Re: [PATCH 2/3] ARM: mvebu: Create a NSA3xx hardware monitoring driver

[Guenter Roeck <linux-0h96xk9xTtrk1uMJSBkQmQ@public.gmane.org>]

Hi,

On Wed, Feb 10, 2016 at 11:33:43PM +0000, Adam Baker wrote:
> Create a driver to support the hardware monitoring chip present in
> several models of Zyxel NSA3xx NAS devices.
> 
> The driver reads fan speed and temperature from a suitably programmed
> MCU on the device.
> 
> Signed-off-by: Adam Baker <linux-u8Lxuz9p/S1csCcyGdaD/Q@public.gmane.org>

Please name the driver nsa320. It may ultimately support other devices,
but that doesn't mean it is going to support NSA-35x or other future
similar devices.

Got the headline, please use

hwmon: Add driver for Zyxel NSA-320

> ---
> I've tested this on an NSA-320, I've had someone offer to test it on
> NSA-325 so hopefully I will get a tested by back.
> ---
>  drivers/hwmon/Kconfig        |  13 +++
>  drivers/hwmon/Makefile       |   1 +
>  drivers/hwmon/nsa3xx-hwmon.c | 223 +++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 237 insertions(+)
>  create mode 100644 drivers/hwmon/nsa3xx-hwmon.c
> 
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index 80a73bf..8801b78 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -1757,6 +1757,19 @@ config SENSORS_ULTRA45
>  	  This driver provides support for the Ultra45 workstation environmental
>  	  sensors.
>  
> +config SENSORS_NSA3XX

Please try to stick with alphabetic order.

> +	tristate "ZyXEL NSA3xx fan speed and temperature sensors"

... NSA-320 and compatible ...

> +	depends on GPIOLIB && OF

Can you add some additional dependencies ?

It is quite unlikely that the driver is needed on an X86.
Please use at least

	depends on ARM || COMPILE_TEST

or a more specific dependency if the SoC is known. MACH_KIRKWOOD, maybe ?

> +	help
> +	  If you say yes here you get support for hardware monitoring
> +	  for the ZyXEL NSA3XX Media Servers.

Please be specific which devices are supported. Feel free to add "and
compatibles", but please no generic and misleading statements such as 3XX.

> +
> +	  The sensor data is taken from a Holtek HT46R065 microcontroller
> +	  connected to GPIO lines.
> +
Is that relevant ?

> +	  This driver can also be built as a module. If so, the module
> +	  will be called nsa3xx-hwmon.
> +
>  if ACPI
>  
>  comment "ACPI drivers"
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index 12a3239..e85414a 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -118,6 +118,7 @@ obj-$(CONFIG_SENSORS_MAX6650)	+= max6650.o
>  obj-$(CONFIG_SENSORS_MAX6697)	+= max6697.o
>  obj-$(CONFIG_SENSORS_MAX31790)	+= max31790.o
>  obj-$(CONFIG_SENSORS_MC13783_ADC)+= mc13783-adc.o
> +obj-$(CONFIG_SENSORS_NSA3XX)	+= nsa3xx-hwmon.o

Please try to stick with alphabetic order.

>  obj-$(CONFIG_SENSORS_MCP3021)	+= mcp3021.o
>  obj-$(CONFIG_SENSORS_MENF21BMC_HWMON) += menf21bmc_hwmon.o
>  obj-$(CONFIG_SENSORS_NCT6683)	+= nct6683.o
> diff --git a/drivers/hwmon/nsa3xx-hwmon.c b/drivers/hwmon/nsa3xx-hwmon.c
> new file mode 100644
> index 0000000..0fbf118
> --- /dev/null
> +++ b/drivers/hwmon/nsa3xx-hwmon.c

nsa320-hwmon.c

> @@ -0,0 +1,223 @@
> +/*
> + * drivers/hwmon/nsa3xx-hwmon.c
> + *
> + * ZyXEL NSA3xx Media Servers

Please be specific.

> + * hardware monitoring
> + *
> + * Copyright (C) 2016 Adam Baker <linux-u8Lxuz9p/S1csCcyGdaD/Q@public.gmane.org>
> + * based on a board file driver
> + * Copyright (C) 2012 Peter Schildmann <linux-7ojh7neMhKcWe5+p9BQgIQ@public.gmane.org>
> + *
> + * This program is free software; you can redistribute it and/or modify it
> + * under the terms of the GNU General Public License v2 as published by the
> + * Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but WITHOUT
> + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> + * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
> + * more details.
> + */
> +
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include <linux/platform_device.h>
> +#include <linux/err.h>
> +#include <linux/gpio/consumer.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/jiffies.h>
> +#include <linux/delay.h>
> +#include <linux/of.h>
> +#include <linux/of_device.h>
> +#include <linux/of_platform.h>
> +
Please order include files alphabetically.

