[PATCH] ARM: dts: stm32: Enable thermal sensor support on stm32mp15xx-dhcor

[Marek Vasut]

On 9/25/20 6:20 PM, Manivannan Sadhasivam wrote:
> On Fri, Sep 25, 2020 at 01:12:12PM +0200, Marek Vasut wrote:
>> On 9/25/20 4:21 AM, Manivannan Sadhasivam wrote:
>>>
>>>
>>> On 24 September 2020 4:11:11 PM IST, Marek Vasut <marex at denx.de> wrote:
>>>> On 9/24/20 7:16 AM, Manivannan Sadhasivam wrote:
>>>>>
>>>>>
>>>>> On 24 September 2020 4:55:35 AM IST, Marek Vasut <marex at denx.de>
>>>> wrote:
>>>>>> Enable STM32 Digital Thermal Sensor driver for stm32mp15xx-dhcor
>>>> SoMs.
>>>>>>
>>>>>> Fixes: 94cafe1b6482 ("ARM: dts: stm32: Add Avenger96 devicetree
>>>> support
>>>>>> based on STM32MP157A")
>>>>>
>>>>> The change looks good but what does this patch fixes? 
>>>>
>>>> The missing temp sensor, which helps you detect overheat of the SoC.
>>>> That is esp. important on the 800 MHz AV96.
>>>
>>> This doesn't quality as a "fix". Essentially you're just adding a missing feature and not fixing any issues. So please remove the fixes tag and resubmit. 
>>
>> I would argue that if the system overheats and crashes, we want to know
>> about that, possibly in advance so thermal throttling can be applied.
>> Currently this is not possible and I think that is a bug.
> 
> No, this is not a _bug_. This is a missing feature that the current kernel
> doesn't support and you know about that! The fact that you can trigger a crash
> due to hw limitation doesn't qualify as a bug IMO. And you can do that by other
> means also (CPU throttling without CPUFreq support etc...)

I would say that makes the hardware pretty useless, which is not the
expectation I would have of a system I would like to use in production.
Just assume your laptop would randomly overheat and crash, would you
consider that a missing feature or a bug that should be fixed ?

> Anyway, I'll stop here and let Alex to make a call.

I agree.