Enabling enhanced acoustics on adt7475

Enabling enhanced acoustics on adt7475

[Chris Packham]

Hi,

I'm looking at adding code to configure the enhanced acoustics on the 
ADT7475 and compatible devices.

I'm wondering if I should be adding these as device-tree properties or 
as sysfs attributes. And in either case how these properties/attributes 
should behave. The main impact of this would be to provide a mechanism 
for configuring how fast the pwm output will change.

The fact that I wrote "configuring" above makes me lean towards sysfs 
attributes. Nothing in Documentation/hwmon/sysfs-interface really covers 
this. dme1737.c has a pwmN_ramp_rate which would seem close to what I 
want. Is there any other precedent I should follow?

Thanks in advance,
Chris

Re: Enabling enhanced acoustics on adt7475

[Guenter Roeck]

On Mon, May 01, 2017 at 08:45:14AM +0000, Chris Packham wrote:
> Hi,
> 
> I'm looking at adding code to configure the enhanced acoustics on the 
> ADT7475 and compatible devices.
> 
> I'm wondering if I should be adding these as device-tree properties or 
> as sysfs attributes. And in either case how these properties/attributes 
> should behave. The main impact of this would be to provide a mechanism 
> for configuring how fast the pwm output will change.
> 
> The fact that I wrote "configuring" above makes me lean towards sysfs 
> attributes. Nothing in Documentation/hwmon/sysfs-interface really covers 

Definitely. This does not describe hardware.

> this. dme1737.c has a pwmN_ramp_rate which would seem close to what I 
> want. Is there any other precedent I should follow?
> 
There is pwmX_step but that does not describe a ramp rate in time.
dme1737.c indeed seems to be the closest match. Nothing else I can think of.

Guenter

Re: Enabling enhanced acoustics on adt7475

[Chris Packham]

On 02/05/17 07:16, Guenter Roeck wrote:
> On Mon, May 01, 2017 at 08:45:14AM +0000, Chris Packham wrote:
>> Hi,
>>
>> I'm looking at adding code to configure the enhanced acoustics on the
>> ADT7475 and compatible devices.
>>
>> I'm wondering if I should be adding these as device-tree properties or
>> as sysfs attributes. And in either case how these properties/attributes
>> should behave. The main impact of this would be to provide a mechanism
>> for configuring how fast the pwm output will change.
>>
>> The fact that I wrote "configuring" above makes me lean towards sysfs
>> attributes. Nothing in Documentation/hwmon/sysfs-interface really covers
>
> Definitely. This does not describe hardware.
>
>> this. dme1737.c has a pwmN_ramp_rate which would seem close to what I
>> want. Is there any other precedent I should follow?
>>
> There is pwmX_step but that does not describe a ramp rate in time.
> dme1737.c indeed seems to be the closest match. Nothing else I can think of.
>

I also found tempN_smoothing_enable and tempN_smoothing_time in 
asc7621.c which lines up quite closely with the adt7475.

Re: Enabling enhanced acoustics on adt7475

[Guenter Roeck]

On 05/01/2017 08:45 PM, Chris Packham wrote:
> On 02/05/17 07:16, Guenter Roeck wrote:
>> On Mon, May 01, 2017 at 08:45:14AM +0000, Chris Packham wrote:
>>> Hi,
>>>
>>> I'm looking at adding code to configure the enhanced acoustics on the
>>> ADT7475 and compatible devices.
>>>
>>> I'm wondering if I should be adding these as device-tree properties or
>>> as sysfs attributes. And in either case how these properties/attributes
>>> should behave. The main impact of this would be to provide a mechanism
>>> for configuring how fast the pwm output will change.
>>>
>>> The fact that I wrote "configuring" above makes me lean towards sysfs
>>> attributes. Nothing in Documentation/hwmon/sysfs-interface really covers
>>
>> Definitely. This does not describe hardware.
>>
>>> this. dme1737.c has a pwmN_ramp_rate which would seem close to what I
>>> want. Is there any other precedent I should follow?
>>>
>> There is pwmX_step but that does not describe a ramp rate in time.
>> dme1737.c indeed seems to be the closest match. Nothing else I can think of.
>>
>
> I also found tempN_smoothing_enable and tempN_smoothing_time in
> asc7621.c which lines up quite closely with the adt7475.
>

Your call.

Guenter