linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: [RFC PATCH 4/8] hwmon: Add DA906x hardware monitoring support.
       [not found] ` <201208241505@sw-eng-lt-dc-vm2>
@ 2012-08-24 18:45   ` Guenter Roeck
  2012-08-29 13:25     ` [PATCH] regulator: Fix bug in regulator_mode_to_status() core function Krystian Garbaciak
  0 siblings, 1 reply; 2+ messages in thread
From: Guenter Roeck @ 2012-08-24 18:45 UTC (permalink / raw)
  To: Krystian Garbaciak
  Cc: linux-kernel, rtc-linux, lm-sensors, linux-input, linux-watchdog,
	linux-leds, Samuel Ortiz, Liam Girdwood, Mark Brown,
	Alessandro Zummo, Jean Delvare, Dmitry Torokhov, Ashish Jangam,
	Andrew Jones, Donggeun Kim, Philippe Rétornaz,
	Wim Van Sebroeck, Bryan Wu,
	Richard Purdie <rpurdie@rpsys.net> Anthony Olech

On Fri, Aug 24, 2012 at 03:05:00PM +0100, Krystian Garbaciak wrote:
> DA906x PMIC provides ADC for voltage and temperature monitoring.
> 
Hi Krystian,

DA906x seems to be a bad choice for a name. It covers DA906[0-9] and possibly
even DA906[A-Z]. The only chip really in existence seems to be DA9064.

I personally think it is a bad idea to have an 'x' in a driver name. What if
DA9069 shows up at some point and is completely different ? I think the driver
should be named for the first supported chip; reference the others in the code
and documentation.

Worse, I find no information anywhere in your patch set indicating which chips
are actually supported. I don't know how other subsystems handle this, but for
hwmon this is a no-go.

> The driver provides results of following ADC channels:
>  - in0 - system voltage (2500 - 5500 mV)
>  - in1 - universal voltage channel #1 (0 - 2500 mV)
>  - in2 - universal voltage channel #2 (0 - 2500 mV)
>  - in3 - universal voltage channel #3 (0 - 2500 mV)
>  - in4 - backup battery voltage (0 - 5000 mV)
>  - temp1 - PMIC internal junction temperature (-88 - 333 Celcius degrees)
> 
> Signed-off-by: Krystian Garbaciak <krystian.garbaciak@diasemi.com>
> ---
>  drivers/hwmon/Kconfig        |    6 +
>  drivers/hwmon/Makefile       |    1 +
>  drivers/hwmon/da906x-hwmon.c |  393 ++++++++++++++++++++++++++++++++++++++++++

Also please provide Documentation/hwmon/da906x-hwmon.

>  3 files changed, 400 insertions(+), 0 deletions(-)
>  create mode 100644 drivers/hwmon/da906x-hwmon.c
> 
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index b0a2e4c..7abc9a0 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -1373,6 +1373,12 @@ config SENSORS_WM8350
>  	  This driver can also be built as a module.  If so, the module
>  	  will be called wm8350-hwmon.
>  
> +config SENSORS_DA906X
> +	tristate "DA906X HWMON device drivers"
> +	depends on MFD_DA906X
> +	help
> +	  Support for the HWMON DA906X device driver.
> +

Alphabetical order, please, and describe which chip(s) are supported. And you
don't really support a device driver, the device driver presumably supports a
chip. Since it is tristate, we also expect you to provide the module name if the
driver is built as module.

