Hi,
On 16/10/2019 00:29, Rob Herring wrote:
On Mon, Oct 07, 2019 at 02:41:41PM +0300, Roger Quadros wrote:
TI platforms have a wrapper module around the Cadence USB3
controller. Add binding information for that.
Signed-off-by: Roger Quadros <rogerq@xxxxxx>
Signed-off-by: Sekhar Nori <nsekhar@xxxxxx>
---
.../devicetree/bindings/usb/cdns-usb3-ti.txt | 59 +++++++++++++++++++
1 file changed, 59 insertions(+)
create mode 100644 Documentation/devicetree/bindings/usb/cdns-usb3-ti.txt
Please convert to DT schema.
diff --git a/Documentation/devicetree/bindings/usb/cdns-usb3-ti.txt b/Documentation/devicetree/bindings/usb/cdns-usb3-ti.txt
new file mode 100644
index 000000000000..12c7c903e6da
--- /dev/null
+++ b/Documentation/devicetree/bindings/usb/cdns-usb3-ti.txt
@@ -0,0 +1,59 @@
+Binding for the TI specific wrapper for the Cadence USBSS-DRD controller
+
+Required properties:
+ - compatible: Should contain "ti,j721e-usb"
+ - reg: Physical base address and size of the wrappers register area.
+ - power-domains: Should contain a phandle to a PM domain provider node
+ and an args specifier containing the USB device id
+ value. This property is as per the binding documentation:
+ Documentation/devicetree/bindings/soc/ti/sci-pm-domain.txt
+ - clocks: Clock phandles to usb2_refclk and lpm_clk
+ - clock-names: Should contain "usb2_refclk" and "lpm_clk"
_clk is redundant. 'ref' amd 'lpm' is sufficient.
+
+Optional properties:
+ - ti,usb2-only: If present, it restricts the controller to USB2.0 mode of
+ operation. Must be present if USB3 PHY is not available
+ for USB.
Seems like this should be discoverable based on describing the phy
connections.
I don't think so. the PHY connections are in the child node.
+ - ti,modestrap-host: Set controller modestrap to HOST mode.
+ - ti,modestrap-peripheral: Set controller modestrap to PERIPHERAL mode.
What does modestrap mean? Fixed to the mode or that's the default? For
default, John Stultz sent a similar binding. Seems we need something
common.
It means if the controller needs to be hard-wired (using a register bit)
to be either host or peripheral.
I think that this will not really be used so I'll get rid of it. It was mostly
used for initial debug.
The controller is dual-role and the role can be set to either host or peripheral
using the operational registers.
+ - ti,vbus-divider: Should be present if USB VBUS line is connected to the
+ VBUS pin of the SoC via a 1/3 voltage divider.
+
+Sub-nodes:
+The USB2 PHY and the Cadence USB3 controller should be the sub-nodes.
+
+Example:
+
+ ti_usb0: cdns_usb@4104000 {
+ compatible = "ti,j721e-usb";
+ reg = <0x00 0x4104000 0x00 0x100>;
+ power-domains = <&k3_pds 288 TI_SCI_PD_EXCLUSIVE>;
+ clocks = <&k3_clks 288 15>, <&k3_clks 288 3>;
+ clock-names = "usb2_refclk", "lpm_clk";
+ assigned-clocks = <&k3_clks 288 15>; /* USB2_REFCLK */
+ assigned-clock-parents = <&k3_clks 288 16>; /* HFOSC0 */
+ #address-cells = <2>;
+ #size-cells = <2>;
+ ranges;
+
+ phy@4108000 {
+ compatible = "ti,j721e-usb2-phy";
+ reg = <0x00 0x4108000 0x00 0x400>;
+ };
Why is this a child node? Use the phy binding.
This is a USB2 PHY that physically sits within the ti_usb0 wrapper module that
controls power and clock to it.
There isn't a device driver required for it so should I just get rid of it?
Or can it just live there? what can the compatible be?
Or we can add it later if at all a driver is needed.
+
+ usb0: usb@6000000 {
+ compatible = "cdns,usb3-1.0.1";
Not documented.
should be
compatible = "cdns,usb3";
+ reg = <0x00 0x6000000 0x00 0x10000>,
+ <0x00 0x6010000 0x00 0x10000>,
+ <0x00 0x6020000 0x00 0x10000>;
+ reg-names = "otg", "xhci", "dev";
+ interrupts = <GIC_SPI 96 IRQ_TYPE_LEVEL_HIGH>, /* irq.0 */
+ <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>, /* irq.6 */
+ <GIC_SPI 120 IRQ_TYPE_LEVEL_HIGH>; /* otgirq.0 */
+ interrupt-names = "host",
+ "peripheral",
+ "otg";
+ maximum-speed = "super-speed";
+ dr_mode = "otg";
+ };
+ };
--
cheers,
-roger
Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki.
Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki
