On 2020-08-17 17:24, Stephen Boyd wrote:
Quoting Tanmay Shah (2020-08-17 15:53:00)
diff --git a/arch/arm64/boot/dts/qcom/sc7180.dtsi
b/arch/arm64/boot/dts/qcom/sc7180.dtsi
index 31b9217bb5bf..bf2f2bb1aa79 100644
--- a/arch/arm64/boot/dts/qcom/sc7180.dtsi
+++ b/arch/arm64/boot/dts/qcom/sc7180.dtsi
@@ -2440,6 +2447,71 @@ dsi_phy: dsi-phy@ae94400 {
status = "disabled";
};
+
+ msm_dp: displayport-controller@ae90000 {
This should come before dsi-phy and dsi node. It should be sorted by
the
address (0xae90000).
+ status = "disabled";
+ compatible = "qcom,sc7180-dp";
+
+ reg = <0 0x0ae90000 0 0x1400>;
+
+ interrupt-parent = <&mdss>;
+ interrupts = <12>;
+
[...]
};
dispcc: clock-controller@af00000 {
@@ -2449,8 +2521,8 @@ dispcc: clock-controller@af00000 {
<&gcc GCC_DISP_GPLL0_CLK_SRC>,
<&dsi_phy 0>,
<&dsi_phy 1>,
- <0>,
- <0>;
+ <&msm_dp 0>,
+ <&msm_dp 1>;
Don't think we should apply this still because the binding will change
when the phy is split out to qmp node. Maybe just leave this part off
for now?
Ok fine. But, that will break DP driver functionality.