>  config SENSORS_ULTRA45
>  	tristate "Sun Ultra45 PIC16F747"
>  	depends on SPARC64
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index 7aa9811..ffbe151 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -43,6 +43,7 @@ obj-$(CONFIG_SENSORS_ASC7621)	+= asc7621.o
>  obj-$(CONFIG_SENSORS_ATXP1)	+= atxp1.o
>  obj-$(CONFIG_SENSORS_CORETEMP)	+= coretemp.o
>  obj-$(CONFIG_SENSORS_DA9052_ADC)+= da9052-hwmon.o
> +obj-$(CONFIG_SENSORS_DA906X)	+= da906x-hwmon.o
>  obj-$(CONFIG_SENSORS_DME1737)	+= dme1737.o
>  obj-$(CONFIG_SENSORS_DS620)	+= ds620.o
>  obj-$(CONFIG_SENSORS_DS1621)	+= ds1621.o
> diff --git a/drivers/hwmon/da906x-hwmon.c b/drivers/hwmon/da906x-hwmon.c
> new file mode 100644
> index 0000000..8ece931
> --- /dev/null
> +++ b/drivers/hwmon/da906x-hwmon.c
> @@ -0,0 +1,393 @@
> +/*
> + * HW Monitor support for Dialog DA906X PMIC series
> + *
> + * Copyright 2012 Dialog Semiconductor Ltd.
> + *
> + * Author: Krystian Garbaciak <krystian.garbaciak@diasemi.com>,
> + *         Michal Hajduk <michal.hajduk@diasemi.com>
> + *
> + *  This program is free software; you can redistribute  it and/or modify it
> + *  under  the terms of  the GNU General  Public License as published by the
> + *  Free Software Foundation;  either version 2 of the  License, or (at your
> + *  option) any later version.
> + *
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/err.h>
> +#include <linux/delay.h>
> +#include <linux/init.h>
> +#include <linux/slab.h>
> +#include <linux/string.h>
> +#include <linux/platform_device.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/mfd/da906x/core.h>
> +#include <linux/mfd/da906x/pdata.h>
> +
> +/* ADC resolutions for manual and auto modes */
> +#define DA906X_ADC_RES		\
> +		(1 << (DA906X_ADC_RES_L_BITS + DA906X_ADC_RES_M_BITS))
> +#define DA906X_ADC_MAX		(DA906X_ADC_RES - 1)
> +#define	DA906X_ADC_AUTO_RES	(1 << DA906X_ADC_RES_M_BITS)
> +#define	DA906X_ADC_AUTO_MAX	(DA906X_ADC_AUTO_RES - 1)
> +
> +/* Define interpolation table to calculate ADC values  */
> +struct i_table {
> +	int x0;
> +	int a;
> +	int b;
> +};
> +#define ILINE(x1, x2, y1, y2)	{ \
> +		.x0 = (x1), \
> +		.a = ((y2) - (y1)) * DA906X_ADC_RES / ((x2) - (x1)), \
> +		.b = (y1) - ((y2) - (y1)) * (x1) / ((x2) - (x1)), \
> +	}
> +
> +struct channel_info {
> +	const char *name;
> +	const struct i_table *tbl;
> +	int tbl_max;
> +	u16 reg_auto_en;
> +};
> +
> +enum da906x_adc {
> +	DA906X_VSYS,
> +	DA906X_ADCIN1,
> +	DA906X_ADCIN2,
> +	DA906X_ADCIN3,
> +	DA906X_TJUNC,
> +	DA906X_VBBAT,
> +
> +	DA906X_CHAN_NUM
> +};
> +
> +static const struct i_table vsys_tbl[] = {
> +	ILINE(0, DA906X_ADC_MAX, 2500, 5500)
> +};
> +
> +static const struct i_table adcin_tbl[] = {
> +	ILINE(0, DA906X_ADC_MAX, 0, 2500)
> +};
> +
> +static const struct i_table tjunc_tbl[] = {
> +	ILINE(0, DA906X_ADC_MAX, 333, -86)
> +};
> +
> +static const struct i_table vbbat_tbl[] = {
> +	ILINE(0, DA906X_ADC_MAX, 0, 5000)
> +};

Since the first parameter to ILINE is always 0, it is not needed. This means
that x0 in itable is also always 0 and thus not needed. 

