bmp380_buffer_handler(int irq, void *p)
> +{
> +	struct iio_poll_func *pf = p;
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct bmp280_data *data = iio_priv(indio_dev);
> +	s32 adc_temp, adc_press;
> +	int ret, chan_value;
> +
> +	guard(mutex)(&data->lock);
> +
> +	/* Burst read data registers */
> +	ret = regmap_bulk_read(data->regmap, BMP380_REG_PRESS_XLSB,
> +			       data->buf, 6);
> +	if (ret) {
> +		dev_err(data->dev, "failed to read sensor data\n");
> +		return ret;
> +	}
> +
> +	/* Temperature calculations */
> +	memcpy(&chan_value, &data->buf[3], 3);
> +
> +	adc_temp = get_unaligned_le24(&chan_value);

Same comments as below.

> +	if (adc_temp == BMP380_TEMP_SKIPPED) {
> +		/* reading was skipped */
> +		dev_err(data->dev, "reading temperature skipped\n");
> +		return -EIO;
> +	}
> +	data->iio_buf.sensor_data[0] = bmp380_compensate_temp(data, adc_temp);
> +
> +	/* Pressure calculations */
> +	memcpy(&chan_value, &data->buf[0], 3);
> +
> +	adc_press = get_unaligned_le24(&chan_value);
> +	if (adc_press == BMP380_PRESS_SKIPPED) {
> +		/* reading was skipped */
> +		dev_err(data->dev, "reading pressure skipped\n");
> +		return -EIO;
> +	}
> +	data->iio_buf.sensor_data[1] = bmp380_compensate_press(data, adc_press);
> +
> +	iio_push_to_buffers_with_timestamp(indio_dev, &data->iio_buf,
> +					   iio_get_time_ns(indio_dev));
> +
> +	iio_trigger_notify_done(indio_dev->trig);
> +
> +	return IRQ_HANDLED;
> +}

> +static irqreturn_t bmp580_buffer_handler(int irq, void *p)
> +{
> +	struct iio_poll_func *pf = p;
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct bmp280_data *data = iio_priv(indio_dev);
> +	s32 adc_temp, adc_press;
> +	int ret, chan_value;
> +
> +	guard(mutex)(&data->lock);
> +
> +	/* Burst read data registers */
> +	ret = regmap_bulk_read(data->regmap, BMP580_REG_TEMP_XLSB,
> +			       data->buf, 6);
> +	if (ret) {
> +		dev_err(data->dev, "failed to read sensor data\n");
> +		return ret;
> +	}
> +
> +	/* Temperature calculations */
> +	memcpy(&chan_value, &data->buf[3], 3);
> +
> +	adc_temp = get_unaligned_le24(&chan_value);

As in other thread branch, get it directly from &data->buf[3]
rather than bouncing it via another buffer.

> +	if (adc_temp == BMP580_TEMP_SKIPPED) {
> +		/* reading was skipped */
> +		dev_err(data->dev, "reading temperature skipped\n");
> +		return -EIO;
-EIO isn't irqreturn_t 

There isn't a good way to return errors from interrupt handlers,
so just return IRQ_HANDLED after printing the error message.

> +	}
> +	data->iio_buf.sensor_data[0] = adc_temp;
> +
> +	/* Pressure calculations */
> +	memcpy(&chan_value, &data->buf[0], 3);
> +
> +	adc_press = get_unaligned_le24(&chan_value);
As above, get it directly.

> +	if (adc_press == BMP580_PRESS_SKIPPED) {
> +		/* reading was skipped */
> +		dev_err(data->dev, "reading pressure skipped\n");
> +		return -EIO;
return IRQ_HANDLED;

> +	}
> +	data->iio_buf.sensor_data[1] =  adc_press;
Extra space.

> +
> +	iio_push_to_buffers_with_timestamp(indio_dev, &data->iio_buf,
> +					   iio_get_time_ns(indio_dev));
> +
> +	iio_trigger_notify_done(indio_dev->trig);
> +
> +	return IRQ_HANDLED;
> +}
> +
>  static const int bmp580_oversampling_avail[] = { 1, 2, 4, 8, 16, 32, 64, 128 };
>  static const u8 bmp580_chip_ids[] = { BMP580_CHIP_ID, BMP580_CHIP_ID_ALT };
>  static const int bmp580_temp_coeffs[] = { 1000, 16 };
> @@ -1903,6 +2075,8 @@ const struct bmp280_chip_info bmp580_chip_info = {
>  	.read_temp = bmp580_read_temp,
>  	.read_press = bmp580_read_press,
>  	.preinit = bmp580_preinit,
> +
> +	.buffer_handler = bmp580_buffer_handler
>  };
>  EXPORT_SYMBOL_NS(bmp580_chip_info, IIO_BMP280);
>  
> @@ -2054,7 +2228,7 @@ static int bmp180_read_adc_press(struct bmp280_data *data, int *val)
>  		return ret;
>  
>  	ret = regmap_bulk_read(data->regmap, BMP180_REG_OUT_MSB,
> -			       data->buf, sizeof(data->buf));
> +			       data->buf, BMP280_NUM_PRESS_BYTES);
>  	if (ret)
>  		return ret;
>  
> @@ -2122,6 +2296,35 @@ static int bmp180_chip_config(struct bmp280_data *data)
>  	return 0;
>  }
>  
> +static irqreturn_t bmp180_buffer_handler(int irq, void *p)
> +{
> +	struct iio_poll_func *pf = p;
> +	struct iio_dev *indio_dev = pf->indio_dev;
> +	struct bmp280_data *data = iio_priv(indio_dev);
> +	int ret, chan_value;
> +
> +	guard(mutex)(&data->lock);
> +
> +	ret = data->chip_info->read_temp(data, &chan_value);
> +	if (ret)
> +		return ret;
> +
> +	data->iio_buf.sensor_data[0] = chan_value;
> +
> +	ret = data->chip_info->read_press(data, &chan_value);

