On 10/22/26 12:52 PM, Krzysztof Kozlowski wrote:
On 22/10/2026 09:08, Mehmet Fide wrote:
- phy_addr = ofnode_get_addr(phy_node);
- for (i = 0; i < ARRAY_SIZE(phy_bases); i++) {
- if (phy_addr == phy_bases[i])
- break;
- }
- if (i == ARRAY_SIZE(phy_bases))
+ ret = ofnode_get_alias_seq(phy.node, "usbphy", &seq);
Just for the record, this looks like undocumented ABI which would not be
accepted in Linux kernel, so you won't be able to use OF_UPSTREAM.
This does indeed interact with the enumeration of USB controllers in
U-Boot, which itself is a legacy crutch to prevent the command line ABI
breakage. It seems to resolve PHY node using the /aliases { usbphy =
&... } which is already in Linux ./arch/arm/boot/dts/nxp/vf/vfxxx.dtsi :
27 usbphy0 = &usbphy0;
28 usbphy1 = &usbphy1;
Mehmet, would it be possible to parse fsl,usbphy = <&usbphy0>; DT
property to arrive at the PHY node instead ?
441 usbdev0: usb@40034000 {
442 compatible = "fsl,vf610-usb",
"fsl,imx27-usb";
443 reg = <0x40034000 0x800>;
444 interrupts = <75 IRQ_TYPE_LEVEL_HIGH>;
445 clocks = <&clks VF610_CLK_USBC0>;
446 fsl,usbphy = <&usbphy0>;
<----------- HERE
447 fsl,usbmisc = <&usbmisc0 0>;
448 dr_mode = "peripheral";
449 status = "disabled";
450 };
I think you might still need the portnr sequence number, or could you
get rid of that too ? (I would be fine with a follow up series too) ?
