Message-ID: <b668c7f1-99ac-f2ba-eaab-f3c0f9e1cfb2@free-electrons.com>
Date:   Tue, 20 Dec 2016 16:20:04 +0100
From:   Quentin Schulz <quentin.schulz@...e-electrons.com>
To:     Maxime Ripard <maxime.ripard@...e-electrons.com>
Cc:     jic23@...nel.org, knaack.h@....de, lars@...afoo.de,
        pmeerw@...erw.net, robh+dt@...nel.org, mark.rutland@....com,
        wens@...e.org, lee.jones@...aro.org, linux@...linux.org.uk,
        stefan.mavrodiev@...il.com, linux-iio@...r.kernel.org,
        devicetree@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-kernel@...r.kernel.org, thomas.petazzoni@...e-electrons.com
Subject: Re: [PATCH 3/7] iio: adc: sun4i-gpadc-iio: add support for A33
 thermal sensor

Hi,

On 20/12/2016 15:44, Maxime Ripard wrote:
> On Tue, Dec 20, 2016 at 11:27:05AM +0100, Quentin Schulz wrote:
>> This adds support for the Allwinner A33 thermal sensor.
>>
>> Unlike the A10, A13 and A31, the Allwinner A33 only has one channel
>> which is dedicated to the thermal sensor. Moreover, its thermal sensor
>> does not generate interruptions, thus we only need to directly read the
>> register storing the temperature value.
>>
>> The MFD used by the A10, A13 and A31, was created to avoid breaking the
>> DT binding, but since the nodes for the ADC weren't there for the A33,
>> it is not needed.
>>
>> Signed-off-by: Quentin Schulz <quentin.schulz@...e-electrons.com>
>> ---
>>  drivers/iio/adc/Kconfig           |  21 ++--
>>  drivers/iio/adc/sun4i-gpadc-iio.c | 204 ++++++++++++++++++++++++++++----------
>>  include/linux/mfd/sun4i-gpadc.h   |   4 +
>>  3 files changed, 172 insertions(+), 57 deletions(-)
>>
>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>> index 6a6d369..06041ff 100644
>> --- a/drivers/iio/adc/Kconfig
>> +++ b/drivers/iio/adc/Kconfig
>> @@ -437,17 +437,24 @@ config STX104
>>  config SUN4I_GPADC
>>  	tristate "Support for the Allwinner SoCs GPADC"
>>  	depends on IIO
>> -	depends on MFD_SUN4I_GPADC
>> -	help
>> -	  Say yes here to build support for Allwinner (A10, A13 and A31) SoCs
>> -	  GPADC. This ADC provides 4 channels which can be used as an ADC or as
>> -	  a touchscreen input and one channel for thermal sensor.
>> -
>> -	  The thermal sensor slows down ADC readings and can be disabled by
>> +# MFD_SUN4I_GPADC is needed for sun4i, sun5i and sun6i but not for sun8i
> 
> The indentation is wrong here, and I wouldn't list all the families
> there, this is just going to be always amended, for no particular
> reason.
> 

ACK.

>> +	select MFD_SUN4I_GPADC if MACH_SUN4I || MACH_SUN5I || MACH_SUN6I
> 
> Why did you change the depends on to a select?
> 

The "depends on" does not have an if condition but "select" has. See:
https://www.kernel.org/doc/Documentation/kbuild/kconfig-language.txt

> And isn't that redundant with the comment just above?
> 
>> +# THERMAL_OF can be disabled on sun4i, sun5i and sun6i to quicken ADC readings
> 
> I'm not sure this is worth adding either. Any option can be added with
> any number of side effects, I'm not sure we want to list all of them.
> 

ACK.

>> +	depends on THERMAL_OF || MACH_SUN4I || MACH_SUN5I || MACH_SUN6I
> 
> So you can't disable THERMAL_OF on MACH_SUN8I?
> 

Not in the current state. devm_thermal_zone_of_sensor_register from
sun4i_gpadc_probe_dt returns an error if THERMAL_OF is disabled and
thus, the probe fails. Maybe it should rather fail silently and let the
user choose whether (s)he wants the thermal framework to be able to read
data from this driver?

