Re: [PATCH v3 7/8] hwmon: pwm-fan: Read PWM FAN configuration from device tree

[Guenter Roeck <linux@roeck-us.net>]

On 02/08/2015 01:36 PM, Lukasz Majewski wrote:
> On Fri, 6 Feb 2015 10:36:57 -0800
> Guenter Roeck <linux@roeck-us.net> wrote:
>
>> On Fri, Feb 06, 2015 at 05:59:07PM +0100, Lukasz Majewski wrote:
>>> This patch provides code for reading PWM FAN configuration data via
>>> device tree. The pwm-fan can work with full speed when configuration
>>> is not provided. However, errors are propagated when wrong DT
>>> bindings are found.
>>> Additionally the struct pwm_fan_ctx has been extended.
>>>
>>> Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
>>> ---
>>> Changes for v2:
>>> - Rename pwm_fan_max_states to pwm_fan_cooling_levels
>>> - Moving pwm_fan_of_get_cooling_data() call after setting end
>>> enabling PWM FAN
>>> - pwm_fan_of_get_cooling_data() now can fail - preserving old
>>> behaviour
>>> - Remove unnecessary dev_err() call
>>> Changes for v3:
>>> - Patch's headline has been reedited
>>> - pwm_fan_of_get_cooling_data() return code is now being checked.
>>> - of_property_count_elems_of_size() is now used instead
>>> of_find_property()
>>> - More verbose patch description added
>>> ---
>>>   drivers/hwmon/pwm-fan.c | 54
>>> ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed,
>>> 53 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
>>> index 870e100..f3f5843 100644
>>> --- a/drivers/hwmon/pwm-fan.c
>>> +++ b/drivers/hwmon/pwm-fan.c
>>> @@ -30,7 +30,10 @@
>>>   struct pwm_fan_ctx {
>>>   	struct mutex lock;
>>>   	struct pwm_device *pwm;
>>> -	unsigned char pwm_value;
>>> +	unsigned int pwm_value;
>>> +	unsigned int pwm_fan_state;
>>> +	unsigned int pwm_fan_max_state;
>>> +	unsigned int *pwm_fan_cooling_levels;
>>>   };
>>>
>>>   static int  __set_pwm(struct pwm_fan_ctx *ctx, unsigned long pwm)
>>> @@ -100,6 +103,50 @@ static struct attribute *pwm_fan_attrs[] = {
>>>
>>>   ATTRIBUTE_GROUPS(pwm_fan);
>>>
>>> +int pwm_fan_of_get_cooling_data(struct device *dev, struct
>>> pwm_fan_ctx *ctx) +{
>>> +	struct device_node *np = dev->of_node;
>>> +	int num, i, ret;
>>> +
>>> +	ret = of_property_count_elems_of_size(np, "cooling-levels",
>>> +					      sizeof(u32));
>>> +
>>> +	if (ret == -EINVAL) {
>>> +		dev_err(dev, "Property 'cooling-levels' not
>>> found\n");
>>> +		return 0;
>>
>> Returning 0 is obviously not an error, otherwise you would not return
>> 0 here. So dev_err is wrong.
>
> pr_info would be more appropriate here.
>
>> Also, the message itself is wrong; the
>> property may well be there but have a wrong size.
>
> As fair as I remember the -EINVAL is set only when the property is not
> defined. Such situation is correct and our pwm-fan driver should work
> with or without it.
>

of_property_count_elems_of_size returns -EINVAL if np is NULL or if
there is no peoperty or prop->length % elem_size != 0, and -ENODATA
if there is no value associated with the property.

If -EINVAL is not an error, please no message. Noisy drivers are
just that, noisy.

>>
>>> +	}
>>> +
>>> +	if (ret <= 0) {
>>> +		dev_err(dev, "Wrong data!\n");
>>> +		return ret;
>>> +	}
>>
>> This will result in the driver failing to load if devicetree is not
>> configured (of_property_count_elems_of_size will return -ENOSYS).
>
> As you pointed out in the comment to v2:
>
> It is OK if the "cooling-levels" is not defined in DT. However, if it
> has broken definition, then we should return error. This is what we do
> here.
>
You don't return an error, you return 0, at least in some circumstances.

>> This is not acceptable. Also, if the call returns 0 you don't return
>> an error but display a "Wrong data!" error message. Either it is an
>> error or it is not, but not both.
>
> This is an error. "cooling-levels" with zero elements is regarded as a
> broken property.
>
It returns -ENOSYS if DT is not configured, which is still unacceptable.
And, again, if ret == 0 is an error, you should return an error code,
not display an error message and return 0.

Guenter