On 01/10/2024 15:01, Francesco Dolcini wrote:
> Hello Krzysztof,
>
> On Tue, Oct 01, 2024 at 01:54:56PM +0200, Krzysztof Kozlowski wrote:
>> On 01/10/2024 13:14, João Paulo Gonçalves wrote:
>>> From: João Paulo Gonçalves <joao.goncalves@xxxxxxxxxxx>
>>>
>>> With commit f1c9ce0ced2d ("iio: adc: ti-ads1015: Add TLA2024 support") a
>>> new compatible was introduced for TLA2024 ADC. Update the device
>>> tree to use the correct compatible for the Verdin-AM62 hardware.
>>>
>>> Signed-off-by: João Paulo Gonçalves <joao.goncalves@xxxxxxxxxxx>
>>> ---
>>> arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> index 5bef31b8577b..f201722d81b3 100644
>>> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> @@ -1220,7 +1220,7 @@ sensor@48 {
>>> };
>>>
>>> adc@49 {
>>> - compatible = "ti,ads1015";
>>> + compatible = "ti,tla2024";
>>
>> So it is not always TI, who breaks their users. :) (as pointed out in
>> LPC DT BoF).
>
> So, let's adjust what I said at that time, I think is important, and I
> appreciate you giving me an excuse for doing that :-)
>
> Lately as Toradex we are working a lot with TI, and one of the reasons is
> that they have a great software support, backed-up by a great strategy
> on the way they contribute to the various upstream projects they build
> their SDK on top (Linux, U-Boot, and more).
>
> With that is normal that while working so closely with them we find
> issues, everybody have those, it's just that those are the one we
> care the most at the moment :-). Not to mention that we started working
> with TI a couple of years ago, so TI is still somehow "new" to us and we
> are still "learning".
>
> On this regards I was recently working on updating our BSP to the
> latest SDK from TI, that is based on a v6.6 stable kernel and looking at
> the patches we had to apply on top, the total counts of the patches we
> do not have in mainline to support the board subject of this patch is
> just _zero_. This to me is a great achievement.
>
> Nishant: this is also for you, and feel free to "market" this
> internally/externally :-)
>
>
>> If you want to break users, sure, but at least explain in commit msg why.
>
> Now, on this specific topic, the actual device that is assembled on this
> board is a TI TLA2024, and it's like that since ever, the board never
> changed. The current compatible is not matching what is assembled on
> board. It works because the device is close enough to TI ADS1015.
>
> With that said, I do not think this is breaking any actual compatibility
> issue.
>
> - The old DTB will keep working with old and new kernel.
New DTB stops working with old kernel and this is what we talked about
during LPC.
All out-of-tree users of this DTS, like other operating systems, will be
affected as well probably.
> - The commit adding support for TI TLA2024 is in v5.19, and this board
> is supported only from kernel v6.5, so you cannot run an older kernel
> with this board. In addition to that I do not think that is
> reasonable to have someone using a "new" dtb with an "old" kernel.
> - The firmware, U-Boot, the only one that currently supports this board,
> is not making any use of this compatible nor of this ADC.
Best regards,
Krzysztof
