Re: [PATCH v2 5/6] dt-bindings: memory: tegra20: emc: Convert to schema

[Dmitry Osipenko]

02.04.2021 17:45, Dmitry Osipenko пишет:
> 01.04.2021 18:55, Rob Herring пишет:
>> On Wed, Mar 31, 2021 at 05:59:39PM +0300, Dmitry Osipenko wrote:
>>> 31.03.2021 16:40, Rob Herring пишет:
>>>> On Wed, 31 Mar 2021 02:04:44 +0300, Dmitry Osipenko wrote:
>>>>> Convert Tegra20 External Memory Controller binding to schema.
>>>>>
>>>>> Signed-off-by: Dmitry Osipenko <digetx@xxxxxxxxx>
>>>>> ---
>>>>>  .../memory-controllers/nvidia,tegra20-emc.txt | 130 --------
>>>>>  .../nvidia,tegra20-emc.yaml                   | 294 ++++++++++++++++++
>>>>>  2 files changed, 294 insertions(+), 130 deletions(-)
>>>>>  delete mode 100644 Documentation/devicetree/bindings/memory-controllers/nvidia,tegra20-emc.txt
>>>>>  create mode 100644 Documentation/devicetree/bindings/memory-controllers/nvidia,tegra20-emc.yaml
>>>>>
>>>>
>>>> My bot found errors running 'make dt_binding_check' on your patch:
>>>>
>>>> yamllint warnings/errors:
>>>>
>>>> dtschema/dtc warnings/errors:
>>>> Documentation/devicetree/bindings/memory-controllers/nvidia,tegra20-emc.example.dts:33.26-55.15: Warning (unit_address_vs_reg): /example-0/external-memory-controller@7000f400/emc-tables@0: node has a unit name, but no reg or ranges property
>>>>
>>>> See https://patchwork.ozlabs.org/patch/1460288
>>>>
>>>> This check can fail if there are any dependencies. The base for a patch
>>>> series is generally the most recent rc1.
>>>>
>>>> If you already ran 'make dt_binding_check' and didn't see the above
>>>> error(s), then make sure 'yamllint' is installed and dt-schema is up to
>>>> date:
>>>>
>>>> pip3 install dtschema --upgrade
>>>>
>>>> Please check and re-submit.
>>>>
>>>
>>> FYI, I'm aware about this warning and I think it's fine to skip it in
>>> the case of this binding. 
>>
>> It's not because dt_binding_check should be warning free.
>>
>>> The dtbs_check doesn't bother with that warning.
>>
>> With W=1 it will. It's off by default because there are too many of 
>> these warnings. Patches welcome.
> 
> Such warning could be silenced with some kind of a new pragma option in
> schema which will tell what warnings are inappropriate.
> 
> But since there is no such option today, perhaps indeed should be better
> to just add a dummy reg property and fix the device-trees.
> 

This actually was my bad, the reg property is already specified in this
version of binding as required. Apparently I got confused while was
looking at some device-tree that doesn't have the reg property and it
should be corrected.