Re: [PATCH 4/4] dt-bindings: soc: fsl: qe: Add an interrupt controller for QUICC Engine Ports

[Christophe Leroy <christophe.leroy@csgroup.eu>]

Le 12/08/2025 à 19:16, Rob Herring a écrit :
> On Tue, Aug 12, 2025 at 10:23 AM Krzysztof Kozlowski <krzk@kernel.org> wrote:
>>
>> On 12/08/2025 13:02, Christophe Leroy wrote:
>>> The QUICC Engine provides interrupts for a few I/O ports. This is
>>> handled via a separate interrupt ID and managed via a triplet of
>>> dedicated registers hosted by the SoC.
>>>
>>> Implement an interrupt driver for it for that those IRQs can then
>>> be linked to the related GPIOs.
>>>
>>> Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
>>> ---
>>>   .../soc/fsl/cpm_qe/fsl,qe-ports-ic.yaml       | 63 +++++++++++++++++++
>>>   1 file changed, 63 insertions(+)
>>>   create mode 100644 Documentation/devicetree/bindings/soc/fsl/cpm_qe/fsl,qe-ports-ic.yaml
>>>
>>> diff --git a/Documentation/devicetree/bindings/soc/fsl/cpm_qe/fsl,qe-ports-ic.yaml b/Documentation/devicetree/bindings/soc/fsl/cpm_qe/fsl,qe-ports-ic.yaml
>>> new file mode 100644
>>> index 0000000000000..7c98706d03dd1
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/soc/fsl/cpm_qe/fsl,qe-ports-ic.yaml
>>> @@ -0,0 +1,63 @@
>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>>> +
>>> +title: Freescale QUICC Engine I/O Ports Interrupt Controller
>>> +
>>> +maintainers:
>>> +  - name: Christophe Leroy
>>> +    email: christophe.leroy@csgroup.eu
>>
>> Oh no...
>>
>>> +
>>> +description: |
>>> +  Interrupt controller for the QUICC Engine I/O ports found on some
>>> +  Freescale/NXP PowerQUICC and QorIQ SoCs.
>>> +
>>> +properties:
>>> +  compatible:
>>> +    enum:
>>> +      - fsl,mpc8323-qe-ports-ic
>>> +      - fsl,mpc8360-qe-ports-ic
>>> +      - fsl,mpc8568-qe-ports-ic
>>> +
>>> +  reg:
>>> +    description: Base address and size of the QE I/O Ports Interrupt Controller registers.
>>> +    minItems: 1
>>> +    maxItems: 1
>>
>> This was never tested but more important this and everything further
>> looks like generated by AI. Please don't do that or at least mark it
>> clearly, so I will prioritize accordingly (hint: AI generates poor code
>> and burden to decipher AI slop should not be on open source reviewers
>> but on users of AI, but as one of maintainers probably you already know
>> that, so sorry for lecturing).
> 
> If anyone needs some AI (chatgpt) converted bindings, my "dt-convert"
> branch has ~800 of them. Feeding the warnings back to AI to fix was
> somewhat effective. The result is not the worst I've seen submitted.
> It saves some of the boilerplate, but can't fix things that are just
> wrong or unclear in .txt bindings. Despite my 'prompt engineering'
> attempts, it still tends to get the same things wrong over and over.

By the way, the new binding was not generated from text binding. I fed 
the AI with the driver C source file.

Christophe