[PATCH RFC 0/3] hwmon: pwm-fan: Add RPM support

[PATCH RFC 0/3] hwmon: pwm-fan: Add RPM support

[Stefan Wahren]

Contrary to the gpio-fan the pwm-fan driver isn't easy to setup
with pwmconfig/fancontrol because of the missing hwmon sysfs entry
for actual revolutions per minute. This series adds this feature.

This series based on the recently submitted regulator support.

Stefan Wahren (3):
  dt-bindings: hwmon: Add tachometer interrupt to pwm-fan
  Documentation: pwm-fan: Add description for RPM support
  hwmon: pwm-fan: Add RPM support via external interrupt

 .../devicetree/bindings/hwmon/pwm-fan.txt          |  3 ++
 Documentation/hwmon/pwm-fan                        |  4 ++
 drivers/hwmon/pwm-fan.c                            | 55 ++++++++++++++++++++++
 3 files changed, 62 insertions(+)

-- 
2.7.4

[PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Stefan Wahren]

This adds the tachometer interrupt to the pwm-fan binding, which is
necessary for RPM support.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
 Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
index 49ca5d8..7f69b0b 100644
--- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
+++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
@@ -8,6 +8,9 @@ Required properties:
 
 Optional properties:
 - fan-supply    : phandle to the regulator that provides power to the fan
+- interrupts    : contains a single interrupt specifier which describes the
+                  tachometer pin output of a 2 pulse-per-revolution fan.
+                  See interrupt-controller/interrupts.txt for the format.
 
 Example:
 	fan0: pwm-fan {
-- 
2.7.4

[PATCH RFC 2/3] Documentation: pwm-fan: Add description for RPM support

[Stefan Wahren]

This adds a short description for the new RPM support of the pwm-fan
driver.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
 Documentation/hwmon/pwm-fan | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/Documentation/hwmon/pwm-fan b/Documentation/hwmon/pwm-fan
index 18529d2..a1ba34c 100644
--- a/Documentation/hwmon/pwm-fan
+++ b/Documentation/hwmon/pwm-fan
@@ -15,3 +15,7 @@ The driver implements a simple interface for driving a fan connected to
 a PWM output. It uses the generic PWM interface, thus it can be used with
 a range of SoCs. The driver exposes the fan to the user space through
 the hwmon's sysfs interface.
+
+The measured fan rotation speed returned via 'fan1_input' is derived
+from the sampled interrupts from the tachometer signal by assuming
+a 2 pulse-per-revolution fan.
-- 
2.7.4

[PATCH RFC 3/3] hwmon: pwm-fan: Add RPM support via external interrupt

[Stefan Wahren]

This adds RPM support to the pwm-fan driver in order to use with
fancontrol/pwmconfig. This feature is intended for 2 pulse-per-revolution
fans which provides a tachometer output signal.

Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
---
 drivers/hwmon/pwm-fan.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)

diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
index 9e0591e..731fdc6 100644
--- a/drivers/hwmon/pwm-fan.c
+++ b/drivers/hwmon/pwm-fan.c
@@ -18,6 +18,7 @@
 
 #include <linux/hwmon.h>
 #include <linux/hwmon-sysfs.h>
+#include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/mutex.h>
 #include <linux/of.h>
@@ -26,13 +27,22 @@
 #include <linux/regulator/consumer.h>
 #include <linux/sysfs.h>
 #include <linux/thermal.h>
+#include <linux/timer.h>
 
 #define MAX_PWM 255
+#define SAMPLE_TIME		1 /* seconds */
+#define PULSE_PER_REVOLUTION	2
 
 struct pwm_fan_ctx {
 	struct mutex lock;
 	struct pwm_device *pwm;
 	struct regulator *reg_en;
+
+	int irq;
+	unsigned int pulses;
+	unsigned int rpm;
+	struct timer_list sample_timer;
+
 	unsigned int pwm_value;
 	unsigned int pwm_fan_state;
 	unsigned int pwm_fan_max_state;
@@ -40,6 +50,27 @@ struct pwm_fan_ctx {
 	struct thermal_cooling_device *cdev;
 };
 
