Re: [PATCH] arm64: dts: qcom: sm6115: Add EUD dt node and dwc3 connector

From: Konrad Dybcio
Date: Mon Jan 02 2023 - 12:04:14 EST




On 2.01.2023 17:54, Bhupesh Sharma wrote:
>
> On 1/2/23 4:16 PM, Konrad Dybcio wrote:
>>
>>
>> On 31.12.2022 14:19, Bhupesh Sharma wrote:
>>> Add the Embedded USB Debugger(EUD) device tree node for
>>> SM6115 / SM4250 SoC.
>>>
>>> The node contains EUD base register region and EUD mode
>>> manager register regions along with the interrupt entry.
>>>
>>> Also add the typec connector node for EUD which is attached to
>>> EUD node via port. EUD is also attached to DWC3 node via port.
>>>
>>> Cc: Souradeep Chowdhury <quic_schowdhu@xxxxxxxxxxx>
>>> Signed-off-by: Bhupesh Sharma <bhupesh.sharma@xxxxxxxxxx>
>>> ---
>>> - This patch is based on my earlier sm6115 usb related changes, which can
>>>    be seen here:
>>>    https://lore.kernel.org/linux-arm-msm/20221215094532.589291-1-bhupesh.sharma@xxxxxxxxxx/
>>> - This patch is also dependent on my sm6115 eud dt-binding and driver changes
>>>    sent earlier, which can be seen here:
>>>    https://lore.kernel.org/linux-arm-msm/20221231130743.3285664-1-bhupesh.sharma@xxxxxxxxxx/
>>>
>>>   arch/arm64/boot/dts/qcom/sm6115.dtsi | 37 ++++++++++++++++++++++++++++
>>>   1 file changed, 37 insertions(+)
>>>
>>> diff --git a/arch/arm64/boot/dts/qcom/sm6115.dtsi b/arch/arm64/boot/dts/qcom/sm6115.dtsi
>>> index 030763187cc3f..c775f7fdb7015 100644
>>> --- a/arch/arm64/boot/dts/qcom/sm6115.dtsi
>>> +++ b/arch/arm64/boot/dts/qcom/sm6115.dtsi
>>> @@ -565,6 +565,37 @@ gcc: clock-controller@1400000 {
>>>               #power-domain-cells = <1>;
>>>           };
>>>   +        eud: eud@1610000 {
>>> +            compatible = "qcom,sm6115-eud","qcom,eud";
>> Missing space between entries.
>>
>>> +            reg = <0x01610000 0x2000>,
>>> +                  <0x01612000 0x1000>,
>>> +                  <0x003e5018 0x4>;
>>> +            interrupts = <GIC_SPI 189 IRQ_TYPE_LEVEL_HIGH>;
>>> +            ports {
>> Newline before ports {}.
>>
>> Not sure if debugging hardware should be enabled by default..
>>> +                port@0 {
>>> +                    eud_ep: endpoint {
>>> +                        remote-endpoint = <&usb2_role_switch>;
>>> +                    };
>>> +                };
>> Newline between subsequent nodes.
>>
>>> +                port@1 {
>>> +                    eud_con: endpoint {
>>> +                        remote-endpoint = <&con_eud>;
>>> +                    };
>>> +                };
>>> +            };
>>> +        };
>>> +
>>> +        eud_typec: connector {
>> Non-MMIO nodes don't belong under /soc.
>>
>>> +            compatible = "usb-c-connector";
>> Newline between properties and subnode.
>>
>>
>>> +            ports {
>>> +                port@0 {
>>> +                    con_eud: endpoint {
>>> +                        remote-endpoint = <&eud_con>;
>>> +                    };
>>> +                };
>>> +            };
>>> +        };
>>> +
>>>           usb_hsphy: phy@1613000 {
>>>               compatible = "qcom,sm6115-qusb2-phy";
>>>               reg = <0x01613000 0x180>;
>>> @@ -1064,6 +1095,12 @@ usb_dwc3: usb@4e00000 {
>>>                   snps,has-lpm-erratum;
>>>                   snps,hird-threshold = /bits/ 8 <0x10>;
>>>                   snps,usb3_lpm_capable;
>>> +                usb-role-switch;
>> Same here.
>
> For all the above points, the format is same as suggested in [1] and already used in existing dts [2].
>
> [1]. https://www.kernel.org/doc/Documentation/devicetree/bindings/soc/qcom/qcom%2Ceud.yaml
> [2]. https://github.com/torvalds/linux/blob/master/arch/arm64/boot/dts/qcom/sc7280.dtsi#L3587
The fact that it's landed does not necessarily imply it's 100% correct..
That one seems to have slipped through review and could use some fixing up.

>
>> On a note, this commit + driver-side changes give me a:
>>
>> 1610000.eud     qcom_eud: failed to get role switch
>
> You need to set dr_mode = "otg", for 'usb_dwc3' to make the role switch work.
Thanks, couldn't find that anywhere. This however kicks me into EDL,
so that's one more reason to disable it by default.

Konrad>
> Thanks,
> Bhupesh
>
>>> +                port {
>>> +                    usb2_role_switch: endpoint {
>>> +                        remote-endpoint = <&eud_ep>;
>>> +                    };
>>> +                };
>>>               };
>>>           };
>>>