Hi Marek On 5/17/23 17:41, Marek Vasut wrote: > On 5/17/23 16:35, Raphael Gallais-Pou wrote: > > Hi, > >> diff --git a/arch/arm/boot/dts/stm32mp15xx-dkx.dtsi >> b/arch/arm/boot/dts/stm32mp15xx-dkx.dtsi >> index 0f1110e42c93..a6e2e20f12fa 100644 >> --- a/arch/arm/boot/dts/stm32mp15xx-dkx.dtsi >> +++ b/arch/arm/boot/dts/stm32mp15xx-dkx.dtsi >> @@ -457,8 +457,7 @@ <dc { >> status = "okay"; >> port { >> - ltdc_ep0_out: endpoint@0 { >> - reg = <0>; >> + ltdc_ep0_out: endpoint { >> remote-endpoint = <&sii9022_in>; >> }; >> }; > > This LTDC port/endpoint stuff always scares me, because I always feel I get it > wrong. > > I believe the LTDC does have one "port" , correct. > > But I think (?) that the LTDC has two endpoints, endpoint@0 for DPI (parallel > output out of the SoC) and endpoint@1 for DSI (internal connection into the > DSI serializer) ? You are correct indeed, I rushed the patch and did not thought about this. I agree that this can be confusing, as I also take some time to think through it. > > Only one of the endpoints can be connected at a time, but there are actually > two endpoints in the LTDC port {} node, aren't there ? Yes, they are mutually exclusive. > > So the original description should be OK I think , maybe #address/#size-cells > are missing instead ? Thing is: this file is only included in two device-trees : stm32mp157c-dk1.dts and stm32mp157c-dk2.dts. Among those two files there is only one which adds a second endpoint. Thus if the fields are set higher in the hierarchy, a warning yields. One way to do it would be to make the endpoint@0 go down in the device-tree with its dependencies, so that both endpoints are the same level without generating noise. Raphaël