devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jonathan Cameron <jic23-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
To: "Vianney le Clément de Saint-Marcq"
	<vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org>,
	linux-iio-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Cc: "Arnout Vandecappelle (Essensium/Mind)"
	<arnout-agwphqFHkxc@public.gmane.org>
Subject: Re: [PATCH v2 6/7] iio: mlx90614: Add power management
Date: Sat, 28 Mar 2015 12:54:43 +0000	[thread overview]
Message-ID: <5516A493.7090406@kernel.org> (raw)
In-Reply-To: <1427212459-31585-7-git-send-email-vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org>

On 24/03/15 15:54, Vianney le Clément de Saint-Marcq wrote:
> Add support for system sleep and runtime power management.
> 
> To wake up the device, the SDA line should be held low for at least 33ms
> while SCL is high.  As this is not possible using the i2c API (and not
> supported by all i2c adapters), a GPIO connected to the SDA line is
> needed.  The GPIO is named "wakeup" and can be specified in a device
> tree with the "wakeup-gpios" binding.
> 
> If the wake-up GPIO is not given, disable power management for the
> device.  Entering sleep requires an SMBus byte access, hence power
> management is also disabled if byte access is not supported by the
> adapter.
> 
> Introduce "melexis" as a vendor prefix for device tree bindings.
> 
> Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org>
> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout-agwphqFHkxc@public.gmane.org>
> 
A couple of bits inline.  In particularly please break the docs into
a separate patch and only add the new element in this one.

Also, your probe and remove are not mirror images and should be.

