Re: [PATCHv4 7/7] ARM: dts: Add device tree sources for Exynos3250

[Chanwoo Choi]

Hi Tomasz,

On 05/09/2014 02:02 PM, Tomasz Figa wrote:
> Hi Chanwoo,
> 
> On 09.05.2014 03:06, Chanwoo Choi wrote:
>> On 04/26/2014 09:51 AM, Tomasz Figa wrote:
>>> On 25.04.2014 03:16, Chanwoo Choi wrote:
> 
> [snip]
> 
>>>> +    cpus {
>>>> +        #address-cells = <1>;
>>>> +        #size-cells = <0>;
>>>> +
>>>> +        cpu@0 {
>>>> +            device_type = "cpu";
>>>> +            compatible = "arm,cortex-a7";
>>>> +            reg = <0>;
>>>> +            clock-frequency = <1000000000>;
>>>> +        };
>>>
>>> Why only one CPU? I believe Exynos3250 is dual core.
>>
>> I'll add cpu1 information.
>>
>>> Also are physical IDs of the cores really 0 and 1? On Exynos4210 for example they are 0x900 and 0x901, while on Exynos4212 they are 0xa00 and 0xa01. Please check this.
>>
>> The 'reg' property means only hardware id(hwid) of CPU.
>> You can check it on arm_dt_init_cpu_maps() in arch/arm/kernel/devtree.c.h.
>> or Documentation/devicetree/bindings/arm/cpus.txt.
>>
> 
> Well, as described in Documentation/devicetree/bindings/arm/cpus.txt, on 32-bit ARM v7 or later CPUs the "reg" property should be equal to the lower 24-bits of MPIDR value of given CPU, which in addition to core ID includes also cluster ID, which can be non-zero, even on single cluster SoCs (like it is on Exynos4210 and 4x12).

I checked the lower 24-bit of MPIDR value for Exynos3250 in arm_dt_init_cpu_maps().
- the lower 24-bit of MPIDR for CPU0 is '0x0'.

> 
>>>> +    };
>>>> +
>>>> +    fixed-rate-clocks {
>>>> +        compatible = "simple-bus";
>>>> +        #address-cells = <1>;
>>>> +        #size-cells = <0>;
> 
> [snip]
> 
>>>> +    cmu: clock-controller@10030000 {
>>>> +        compatible = "samsung,exynos3250-cmu";
>>>> +        reg = <0x10030000 0x20000>;
>>>> +        #clock-cells = <1>;
>>>> +    };
>>>> +
>>>> +    rtc@10070000 {
>>>
>>> Please add label to the node, so it can be referenced from board dts files added later (using the method I explained above).
>>
>> OK, I'll add lable as following:
>>
>>     rtc_0: rtc@10070000 {
> 
> There is no need to suffix the RTC with _0, as there is just one RTC in the SoC. So in this case rtc: rtc@10070000 will be enough.

OK, I'll modify it without prefix('_0).

Best Regards,
Chanwoo Choi
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/