[PATCH v2 00/38] Thirty eight patches big Exynos cleanup

[Kukjin Kim]

On 06/18/13 20:11, Tomasz Figa wrote:
> On Tuesday 18 of June 2013 19:23:29 Kukjin Kim wrote:
>> Tomasz Figa wrote:
>>> Hi All,
>>>
>>> As I promised, I am doing some spring (or rather summer) cleaning
>>> of Exynos- and Samsung- related code.
>>>
>>> This first part consists mostly of removing dead code remaining after
>>> removal of ATAGS support for Exynos, but several patches cleans up
>>> other things found by the way.
>>>
>>> See particular patches for more detailed description.
>>>
>>> On Exynos4210-based Trats board:
>>> Tested-by: Tomasz Figa<t.figa at samsung.com>
>>>
>>> Build tested every patch with s3c6400_defconfig, s5pv210_defconfig
>>> and exynos_defconfig.
>>>
>>> Changes since v1:
>>>   - Split patches removing SOC_EXYNOS4412
>>>   - Adjusted drivers to check for ARCH_EXYNOS in addition to PLAT_S5P
>>>   - Fixed compilation issue on s5pc100
>>>   - Removed empty line from mach-exynos/Makefile
>>>   - Removed SAMSUNG_GPIOLIB_4BIT selection from ARCH_EXYNOS
>>>   - Sorted Kconfig and Makefile entries
>>>   - Added additional cleanup from Arnd (split and slightly modified)
>>
>> Tomasz, good job.
>
> Thanks.
>
>> After looking at, almost same with your previous version but modifying 16,
>> 19, 21, 25(just Cc) and adding some patches 17, 18, 29 and 35 to 38. Would
>> be helpful to  me if you could note about that in this series ;-)
>
> OK, my bad, haven't though about it, sorry. Will remember to do it next time
> such series update happens.
>
> By the way, patch 14 has been updated as well.
>
Thanks ;)

> Best regards,
> Tomasz
>
>> Anyway, looks good I will apply this series
>
> OK, thanks.
>
>> except removing
>> CONFIG_SOC_EXYNOS4412 patches.
>
> I don't really understand this decision, but if you really insist...
>
Sorry about that, but exynos4412 is not same with exynos4212 except 
number of cores and exynos4412 is not one kind...

> Still, I think current setup is wrong, because you need to specify both SoCs
> in drivers and both Kconfig entries must have "select" and "depends on" clauses
> duplicated, while there are places currently where only one has been specified
> by mistake.
>
> So if we can't get this duplication removed, I'd suggest introducing
> CONFIG_SOC_EXYNOS4X12 which would represent the whole SoC series and making
> existing two entries just simple bools selecting the new one.
>
Could be a solution but let's wait until something will be implemented 
in mainline for exynos4412 not exynos4212 :)

Thanks,
- Kukjin