Jonathan
> ---
> 
> The default autosleep delay (5s) is chosen arbitrarily, trying to take
> into account the long startup time (250ms).  Feel free to change it to
> another value if you think it is saner.
> 
> v2: return -EINTR when the startup delay is interrupted
> ---
>  .../bindings/iio/temperature/mlx90614.txt          |  24 ++
>  .../devicetree/bindings/vendor-prefixes.txt        |   1 +
>  drivers/iio/temperature/mlx90614.c                 | 246 ++++++++++++++++++++-
>  3 files changed, 269 insertions(+), 2 deletions(-)
>  create mode 100644 Documentation/devicetree/bindings/iio/temperature/mlx90614.txt
> 
> diff --git a/Documentation/devicetree/bindings/iio/temperature/mlx90614.txt b/Documentation/devicetree/bindings/iio/temperature/mlx90614.txt
> new file mode 100644
> index 0000000..9be57b0
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/iio/temperature/mlx90614.txt
Please split out an initial device tree patch adding the device and docs
then a followup one adding the new stuff.
> @@ -0,0 +1,24 @@
> +* Melexis MLX90614 contactless IR temperature sensor
> +
> +http://melexis.com/Infrared-Thermometer-Sensors/Infrared-Thermometer-Sensors/MLX90614-615.aspx
> +
> +Required properties:
> +
> +  - compatible: should be "melexis,mlx90614"
> +  - reg: the I2C address of the sensor
> +
> +Optional properties:
> +
> +  - wakeup-gpios: device tree identifier of the GPIO connected to the SDA line
> +      to hold low in order to wake up the device.  In normal operation, the
> +      GPIO is set as input and will not interfere in I2C communication.  There
> +      is no need for a GPIO driving the SCL line.  If no GPIO is given, power
> +      management is disabled.
> +
> +Example:
> +
> +mlx90614@5a {
> +	compatible = "melexis,mlx90614";
> +	reg = <0x5a>;
> +	wakeup-gpios = <&gpio0 2 GPIO_ACTIVE_HIGH>;
> +};
> diff --git a/Documentation/devicetree/bindings/vendor-prefixes.txt b/Documentation/devicetree/bindings/vendor-prefixes.txt
> index fae26d0..c8db30c 100644
> --- a/Documentation/devicetree/bindings/vendor-prefixes.txt
> +++ b/Documentation/devicetree/bindings/vendor-prefixes.txt
> @@ -110,6 +110,7 @@ lltc	Linear Technology Corporation
>  marvell	Marvell Technology Group Ltd.
>  maxim	Maxim Integrated Products
>  mediatek	MediaTek Inc.
> +melexis	Melexis N.V.
>  merrii	Merrii Technology Co., Ltd.
>  micrel	Micrel Inc.
>  microchip	Microchip Technology Inc.
> diff --git a/drivers/iio/temperature/mlx90614.c b/drivers/iio/temperature/mlx90614.c
> index b981d1e..6b9bfcd 100644
> --- a/drivers/iio/temperature/mlx90614.c
> +++ b/drivers/iio/temperature/mlx90614.c
> @@ -12,13 +12,24 @@
>   *
>   * (7-bit I2C slave address 0x5a, 100KHz bus speed only!)
>   *
> - * TODO: sleep mode, filter configuration
> + * To wake up from sleep mode, the SDA line must be held low while SCL is high
> + * for at least 33ms.  This is achieved with an extra GPIO that can be connected
> + * directly to the SDA line.  In normal operation, the GPIO is set as input and
> + * will not interfere in I2C communication.  While the GPIO is driven low, the
> + * i2c adapter is locked since it cannot be used by other clients.  The SCL line
> + * always has a pull-up so we do not need an extra GPIO to drive it high.  If
> + * the "wakeup" GPIO is not given, power management will be disabled.
> + *
> + * TODO: filter configuration
>   */
>  
>  #include <linux/err.h>
>  #include <linux/i2c.h>
>  #include <linux/module.h>
>  #include <linux/delay.h>
> +#include <linux/jiffies.h>
> +#include <linux/gpio/consumer.h>
> +#include <linux/pm_runtime.h>
>  
>  #include <linux/iio/iio.h>
>  
> @@ -52,9 +63,13 @@
>  #define MLX90614_TIMING_WAKEUP 34 /* time to hold SDA low for wake-up */
>  #define MLX90614_TIMING_STARTUP 250 /* time before first data after wake-up */
>  
> +#define MLX90614_AUTOSLEEP_DELAY 5000 /* default autosleep delay */
> +
>  struct mlx90614_data {
>  	struct i2c_client *client;
>  	struct mutex lock; /* for EEPROM access only */
> +	struct gpio_desc *wakeup_gpio; /* NULL to disable sleep/wake-up */
> +	unsigned long ready_timestamp; /* in jiffies */
>  };
>  
>  /*
> @@ -95,6 +110,54 @@ static s32 mlx90614_write_word(const struct i2c_client *client, u8 command,
>  	return ret;
>  }
>  
> +#ifdef CONFIG_PM
> +/*
> + * If @startup is true, make sure MLX90614_TIMING_STARTUP ms have elapsed since
> + * the last wake-up.  This is normally only needed to get a valid temperature
> + * reading.  EEPROM access does not need such delay.
> + * Return 0 on success, <0 on error.
> + */
> +static int mlx90614_power_get(struct mlx90614_data *data, bool startup)
> +{
> +	unsigned long now;
> +
> +	if (!data->wakeup_gpio)
> +		return 0;
> +
> +	pm_runtime_get_sync(&data->client->dev);
> +
> +	if (startup) {
> +		now = jiffies;
> +		if (time_before(now, data->ready_timestamp) &&
> +		    msleep_interruptible(jiffies_to_msecs(
> +				data->ready_timestamp - now)) != 0) {
> +			pm_runtime_put_autosuspend(&data->client->dev);
> +			return -EINTR;
> +		}
> +	}
> +
> +	return 0;
> +}
> +
> +static void mlx90614_power_put(struct mlx90614_data *data)
> +{
> +	if (!data->wakeup_gpio)
> +		return;
> +
> +	pm_runtime_mark_last_busy(&data->client->dev);
> +	pm_runtime_put_autosuspend(&data->client->dev);
> +}
> +#else
> +static inline int mlx90614_power_get(struct mlx90614_data *data, bool startup)
> +{
> +	return 0;
> +}
> +
> +static inline void mlx90614_power_put(struct mlx90614_data *data)
> +{
> +}
> +#endif
> +
>  static int mlx90614_read_raw(struct iio_dev *indio_dev,
>  			    struct iio_chan_spec const *channel, int *val,
>  			    int *val2, long mask)
> @@ -125,7 +188,12 @@ static int mlx90614_read_raw(struct iio_dev *indio_dev,
>  			return -EINVAL;
>  		}
>  
> +		ret = mlx90614_power_get(data, true);
> +		if (ret < 0)
> +			return ret;
>  		ret = i2c_smbus_read_word_data(data->client, cmd);
> +		mlx90614_power_put(data);
> +
>  		if (ret < 0)
>  			return ret;
>  		*val = ret;
> @@ -138,10 +206,12 @@ static int mlx90614_read_raw(struct iio_dev *indio_dev,
>  		*val = 20;
>  		return IIO_VAL_INT;
>  	case IIO_CHAN_INFO_EMISSIVITY: /* 1/65535 / LSB */
> +		mlx90614_power_get(data, false);
>  		mutex_lock(&data->lock);
>  		ret = i2c_smbus_read_word_data(data->client,
>  					       MLX90614_EMISSIVITY);
>  		mutex_unlock(&data->lock);
> +		mlx90614_power_put(data);
>  
>  		if (ret < 0)
>  			return ret;
> @@ -172,10 +242,12 @@ static int mlx90614_write_raw(struct iio_dev *indio_dev,
>  			return -EINVAL;
>  		val = val * 65535 + val2 / 15259; /* 1/65535 ~ 0.000015259 */
>  
> +		mlx90614_power_get(data, false);
>  		mutex_lock(&data->lock);
>  		ret = mlx90614_write_word(data->client, MLX90614_EMISSIVITY,
>  					  val);
>  		mutex_unlock(&data->lock);
> +		mlx90614_power_put(data);
>  
>  		if (ret < 0)
>  			return ret;
> @@ -235,6 +307,98 @@ static const struct iio_info mlx90614_info = {
>  	.driver_module = THIS_MODULE,
>  };
>  
> +#ifdef CONFIG_PM
> +static int mlx90614_sleep(struct mlx90614_data *data)
> +{
> +	s32 ret;
> +
> +	if (!data->wakeup_gpio) {
> +		dev_dbg(&data->client->dev, "Sleep disabled");
> +		return -ENOSYS;
> +	}
> +
> +	dev_dbg(&data->client->dev, "Requesting sleep");
> +
> +	mutex_lock(&data->lock);
> +	ret = i2c_smbus_xfer(data->client->adapter, data->client->addr,
> +			     data->client->flags | I2C_CLIENT_PEC,
> +			     I2C_SMBUS_WRITE, MLX90614_OP_SLEEP,
> +			     I2C_SMBUS_BYTE, NULL);
> +	mutex_unlock(&data->lock);
> +
> +	return ret;
> +}
> +
> +static int mlx90614_wakeup(struct mlx90614_data *data)
> +{
> +	if (!data->wakeup_gpio) {
> +		dev_dbg(&data->client->dev, "Wake-up disabled");
> +		return -ENOSYS;
> +	}
> +
> +	dev_dbg(&data->client->dev, "Requesting wake-up");
> +
> +	i2c_lock_adapter(data->client->adapter);
> +	gpiod_direction_output(data->wakeup_gpio, 0);
> +	msleep(MLX90614_TIMING_WAKEUP);
> +	gpiod_direction_input(data->wakeup_gpio);
> +	i2c_unlock_adapter(data->client->adapter);
> +
> +	data->ready_timestamp = jiffies +
> +			msecs_to_jiffies(MLX90614_TIMING_STARTUP);
> +
> +	/*
> +	 * Quirk: the i2c controller may get confused right after the
> +	 * wake-up signal has been sent.  As a workaround, do a dummy read.
> +	 * If the read fails, the controller will probably be reset so that
> +	 * further reads will work.
> +	 */
> +	i2c_smbus_read_word_data(data->client, MLX90614_CONFIG);
That wins the nasty award for the day!  It'll probably not work
on lots of boards, but there we are...  No obvious alternatives exist.

