[PATCH 2/5] dt-bindings: add Canaan K230 boards compatible strings

Krzysztof Kozlowski krzysztof.kozlowski at linaro.org
Mon Mar 4 02:27:19 PST 2024


On 04/03/2024 09:51, Yangyu Chen wrote:
> On 2024/3/4 16:11, Krzysztof Kozlowski wrote:
>> On 03/03/2024 14:26, Yangyu Chen wrote:
>>> Since K230 was released, K210 is no longer the only SoC in the Kendryte
>>> series, so remove the K210 string from the description. Also, add two
>>> boards based on k230 to compatible strings to allow them to be used in the
>>> dt.
>>>
>>> Signed-off-by: Yangyu Chen <cyy at cyyself.name>
>>> ---
>>>   Documentation/devicetree/bindings/riscv/canaan.yaml | 13 ++++++++++++-
>>>   1 file changed, 12 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/Documentation/devicetree/bindings/riscv/canaan.yaml b/Documentation/devicetree/bindings/riscv/canaan.yaml
>>> index 41fd11f70a49..444758db964e 100644
>>> --- a/Documentation/devicetree/bindings/riscv/canaan.yaml
>>> +++ b/Documentation/devicetree/bindings/riscv/canaan.yaml
>>> @@ -10,7 +10,7 @@ maintainers:
>>>     - Damien Le Moal <dlemoal at kernel.org>
>>>   
>>>   description:
>>> -  Canaan Kendryte K210 SoC-based boards
>>> +  Canaan Kendryte SoC-based boards
>>>   
>>>   properties:
>>>     $nodename:
>>> @@ -42,6 +42,17 @@ properties:
>>>         - items:
>>>             - const: canaan,kendryte-k210
>>>   
>>> +      - items:
>>> +          - const: canaan,k230-usip-lp3-evb
>>> +          - const: canaan,kendryte-k230
>>> +
>>> +      - items:
>>> +          - const: canaan,canmv-k230
>>
>> Why this is not part of previous entry in an enum?
>>
>>> +          - const: canaan,kendryte-k230
>>> +
>>> +      - items:
>>> +          - const: canaan,kendryte-k230
>>
>> Usually you cannot run SoCs alone. What does it represent (in real life)?
>>
> 
> I'm not sure what it means.
> 
> If you wonder why should I add a compatible string for soc, that is 
> although we cannot run SoCs alone, adding a soc compatible will allow 
> some bootloaders or SBI on RISC-V to choose an errata for a soc. Such as 
> this opensbi patch. [1]

No, this piece of code will not allow this. They choose errata
regardless of this change.

> 
> If you wonder why I should allow a soc-compatible string with soc alone, 
> that is because k210 did it previously. And provide a k210_generic.dts 

I don't remember background behind k210_generic. Any SoC-compatible
alone is exception, so needs serious justification. Drop it or provide
proper rationale.

> to use it. I haven't provided generic dts now but allowing only 
> soc-compatible string alone would also be acceptable I think.

No, it is not. Stop making own rules.


Best regards,
Krzysztof




More information about the linux-riscv mailing list