Given my earlier suggestion to stop hiding t_fine in the
structure, these callbacks will need to bring it all the way out
here (from the temperature read) and pass it back into the
pressure read.

That will have the pleasing side effect of making it obvious why
we aren't handling subsets of channels (as they aren't
independent)


> +	if (ret)
> +		return ret;
> +
> +	data->iio_buf.sensor_data[1] = chan_value;
> +
> +	iio_push_to_buffers_with_timestamp(indio_dev, &data->iio_buf,
> +					   iio_get_time_ns(indio_dev));
> +
> +	iio_trigger_notify_done(indio_dev->trig);
> +
> +	return IRQ_HANDLED;
> +}
> +

> diff --git a/drivers/iio/pressure/bmp280-spi.c b/drivers/iio/pressure/bmp280-spi.c
> index a444d4b2978b..dc297583cac1 100644
> --- a/drivers/iio/pressure/bmp280-spi.c
> +++ b/drivers/iio/pressure/bmp280-spi.c
> @@ -40,14 +40,14 @@ static int bmp380_regmap_spi_read(void *context, const void *reg,
>  				  size_t reg_size, void *val, size_t val_size)
>  {
>  	struct spi_device *spi = to_spi_device(context);
> -	u8 rx_buf[4];
> +	u8 rx_buf[9];
>  	ssize_t status;
>  
>  	/*
> -	 * Maximum number of consecutive bytes read for a temperature or
> -	 * pressure measurement is 3.
> +	 * Maximum number of a burst read for temperature, pressure, humidity
> +	 * is 8 bytes.

Once this 8 is expressed as the sum of the 3 types, you can drop the comment
as it will add nothing useful.

>  	 */
> -	if (val_size > 3)
> +	if (val_size > 8)
>  		return -EINVAL;
>  
>  	/*
> diff --git a/drivers/iio/pressure/bmp280.h b/drivers/iio/pressure/bmp280.h
> index 8cc3eed70c18..32155567faf6 100644
> --- a/drivers/iio/pressure/bmp280.h
> +++ b/drivers/iio/pressure/bmp280.h
> @@ -301,6 +301,10 @@
>  #define BMP280_PRESS_SKIPPED		0x80000
>  #define BMP280_HUMIDITY_SKIPPED		0x8000
>  
> +#define BMP280_NUM_TEMP_BYTES		3
> +#define BMP280_NUM_PRESS_BYTES		3
> +#define BME280_NUM_HUMIDITY_BYTES	2
> +
>  /* Core exported structs */
>  
>  static const char *const bmp280_supply_names[] = {
> @@ -400,13 +404,19 @@ struct bmp280_data {
>  	 */
>  	s32 t_fine;
>  
> +	/* Data to push to userspace */
> +	struct {
> +		s32 sensor_data[3];

This doesn't work (as explanation of data) for all cases.
 If you only have 2 channels, the packing needs to be.

		s32 sensor_data[2];
//no padding
		s64 timestamp __aligned(8);
> +		s64 timestamp __aligned(8);
> +	} iio_buf;
> +
So using a structure for this definition isn't a bug as such as the core
doesn't care that you provide a bigger buffer than needed, but it is misleading.
Use something along lines of

	/* Up to 3 channels and aligned s64 timestamp */
	s32 sensor_data[6] __aligned(8);

or use a union of two structures to cover the two layouts making
sure to write and read from the correct one in each callback.

>  	/*
>  	 * DMA (thus cache coherency maintenance) may require the
>  	 * transfer buffers to live in their own cache lines.
>  	 */
>  	union {
>  		/* Sensor data buffer */
> -		u8 buf[3];
> +		u8 buf[8];
As in previous discussion - build this up as a sum of what can go in it.
Maybe via a define for BMP280_BURST_READ_MAX 


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