+static irqreturn_t pulse_handler(int irq, void *dev_id)
+{
+	struct pwm_fan_ctx *ctx = dev_id;
+
+	if (ctx->pulses < INT_MAX / 2)
+		ctx->pulses++;
+
+	return IRQ_HANDLED;
+}
+
+static void sample_timer(struct timer_list *t)
+{
+	struct pwm_fan_ctx *ctx = from_timer(ctx, t, sample_timer);
+	unsigned int pulses;
+
+	pulses = ctx->pulses;
+	ctx->pulses -= pulses;
+	ctx->rpm = pulses * 60 / SAMPLE_TIME / PULSE_PER_REVOLUTION;
+	mod_timer(&ctx->sample_timer, jiffies + (HZ * SAMPLE_TIME));
+}
+
 static int  __set_pwm(struct pwm_fan_ctx *ctx, unsigned long pwm)
 {
 	unsigned long period;
@@ -100,11 +131,20 @@ static ssize_t pwm_show(struct device *dev, struct device_attribute *attr,
 	return sprintf(buf, "%u\n", ctx->pwm_value);
 }
 
+static ssize_t rpm_show(struct device *dev,
+			struct device_attribute *attr, char *buf)
+{
+	struct pwm_fan_ctx *ctx = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%u\n", ctx->rpm);
+}
 
 static SENSOR_DEVICE_ATTR_RW(pwm1, pwm, 0);
+static SENSOR_DEVICE_ATTR_RO(fan1_input, rpm, 0);
 
 static struct attribute *pwm_fan_attrs[] = {
 	&sensor_dev_attr_pwm1.dev_attr.attr,
+	&sensor_dev_attr_fan1_input.dev_attr.attr,
 	NULL,
 };
 
@@ -263,6 +303,19 @@ static int pwm_fan_probe(struct platform_device *pdev)
 		}
 	}
 
+	timer_setup(&ctx->sample_timer, sample_timer, 0);
+
+	ctx->irq = platform_get_irq(pdev, 0);
+	if (ctx->irq >= 0) {
+		ret = devm_request_irq(&pdev->dev, ctx->irq, pulse_handler, 0,
+				       pdev->name, ctx);
+		if (ret) {
+			dev_err(&pdev->dev, "Can't get interrupt working.\n");
+			return ret;
+		}
+		mod_timer(&ctx->sample_timer, jiffies + (HZ * SAMPLE_TIME));
+	}
+
 	hwmon = devm_hwmon_device_register_with_groups(&pdev->dev, "pwmfan",
 						       ctx, pwm_fan_groups);
 	if (IS_ERR(hwmon)) {
@@ -304,6 +357,8 @@ static int pwm_fan_remove(struct platform_device *pdev)
 	struct pwm_fan_ctx *ctx = platform_get_drvdata(pdev);
 
 	thermal_cooling_device_unregister(ctx->cdev);
+	del_timer_sync(&ctx->sample_timer);
+
 	if (ctx->pwm_value)
 		pwm_disable(ctx->pwm);
 
-- 
2.7.4

Re: [PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Guenter Roeck]

On Wed, Jan 30, 2019 at 04:07:05PM +0100, Stefan Wahren wrote:
> This adds the tachometer interrupt to the pwm-fan binding, which is
> necessary for RPM support.
> 
> Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
> ---
>  Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> index 49ca5d8..7f69b0b 100644
> --- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> +++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> @@ -8,6 +8,9 @@ Required properties:
>  
>  Optional properties:
>  - fan-supply    : phandle to the regulator that provides power to the fan
> +- interrupts    : contains a single interrupt specifier which describes the
> +                  tachometer pin output of a 2 pulse-per-revolution fan.
> +                  See interrupt-controller/interrupts.txt for the format.

So a hypothetical {1,4} pulse-per-revolution fan would explicitly not be
supported ? Why ?