> +
> +	return 0;
> +}
> +
> +/* Return wake-up GPIO or NULL if sleep functionality should be disabled. */
> +static struct gpio_desc *mlx90614_probe_wakeup(struct i2c_client *client)
> +{
> +	struct gpio_desc *gpio;
> +
> +	if (!i2c_check_functionality(client->adapter,
> +						I2C_FUNC_SMBUS_WRITE_BYTE)) {
> +		dev_info(&client->dev,
> +			 "i2c adapter does not support SMBUS_WRITE_BYTE, sleep disabled");
> +		return NULL;
> +	}
> +
> +	gpio = devm_gpiod_get_optional(&client->dev, "wakeup", GPIOD_IN);
> +
> +	if (IS_ERR(gpio)) {
> +		dev_warn(&client->dev,
> +			 "gpio acquisition failed with error %ld, sleep disabled",
> +			 PTR_ERR(gpio));
> +		return NULL;
> +	} else if (!gpio) {
> +		dev_info(&client->dev,
> +			 "wakeup-gpio not found, sleep disabled");
> +	}
> +
> +	return gpio;
> +}
> +#else
> +static inline int mlx90614_sleep(struct mlx90614_data *data)
> +{
> +	return -ENOSYS;
> +}
> +static inline int mlx90614_wakeup(struct mlx90614_data *data)
> +{
> +	return -ENOSYS;
> +}
> +static inline struct gpio_desc *mlx90614_probe_wakeup(struct i2c_client *client)
> +{
> +	return NULL;
> +}
> +#endif
> +
>  /* Return 0 for single sensor, 1 for dual sensor, <0 on error. */
>  static int mlx90614_probe_num_ir_sensors(struct i2c_client *client)
>  {
> @@ -266,6 +430,9 @@ static int mlx90614_probe(struct i2c_client *client,
>  	i2c_set_clientdata(client, indio_dev);
>  	data->client = client;
>  	mutex_init(&data->lock);
> +	data->wakeup_gpio = mlx90614_probe_wakeup(client);
> +
> +	mlx90614_wakeup(data);
>  
>  	indio_dev->dev.parent = &client->dev;
>  	indio_dev->name = id->name;
> @@ -288,12 +455,30 @@ static int mlx90614_probe(struct i2c_client *client,
>  		return ret;
>  	}
>  
> +	if (data->wakeup_gpio) {
> +		pm_runtime_set_autosuspend_delay(&client->dev,
> +						 MLX90614_AUTOSLEEP_DELAY);
> +		pm_runtime_use_autosuspend(&client->dev);
> +		pm_runtime_set_active(&client->dev);
> +		pm_runtime_enable(&client->dev);
> +	}
> +
>  	return iio_device_register(indio_dev);
>  }
>  
>  static int mlx90614_remove(struct i2c_client *client)
>  {
> -	iio_device_unregister(i2c_get_clientdata(client));
> +	struct iio_dev *indio_dev = i2c_get_clientdata(client);
> +	struct mlx90614_data *data = iio_priv(indio_dev);
> +
> +	if (data->wakeup_gpio) {
> +		pm_runtime_disable(&client->dev);
> +		if (!pm_runtime_status_suspended(&client->dev))
> +			mlx90614_sleep(data);
> +		pm_runtime_set_suspended(&client->dev);
> +	}
> +
> +	iio_device_unregister(indio_dev);
Your ordering is not the reverse of probe.  The unregister should
be before the pm stuff.  While it may well not actually matter,
keeping this clean remove -> reverse operations of probe makes
for more obviously correct code so is good to maintain where
possible (and add comments where it actually needs to be different!)

