Re: [PATCH 1/7] ARM: dts: exynos: Add missing memory node for Exynos5440 boards

[Krzysztof Kozlowski]

On 09/01/2016 11:36 AM, Javier Martinez Canillas wrote:
> Hello Krzysztof,
> 
> On 09/01/2016 11:28 AM, Krzysztof Kozlowski wrote:
>> On 09/01/2016 11:19 AM, Javier Martinez Canillas wrote:
>>> Hello Krzysztof,
>>>
>>> On 09/01/2016 11:05 AM, Krzysztof Kozlowski wrote:
>>>> On 09/01/2016 10:43 AM, Javier Martinez Canillas wrote:
>>>>>>> --- a/arch/arm/boot/dts/exynos5440-ssdk5440.dts
>>>>>>> +++ b/arch/arm/boot/dts/exynos5440-ssdk5440.dts
>>>>>>> @@ -21,6 +21,11 @@
>>>>>>>  		bootargs = "root=/dev/sda2 rw rootwait ignore_loglevel earlyprintk no_console_suspend mem=2048M@0x80000000 mem=6144M@0x100000000 console=ttySAC0,115200";
>>>>>>>  	};
>>>>>>>  
>>>>>>> +	memory {
>>>>>>> +		device_type = "memory";
>>>>>>> +		reg = <0 0>;
>>>>>>> +	};
>>>>>>
>>>>>> This looks okay, but the default size and address cells won't be
>>>>>> present. Won't that be a problem?
>>>>>>
>>>>>
>>>>> They will be present since at this point exynos5440.dtsi is still including
>>>>> exynos5440.dtsi, we are just overriding the memory node with the same values
>>>>> than skeleton.dtsi so this can be removed later in patch 7/7.
>>>>
>>>> I don't get it. exynos5440.dtsi does not set size/address cells in top node.
>>>>
>>>
>>> Sorry, maybe I'm misunderstanding your question. The size/address cells are set
>>> in the top node by exynos5440.dtsi including the skeleton.dtsi (since these are
>>> set in that included dtsi).
>>
>> Where? Nothing in my sources...
>>
>> 15 / {
>> 16         compatible = "samsung,exynos5440", "samsung,exynos5";
>> 17
>> 18         interrupt-parent = <&gic>;
>>
> 
> No, I meant that these are set by the fact that exynos5440.dtsi is including
> the skeleton.dtsi, since the dtsi set these properties in the top node.
> 
>>>
>>> IOW, scripts/dtc/dtx_diff doesn't show differences for these exynos5440 DTB
>>> before and after $SUBJECT.
>>
>> Maybe these are also the default values so no differences... stating
>> them explicitly would be useful.
>>
> 
> I could do it in this patch... but I don't see why that should be needed since:
> 
> a) All the others Exynos SoC dtsi file don't explicitly set these properties
>    since are set by skeleton.dtsi when including it. So not adding them keeps
>    exynos5440.dtsi consistent with the other SoC dtsi.
> 
> b) Adding those to the Exynos SoC dtsi when removing the skeleton.dtsi is done
>    in a later patch of this series. Again, making it consistent with all the
>    Exynos SoC dtsi since none will include skeleton.dtsi anymore.

Ah, I missed it. The 'b' is the solution I wanted so everything is fine.

Best regards,
Krzysztof