Re: [PATCH 1/7] ARM: dts: s5pv210: Split memory nodes to match spec

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 2022-03-23 8:06 a.m., Krzysztof Kozlowski wrote:
> On 23/03/2022 15:59, Jonathan Bakker wrote:
>> Hi Krzysztof,
>>
>> On 2022-03-23 3:54 a.m., Krzysztof Kozlowski wrote:
>>> On 22/03/2022 21:11, Jonathan Bakker wrote:
>>>> Memory nodes should only have a singular reg property in them, so
>>>> split the memory nodes such that there is only per node.
>>>>
>>>> Signed-off-by: Jonathan Bakker <xc-racer2@xxxxxxx>
>>>> ---
>>>>  arch/arm/boot/dts/s5pv210-aquila.dts |  8 ++++++--
>>>>  arch/arm/boot/dts/s5pv210-aries.dtsi | 14 +++++++++++---
>>>>  arch/arm/boot/dts/s5pv210-goni.dts   | 14 +++++++++++---
>>>>  3 files changed, 28 insertions(+), 8 deletions(-)
>>>>
>>>> diff --git a/arch/arm/boot/dts/s5pv210-aquila.dts b/arch/arm/boot/dts/s5pv210-aquila.dts
>>>> index 6423348034b6..6984479ddba3 100644
>>>> --- a/arch/arm/boot/dts/s5pv210-aquila.dts
>>>> +++ b/arch/arm/boot/dts/s5pv210-aquila.dts
>>>> @@ -29,8 +29,12 @@
>>>>  
>>>>  	memory@30000000 {
>>>>  		device_type = "memory";
>>>> -		reg = <0x30000000 0x05000000
>>>> -			0x40000000 0x18000000>;
>>>> +		reg = <0x30000000 0x05000000>;
>>>> +	};
>>>> +
>>>> +	memory@40000000 {
>>>> +		device_type = "memory";
>>>> +		reg = <0x40000000 0x18000000>;
>>>>  	};
>>>>  
>>>>  	pmic_ap_clk: clock-0 {
>>>> diff --git a/arch/arm/boot/dts/s5pv210-aries.dtsi b/arch/arm/boot/dts/s5pv210-aries.dtsi
>>>> index 160f8cd9a68d..70ff56daf4cb 100644
>>>> --- a/arch/arm/boot/dts/s5pv210-aries.dtsi
>>>> +++ b/arch/arm/boot/dts/s5pv210-aries.dtsi
>>>> @@ -24,9 +24,17 @@
>>>>  
>>>>  	memory@30000000 {
>>>>  		device_type = "memory";
>>>> -		reg = <0x30000000 0x05000000
>>>> -			0x40000000 0x10000000
>>>> -			0x50000000 0x08000000>;
>>>
>>> 0x40000000 to 0x58000000 is continues, so I wonder why it is split? Look
>>> at Aquila DTS.
>>>
>>>
>>
>> Yes, it was split in the vendor kernel as well [1], and that's been continued along
>> here.  I personally don't see a reason to keep it split, but there might be something
>> I'm not aware of.
>>
> 
> I guess they wanted maybe to express the physical banks. Fine with me.
> Just your explanation is not entirely correct:
>> Memory nodes should only have a singular reg property in them,
> one reg but it can have multiple items. Why do think multiple reg items
> is not allowed?
> 

I was basing it off of this warning when running make dtbs_check

rch/arm/boot/dts/s5pv210-aquila.dt.yaml: /: memory@30000000:reg:0: [805306368, 83886080, 1073741824, 402653184] is too long
	From schema: /home/jon/.local/lib/python3.7/site-packages/dtschema/schemas/reg.yaml

and this solved the warning, booted, and I still had the correct
amount of memory.

Would

	memory@30000000 {
		device_type = "memory";
		reg = <0x30000000 0x05000000>,
			<0x40000000 0x18000000>;
	};

be equally correct?  (untested).

Thanks,
Jonathan

> 
> Best regards,
> Krzysztof
> 



[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]


  Powered by Linux