>  
>  	return 0;
>  }
> @@ -304,10 +489,67 @@ static const struct i2c_device_id mlx90614_id[] = {
>  };
>  MODULE_DEVICE_TABLE(i2c, mlx90614_id);
>  
> +#ifdef CONFIG_PM_SLEEP
> +static int mlx90614_pm_suspend(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
> +	struct mlx90614_data *data = iio_priv(indio_dev);
> +
> +	if (data->wakeup_gpio && pm_runtime_active(dev))
> +		return mlx90614_sleep(data);
> +
> +	return 0;
> +}
> +
> +static int mlx90614_pm_resume(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
> +	struct mlx90614_data *data = iio_priv(indio_dev);
> +	int err;
> +
> +	if (data->wakeup_gpio) {
> +		err = mlx90614_wakeup(data);
> +		if (err < 0)
> +			return err;
> +
> +		pm_runtime_disable(dev);
> +		pm_runtime_set_active(dev);
> +		pm_runtime_enable(dev);
> +	}
> +
> +	return 0;
> +}
> +#endif
> +
> +#ifdef CONFIG_PM
> +static int mlx90614_pm_runtime_suspend(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
> +	struct mlx90614_data *data = iio_priv(indio_dev);
> +
> +	return mlx90614_sleep(data);
> +}
> +
> +static int mlx90614_pm_runtime_resume(struct device *dev)
> +{
> +	struct iio_dev *indio_dev = i2c_get_clientdata(to_i2c_client(dev));
> +	struct mlx90614_data *data = iio_priv(indio_dev);
> +
> +	return mlx90614_wakeup(data);
Could have rolled this into one line, but it doesn't really matter.
        return mlx90614_wakeup(iio_priv(i2c_get_clientdata(to_i2c_client(dev))));
> +}
> +#endif
> +
> +static const struct dev_pm_ops mlx90614_pm_ops = {
> +	SET_SYSTEM_SLEEP_PM_OPS(mlx90614_pm_suspend, mlx90614_pm_resume)
> +	SET_RUNTIME_PM_OPS(mlx90614_pm_runtime_suspend,
> +			   mlx90614_pm_runtime_resume, NULL)
> +};
> +
>  static struct i2c_driver mlx90614_driver = {
>  	.driver = {
>  		.name	= "mlx90614",
>  		.owner	= THIS_MODULE,
> +		.pm	= &mlx90614_pm_ops,
>  	},
>  	.probe = mlx90614_probe,
>  	.remove = mlx90614_remove,
> 

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

      parent reply	other threads:[~2015-03-28 12:54 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-24 15:54 [PATCH v2 0/7] iio: mlx90614 enhancements Vianney le Clément de Saint-Marcq
     [not found] ` <1427212459-31585-1-git-send-email-vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org>
2015-03-24 15:54   ` [PATCH v2 6/7] iio: mlx90614: Add power management Vianney le Clément de Saint-Marcq
     [not found]     ` <1427212459-31585-7-git-send-email-vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org>
2015-03-28 12:54       ` Jonathan Cameron [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5516A493.7090406@kernel.org \
    --to=jic23-dgejt+ai2ygdnm+yrofe0a@public.gmane.org \
    --cc=arnout-agwphqFHkxc@public.gmane.org \
    --cc=devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-iio-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=vianney.leclement-buIOx9BAs4sybS5Ee8rs3A@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).