> +
> +static const struct channel_info da906x_channels[] = {
> +	[DA906X_VSYS]	= { "VSYS",
> +			    vsys_tbl, ARRAY_SIZE(vsys_tbl) - 1,
> +			    DA906X_REG_VSYS_RES },
> +	[DA906X_ADCIN1]	= { "ADCIN1",
> +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> +			    DA906X_REG_ADCIN1_RES },
> +	[DA906X_ADCIN2]	= { "ADCIN2",
> +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> +			    DA906X_REG_ADCIN2_RES },
> +	[DA906X_ADCIN3]	= { "ADCIN3",
> +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> +			    DA906X_REG_ADCIN3_RES },
> +	[DA906X_TJUNC]	= { "TJUNC",
> +			    tjunc_tbl,	ARRAY_SIZE(tjunc_tbl) - 1 },
> +	[DA906X_VBBAT]	= { "VBBAT",
> +			    vbbat_tbl,	ARRAY_SIZE(vbbat_tbl) - 1}

	s/1}/1 }/

> +};

You lost me here a bit (I am missing something ?).

Seems to be each table has exactly one entry. Since the table size is 1,
ARRAY_SIZE(vbbat_tbl) - 1 is 0, and ...

> +#define DA906X_ADC_AUTO_MODE_SUPPORT_MASK	(DA906X_ADC_AUTO_VSYS_EN | \
> +						 DA906X_ADC_AUTO_AD1_EN | \
> +						 DA906X_ADC_AUTO_AD2_EN | \
> +						 DA906X_ADC_AUTO_AD3_EN)
> +
> +struct da906x_hwmon {
> +	struct da906x *da906x;
> +	struct device *class_dev;
> +	struct completion man_adc_rdy;	/* Manual read completion flag */
> +	struct mutex hwmon_mutex;	/* Queue concurent manual reads */
> +	int irq;
> +	u8 adc_auto_en;     /* Bitmask of channels with auto mode enabled */
> +	s8 tjunc_offset;    /* Calibration offset for junction temperature */
> +};
> +
> +int da906x_adc_convert(int channel, int x)
> +{
> +	const struct channel_info *info = &da906x_channels[channel];
> +	int i, ret;
> +
> +	for (i = info->tbl_max; i > 0; i--)
> +		if (info->tbl[i].x0 <= x)
> +			break;

... this loop never does anything because info->tbl_max is always 0.
Besides, even if the loop was used, x0 is always 0 anyway, so you might well
compare against 0 instead which doesn't make sense.

So what is the point for making the code that complex ? For me it just adds a
lot of confusion.

> +
> +	ret = info->tbl[i].a * x;
> +	if (ret >= 0)
> +		ret += DA906X_ADC_RES / 2;
> +	else
> +		ret -= DA906X_ADC_RES / 2;
> +	ret = ret / DA906X_ADC_RES + info->tbl[i].b;

	ret = DIV_ROUND_CLOSEST(ret, DA906X_ADC_RES) + info->tbl[i].b;

has the same effect as the 5 lines above and is simpler.

> +	return ret;
> +}
> +
> +static int da906x_adc_manual_read(struct da906x_hwmon *hwmon, int channel)
> +{
> +	int ret;
> +	u8 data[2];
> +
> +	mutex_lock(&hwmon->hwmon_mutex);
> +
> +	init_completion(&hwmon->man_adc_rdy);
> +
> +	/* Start measurment on selected channel */
> +	data[0] = (channel << DA906X_ADC_MUX_SHIFT) & DA906X_ADC_MUX_MASK;
> +	data[0] |= DA906X_ADC_MAN;
> +	ret = da906x_reg_update(hwmon->da906x, DA906X_REG_ADC_MAN,
> +				DA906X_ADC_MUX_MASK | DA906X_ADC_MAN, data[0]);
> +	if (ret < 0)
> +		goto out;
> +
> +	/* Wait for interrupt from ADC */
> +	ret = wait_for_completion_timeout(&hwmon->man_adc_rdy,
> +					  msecs_to_jiffies(1000));
> +	if (ret == 0) {
> +		ret = -EBUSY;

Should this be -ETIMEDOUT ?

