Re: [PATCH 1/3] dt-bindings: interconnect: qcom,x1e80100-rpmh: add clocks property to enable QoS
From: Raviteja Laggyshetty
Date: Sun May 24 2026 - 14:44:58 EST
On 5/21/2026 5:26 AM, Dmitry Baryshkov wrote:
> On Wed, May 20, 2026 at 09:33:52PM +0200, Krzysztof Kozlowski wrote:
>> On 20/05/2026 21:27, Dmitry Baryshkov wrote:
>>> On Wed, May 20, 2026 at 10:11:47PM +0300, Georgi Djakov wrote:
>>>> On 5/20/26 9:51 PM, Krzysztof Kozlowski wrote:
>>>>> On 23/04/2026 10:38, Krzysztof Kozlowski wrote:
>>>>>> On Wed, Apr 22, 2026 at 02:05:11AM +0000, Raviteja Laggyshetty wrote:
>>>>>>> Some interconnect nodes on X1E80100 have QoS registers located inside
>>>>>>> a block whose interface is clock-gated. For those nodes, driver
>>>>>>> must enable the corresponding clock(s) before accessing the
>>>>>>> registers. Add the 'clocks' property so the driver can obtain
>>>>>>> and enable the required clock(s).
>>>>>>>
>>>>>>> Only interconnects that have clock-gated QoS register interface
>>>>>>> use this property; it is not applicable to all interconnect nodes.
>>>>>>>
>>>>>>> Signed-off-by: Raviteja Laggyshetty <raviteja.laggyshetty@xxxxxxxxxxxxxxxx>
>>>>>>> ---
>>>>>>> .../bindings/interconnect/qcom,x1e80100-rpmh.yaml | 62 ++++++++++++++++++++++
>>>>>>> 1 file changed, 62 insertions(+)
>>>>>>
>>>>>> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxxxxxxxx>
>>>>>
>>>>> And unreviewed as it breaks users:
>>>>> https://krzk.eu/#/builders/102/builds/70/steps/23/logs/warnings__3_
>>>>>
>>>>
>>>> I will just drop these patches for now, i have put them in a separate branch anyway.
>>>
>>> I think, dropping 'required' clause would be the easiest fix. Or just
>>> wait for Bjorn to pick up the DT changes.
>>
>> Could be, initially I thought this is actual impact on users, but indeed
>> now I recall that driver prints "info" message and continues. So the
>> binding is not correct.
>
> I'd say, the binding was not correct: the hardware has the clocks and
> the requires them to function completely correctly. I think, the problem
> is that we allowed incomplete drivers and incomplete bindings (Without
> QoS support).
>
Thanks for pointing this out, and sorry for the breakage caused.
This is one of the earlier targets where interconnect support was
upstreamed without QoS support. Making the `clocks` property required broke
existing DTs for that target, if the corresponding DT changes are not
picked.
For newer targets, this should not be an issue as QoS support is introduced
along with interconnect.
I will respin the patch following commit
e07f3b8c9e1c ("dt-bindings: interconnect: qcom,qcs615-rpmh: add clocks
property to enable QoS"),
keeping the QoS-related clocks optional to maintain backward compatibility.
Thanks,
Raviteja.