While implementing support for this USB controller in U-Boot, I noticed that the reset line alsp affects they PHY. It looks like most platforms use a separate glue node to represent this, and in fact there is already a compatible for the H6 listed in drivers/usb/dwc3/dwc3-of-simple.c. Since this layout matches the usual way of modeling this hardware, it allows using the existing drivers without adding platform-specific code. I tried to follow the existing DWC3 glue bindings (most of which are still .txt). With this version, `make dt_binding_check` still raises a couple of issues, which I do not know how best to fix: - Warning (unit_address_vs_reg): /example-0/usb@5200000: node has a unit name, but no reg or ranges property => Since there is no MMIO translation, an empty `ranges;` seemed appropriate, but it causes this warning. - usb@5200000: usb@5200000:phy-names:0: 'usb2-phy' was expected => This may be an issue with the snps,dwc3 binding, where the `items` list overrides `minItems`. I believe the intention is that both PHY references are optional. This implementation has only one PHY. Changes from v1 to v2: - Updated the binding to reference the PHY binding by path correctly - Dropped DT updates for Pine H64 Samuel Holland (2): dt-bindings: usb: Document the Allwinner H6 DWC3 glue arm64: dts: allwinner: h6: Wrap DWC3 and PHY in glue layer .../usb/allwinner,sun50i-h6-dwc3.yaml | 75 +++++++++++++++++++ .../dts/allwinner/sun50i-h6-beelink-gs1.dts | 6 +- .../dts/allwinner/sun50i-h6-orangepi-3.dts | 6 +- .../dts/allwinner/sun50i-h6-tanix-tx6.dts | 6 +- arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi | 60 ++++++++------- 5 files changed, 111 insertions(+), 42 deletions(-) create mode 100644 Documentation/devicetree/bindings/usb/allwinner,sun50i-h6-dwc3.yaml -- 2.26.3