Guenter

Re: [PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Stefan Wahren]

Hi Guenter,

> Guenter Roeck <linux@roeck-us.net> hat am 30. Januar 2019 um 18:28 geschrieben:
> 
> 
> On Wed, Jan 30, 2019 at 04:07:05PM +0100, Stefan Wahren wrote:
> > This adds the tachometer interrupt to the pwm-fan binding, which is
> > necessary for RPM support.
> > 
> > Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
> > ---
> >  Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > index 49ca5d8..7f69b0b 100644
> > --- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > +++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > @@ -8,6 +8,9 @@ Required properties:
> >  
> >  Optional properties:
> >  - fan-supply    : phandle to the regulator that provides power to the fan
> > +- interrupts    : contains a single interrupt specifier which describes the
> > +                  tachometer pin output of a 2 pulse-per-revolution fan.
> > +                  See interrupt-controller/interrupts.txt for the format.
> 
> So a hypothetical {1,4} pulse-per-revolution fan would explicitly not be
> supported ? Why ?

i could add an additional property to specify the pulse per revolution and use the 2 as default (according to the Intel spec for 4 pin pwm fan) which should fit in most cases.

> 
> Guenter

Re: [PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Guenter Roeck]

On Wed, Jan 30, 2019 at 09:23:31PM +0100, Stefan Wahren wrote:
> Hi Guenter,
> 
> > Guenter Roeck <linux@roeck-us.net> hat am 30. Januar 2019 um 18:28 geschrieben:
> > 
> > 
> > On Wed, Jan 30, 2019 at 04:07:05PM +0100, Stefan Wahren wrote:
> > > This adds the tachometer interrupt to the pwm-fan binding, which is
> > > necessary for RPM support.
> > > 
> > > Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
> > > ---
> > >  Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
> > >  1 file changed, 3 insertions(+)
> > > 
> > > diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > > index 49ca5d8..7f69b0b 100644
> > > --- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > > +++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> > > @@ -8,6 +8,9 @@ Required properties:
> > >  
> > >  Optional properties:
> > >  - fan-supply    : phandle to the regulator that provides power to the fan
> > > +- interrupts    : contains a single interrupt specifier which describes the
> > > +                  tachometer pin output of a 2 pulse-per-revolution fan.
> > > +                  See interrupt-controller/interrupts.txt for the format.
> > 
> > So a hypothetical {1,4} pulse-per-revolution fan would explicitly not be
> > supported ? Why ?
> 
> i could add an additional property to specify the pulse per revolution and use the 2 as default (according to the Intel spec for 4 pin pwm fan) which should fit in most cases.
> 
That would be a possibility and make sense, but that is not
the point here.  The "interrupts" property does not and should
not care how many pulses per revolution the fan provides.

Guenter

Re: [PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Stefan Wahren]

Hi Guenter,

Am 30.01.19 um 21:35 schrieb Guenter Roeck:
> On Wed, Jan 30, 2019 at 09:23:31PM +0100, Stefan Wahren wrote:
>> Hi Guenter,
>>
>>> Guenter Roeck <linux@roeck-us.net> hat am 30. Januar 2019 um 18:28 geschrieben:
>>>
>>>
>>> On Wed, Jan 30, 2019 at 04:07:05PM +0100, Stefan Wahren wrote:
>>>> This adds the tachometer interrupt to the pwm-fan binding, which is
>>>> necessary for RPM support.
>>>>
>>>> Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
>>>> ---
>>>>  Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
>>>>  1 file changed, 3 insertions(+)
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
>>>> index 49ca5d8..7f69b0b 100644
>>>> --- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
>>>> +++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
>>>> @@ -8,6 +8,9 @@ Required properties:
>>>>  
>>>>  Optional properties:
>>>>  - fan-supply    : phandle to the regulator that provides power to the fan
>>>> +- interrupts    : contains a single interrupt specifier which describes the
>>>> +                  tachometer pin output of a 2 pulse-per-revolution fan.
>>>> +                  See interrupt-controller/interrupts.txt for the format.
>>> So a hypothetical {1,4} pulse-per-revolution fan would explicitly not be
>>> supported ? Why ?
>> i could add an additional property to specify the pulse per revolution and use the 2 as default (according to the Intel spec for 4 pin pwm fan) which should fit in most cases.
>>
> That would be a possibility and make sense, but that is not
> the point here.  The "interrupts" property does not and should
> not care how many pulses per revolution the fan provides.

sorry, i'm not sure what's the problem about. Do you want me to use a
GPIO instead of interrupt?

Or is it the wording here?

