Pull request: u-boot-sunxi/master

[André Przywara]

On 17/11/2020 01:56, Samuel Holland wrote:
> On 11/16/20 10:58 AM, Tom Rini wrote:
>> On Mon, Nov 16, 2020 at 04:51:32PM +0000, André Przywara wrote:
>>> On 16/11/2020 16:13, Tom Rini wrote:
>>>> On Mon, Nov 16, 2020 at 04:09:42PM +0000, André Przywara wrote:
>>>>> On 16/11/2020 15:50, Tom Rini wrote:
>>>>>
>>>>> Hi Tom,
>>>>>
>>>>>> On Mon, Nov 16, 2020 at 12:42:53PM +0530, Jagan Teki wrote:
>>>>>>
>>>>>>> Hi Tom, 
>>>>>>>
>>>>>>> Please pull this PR.
>>>>>>>
>>>>>>> Summary:
>>>>>>> - PinePhone support (Samuel)
>>>>>>> - V3/S3 support (Icenowy)
>>>>>>>
>>>>>>> thanks,
>>>>>>> Jagan.
>>>>>>>
>>>>>>> The following changes since commit de865f7ee1d9b6dff6e265dee44509c8274ea606:
>>>>>>>
>>>>>>>   Merge tag 'efi-2021-01-rc3' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi (2020-11-14 09:47:33 -0500)
>>>>>>>
>>>>>>> are available in the Git repository at:
>>>>>>>
>>>>>>>   https://gitlab.denx.de/u-boot/custodians/u-boot-sunxi master
>>>>>>>
>>>>>>> for you to fetch changes up to 27007e5d4a6c4ac1bbd1bc81fb4c19bc45191f35:
>>>>>>>
>>>>>>>   sunxi: dts: sync Allwinner V3s-related DTs from Linux 5.10-rc1 (2020-11-16 12:34:09 +0530)
>>>>>>>
>>>>>>
>>>>>> NAK.  A large number of sunxi boards fail to build with:
>>>>>> +board/sunxi/board.c:338:13: error: 'set_spl_dt_name' defined but not used [-Werror=unused-function]
>>>>>> +  338 | static void set_spl_dt_name(const char *name)
>>>>>> +      |             ^~~~~~~~~~~~~~~
>>>>>> +cc1: all warnings being treated as errors
>>>>>> +make[2]: *** [board/sunxi/board.o] Error 1
>>>>>> +make[1]: *** [board/sunxi] Error 2
>>>>>> +make: *** [sub-make] Error 2
>>>>>
>>>>> Ouch, thanks for that heads up!
>>>>> But this only happens after this very patch, and is fixed by a later
>>>>> patch, right? ("sunxi: board: Set fdtfile to match the DT chosen by SPL"
>>>>> introduces another user of set_spl_dt_name(), outside of any #ifdef's).
>>>>
>>>> No, that's from the PR itself.
>>>
>>> Right, sorry, looked at the wrong function.
>>> I found the culprit and have a fix, but I will run some actual tests on
>>> my boards tonight to make sure it really works.
>>> Jagan or me will send a new PR later then.
>>
>> OK, thanks for the quick follow-up!
> 
> Thank you, Andre, and sorry for the breakage.

No worries, I should have spotted it myself during the review. But it
just points to the fact that board.c is "ifdef hell" at its best, we
should clean that up one day.

And that taught me a lesson to not forget about those poor 32-bit boards ;-)

> For what it's worth, I am happy with the fixups you made.

Thanks for the confirmation!

Cheers,
Andre