Re: [PATCH v2 2/2] arm64: dts: qcom: x1e78100-t14s: Add audio playback over DisplayPort

[Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>]

On 12/17/25 2:00 PM, Krzysztof Kozlowski wrote:
> On 17/12/2025 13:51, Konrad Dybcio wrote:
>> On 12/17/25 1:38 PM, Krzysztof Kozlowski wrote:
>>> On 17/12/2025 13:33, Konrad Dybcio wrote:
>>>> On 12/17/25 1:00 PM, Krzysztof Kozlowski wrote:
>>>>> Add necessary DAI links and DAI name prefixes to enable audio playback
>>>>> over USB/DisplayPort and HDMI.  The HDMI port is not yet enabled, but it
>>>>> should carry respective DAI name prefix regardless.
>>>>>
>>>>> Reviewed-by: Abel Vesa <abel.vesa@linaro.org>
>>>>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
>>>>>
>>>>> ---
>>>>
>>>> (something's inserting a \n before --- in your latest patches but I
>>>
>>> No, it is not. It was always like that and only recently git started
>>> dropping it, but it is irrelevant.
>>>
>>>> don't know if this is a problem)
>>>>
>>>> [...]
>>>>
>>>>> +		displayport-2-dai-link {
>>>>> +			link-name = "DisplayPort2 Playback";
>>>>> +
>>>>> +			codec {
>>>>> +				sound-dai = <&mdss_dp2>;
>>>>
>>>> How does this work out with fw_devlink?
>>>
>>> How is this related to this patchset? Please stop nit-picking irrelevant
>>> things.
>>
>> I'm asking whether this is going to break sync_state because you're not
>> enabling mdss_dp2 - I believe that's a fair question..
> 
> DTS description is independent of driver, thus this is correct code
> regardless whether there is interconnect involved anywhere or not. I
> don't have the answer how this affects interconnect, but I see no reason
> anyone would create such ICC path - between hardware and fake SW construct.

sync_state isn't exclusive to the interconnect framework

In any case, could you please confirm/deny that sound still works as
expected with linux-next/master + this patch alone and we move on?

Konrad