Suggestion:

contains a single interrupt specifier which is describes the tachometer
output of the fan as an input

Stefan

>
> Guenter

Re: [PATCH RFC 1/3] dt-bindings: hwmon: Add tachometer interrupt to pwm-fan

[Guenter Roeck]

On Thu, Jan 31, 2019 at 09:02:33AM +0100, Stefan Wahren wrote:
> Hi Guenter,
> 
> Am 30.01.19 um 21:35 schrieb Guenter Roeck:
> > On Wed, Jan 30, 2019 at 09:23:31PM +0100, Stefan Wahren wrote:
> >> Hi Guenter,
> >>
> >>> Guenter Roeck <linux@roeck-us.net> hat am 30. Januar 2019 um 18:28 geschrieben:
> >>>
> >>>
> >>> On Wed, Jan 30, 2019 at 04:07:05PM +0100, Stefan Wahren wrote:
> >>>> This adds the tachometer interrupt to the pwm-fan binding, which is
> >>>> necessary for RPM support.
> >>>>
> >>>> Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
> >>>> ---
> >>>>  Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
> >>>>  1 file changed, 3 insertions(+)
> >>>>
> >>>> diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> >>>> index 49ca5d8..7f69b0b 100644
> >>>> --- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> >>>> +++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
> >>>> @@ -8,6 +8,9 @@ Required properties:
> >>>>  
> >>>>  Optional properties:
> >>>>  - fan-supply    : phandle to the regulator that provides power to the fan
> >>>> +- interrupts    : contains a single interrupt specifier which describes the
> >>>> +                  tachometer pin output of a 2 pulse-per-revolution fan.
> >>>> +                  See interrupt-controller/interrupts.txt for the format.
> >>> So a hypothetical {1,4} pulse-per-revolution fan would explicitly not be
> >>> supported ? Why ?
> >> i could add an additional property to specify the pulse per revolution and use the 2 as default (according to the Intel spec for 4 pin pwm fan) which should fit in most cases.
> >>
> > That would be a possibility and make sense, but that is not
> > the point here.  The "interrupts" property does not and should
> > not care how many pulses per revolution the fan provides.
> 
> sorry, i'm not sure what's the problem about. Do you want me to use a
> GPIO instead of interrupt?
> 
> Or is it the wording here?
> 

You wording limits the use of interrupts with pwm fans to fans with
2 pulses per revolution. You do not explain why that restriction would
be required or even make sense, or why it should be associated with
the 'interrupts' property.

Logically I assume that is because you expect an interrupt per pulse,
but that is not explained (or what the interrupt is expected to be
used for in the first place - it might, after all, be some kind of
error interrupt).

Guenter

Re: [PATCH RFC 3/3] hwmon: pwm-fan: Add RPM support via external interrupt

[Guenter Roeck]

On Wed, Jan 30, 2019 at 04:07:07PM +0100, Stefan Wahren wrote:
> This adds RPM support to the pwm-fan driver in order to use with
> fancontrol/pwmconfig. This feature is intended for 2 pulse-per-revolution
> fans which provides a tachometer output signal.
> 
> Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
> ---
>  drivers/hwmon/pwm-fan.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 55 insertions(+)
> 
> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
> index 9e0591e..731fdc6 100644
> --- a/drivers/hwmon/pwm-fan.c
> +++ b/drivers/hwmon/pwm-fan.c
> @@ -18,6 +18,7 @@
>  
>  #include <linux/hwmon.h>
>  #include <linux/hwmon-sysfs.h>
> +#include <linux/interrupt.h>
>  #include <linux/module.h>
>  #include <linux/mutex.h>
>  #include <linux/of.h>
> @@ -26,13 +27,22 @@
>  #include <linux/regulator/consumer.h>
>  #include <linux/sysfs.h>
>  #include <linux/thermal.h>
> +#include <linux/timer.h>
>  
>  #define MAX_PWM 255
> +#define SAMPLE_TIME		1 /* seconds */
> +#define PULSE_PER_REVOLUTION	2
>  
>  struct pwm_fan_ctx {
>  	struct mutex lock;
>  	struct pwm_device *pwm;
>  	struct regulator *reg_en;
> +
> +	int irq;
> +	unsigned int pulses;
> +	unsigned int rpm;
> +	struct timer_list sample_timer;
> +
>  	unsigned int pwm_value;
>  	unsigned int pwm_fan_state;
>  	unsigned int pwm_fan_max_state;
> @@ -40,6 +50,27 @@ struct pwm_fan_ctx {
>  	struct thermal_cooling_device *cdev;
>  };
>  
> +static irqreturn_t pulse_handler(int irq, void *dev_id)
> +{
> +	struct pwm_fan_ctx *ctx = dev_id;
> +
> +	if (ctx->pulses < INT_MAX / 2)
> +		ctx->pulses++;