> +		goto out;
> +	}
> +
> +	/* Get results */
> +	ret = da906x_block_read(hwmon->da906x, DA906X_REG_ADC_RES_L, 2, data);
> +	if (ret < 0)
> +		goto out;
> +	ret = (data[0] & DA906X_ADC_RES_L_MASK) >> DA906X_ADC_RES_L_SHIFT;
> +	ret |= data[1] << DA906X_ADC_RES_L_BITS;
> +out:
> +	mutex_unlock(&hwmon->hwmon_mutex);
> +	return ret;
> +}
> +
> +static int da906x_adc_auto_read(struct da906x *da906x, int channel)
> +{
> +	const struct channel_info *info = &da906x_channels[channel];
> +
> +	return da906x_reg_read(da906x, info->reg_auto_en);
> +}
> +
> +static irqreturn_t da906x_hwmon_irq_handler(int irq, void *irq_data)
> +{
> +	struct da906x_hwmon *hwmon = irq_data;
> +
> +	complete(&hwmon->man_adc_rdy);
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static ssize_t da906x_adc_read(struct device *dev,
> +			       struct device_attribute *devattr, char *buf)
> +{
> +	struct da906x_hwmon *hwmon = dev_get_drvdata(dev);
> +	int channel = to_sensor_dev_attr(devattr)->index;
> +	int val;
> +
> +	if (hwmon->adc_auto_en & (1 << channel)) {
> +		val = da906x_adc_auto_read(hwmon->da906x, channel);
> +		if (val < 0)
> +			return val;
> +
> +		val *= DA906X_ADC_RES / DA906X_ADC_AUTO_RES;
> +		val = da906x_adc_convert(channel, val);
> +	} else {
> +		val = da906x_adc_manual_read(hwmon, channel);
> +		if (val < 0)
> +			return val;
> +
> +		if (channel == DA906X_TJUNC)
> +			val += hwmon->tjunc_offset;
> +		val = da906x_adc_convert(channel, val);

This call is really the same for both the if and else path. Might as well put it
after the conditional code.

> +	}
> +
> +	return sprintf(buf, "%d\n", val);
> +}
> +
> +static ssize_t da906x_show_name(struct device *dev,
> +	struct device_attribute *attr, char *buf)


Please align all multi-line function parameters with the opening (.

