On 9/22/2025 1:14 PM, Rob Herring wrote:

On Fri, Sep 19, 2025 at 08:21:02PM -0700, Wesley Cheng wrote:

quoted

Add the Glymur compatible to the M31 eUSB2 PHY, and use the SM8750 as
the fallback.

Signed-off-by: Wesley Cheng <redacted>
---
  .../devicetree/bindings/phy/qcom,m31-eusb2-phy.yaml   | 11 ++++++-----
  1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/Documentation/devicetree/bindings/phy/qcom,m31-eusb2-phy.yaml b/Documentation/devicetree/bindings/phy/qcom,m31-eusb2-phy.yaml
index c84c62d0e8cb..b96b1ee80257 100644
--- a/Documentation/devicetree/bindings/phy/qcom,m31-eusb2-phy.yaml
+++ b/Documentation/devicetree/bindings/phy/qcom,m31-eusb2-phy.yaml

@@ -15,9 +15,12 @@ description:
  
  properties:
    compatible:
-    items:
-      - enum:
-          - qcom,sm8750-m31-eusb2-phy
+    oneOf:
+      - items:
+          - enum:
+              - qcom,glymur-m31-eusb2-phy
+          - const: qcom,sm8750-m31-eusb2-phy
+      - const: qcom,sm8750-m31-eusb2-phy
  
    reg:
      maxItems: 1

@@ -53,8 +56,6 @@ required:
    - compatible
    - reg
    - "#phy-cells"
-  - clocks
-  - clock-names

How is it compatible if clocks aren't required now? And clocks are
suddenly no longer required on sm8750?

Hi Rob,

It depends on the clock subsystem.  On SM8750, we still need the clock 
entry, because we need to control the output of our CXO/reference clock 
to our HS PHY.  However, on chipsets like Glymur, some HS PHYs in our 
USB subsystem doesn't have this refclk output control.

Thanks
Wesley Cheng

quoted

    - resets
    - vdd-supply
    - vdda12-supply

-- 
linux-phy mailing list
linux-phy@lists.infradead.org
https://lists.infradead.org/mailman/listinfo/linux-phy