[PATCH v6 00/14] arm: am335x: Switch to upstream devicetree

[Markus Schneider-Pargmann]

Hi Judith,

On Thu Apr 16, 2026 at 10:33 PM CEST, Judith Mendez wrote:
> Hi Markus,
>
> On 4/16/26 2:55 PM, Judith Mendez wrote:
>> Hi Markus,
>> 
>> On 4/15/26 11:02 AM, Markus Schneider-Pargmann (TI) wrote:
>>> The series is split into two logical groups:
>>>
>>> - Patches 1-3: Fixes for am335x-evm to boot and support the LCD panel
>>>    with the current u-boot devicetree.
>>> - Patches 4-14: Make upstream devicetree working with uboot. This
>>>    adds tick-timer, adds compatibility of the board code for different
>>>    USB probing, makes SPL smaller and updates the defconfig.
>>>
>>> The series has been tested on am335x-evm. Note that I don't have all
>>> other boards built with the am335x_*evm_*defconfig, so tests are
>>> appreciated. buildman reports builds are working.
>>>
>>> Dependencies:
>>>    - net: cpsw: Add cpsw-switch DT binding support
>>>        https://lore.kernel.org/r/20260410-topic-cpsw-switch-v2026-07- 
>>> next-v1-0-c914724b13e4 at baylibre.com
>>>
>> 
>> I have been planning to test this series on a couple of AM335 boards,
>> So far :
>> 
>> - am335 EVM: Pass
>> - am335 SK: Fail (at least SD boot regression)
>> 
>> Pending to test:
>> - beaglebone black and am335 ice boards.
>> 
>> 
>> am335 sk:
>> without of_upstream: https://gist.github.com/jmenti/ 
>> e51e1fc1b25c19f94786ad39a38737b7
>> with of_upstream: https://gist.github.com/jmenti/ 
>> fc17347eff33991a78ce365c42a76a25
>> 
> BBB boot without of_upstream: 
> https://gist.github.com/jmenti/b8f7151e7666ca16b9bec7ec5b2efe17
> BBB with of_upstream: 
> https://gist.github.com/jmenti/c22e017605d3d6c20189eaf3cfce6231

Thank you for testing!

> Seems like MMC enumeration is switched, BBB now boots to emergency shell.
>
> Also forgot to mentioned that there are more failures on AM335 SK
> than just MMC, related to USB and NAND I think.

Yes, the mmc node names are different now and it seems that's how uboot
enumerates the devices. With the next version I have
CONFIG_PHANDLE_CHECK_SEQ set which should use phandles to distinguish
them.

I found an issue with the USB setup which was different from the old
behaviour. The USB drivers were always probed which was only done on
demand before OF_UPSTREAM. I fixed that as well.

Then I noticed in the logs that MMC doesn't properly populate PARTUUID
for the one board because of the mmc numbering change and it boots from
the wrong mmc on the other board booting an old kernel and dropping you
to the emergency shell.

I didn't see NAND issues in the log I think. Can you point me to it
please?

Also I noticed ethernet is not there, probably you are missing the cpsw
series here:
  https://lore.kernel.org/r/20260410-topic-cpsw-switch-v2026-07-next-v1-0-c914724b13e4@baylibre.com

Best
Markus
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 289 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20260422/7a0722c3/attachment.sig>