> +{
> +	return sprintf(buf, DA906X_DRVNAME_HWMON "\n");
> +}
> +
> +static ssize_t da906x_show_label(struct device *dev,
> +				 struct device_attribute *devattr, char *buf)
> +{
> +	const struct channel_info *info;
> +
> +	info = &da906x_channels[to_sensor_dev_attr(devattr)->index];
> +	return sprintf(buf, "%s\n", info->name);
> +}
> +
> +/* Vsys voltage */
> +static SENSOR_DEVICE_ATTR(in0_input, S_IRUGO,
> +			  da906x_adc_read, NULL, DA906X_VSYS);
> +static SENSOR_DEVICE_ATTR(in0_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_VSYS);
> +
> +/* Universal ADC channel #1 */
> +static SENSOR_DEVICE_ATTR(in1_input, S_IRUGO,
> +			  da906x_adc_read, NULL, DA906X_ADCIN1);
> +static SENSOR_DEVICE_ATTR(in1_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_ADCIN1);
> +
> +/* Universal ADC channel #2 */
> +static SENSOR_DEVICE_ATTR(in2_input, S_IRUGO,
> +			  da906x_adc_read, NULL,
> +			  DA906X_ADCIN2);
> +static SENSOR_DEVICE_ATTR(in2_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_ADCIN2);
> +
> +/* Universal ADC channel #3 */
> +static SENSOR_DEVICE_ATTR(in3_input, S_IRUGO,
> +			  da906x_adc_read, NULL, DA906X_ADCIN3);
> +static SENSOR_DEVICE_ATTR(in3_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_ADCIN3);
> +
> +/* Backup battery voltage */
> +static SENSOR_DEVICE_ATTR(in4_input, S_IRUGO,
> +			  da906x_adc_read, NULL, DA906X_VBBAT);
> +static SENSOR_DEVICE_ATTR(in4_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_VBBAT);
> +
> +/* Junction temperature */
> +static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO,
> +			  da906x_adc_read, NULL, DA906X_TJUNC);
> +
> +static SENSOR_DEVICE_ATTR(temp1_label, S_IRUGO,
> +			  da906x_show_label, NULL, DA906X_TJUNC);
> +
> +/* Device name */
> +static DEVICE_ATTR(name, S_IRUGO, da906x_show_name, NULL);
> +
> +static struct attribute *da906x_attributes[] = {
> +	&dev_attr_name.attr,
> +	&sensor_dev_attr_in0_input.dev_attr.attr,
> +	&sensor_dev_attr_in0_label.dev_attr.attr,
> +	&sensor_dev_attr_in1_input.dev_attr.attr,
> +	&sensor_dev_attr_in1_label.dev_attr.attr,
> +	&sensor_dev_attr_in2_input.dev_attr.attr,
> +	&sensor_dev_attr_in2_label.dev_attr.attr,
> +	&sensor_dev_attr_in3_input.dev_attr.attr,
> +	&sensor_dev_attr_in3_label.dev_attr.attr,
> +	&sensor_dev_attr_in4_input.dev_attr.attr,
> +	&sensor_dev_attr_in4_label.dev_attr.attr,
> +	&sensor_dev_attr_temp1_input.dev_attr.attr,
> +	&sensor_dev_attr_temp1_label.dev_attr.attr,
> +	NULL
> +};
> +
> +static const struct attribute_group da906x_attr_group = {
> +	.attrs = da906x_attributes,
> +};
> +
> +static int __devinit da906x_hwmon_probe(struct platform_device *pdev)
> +{
> +	struct da906x *da906x = dev_get_drvdata(pdev->dev.parent);
> +	struct da906x_pdata *da906x_pdata = dev_get_platdata(da906x->dev);
> +	struct da906x_hwmon *hwmon;
> +	int ret;
> +
> +	hwmon = devm_kzalloc(&pdev->dev, sizeof(struct da906x_hwmon),
> +			     GFP_KERNEL);
> +	if (!hwmon)
> +		return -ENOMEM;
> +
> +	mutex_init(&hwmon->hwmon_mutex);
> +	init_completion(&hwmon->man_adc_rdy);
> +	hwmon->da906x = da906x;
> +
> +	ret = da906x_reg_read(da906x, DA906X_REG_ADC_CONT);
> +	if (ret < 0)
> +		return ret;
> +	hwmon->adc_auto_en = ret & DA906X_ADC_AUTO_MODE_SUPPORT_MASK;
> +
> +	if (da906x_pdata->flags & DA906X_FLG_FORCE_IN0_MANUAL_MODE)

You should check if da906x_pdata is NULL before using it.

