On 13/08/2021 17:10, Peter Geis wrote:
> On Fri, Aug 13, 2021 at 10:54 AM Daniel Lezcano
> <daniel.lezcano@xxxxxxxxxx> wrote:
>>
>>
>> Hi Robin,
>>
>>
>> On 13/08/2021 15:51, Robin Murphy wrote:
>>> On 2021-08-13 13:59, Daniel Lezcano wrote:
>>>> On 30/07/2021 17:17, Peter Geis wrote:
>>>>> The rockpro64 had a fan node since
>>>>> commit 5882d65c1691 ("arm64: dts: rockchip: Add PWM fan for RockPro64")
>>>>> however it was never tied into the thermal driver for automatic control.
>>>>>
>>>>> Add the links to the thermal node to permit the kernel to handle this
>>>>> automatically.
>>>>> Borrowed from the (rk3399-khadas-edge.dtsi).
>>>>>
>>>>> Signed-off-by: Peter Geis <pgwipeout@xxxxxxxxx>
>>
>> [ ... ]
>>
>>>>> +&cpu_thermal {
>>>>> + trips {
>>>>> + cpu_warm: cpu_warm {
>>>>> + temperature = <55000>;
>>>>> + hysteresis = <2000>;
>>>>> + type = "active";
>>>>> + };
>>>>> +
>>>>> + cpu_hot: cpu_hot {
>>>>> + temperature = <65000>;
>>>>> + hysteresis = <2000>;
>>>>> + type = "active";
>>>>> + };
>>>>> + };
>>>>> +
>>>>
>>>> Why two trip points ?
>>>>
>>>> Why not one functioning temperature and no lower / upper limits for the
>>>> cooling maps ?
>>>
>>> Certainly when I first did this for NanoPC-T4, IIRC it was to avoid the
>>> fan ramping up too eagerly, since level 1 for my fan is effectively
>>> silent but still cools enough to let a moderate load eventually settle
>>> to a steady state below the second trip.
>
> That's the same issue I had on the rockpro64.
>
>>
>> Thanks for your answer.
>>
>> What would be the governor for this setup ?
>>
>
> The default governor when using arm64_defconfig is step_wise.
Ok, thanks
--
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs
Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog
