[PATCH v5 3/7] ARM: dts: Exynos542x/5800: add CPU OPP properties

Krzysztof Kozlowski k.kozlowski at samsung.com
Thu Dec 10 20:53:26 PST 2015


On 11.12.2015 13:38, Viresh Kumar wrote:
> On 11-12-15, 13:18, Krzysztof Kozlowski wrote:
>> We had such configuration before (before df09df6f9ac3). I don't see any
>> benefit in what you described. Where is the "thing" to be fixed? It is
>> mixed up. The contiguous ordering is easier to read and more natural.
> 
> This is what you are doing today (keeping on one CPU per cluster to
> simplify it):
> 
> 		cpu0: cpu at 0 {
> 			device_type = "cpu";
> 			compatible = "arm,cortex-a15";
> 			reg = <0x0>;
> 			clock-frequency = <1800000000>;
> 			cci-control-port = <&cci_control1>;
> 		};
> 
> 		cpu4: cpu at 100 {
> 			device_type = "cpu";
> 			compatible = "arm,cortex-a7";
> 			reg = <0x100>;
> 			clock-frequency = <1000000000>;
> 			cci-control-port = <&cci_control0>;
> 		};
> 
> 
> Then you overwrite it with:
> 
>                 &cpu0 {
>                 	device_type = "cpu";
>                 	compatible = "arm,cortex-a7";
>                 	reg = <0x100>;
>                 	clock-frequency = <1000000000>;
>                 	cci-control-port = <&cci_control0>;
>                 };
>                 
>                 &cpu4 {
>                 	device_type = "cpu";
>                 	compatible = "arm,cortex-a15";
>                 	reg = <0x0>;
>                 	clock-frequency = <1800000000>;
>                 	cci-control-port = <&cci_control1>;
>                 };
> 
> 
> Don't you think this isn't the right way of solving problems?
> 
> The DT overwrite feature isn't there to do such kind of stuff, though
> it doesn't stop you from doing that.

This is quite ugly, indeed, and it is getting uglier :)... but it does
not violate the idea of DT to describe the hardware. Both hardware
descriptions - the 5420 and overridden - are entirely correct... because
the CPU ordering comes from booting sequence (actually code in IROM
decides according to pulled up GPIO gpg2-1).


> Either you should keep separate paths for both the SoCs,

I like that idea. That makes it much more readable. Thanks for feedback!
I will send a patch for that.


> or can solve
> it the way I suggested earlier.
> 
> This came up because in the current series you are doing this:
> 
> 		cpu0: cpu at 0 {
> 			compatible = "arm,cortex-a15";
>                         operating-points-v2 = <&cpu0_opp_table>;
> 		};
> 
> 		cpu4: cpu at 100 {
> 			device_type = "cpu";
> 			compatible = "arm,cortex-a7";
>                         operating-points-v2 = <&cpu1_opp_table>;
> 		};
> 
> 
> Then you overwrite it with:
> 
>                 &cpu0 {
>                 	compatible = "arm,cortex-a7";
>                         operating-points-v2 = <&cpu1_opp_table>;
>                 };
>                 
>                 &cpu4 {
>                 	compatible = "arm,cortex-a15";
>                         operating-points-v2 = <&cpu0_opp_table>;
>                 };

Yes, it is getting uglier with each change...

Best regards,
Krzysztof





More information about the linux-arm-kernel mailing list