>> +	depends on !TOUCHSCREEN_SUN4I
> 
> This should be a different patch.
> 
>> +	help
>> +	  Say yes here to build support for Allwinner (A10, A13, A31 and A33)
>> +	  SoCs GPADC.
>> +
>> +	  The ADC on A10, A13 and A31 provides 4 channels which can be used as
>> +	  an ADC or as a touchscreen input and one channel for thermal sensor.
>> +	  Their thermal sensor slows down ADC readings and can be disabled by
> 
> Again, I'm not sure putting all those details in the Kconfig help
> really helps. This is only going to grow with more and more cases, and
> this isn't something really helpful anyway.
> 

Are you suggesting to remove completely the paragraph on why it is
possible to disable CONFIG_THERMAL_OF for the A10, A13 and A31 or only
to remove the mention to SoCs?

>>  	  disabling CONFIG_THERMAL_OF. However, the thermal sensor should be
>>  	  enabled by default since the SoC temperature is usually more critical
>>  	  than ADC readings.
>>  
>> +	  The ADC on A33 provides one channel for thermal sensor.
>> +
>>  	  To compile this driver as a module, choose M here: the module will be
>>  	  called sun4i-gpadc-iio.
>>  
>> diff --git a/drivers/iio/adc/sun4i-gpadc-iio.c b/drivers/iio/adc/sun4i-gpadc-iio.c
>> index a8e134f..8be694e 100644
>> --- a/drivers/iio/adc/sun4i-gpadc-iio.c
>> +++ b/drivers/iio/adc/sun4i-gpadc-iio.c
>> @@ -1,4 +1,4 @@
>> -/* ADC driver for sunxi platforms' (A10, A13 and A31) GPADC
>> +/* ADC driver for sunxi platforms' (A10, A13, A31 and A33) GPADC
>>   *
>>   * Copyright (c) 2016 Quentin Schulz <quentin.schulz@...e-electrons.com>
>>   *
>> @@ -85,6 +85,12 @@ static const struct gpadc_data sun6i_gpadc_data = {
>>  	.adc_chan_mask = SUN6I_GPADC_CTRL1_ADC_CHAN_MASK,
>>  };
>>  
>> +static const struct gpadc_data sun8i_gpadc_data = {
>> +	.temp_offset = -1662,
>> +	.temp_scale = 162,
>> +	.tp_mode_en = SUN8I_GPADC_CTRL1_CHOP_TEMP_EN,
>> +};
>> +
>>  struct sun4i_gpadc_iio {
>>  	struct iio_dev			*indio_dev;
>>  	struct completion		completion;
>> @@ -96,6 +102,7 @@ struct sun4i_gpadc_iio {
>>  	unsigned int			temp_data_irq;
>>  	atomic_t			ignore_temp_data_irq;
>>  	const struct gpadc_data		*data;
>> +	bool				use_dt;
>>  	/* prevents concurrent reads of temperature and ADC */
>>  	struct mutex			mutex;
>>  };
>> @@ -138,6 +145,23 @@ static const struct iio_chan_spec sun4i_gpadc_channels_no_temp[] = {
>>  	SUN4I_GPADC_ADC_CHANNEL(3, "adc_chan3"),
>>  };
>>  
>> +static const struct iio_chan_spec sun8i_gpadc_channels[] = {
> 
> sun8i_a33. Other SoCs from that same family might have different
> channels.
> 

ACK.

