[PATCH V2 2/8] dt-bindings: usb: Add Tegra264 XUSB host support
Wayne Chang
waynec at nvidia.com
Thu Sep 3 00:11:05 PDT 2026
Hi Krzysztof,
Thanks for the review. I'll split the changes into three parts and
address all the comments.
- phy: tegra: xusb: dt-bindings/phy + drivers/phy/tegra
- usb: host: xhci-tegra + its dt-bindings/usb
- arm64: DTS enablement
thanks,
Wayne.
On 9/1/26 18:10, Krzysztof Kozlowski wrote:
> External email: Use caution opening links or attachments
>
>
> On Tue, Sep 01, 2026 at 02:58:29AM +0800, Wayne Chang wrote:
>> Extend the Tegra234 XUSB host binding to cover Tegra264:
>>
>> - Add nvidia,tegra264-xusb compatible string
>> - Document optional USB wake interrupts for Tegra264
>> - Document Tegra264 PMC wake event to port mapping
>> - Allow up to five IOMMU specifiers for the additional XUSB host
>> stream IDs
>>
>> Signed-off-by: Wayne Chang <waynec at nvidia.com>
> Why are USB patches mixed with DTS and phy ones? That's like three
> separate subsystems, of which Greg explicitly said many, many times he
> takes everything or nothing. Please read submitting patches in DT.
>
>
>
>> ---
>> V1 -> V2: Property definitions are back at the top level with the widest
>> constraints, The allOf conditional only tightens Tegra234 to interrupts
>> maxItems 9 and iommus maxItems 1.
>>
>> .../bindings/usb/nvidia,tegra234-xusb.yaml | 35 ++++++++++++++++---
>> 1 file changed, 30 insertions(+), 5 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/usb/nvidia,tegra234-xusb.yaml b/Documentation/devicetree/bindings/usb/nvidia,tegra234-xusb.yaml
>> index ec0993497fbb..d63f0db4224c 100644
>> --- a/Documentation/devicetree/bindings/usb/nvidia,tegra234-xusb.yaml
>> +++ b/Documentation/devicetree/bindings/usb/nvidia,tegra234-xusb.yaml
>> @@ -17,7 +17,9 @@ description: |
>>
>> properties:
>> compatible:
>> - const: nvidia,tegra234-xusb
>> + enum:
>> + - nvidia,tegra234-xusb
>> + - nvidia,tegra264-xusb
>>
>> reg:
>> items:
>> @@ -43,6 +45,7 @@ properties:
>> - description: USB wake event 4
>> - description: USB wake event 5
>> - description: USB wake event 6
>> + - description: USB wake event 7
>> description: |
>> The first two interrupts are required for the USB host controller. The
>> remaining USB wake event interrupts are optional. Each USB wake event is
>> @@ -50,10 +53,9 @@ properties:
>> platform. The USB host controller can function even if no wake-up events
>> are defined. The USB wake event interrupts are handled by the Tegra PMC;
>> hence, the interrupt controller for these is the PMC and the interrupt
>> - IDs correspond to the PMC wake event IDs. A complete list of wake event
>> - IDs is provided below, and this information is also present in the Tegra
>> - TRM document.
>> + IDs correspond to the PMC wake event IDs.
>>
>> + Tegra234 PMC wake events:
>> PMC wake-up 76 for USB3 port 0 wakeup
>> PMC wake-up 77 for USB3 port 1 wakeup
>> PMC wake-up 78 for USB3 port 2 and port 3 wakeup
>> @@ -62,6 +64,16 @@ properties:
>> PMC wake-up 81 for USB2 port 2 wakeup
>> PMC wake-up 82 for USB2 port 3 wakeup
>>
>> + Tegra264 PMC wake events:
>> + PMC wake-up 79 for USB3 port 0 wakeup
>> + PMC wake-up 80 for USB3 port 1 wakeup
>> + PMC wake-up 81 for USB3 port 2 wakeup
>> + PMC wake-up 82 for USB3 port 3 wakeup
>> + PMC wake-up 83 for USB2 port 0 wakeup
>> + PMC wake-up 84 for USB2 port 1 wakeup
>> + PMC wake-up 85 for USB2 port 2 wakeup
>> + PMC wake-up 86 for USB2 port 3 wakeup
>> +
>> clocks:
>> items:
>> - description: XUSB host clock
>> @@ -97,7 +109,8 @@ properties:
>> - const: write
>>
>> iommus:
>> - maxItems: 1
>> + minItems: 1
>> + maxItems: 5
> You should list the items instead. Flexible array means now the order
> have a meaning.
>
>> nvidia,xusb-padctl:
>> $ref: /schemas/types.yaml#/definitions/phandle
>> @@ -137,6 +150,18 @@ properties:
>> allOf:
>> - $ref: usb-xhci.yaml
>>
>> + - if:
>> + properties:
>> + compatible:
>> + contains:
>> + const: nvidia,tegra234-xusb
>> + then:
>> + properties:
>> + interrupts:
>> + maxItems: 9
>> + iommus:
>> + maxItems: 1
> And the new device has flexible interrpts and iommus?
>
> Best regards,
> Krzysztof
>
>
More information about the linux-phy
mailing list