[U-Boot] [PATCH 4/4 V3] EXYNOS: Move files from board/samsung to arch/arm

[Minkyu Kang <mk7.kang@samsung.com>]

>>> diff --git a/arch/arm/include/asm/arch-exynos/cpu.h b/arch/arm/include/asm/arch-exynos/cpu.h
>>> index 36b98c8..75dbe26 100644
>>> --- a/arch/arm/include/asm/arch-exynos/cpu.h
>>> +++ b/arch/arm/include/asm/arch-exynos/cpu.h
>>> @@ -115,7 +115,7 @@
>>>  #define EXYNOS5_DMC_PHY0_BASE                0x10C00000
>>>  #define EXYNOS5_DMC_PHY1_BASE                0x10C10000
>>
>> maybe it can be dmc0 and dmc1
>> If you want to keep the interface with exynos4.
> EXYNOS4 donot have seperate Phy controller it is a part of  DMC controller.
> then I need to add following for EXYNOS4
> EXYNOS4_DMC_PHY1_BASE    DEVICE_NOT_AVAILABLE
> EXYNOS4_DMC_PHY_BASE       DEVICE_NOT_AVAILABLE
> 
>>
>>>  #define EXYNOS5_GPIO_PART3_BASE              0x10D10000
>>> -#define EXYNOS5_DMC_CTRL_BASE                0x10DD0000
>>> +#define EXYNOS5_DMC0_BASE            0x10DD0000
>>
>> why?
> if we want to make funstion to get the DMC base address it is needed
> to be added for EXYNOS5 as well
> hence made this change.
>>
>>>  #define EXYNOS5_GPIO_PART1_BASE              0x11400000
>>>  #define EXYNOS5_MIPI_DSIM_BASE               0x11D00000
>>>  #define EXYNOS5_USB_HOST_EHCI_BASE   0x12110000
>>> @@ -135,6 +135,7 @@
>>>
>>>  #define EXYNOS5_ADC_BASE             DEVICE_NOT_AVAILABLE
>>>  #define EXYNOS5_MODEM_BASE           DEVICE_NOT_AVAILABLE
>>> +#define EXYNOS5_DMC1_BASE            DEVICE_NOT_AVAILABLE
>>
>> It looks weird.
> Since EXYNOS4 has 2 DMC controllers and EXYNOS5 has only 1 DMC
> controller I had to add this to avoid compilation error.
> Please do let me know if  you are fine with these changes.
>>

I suggest following.

EXYNOS4_DMC_BASE	0x-
EXYNOS4_DMC_PHY_BASE	DEVICE_NOT_AVAILABLE

EXYNOS4X12_DMC_BASE	0x-
EXYNOS4X12_DMC_PHY_BASE	DEVICE_NOT_AVAILABLE

EXYNOS5_DMC_PHY_BASE	0x-
EXYNOS5_DMC_BASE	0x-

then we need to add 2 SAMSUNG_BASE.

SAMSUNG_BASE(dmc, DMC_BASE)
SAMSUNG_BASE(dmc_phy, DMC_PHY_BASE)

and we can access dmc1 and dmc phy1 by offset.

What do you think about it?

Thanks,
Minkyu Kang.