[Freedreno] [PATCH v4 13/13] arm64: dts: qcom: sa8295-adp: Enable DP instances

Konrad Dybcio konrad.dybcio at linaro.org
Mon Dec 5 20:09:51 UTC 2022



On 05/12/2022 21:02, Bjorn Andersson wrote:
> On Mon, Dec 05, 2022 at 07:09:45PM +0100, Konrad Dybcio wrote:
>> On 05/12/2022 18:44, Bjorn Andersson wrote:
>>> diff --git a/arch/arm64/boot/dts/qcom/sa8295p-adp.dts b/arch/arm64/boot/dts/qcom/sa8295p-adp.dts
> [..]
>>> +&mdss0_dp2 {
>>> +	status = "okay";
>> status should go last.
>>
> 
> Thanks for pointing that out. Would be nice if the computer told me
> that...somehow...
> 
>>> +
>>> +	data-lanes = <0 1 2 3>;
>>> +
>>> +	ports {
>>> +		port at 1 {
>>> +			reg = <1>;
>>> +			mdss0_dp2_phy_out: endpoint {
>> That's quite a lot of indentation.. couldn't these endpoints be defined in
>> the SoC DT?
>>
> 
> The alternative would be to have the description of each DP controller
> split over multiple nodes and rely on the reader to stitch together the
> view of the node based on the label.
> Based on the naming of these labels they would at least be adjacent, so
> it wouldn't be that bad.
> 
> But I feel that there is enough DP-controller nodes in this board
> already; I've yet to describe the two USB Type-C controllers or the two
> DSI-DP bridges.
> So I don't know if it's worth optimizing indentation-level within each
> node like this.
> 
> 
> And we will end up mixing this optimization between DP controllers, USB
> Type-C nodes, QMP nodes, DSI-DP bridges.
Oh okay, I see, let's keep it as-is then.

Konrad
> 
> Regards,
> Bjorn


More information about the Freedreno mailing list