Re: [PATCH 0/3] arm64: dts: qcom: rename dtsi files for sm6150, x1e80100 and qcs8300

[Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>]

On 9/23/25 5:17 PM, Dmitry Baryshkov wrote:
> On Tue, Sep 23, 2025 at 03:00:19PM +0200, Luca Weiss wrote:
>> On Tue Sep 23, 2025 at 1:01 PM CEST, Dmitry Baryshkov wrote:
>>> Bjorn said to me that he disliked the conseqeuences of renaming
>>> qcs615.dtsi to sm6150.dtsi. Let's increase entropy even more and rename
>>> even more base DTSI files.
>>>
>>> Leaving jokes aside. It's not uncommon to see different names for the
>>> same SoC (or almost same SoC, with no visible differences from the Linux
>>> side). The platform now known as 'lemans' is a good example, because it
>>> had been using SA8775P, QCS9100 and QCS9075 in different contexts for
>>> slightly different modifications. QCS8300 / QCS8275 is another example.
>>> All such names cause a lot of confusion when somebody tries to follow
>>> the actual SoC used by the platform.
>>>
>>> For 'lemans' after a lot of trial, error, Naks and reviews we've settled
>>> upon having 'lemans.dtsi', the core DT file for the platform and then
>>> naming individual DT files following the marketing name for the platform
>>> or for the SoC.
>>>
>>> Apply the same approach to several other platforms, renaming the base
>>> DTSI and keeping the DT names as is.
>>
>> If we're doing this already, sc7280 -> kodiak? That also covers sc7280,
>> qc{m,s}6490 & 5430 and sm7325.
> 
> And few other platform names with multiple SoC names per platform.
> 
>>
>> Also, does this mean that milos-based Fairphone 6 the dtsi should be
>> milos.dtsi while dts should be sm7635-fairphone-fp6? The latest patch
>> series uses milos-fairphone-fp6.dts.
> 
> I'd leave this to Bjorn's discretion.

I think this only makes sense if milos-nos-sm7635-fairphone-fp6 has
physical differences that stem from the SKU being equipped with a
differently capable SoC (which remains mostly software compatible
otherwise)

Konrad