>> +	{
>> +		.type = IIO_TEMP,
>> +		.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
>> +				      BIT(IIO_CHAN_INFO_SCALE) |
>> +				      BIT(IIO_CHAN_INFO_OFFSET),
>> +		.datasheet_name = "temp_adc",
>> +	},
>> +};
>> +
>> +static const struct regmap_config sun4i_gpadc_regmap_config = {
>> +	.reg_bits = 32,
>> +	.val_bits = 32,
>> +	.reg_stride = 4,
>> +	.fast_io = true,
>> +};
>> +
>>  static int sun4i_prepare_for_irq(struct iio_dev *indio_dev, int channel,
>>  				 unsigned int irq)
>>  {
>> @@ -231,7 +255,6 @@ static int sun4i_gpadc_read(struct iio_dev *indio_dev, int channel, int *val,
>>  err:
>>  	pm_runtime_put_autosuspend(indio_dev->dev.parent);
>>  	mutex_unlock(&info->mutex);
>> -
>>  	return ret;
>>  }
>>  
>> @@ -246,6 +269,19 @@ static int sun4i_gpadc_adc_read(struct iio_dev *indio_dev, int channel,
>>  static int sun4i_gpadc_temp_read(struct iio_dev *indio_dev, int *val)
>>  {
>>  	struct sun4i_gpadc_iio *info = iio_priv(indio_dev);
>> +	int ret;
>> +
>> +	if (info->use_dt) {
>> +		pm_runtime_get_sync(indio_dev->dev.parent);
>> +
>> +		ret = regmap_read(info->regmap, SUN4I_GPADC_TEMP_DATA, val);
>> +		if (!ret)
>> +			pm_runtime_mark_last_busy(indio_dev->dev.parent);
>> +
>> +		pm_runtime_put_autosuspend(indio_dev->dev.parent);
>> +
>> +		return 0;
>> +	}
> 
> Why is runtime_pm linked to the DT support or not?
> 
>>  
>>  	return sun4i_gpadc_read(indio_dev, 0, val, info->temp_data_irq);
>>  }

The same runtime_pm functions are called when the driver is not probed
from DT.

sun4i_gpadc_read will call sun4i_prepare_for_irq which does a
pm_runtime_get_sync and then at the end of sun4i_gpadc_read,
pm_runtime_mark_last_busy and pm_runtime_put_autosuspend are called.

I just noticed I forgot to add a comment on this one. The A33
documentation tells us there is an interrupt for the thermal sensor but
after struggling with it, it is just false. I validated my guess with
Allwinner Linux kernel which does not wait an interrupt to read the data
register of the thermal sensor.

sun4i_gpadc_read always wait for an interrupt before reading data regs,
so I'm just not calling it and doing the "logic" (reading the data reg
and interfacing with runtime_pm) directly here in sun4i_gpadc_temp.

I'll add a comment on this one. Maybe I should create a function just
for the "logic" of the A33 thermal sensor, so it is less weird than
currently.

