Re: [PATCH v4 2/4] dt-bindings: PCI: qcom: Document the Glymur PCIe Controller
From: Abel Vesa <hidden>
Date: 2025-10-11 12:16:03
Also in:
linux-arm-msm, linux-pci, linux-phy, lkml
On 25-09-03 23:22:03, Wenbin Yao wrote:
quoted hunk ↗ jump to hunk
From: Prudhvi Yarlagadda <redacted> On the Qualcomm Glymur platform the PCIe host is compatible with the DWC controller present on the X1E80100 platform. So document the PCIe controllers found on Glymur and use the X1E80100 compatible string as a fallback in the schema. Signed-off-by: Prudhvi Yarlagadda <redacted> Signed-off-by: Wenbin Yao <redacted> Acked-by: Rob Herring (Arm) <robh@kernel.org> --- Documentation/devicetree/bindings/pci/qcom,pcie-x1e80100.yaml | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie-x1e80100.yaml b/Documentation/devicetree/bindings/pci/qcom,pcie-x1e80100.yaml index 257068a1826492a7071600d03ca0c99babb75bd9..8600f2c74cb81bcb924fa2035d992c3bd147db31 100644 --- a/Documentation/devicetree/bindings/pci/qcom,pcie-x1e80100.yaml +++ b/Documentation/devicetree/bindings/pci/qcom,pcie-x1e80100.yaml@@ -16,7 +16,12 @@ description: properties: compatible: - const: qcom,pcie-x1e80100 + oneOf: + - const: qcom,pcie-x1e80100 + - items: + - enum: + - qcom,glymur-pcie + - const: qcom,pcie-x1e80100
The cnoc_sf_axi clock is not found on Glymur, at least according to this: https://lore.kernel.org/all/20250925-v3_glymur_introduction-v1-19-24b601bbecc0@oss.qualcomm.com/ (local) And dtbs_check reports the following: arch/arm64/boot/dts/qcom/glymur-crd.dtb: pci@1b40000 (qcom,glymur-pcie): clock-names: ['aux', 'cfg', 'bus_master', 'bus_slave', 'slave_q2a', 'noc_aggr'] is too short from schema $id: http://devicetree.org/schemas/pci/qcom,pcie-x1e80100.yaml# One more thing: arch/arm64/boot/dts/qcom/glymur-crd.dtb: pci@1b40000 (qcom,glymur-pcie): max-link-speed: 5 is not one of [1, 2, 3, 4] from schema $id: http://devicetree.org/schemas/pci/qcom,pcie-x1e80100.yaml# max-link-speed = <5> isn't yet supported and of_pci_get_max_link_speed returns -EINVAL and sets pci->max_link_speed to that without checking the error. So I guess fun stuff is happening based on that later on ...