Re: [PATCH v2 2/3] arm64: dts: nuvoton: ma35d1: add display controller support

From: Joey Lu

Date: Fri Feb 06 2026 - 04:25:20 EST


On 2/6/2026 4:56 PM, Krzysztof Kozlowski wrote:
On 06/02/2026 08:12, Joey Lu wrote:
On 2/5/2026 9:23 PM, Krzysztof Kozlowski wrote:
On Thu, Jan 29, 2026 at 12:05:31PM +0800, Joey Lu wrote:
&uart0 {
@@ -129,3 +165,23 @@ &uart16 {
pinctrl-0 = <&pinctrl_uart16>;
status = "okay";
};
+
+&panel {
+ port {
+ panel_in: endpoint@0 {
+ remote-endpoint = <&dpi_out>;
+ };
+ };
+};
+
+&display {
What sort of ordering rule is followed in Nuvoton? Why is it different
than DTS coding style? Why do you choose other style?

+ pinctrl-names = "default";
+ pinctrl-0 = <&pinctrl_display>;
+ status = "okay";
+
+ port {
+ dpi_out: endpoint@0 {
+ remote-endpoint = <&panel_in>;
+ };
+ };
+};
diff --git a/arch/arm64/boot/dts/nuvoton/ma35d1.dtsi b/arch/arm64/boot/dts/nuvoton/ma35d1.dtsi
index e51b98f5bdce..7d9d077f12b2 100644
--- a/arch/arm64/boot/dts/nuvoton/ma35d1.dtsi
+++ b/arch/arm64/boot/dts/nuvoton/ma35d1.dtsi
@@ -379,5 +379,19 @@ uart16: serial@40880000 {
clocks = <&clk UART16_GATE>;
status = "disabled";
};
+
+ panel: panel {
No, there is no way your SoC has a panel.

Don't add fake stuff to your DTS.

Best regards,
Krzysztof
I'll move panel nodes out of dtsi into board dts.
You did not respond to several comments in total, it's fine, but if you
just ignored them, then it would not be fine.


Best regards,
Krzysztof
Thanks for the review. I've gone through all comments carefully.

Some points were addressed directly in the next revision,so I grouped a few replies together since the original code will no longer appear after the update.

I will also restructure the DTS so the panel is defined under the root node to follow the standard hierarchy and ordering conventions.

For remaining items that need discussion, I'll reply inline one by one.

Joey