>> @@ -410,7 +446,7 @@ static int sun4i_irq_init(struct platform_device *pdev, const char *name,
>>  			  unsigned int *irq, atomic_t *atomic)
>>  {
>>  	int ret;
>> -	struct sun4i_gpadc_dev *mfd_dev = dev_get_drvdata(pdev->dev.parent);
>> +	struct sun4i_gpadc_dev *mfd_dev;
>>  	struct sun4i_gpadc_iio *info = iio_priv(dev_get_drvdata(&pdev->dev));
>>  
>>  	/*
>> @@ -427,6 +463,8 @@ static int sun4i_irq_init(struct platform_device *pdev, const char *name,
>>  	 */
>>  	atomic_set(atomic, 1);
>>  
>> +	mfd_dev = dev_get_drvdata(pdev->dev.parent);
>> +
> 
> Why is that change necessary?
> 
>>  	ret = platform_get_irq_byname(pdev, name);
>>  	if (ret < 0) {
>>  		dev_err(&pdev->dev, "no %s interrupt registered\n", name);
>> @@ -454,31 +492,68 @@ static int sun4i_irq_init(struct platform_device *pdev, const char *name,
>>  	return 0;
>>  }
>>  
>> -static int sun4i_gpadc_probe(struct platform_device *pdev)
>> +static const struct of_device_id sun4i_gpadc_of_id[] = {
>> +	{
>> +		.compatible = "allwinner,sun8i-a33-gpadc-iio",
>> +		.data = &sun8i_gpadc_data,
>> +	},
>> +	{ /* sentinel */ }
>> +};
>> +
>> +static int sun4i_gpadc_probe_dt(struct platform_device *pdev,
>> +				struct iio_dev *indio_dev)
>>  {
>> -	struct sun4i_gpadc_iio *info;
>> -	struct iio_dev *indio_dev;
>> +	struct sun4i_gpadc_iio *info = iio_priv(indio_dev);
>> +	const struct of_device_id *of_dev;
>> +	struct thermal_zone_device *tzd;
>> +	struct resource *mem;
>> +	void __iomem *base;
>>  	int ret;
>> -	struct sun4i_gpadc_dev *sun4i_gpadc_dev;
>>  
>> -	sun4i_gpadc_dev = dev_get_drvdata(pdev->dev.parent);
>> +	of_dev = of_match_device(sun4i_gpadc_of_id, &pdev->dev);
>> +	if (!of_dev)
>> +		return -ENODEV;
>> +
>> +	info->use_dt = true;
>> +	info->data = (struct gpadc_data *)of_dev->data;
>> +	indio_dev->num_channels = ARRAY_SIZE(sun8i_gpadc_channels);
>> +	indio_dev->channels = sun8i_gpadc_channels;
>> +
>> +	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>> +	base = devm_ioremap_resource(&pdev->dev, mem);
>> +	if (IS_ERR(base))
>> +		return PTR_ERR(base);
>> +
>> +	info->regmap = devm_regmap_init_mmio(&pdev->dev, base,
>> +					     &sun4i_gpadc_regmap_config);
>> +	if (IS_ERR(info->regmap)) {
>> +		ret = PTR_ERR(info->regmap);
>> +		dev_err(&pdev->dev, "failed to init regmap: %d\n", ret);
>> +		return ret;
>> +	}
>>  
>> -	indio_dev = devm_iio_device_alloc(&pdev->dev, sizeof(*info));
>> -	if (!indio_dev)
>> -		return -ENOMEM;
>> +	tzd = devm_thermal_zone_of_sensor_register(&pdev->dev, 0, info,
>> +						   &sun4i_ts_tz_ops);
>> +	if (IS_ERR(tzd)) {
>> +		dev_err(&pdev->dev, "could not register thermal sensor: %ld\n",
>> +			PTR_ERR(tzd));
>> +		return PTR_ERR(tzd);
>> +	}
>>  
>> -	info = iio_priv(indio_dev);
>> -	platform_set_drvdata(pdev, indio_dev);
>> +	return 0;
>> +}
>>  
>> -	mutex_init(&info->mutex);
>> +static int sun4i_gpadc_probe_mfd(struct platform_device *pdev,
>> +				 struct iio_dev *indio_dev)
>> +{
>> +	struct sun4i_gpadc_iio *info = iio_priv(indio_dev);
>> +	struct sun4i_gpadc_dev *sun4i_gpadc_dev;
>> +	int ret;
>> +
>> +	info->use_dt = false;
>> +	sun4i_gpadc_dev = dev_get_drvdata(pdev->dev.parent);
>>  	info->regmap = sun4i_gpadc_dev->regmap;
>> -	info->indio_dev = indio_dev;
>> -	init_completion(&info->completion);
>> -	indio_dev->name = dev_name(&pdev->dev);
>> -	indio_dev->dev.parent = &pdev->dev;
>> -	indio_dev->dev.of_node = pdev->dev.of_node;
>> -	indio_dev->info = &sun4i_gpadc_iio_info;
>> -	indio_dev->modes = INDIO_DIRECT_MODE;
>> +
> 
> You should have two patches. One to split the MFD part from the probe,
> and a second one to add the DT probing.
> 

ACK.

>>  	indio_dev->num_channels = ARRAY_SIZE(sun4i_gpadc_channels);
>>  	indio_dev->channels = sun4i_gpadc_channels;
>>  
>> @@ -494,7 +569,6 @@ static int sun4i_gpadc_probe(struct platform_device *pdev)
>>  	 * register the sensor if that option is enabled, eventually leaving
>>  	 * that choice to the user.
>>  	 */
>> -
> 
> Spurious change?
> 

Yes, lots of. Sorry, I will be more careful on that.