Please explain the limit. With this code, the maximum number
for pulses is INT_MAX / 2. The calculation below multiplies
that with 60, which is well above UINT_MAX. Either use a 
realistic limit or drop the if statement (it is kind of
unlikely to get that many interrupts per second).

> +
> +	return IRQ_HANDLED;

This assumes a self resetting edge triggered interrupt ?
You might want to mention that somewhere.

> +}
> +
> +static void sample_timer(struct timer_list *t)
> +{
> +	struct pwm_fan_ctx *ctx = from_timer(ctx, t, sample_timer);
> +	unsigned int pulses;
> +
> +	pulses = ctx->pulses;
> +	ctx->pulses -= pulses;

This is racy; the substract operation is not atomic, and neither
is the addition in the interrupt handler.

> +	ctx->rpm = pulses * 60 / SAMPLE_TIME / PULSE_PER_REVOLUTION;

The calculation assumes that sample_timer is called exactly after one
second. Have to tested how much variation this causes on a loaded system ?

> +	mod_timer(&ctx->sample_timer, jiffies + (HZ * SAMPLE_TIME));
> +}
> +
>  static int  __set_pwm(struct pwm_fan_ctx *ctx, unsigned long pwm)
>  {
>  	unsigned long period;
> @@ -100,11 +131,20 @@ static ssize_t pwm_show(struct device *dev, struct device_attribute *attr,
>  	return sprintf(buf, "%u\n", ctx->pwm_value);
>  }
>  
> +static ssize_t rpm_show(struct device *dev,
> +			struct device_attribute *attr, char *buf)
> +{
> +	struct pwm_fan_ctx *ctx = dev_get_drvdata(dev);
> +
> +	return sprintf(buf, "%u\n", ctx->rpm);
> +}
>  
>  static SENSOR_DEVICE_ATTR_RW(pwm1, pwm, 0);
> +static SENSOR_DEVICE_ATTR_RO(fan1_input, rpm, 0);
>  
>  static struct attribute *pwm_fan_attrs[] = {
>  	&sensor_dev_attr_pwm1.dev_attr.attr,
> +	&sensor_dev_attr_fan1_input.dev_attr.attr,

This attribute is visible even if not supported, meaning it will
report a fan speed of 0 rpm in that case. Please make the attribute
conditional.

>  	NULL,
>  };
>  
> @@ -263,6 +303,19 @@ static int pwm_fan_probe(struct platform_device *pdev)
>  		}
>  	}
>  
> +	timer_setup(&ctx->sample_timer, sample_timer, 0);
> +
> +	ctx->irq = platform_get_irq(pdev, 0);
> +	if (ctx->irq >= 0) {
> +		ret = devm_request_irq(&pdev->dev, ctx->irq, pulse_handler, 0,
> +				       pdev->name, ctx);
> +		if (ret) {
> +			dev_err(&pdev->dev, "Can't get interrupt working.\n");
> +			return ret;
> +		}
> +		mod_timer(&ctx->sample_timer, jiffies + (HZ * SAMPLE_TIME));
> +	}
> +
>  	hwmon = devm_hwmon_device_register_with_groups(&pdev->dev, "pwmfan",
>  						       ctx, pwm_fan_groups);
>  	if (IS_ERR(hwmon)) {
> @@ -304,6 +357,8 @@ static int pwm_fan_remove(struct platform_device *pdev)
>  	struct pwm_fan_ctx *ctx = platform_get_drvdata(pdev);
>  
>  	thermal_cooling_device_unregister(ctx->cdev);
> +	del_timer_sync(&ctx->sample_timer);
> +
>  	if (ctx->pwm_value)
>  		pwm_disable(ctx->pwm);
>