> +		hwmon->adc_auto_en &= ~DA906X_ADC_AUTO_VSYS_EN;
> +	else if (da906x_pdata->flags & DA906X_FLG_FORCE_IN0_AUTO_MODE)
> +		hwmon->adc_auto_en |= DA906X_ADC_AUTO_VSYS_EN;
> +
> +	if (da906x_pdata->flags & DA906X_FLG_FORCE_IN1_MANUAL_MODE)
> +		hwmon->adc_auto_en &= ~DA906X_ADC_AUTO_AD1_EN;
> +	else if (da906x_pdata->flags & DA906X_FLG_FORCE_IN1_AUTO_MODE)
> +		hwmon->adc_auto_en |= DA906X_ADC_AUTO_AD1_EN;
> +
> +	if (da906x_pdata->flags & DA906X_FLG_FORCE_IN2_MANUAL_MODE)
> +		hwmon->adc_auto_en &= ~DA906X_ADC_AUTO_AD2_EN;
> +	else if (da906x_pdata->flags & DA906X_FLG_FORCE_IN2_AUTO_MODE)
> +		hwmon->adc_auto_en |= DA906X_ADC_AUTO_AD2_EN;
> +
> +	if (da906x_pdata->flags & DA906X_FLG_FORCE_IN3_MANUAL_MODE)
> +		hwmon->adc_auto_en &= ~DA906X_ADC_AUTO_AD3_EN;
> +	else if (da906x_pdata->flags & DA906X_FLG_FORCE_IN3_AUTO_MODE)
> +		hwmon->adc_auto_en |= DA906X_ADC_AUTO_AD3_EN;
> +
> +	ret = da906x_reg_update(da906x, DA906X_REG_ADC_CONT,
> +				DA906X_ADC_AUTO_MODE_SUPPORT_MASK,
> +				hwmon->adc_auto_en);
> +	if (ret < 0)
> +		return ret;
> +
> +	hwmon->class_dev = hwmon_device_register(&pdev->dev);
> +	if (IS_ERR(hwmon->class_dev))
> +		return PTR_ERR(hwmon->class_dev);
> +
hwmon registration has to be the last call, after sysfs attributes exist.
sysfs entries are expected to exist when the hwmon device is registered.

> +	hwmon->irq = platform_get_irq_byname(pdev, DA906X_DRVNAME_HWMON);

This function can return an error (-ENXIO).

> +	ret = devm_request_threaded_irq(&pdev->dev, hwmon->irq, NULL,
> +					da906x_hwmon_irq_handler,
> +					IRQF_TRIGGER_LOW | IRQF_ONESHOT,
> +					"HWMON", hwmon);
> +	if (ret) {
> +		dev_err(&pdev->dev, "Failed to request IRQ.\n");
> +		goto err;
> +	}
> +
> +	platform_set_drvdata(pdev, hwmon);
> +
> +	ret = da906x_reg_read(da906x, DA906X_REG_T_OFFSET);
> +	if (ret < 0)
> +		dev_warn(&pdev->dev, "Could not read temp1 callibration offset.\n");

	s/callibration/calibration/

> +	else
> +		hwmon->tjunc_offset = (s8)ret;
> +
> +	ret = sysfs_create_group(&pdev->dev.kobj, &da906x_attr_group);
> +	if (ret)
> +		goto err;
> +
> +	return 0;
> +
> +err:
> +	hwmon_device_unregister(hwmon->class_dev);
> +	return ret;
> +}
> +
> +static int __devexit da906x_hwmon_remove(struct platform_device *pdev)
> +{
> +	struct da906x_hwmon *hwmon = platform_get_drvdata(pdev);
> +
> +	hwmon_device_unregister(hwmon->class_dev);
> +	sysfs_remove_group(&pdev->dev.kobj, &da906x_attr_group);
> +
> +	return 0;
> +}
> +
> +static struct platform_driver da906x_hwmon_driver = {
> +	.probe = da906x_hwmon_probe,
> +	.remove = __devexit_p(da906x_hwmon_remove),
> +	.driver = {
> +		.name = DA906X_DRVNAME_HWMON,
> +		.owner = THIS_MODULE,
> +	},
> +};
> +
> +module_platform_driver(da906x_hwmon_driver);
> +
> +MODULE_DESCRIPTION("DA906X Hardware monitoring");
> +MODULE_AUTHOR("Krystian Garbaciak <krystian.garbaciak@diasemi.com>, Michal Hajduk <michal.hajduk@diasemi.com>");
> +MODULE_LICENSE("GPL");
> +MODULE_ALIAS("paltform:" DA906X_DRVNAME_HWMON);