>>  	if (IS_ENABLED(CONFIG_THERMAL_OF)) {
>>  		/*
>>  		 * This driver is a child of an MFD which has a node in the DT
>> @@ -519,8 +593,7 @@ static int sun4i_gpadc_probe(struct platform_device *pdev)
>>  			dev_err(&pdev->dev,
>>  				"could not register thermal sensor: %ld\n",
>>  				PTR_ERR(tzd));
>> -			ret = PTR_ERR(tzd);
>> -			goto err;
>> +			return PTR_ERR(tzd);
>>  		}
>>  	} else {
>>  		indio_dev->num_channels =
>> @@ -528,49 +601,78 @@ static int sun4i_gpadc_probe(struct platform_device *pdev)
>>  		indio_dev->channels = sun4i_gpadc_channels_no_temp;
>>  	}
>>  
>> -	pm_runtime_set_autosuspend_delay(&pdev->dev,
>> -					 SUN4I_GPADC_AUTOSUSPEND_DELAY);
>> -	pm_runtime_use_autosuspend(&pdev->dev);
>> -	pm_runtime_set_suspended(&pdev->dev);
>> -	pm_runtime_enable(&pdev->dev);
>> -
>>  	if (IS_ENABLED(CONFIG_THERMAL_OF)) {
>>  		ret = sun4i_irq_init(pdev, "TEMP_DATA_PENDING",
>>  				     sun4i_gpadc_temp_data_irq_handler,
>>  				     "temp_data", &info->temp_data_irq,
>>  				     &info->ignore_temp_data_irq);
>>  		if (ret < 0)
>> -			goto err;
>> -	}
>> +			return ret;
>>  
>> -	ret = sun4i_irq_init(pdev, "FIFO_DATA_PENDING",
>> -			     sun4i_gpadc_fifo_data_irq_handler, "fifo_data",
>> -			     &info->fifo_data_irq, &info->ignore_fifo_data_irq);
>> -	if (ret < 0)
>> -		goto err;
>> +		ret = sun4i_irq_init(pdev, "FIFO_DATA_PENDING",
>> +				     sun4i_gpadc_fifo_data_irq_handler,
>> +				     "fifo_data", &info->fifo_data_irq,
>> +				     &info->ignore_fifo_data_irq);
>> +		if (ret < 0)
>> +			return ret;
>> +	}
>>  
>> -	if (IS_ENABLED(CONFIG_THERMAL_OF)) {
>> -		ret = iio_map_array_register(indio_dev, sun4i_gpadc_hwmon_maps);
>> -		if (ret < 0) {
>> -			dev_err(&pdev->dev,
>> -				"failed to register iio map array\n");
>> -			goto err;
>> -		}
>> +	ret = iio_map_array_register(indio_dev, sun4i_gpadc_hwmon_maps);
>> +	if (ret < 0) {
>> +		dev_err(&pdev->dev, "failed to register iio map array\n");
>> +		return ret;
>>  	}
>>  
>> +	return 0;
>> +}
>> +
>> +static int sun4i_gpadc_probe(struct platform_device *pdev)
>> +{
>> +	struct sun4i_gpadc_iio *info;
>> +	struct iio_dev *indio_dev;
>> +	int ret;
>> +
>> +	indio_dev = devm_iio_device_alloc(&pdev->dev, sizeof(*info));
>> +	if (!indio_dev)
>> +		return -ENOMEM;
>> +
>> +	info = iio_priv(indio_dev);
>> +	platform_set_drvdata(pdev, indio_dev);
>> +
>> +	mutex_init(&info->mutex);
>> +	info->indio_dev = indio_dev;
>> +	init_completion(&info->completion);
>> +	indio_dev->name = dev_name(&pdev->dev);
>> +	indio_dev->dev.parent = &pdev->dev;
>> +	indio_dev->dev.of_node = pdev->dev.of_node;
>> +	indio_dev->info = &sun4i_gpadc_iio_info;