> +#define MAGIC_NUMBER 0x55
> +
> +/*
> + * The Zyxel hwmon MCU is a Holtek HT46R065 that is factory programmed
> + * to perform temperature and fan speed monitoring. It is read by taking
> + * the active pin low. The 32 bit output word is then clocked onto the
> + * data line. The MSB of the data word is a magic nuber to indicate it
> + * has been read correctly, the next byte is the fan speed (in hundreds
> + * of RPM) and the last two bytes are the temperature (in tenths of a
> + * degree)
> + */
> +
> +struct nsa3xx_hwmon {

s/nsa3xx/nsa320/g for the entire file, please.

> +	struct device		*classdev;
> +	struct mutex		update_lock;	/* lock GPIO operations */
> +	unsigned long		last_updated;	/* jiffies */
> +	u32			mcu_data;

I would suggest to use unsigned long here. See below for reasons.

> +	struct gpio_desc	*act;
> +	struct gpio_desc	*clk;
> +	struct gpio_desc	*data;
> +};
> +
> +enum nsa3xx_inputs {
> +	NSA3XX_FAN = 1,
> +	NSA3XX_TEMP = 0,

Any special reason for the unusual order ? If yes, please explain.

> +};
> +
> +static const char * const nsa3xx_input_names[] = {
> +	[NSA3XX_FAN] = "Chassis Fan",
> +	[NSA3XX_TEMP] = "System Temperature",
> +};
> +
> +/* Although this protocol looks similar to SPI the long delay
> + * between the active (aka chip select) signal and the shorter
> + * delay between clock pulses are needed for reliable operation.
> + * The delays provided are taken from the manufacturer kernel,
> + * testing suggest they probably incorporate a reasonable safety
> + * margin. (The single device tested became unreliable if the
> + * delay was reduced to 1/10th of this value.)
> + */
> +static unsigned long nsa3xx_hwmon_update(struct device *dev)
> +{
> +	u32 mcu_data;
> +	u32 mask;
> +	struct nsa3xx_hwmon *hwmon = dev_get_drvdata(dev);
> +
> +	mutex_lock(&hwmon->update_lock);
> +
> +	mcu_data = hwmon->mcu_data;
> +
> +	if (time_after(jiffies, hwmon->last_updated + HZ) ||
> +							mcu_data == 0) {
> +

Please no blank line here. Also, please align continuation lines with '('.
However, unless I am missing something, the continuation line is not needed
in the first place. Please no unnecessary continuation lines.

> +		gpiod_set_value(hwmon->act, 1);
> +		msleep(100);
> +
> +		for (mask = BIT(31); mask; mask >>= 1) {

Since you are using BIT(), please include linux/bitops.h.

> +			gpiod_set_value(hwmon->clk, 0);
> +			usleep_range(100, 200);
> +			gpiod_set_value(hwmon->clk, 1);
> +			usleep_range(100, 200);
> +			if (gpiod_get_value(hwmon->data))
> +				mcu_data |= mask;
> +		}
> +
> +		gpiod_set_value(hwmon->act, 0);
> +		dev_dbg(dev, "Read raw MCU data %08x\n", mcu_data);
> +
> +		if ((mcu_data >> 24) != MAGIC_NUMBER) {
> +			dev_err(dev, "Read invalid MCU data %08x\n", mcu_data);
> +			mcu_data = 0;

Instead of an error message, it would be better to return an error code
(probably -EIO since we don't know better).

The calling code can then return the error to user space.

> +		}
> +
> +		hwmon->mcu_data = mcu_data;
> +		hwmon->last_updated = jiffies;
> +	}
> +
> +	mutex_unlock(&hwmon->update_lock);
> +
> +	return mcu_data;
> +}
> +
> +static ssize_t show_label(struct device *dev,
> +			  struct device_attribute *attr, char *buf)
> +{
> +	int channel = to_sensor_dev_attr(attr)->index;
> +
> +	return sprintf(buf, "%s\n", nsa3xx_input_names[channel]);
> +}
> +
> +static ssize_t show_value(struct device *dev,
> +			  struct device_attribute *attr, char *buf)
> +{
> +	int channel = to_sensor_dev_attr(attr)->index;
> +	unsigned long mcu_data = nsa3xx_hwmon_update(dev);
> +	unsigned long value = 0;
> +

If nsa3xx_hwmon_update() returns an error code, you can use

	if (IS_ERR_VALUE(mcu_data))
		return (long)mcu_data;

or even better have nsa3xx_hwmon_update() return int directly
to indicate an error.

> +	pr_debug("channel value 0x%02x!\n", channel);

Is this really useful ? I would suggest to drop debug messages
unless really needed. If you really think the message is useful,
please use dev_dbg().

