[PATCH 4/7] ARM: dts: amlogic: meson6: delete undocumented nodes

[Christian Hewitt]

> On 15 Sep 2024, at 11:12 PM, Martin Blumenstingl <martin.blumenstingl at googlemail.com> wrote:
> 
> Hi Neil,
> 
> On Wed, Sep 11, 2024 at 12:19 PM Neil Armstrong
> <neil.armstrong at linaro.org> wrote:
>> 
>> Delete nodes in meson6.dtsi that are clearly undocumented & unused for meson6,
>> fixing the following check errors:
>> audio-controller at 5400: compatible:0: 'amlogic,aiu' is not one of ['amlogic,aiu-gxbb', 'amlogic,aiu-gxl', 'amlogic,aiu-meson8', 'amlogic,aiu-meson8b']
>>        from schema $id: http://devicetree.org/schemas/sound/amlogic,aiu.yaml#
>> audio-controller at 5400: compatible: ['amlogic,aiu'] is too short
>>        from schema $id: http://devicetree.org/schemas/sound/amlogic,aiu.yaml#
>> /soc/bus at c1100000/pwm at 8550: failed to match any schema with compatible: ['amlogic,meson-pwm']
>> /soc/bus at c1100000/pwm at 8650: failed to match any schema with compatible: ['amlogic,meson-pwm']
>> phy at 8800: compatible: 'oneOf' conditional failed, one must be fixed:
>>        ['amlogic,meson-mx-usb2-phy'] is too short
>>        'amlogic,meson-mx-usb2-phy' is not one of ['amlogic,meson8-usb2-phy', 'amlogic,meson8b-usb2-phy', 'amlogic,meson8m2-usb2-phy']
>>        'amlogic,meson-gxbb-usb2-phy' was expected
>>        from schema $id: http://devicetree.org/schemas/phy/amlogic,meson8b-usb2-phy.yaml#
>> phy at 8820: compatible: 'oneOf' conditional failed, one must be fixed:
>>        ['amlogic,meson-mx-usb2-phy'] is too short
>>        'amlogic,meson-mx-usb2-phy' is not one of ['amlogic,meson8-usb2-phy', 'amlogic,meson8b-usb2-phy', 'amlogic,meson8m2-usb2-phy']
>>        'amlogic,meson-gxbb-usb2-phy' was expected
>>        from schema $id: http://devicetree.org/schemas/phy/amlogic,meson8b-usb2-phy.yaml#
>> /soc/bus at c1100000/mmc at 8c20: failed to match any schema with compatible: ['amlogic,meson-mx-sdio']
>> mmc at 8e00: compatible:0: 'amlogic,meson-mx-sdhc' is not one of ['amlogic,meson8-sdhc', 'amlogic,meson8b-sdhc', 'amlogic,meson8m2-sdhc']
>>        from schema $id: http://devicetree.org/schemas/mmc/amlogic,meson-mx-sdhc.yaml#
>> mmc at 8e00: compatible: ['amlogic,meson-mx-sdhc'] is too short
>>        from schema $id: http://devicetree.org/schemas/mmc/amlogic,meson-mx-sdhc.yaml#
>> remoteproc at 1c: compatible:0: 'amlogic,meson-mx-ao-arc' is not one of ['amlogic,meson8-ao-arc', 'amlogic,meson8b-ao-arc']
>>        from schema $id: http://devicetree.org/schemas/remoteproc/amlogic,meson-mx-ao-arc.yaml#
>> remoteproc at 1c: compatible: ['amlogic,meson-mx-ao-arc'] is too short
>>        from schema $id: http://devicetree.org/schemas/remoteproc/amlogic,meson-mx-ao-arc.yaml#
>> 
>> Signed-off-by: Neil Armstrong <neil.armstrong at linaro.org>
> Generally I'm fine with this. I'm wondering if we should go one step
> further by decommissioning Meson6 support?
> To be specific here, I'm thinking of removing MACH_MESON6 from
> arch/arm/mach-meson/Kconfig and deleting meson6{.dtsi,-atv1200.dts}
> 
> I don't have any boards with Meson6 SoC, there's no upstream GPIO,
> pinctrl, clock, MMC, <you name it> driver for that SoC either.
> So all we have upstream is the ability to boot to initramfs and use
> serial console(s).
> If someone is interested in Meson6 support they should bring it back
> along with a few more patches for upstream support of further
> peripherals.

I think that’s a sensible move. LibreELEC was one of the few distros
that supported Meson6 hardware in the past (via the vendor kernel)
but we ended support in 2018. Meson6 devices are mostly/all? cheap
Android boxes dating from 2011-2013 and these have either been been
discarded by users through upgrades or the boxes died long ago. There
is still community interest in supporting Meson8 boards (the 8x core
devices are used for retro-gaming) but not the earlier Meson6 ones.

Christian