>> +	indio_dev->modes = INDIO_DIRECT_MODE;
>> +
>> +	if (pdev->dev.of_node)
>> +		ret = sun4i_gpadc_probe_dt(pdev, indio_dev);
>> +	else
>> +		ret = sun4i_gpadc_probe_mfd(pdev, indio_dev);
>> +
>> +	if (ret)
>> +		return ret;
>> +
>> +	pm_runtime_set_autosuspend_delay(&pdev->dev,
>> +					 SUN4I_GPADC_AUTOSUSPEND_DELAY);
>> +	pm_runtime_use_autosuspend(&pdev->dev);
>> +	pm_runtime_set_suspended(&pdev->dev);
>> +	pm_runtime_enable(&pdev->dev);
>> +
>>  	ret = devm_iio_device_register(&pdev->dev, indio_dev);
>>  	if (ret < 0) {
>>  		dev_err(&pdev->dev, "could not register the device\n");
>> -		goto err_map;
>> +		goto err;
>>  	}
>>  
>>  	return 0;
>>  
>> -err_map:
>> -	if (IS_ENABLED(CONFIG_THERMAL_OF))
>> -		iio_map_array_unregister(indio_dev);
>> -
>>  err:
>> +	if (!info->use_dt && IS_ENABLED(CONFIG_THERMAL_OF))
>> +		iio_map_array_unregister(indio_dev);
>>  	pm_runtime_put(&pdev->dev);
>>  	pm_runtime_disable(&pdev->dev);
>>  
>> @@ -580,10 +682,11 @@ static int sun4i_gpadc_probe(struct platform_device *pdev)
>>  static int sun4i_gpadc_remove(struct platform_device *pdev)
>>  {
>>  	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
>> +	struct sun4i_gpadc_iio *info = iio_priv(indio_dev);
>>  
>>  	pm_runtime_put(&pdev->dev);
>>  	pm_runtime_disable(&pdev->dev);
>> -	if (IS_ENABLED(CONFIG_THERMAL_OF))
>> +	if (!info->use_dt && IS_ENABLED(CONFIG_THERMAL_OF))
>>  		iio_map_array_unregister(indio_dev);
>>  
>>  	return 0;
>> @@ -599,6 +702,7 @@ static const struct platform_device_id sun4i_gpadc_id[] = {
>>  static struct platform_driver sun4i_gpadc_driver = {
>>  	.driver = {
>>  		.name = "sun4i-gpadc-iio",
>> +		.of_match_table = sun4i_gpadc_of_id,
>>  		.pm = &sun4i_gpadc_pm_ops,
>>  	},
>>  	.id_table = sun4i_gpadc_id,
>> diff --git a/include/linux/mfd/sun4i-gpadc.h b/include/linux/mfd/sun4i-gpadc.h
>> index 509e736..139872c 100644
>> --- a/include/linux/mfd/sun4i-gpadc.h
>> +++ b/include/linux/mfd/sun4i-gpadc.h
>> @@ -38,6 +38,10 @@
>>  #define SUN6I_GPADC_CTRL1_ADC_CHAN_SELECT(x)		(GENMASK(3, 0) & BIT(x))
>>  #define SUN6I_GPADC_CTRL1_ADC_CHAN_MASK			GENMASK(3, 0)
>>  
>> +/* TP_CTRL1 bits for sun8i SoCs */
>> +#define SUN8I_GPADC_CTRL1_CHOP_TEMP_EN			BIT(8)
>> +#define SUN8I_GPADC_CTRL1_GPADC_CALI_EN			BIT(7)
>> +
>>  #define SUN4I_GPADC_CTRL2				0x08
>>  
>>  #define SUN4I_GPADC_CTRL2_TP_SENSITIVE_ADJUST(x)	((GENMASK(3, 0) & (x)) << 28)
>> -- 
>> 2.9.3
>>
> 
> Thanks,
> Maxime
> 

Thanks,
Quentin

-- 
Quentin Schulz, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com



Download attachment "signature.asc" of type "application/pgp-signature" (802 bytes)

Powered by blists - more mailing lists