s/paltform/platform/

> -- 
> 1.7.0.4
> 
> 

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] regulator: Fix bug in regulator_mode_to_status() core function.
  2012-08-24 18:45   ` [RFC PATCH 4/8] hwmon: Add DA906x hardware monitoring support Guenter Roeck
@ 2012-08-29 13:25     ` Krystian Garbaciak
  0 siblings, 0 replies; 2+ messages in thread
From: Krystian Garbaciak @ 2012-08-29 13:25 UTC (permalink / raw)
  To: Guenter Roeck
  Cc: linux-kernel, rtc-linux, lm-sensors, linux-input, linux-watchdog,
	linux-leds, Samuel Ortiz, Liam Girdwood, Mark Brown,
	Alessandro Zummo, Jean Delvare, Dmitry Torokhov, Ashish Jangam,
	Andrew Jones, Donggeun Kim, Philippe Rétornaz,
	Wim Van Sebroeck, Bryan Wu, Richard Purdie, Anthony Olech

> +static const struct i_table vsys_tbl[] = {
> > +	ILINE(0, DA906X_ADC_MAX, 2500, 5500)
> > +};
> > +
> > +static const struct i_table adcin_tbl[] = {
> > +	ILINE(0, DA906X_ADC_MAX, 0, 2500)
> > +};
> > +
> > +static const struct i_table tjunc_tbl[] = {
> > +	ILINE(0, DA906X_ADC_MAX, 333, -86)
> > +};
> > +
> > +static const struct i_table vbbat_tbl[] = {
> > +	ILINE(0, DA906X_ADC_MAX, 0, 5000)
> > +};
> 
> Since the first parameter to ILINE is always 0, it is not needed. This means
> that x0 in itable is also always 0 and thus not needed.
> 
> > +
> > +static const struct channel_info da906x_channels[] = {
> > +	[DA906X_VSYS]	= { "VSYS",
> > +			    vsys_tbl, ARRAY_SIZE(vsys_tbl) - 1,
> > +			    DA906X_REG_VSYS_RES },
> > +	[DA906X_ADCIN1]	= { "ADCIN1",
> > +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> > +			    DA906X_REG_ADCIN1_RES },
> > +	[DA906X_ADCIN2]	= { "ADCIN2",
> > +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> > +			    DA906X_REG_ADCIN2_RES },
> > +	[DA906X_ADCIN3]	= { "ADCIN3",
> > +			    adcin_tbl,	ARRAY_SIZE(adcin_tbl) - 1,
> > +			    DA906X_REG_ADCIN3_RES },
> > +	[DA906X_TJUNC]	= { "TJUNC",
> > +			    tjunc_tbl,	ARRAY_SIZE(tjunc_tbl) - 1 },
> > +	[DA906X_VBBAT]	= { "VBBAT",
> > +			    vbbat_tbl,	ARRAY_SIZE(vbbat_tbl) - 1}
> 
> 	s/1}/1 }/
> 
> > +};
> 
> You lost me here a bit (I am missing something ?).
> 
> Seems to be each table has exactly one entry. Since the table size is 1,
> ARRAY_SIZE(vbbat_tbl) - 1 is 0, and ...

An initial idea was to make the interpolation of the channel values using more
ILINE segments. Eventually, it ended up with one linear segment for every
channel, so it makes sense to reduce the code as you propose.

As suggested, driver name will be changed from "da906x" to "da9063".
I will adapt proposed changes and fixes.

Thank you for your comments,
Krystian


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2012-08-29 13:32 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <201208241500@sw-eng-lt-dc-vm2>
     [not found] ` <201208241505@sw-eng-lt-dc-vm2>
2012-08-24 18:45   ` [RFC PATCH 4/8] hwmon: Add DA906x hardware monitoring support Guenter Roeck
2012-08-29 13:25     ` [PATCH] regulator: Fix bug in regulator_mode_to_status() core function Krystian Garbaciak

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).