> +	switch (channel) {
> +	case NSA3XX_TEMP:
> +		value = (mcu_data & 0xffff) * 100;
> +		break;
> +	case NSA3XX_FAN:
> +		value = ((mcu_data & 0xff0000) >> 16) * 100;
> +		break;
> +	}
> +	return sprintf(buf, "%lu\n", value);
> +}
> +
> +static SENSOR_DEVICE_ATTR(temp1_label, S_IRUGO, show_label, NULL, NSA3XX_TEMP);
> +static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, show_value, NULL, NSA3XX_TEMP);
> +static SENSOR_DEVICE_ATTR(fan1_label, S_IRUGO, show_label, NULL, NSA3XX_FAN);
> +static SENSOR_DEVICE_ATTR(fan1_input, S_IRUGO, show_value, NULL, NSA3XX_FAN);
> +
> +static struct attribute *nsa3xx_attrs[] = {
> +	&sensor_dev_attr_temp1_label.dev_attr.attr,
> +	&sensor_dev_attr_temp1_input.dev_attr.attr,
> +	&sensor_dev_attr_fan1_label.dev_attr.attr,
> +	&sensor_dev_attr_fan1_input.dev_attr.attr,
> +	NULL
> +};
> +
> +ATTRIBUTE_GROUPS(nsa3xx);
> +
> +static const struct of_device_id of_nsa3xx_hwmon_match[] = {
> +	{ .compatible = "zyxel,nsa320-mcu", },
> +	{ },
> +};
> +
> +static int nsa3xx_hwmon_probe(struct platform_device *pdev)
> +{
> +	struct nsa3xx_hwmon *hwmon;
> +
> +	hwmon = devm_kzalloc(&pdev->dev, sizeof(*hwmon), GFP_KERNEL);
> +	if (!hwmon)
> +		return -ENOMEM;
> +
> +	platform_set_drvdata(pdev, hwmon);

Unnecessary.

> +
> +	/* Look up the GPIO pins to use */
> +	hwmon->act = devm_gpiod_get(&pdev->dev, "act", GPIOD_OUT_LOW);
> +	hwmon->clk = devm_gpiod_get(&pdev->dev, "clk", GPIOD_OUT_HIGH);
> +	hwmon->data = devm_gpiod_get(&pdev->dev, "data", GPIOD_IN);
> +
> +	if (IS_ERR(hwmon->act))
> +		return PTR_ERR(hwmon->act);
> +	if (IS_ERR(hwmon->clk))
> +		return PTR_ERR(hwmon->clk);
> +	if (IS_ERR(hwmon->data))
> +		return PTR_ERR(hwmon->data);

Please reorder to have the error checks immediately after the calls
to devm_gpiod_get().

> +
> +	mutex_init(&hwmon->update_lock);
> +
> +	hwmon->classdev = devm_hwmon_device_register_with_groups(&pdev->dev,
> +					"nsa3xx", hwmon, nsa3xx_groups);

classdev is only used in this function and thus does not have to be kept
in struct nsa3xx_hwmon.

> +	if (IS_ERR(hwmon->classdev))
> +		return PTR_ERR(hwmon->classdev);
> +
> +	dev_dbg(&pdev->dev, "initialized\n");
> +

This message is really unnecessary. Please just use
	return PTR_ERR_OR_ZERO(classdev);

> +	return 0;
> +}
> +
> +static int nsa3xx_hwmon_remove(struct platform_device *pdev)
> +{
> +	/* All resources are allocated with devm_*() so
> +	 * there is nothing to do here.
> +	 */
> +
> +	return 0;
> +}

If the remove function is not needed, it is not needed and should not
exist in the first place. Please no dummy functions.

> +
> +static struct platform_driver nsa3xx_hwmon_driver = {
> +	.probe = nsa3xx_hwmon_probe,
> +	.remove = nsa3xx_hwmon_remove,
> +	.driver = {
> +		.name = "nsa3xx-hwmon",
> +		.owner = THIS_MODULE,
> +		.of_match_table = of_match_ptr(of_nsa3xx_hwmon_match),
> +	},
> +};
> +
> +module_platform_driver(nsa3xx_hwmon_driver);
> +
> +MODULE_DEVICE_TABLE(of, of_nsa3xx_hwmon_match);
> +MODULE_AUTHOR("Peter Schildmann <linux-7ojh7neMhKcWe5+p9BQgIQ@public.gmane.org>");
> +MODULE_AUTHOR("Adam Baker <linux-u8Lxuz9p/S1csCcyGdaD/Q@public.gmane.org>");
> +MODULE_DESCRIPTION("NSA3XX Hardware Monitoring");
> +MODULE_LICENSE("GPL");

GPL v2

> +MODULE_ALIAS("platform:nsa3xx-hwmon");
> -- 
> 2.1.4
> 
--
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