Re: [PATCH 2/2] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441

[Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>]

On 1/28/2025 3:54 PM, Krzysztof Kozlowski wrote:
> On 28/01/2025 10:41, Raj Kumar Bhagat wrote:
>>>> +
>>>> +	reserved-memory {
>>>> +		#address-cells = <2>;
>>>> +		#size-cells = <2>;
>>>> +		ranges;
>>>> +
>>>> +		q6_region: wcss@4a900000 {
>>>> +			reg = <0x0 0x4a900000 0x0 0x02b00000>;
>>>> +			no-map;
>>>> +		};
>>>> +
>>>> +		m3_dump: m3-dump@4d400000 {
>>>
>>> This fails with your wireless patchset.
>>>
>>
>> Yes, this will fail with v4 wireless patch.
>> We have v5 (yet to be sent) to read the correct reserved memory. Since,
>> in v4 I was asked to address and send DTS patch separately, The corresponding
>> driver patches are yet to be sent.
> 
> 
> So you send it to prove my point? OK, fine. So this patch confirms your
> driver changes should be rejected.
> 

There seems to be some confusion here.

Based on the comments in v4, as quoted below:
"Please post the updated DTS in a separate series after fixing all the issues
pointed out by the DTS coding style."

I understood that you wanted to review only the DT-related patches first
before the driver implementation. Hence, this is posted separately before the
v5 driver is posted.

If I misunderstood, please let me know. We will continue with v5 of the original
series 'wifi: ath12k: add Ath12k AHB driver support for IPQ5332'. The v5 will
include dt-binding, driver, and reference DTS as well.

>>
>>>> +			reg = <0x0 0x4D400000 0x0 0x100000>;
>>>> +			no-map;
>>>> +		};
>>>> +	};
>>>>  };
>>>>  
>>>>  &blsp1_i2c1 {
>>>> @@ -63,3 +104,16 @@ data-pins {
>>>>  		};
>>>>  	};
>>>>  };
>>>> +
>>>> +&wifi0 {
>>>> +	memory-region = <&q6_region>, <&m3_dump>;
>>>> +	memory-region-names = "q6-region", "m3-dump";
>>>
>>> Binding said you have four items. I don't understand why this varies and
>>> why this is 2 items instead.
>>>
>>
>> In total we have 4 items, with minItems as 2.
>> The other 2 items (q6-caldb & mlo-global-mem) functionalities are not yet enabled
>> in ath12k driver. Hence these memory are not reserved yet.
> 
> That's not a valid reason to skip valid items. DTS is not dependent on
> driver. Submit proper hardware description, not something carved to
> match your driver.
> 

Sure, we can have all four reserved memory sections in the DTS. However, ath12k
will not use all four reserved memory sections as of now. Once the functionality
corresponding to q6-caldb and mlo-global-mem is enabled, these reserved memory
sections will be used.

In that case, we can also drop 'minItems' to 2 in the binding?