Re: [PATCH 3/4] arm64: dts: renesas: Add R8A78000 X5H DTs

[Krzysztof Kozlowski <krzk@kernel.org>]

On 09/09/2025 10:50, Geert Uytterhoeven wrote:
> Hi Krzysztof,
> 
> On Tue, 9 Sept 2025 at 10:38, Krzysztof Kozlowski <krzk@kernel.org> wrote:
>> On 09/09/2025 09:51, Geert Uytterhoeven wrote:
>>> On Tue, 9 Sept 2025 at 09:46, Krzysztof Kozlowski <krzk@kernel.org> wrote:
>>>> On Tue, Sep 09, 2025 at 01:45:09AM +0000, Kuninori Morimoto wrote:
>>>
>>>>> +     extal_clk: extal {
>>>>
>>>> Use some sane prefix.
>>>>
>>>> https://web.git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/devicetree/bindings/clock/fixed-clock.yaml
>>>>
>>>>> +             compatible = "fixed-clock";
>>>>> +             #clock-cells = <0>;
>>>>> +             /* This value must be overridden by the board */
>>>>> +             clock-frequency = <0>;
>>>>
>>>> Drop instead
>>>
>>> clock-frequency is a required property?
>>
>> And it should be provided by the board or fail the DTS. I think now it
>> hides the dtbs_check warnings for no real gain/reason.
> 
> In this particular case, the clock-frequency could indeed be omitted
> n the .dtsi, as the extal clock is mandatory, thus must always be
> augmented in board .dts.
> In case of optional clocks (e.g. serial, PCIe, or CAN external clock
> oscillators may not be populated on all boards), it is different.

1. That is not such clock.
2. So putting invalid value is correct? No, devices which do not exist
should not be in DTSI or should be disabled. Having wrong values is
never correct.

Best regards,
Krzysztof