[U-Boot] [PATCH v2 2/6] apalis/colibri_t20/t30: deactivate displaying board info

[Stephen Warren]

On 10/05/2016 10:12 AM, Stefan Agner wrote:
> On 2016-10-05 08:53, Stephen Warren wrote:
>> On 10/03/2016 02:27 PM, Stefan Agner wrote:
>>> On 03.10.2016 10:28, Stephen Warren wrote:
>>>> On 09/30/2016 04:00 AM, Marcel Ziswiler wrote:
>>>>> On Wed, 2016-09-28 at 12:00 -0600, Stephen Warren wrote:
>>>>>> On 09/28/2016 03:35 AM, Marcel Ziswiler wrote:
>>>>>>>
>>>>>>> Avoid a checkboard() name clash with our upcoming custom
>>>>>>> implementation
>>>>>>> thereof.
>>>>>> If you want to avoid naming conflicts, please simply name your new
>>>>>> function something that doesn't conflict. That way it will avoid
>>>>>> confusion is someone actually wants to enable the
>>>>>> CONFIG_DISPLAY_BOARDINFO option themselves, plus it avoids taking
>>>>>> the
>>>>>> current feature set away.
>>>>>
>>>>> No, it is not just any function. We do want our custom checkboard() to
>>>>> be called upon boot and not the Tegra generic one just printing a hard
>>>>> coded string.
>>>>>
>>>>> I guess alternatively we could gate the checkboard() call
>>>>> in arch/arm/mach-tegra/board2.c with a
>>>>>
>>>>> #if !defined(CONFIG_CUSTOM_BOARDINFO)
>>>>>
>>>>> just as introduced a while ago in common/board_info.c
>>>>>
>>>>> http://git.denx.de/?p=u-boot.git;a=blob;f=common/board_info.c;h=bd5dcfa
>>>>> 066358c2cc44ce5d19fcc3e77d555cd09;hb=HEAD#l20
>>>>>
>>>>> in order to not print the hard coded name from the device tree.
>>>>
>>>> I'd prefer to keep the behaviour standard across all Tegra boards. If
>>>> you want to do additional actions in the checkboard() function, I
>>>> suggest making it call an optional additional function:
>>>>
>>>> __weak int tegra_checkboard(void)
>>>> {
>>>>         return 0;
>>>> }
>>>>
>>>> int checkboard(void)
>>>> {
>>>>     ...
>>>>     return tegra_checkboard();
>>>> }
>>>
>>> Well that would print a message "Board: " ... twice, which is rather
>>> strange.
>>
>> Surely you simply make tegra_checkboard() not contain duplicate code?
>>
>>> What do you think of my idea?
>>
>> I'd rather not introduce any more ifdefs, but instead have a single
>> path through the code-base.
>
> Sorry, I was a bit unclear, with my other idea I meant the answer I sent
> to patch 3/6 of this patchset:
> http://lists.denx.de/pipermail/u-boot/2016-October/268669.html
>
> It does remove a ifdef...

That's probably better than adding more ifdefs. It does have the 
disadvantage of not using the common show_board_info() or Tegra-wide 
checkboard() implementations though, which means that anything added 
there won't execute on all Tegras, which could potentially be confusing.