[PATCH v3 1/5] riscv: dts: spacemit: k3: move USB3 phy to board level
Aurelien Jarno
aurelien at aurel32.net
Sun Sep 27 14:18:11 PDT 2026
Hi,
On 2026-08-31 04:54, sashiko-bot at kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Low] Removing 'phys' and 'phy-names' from the base k3.dtsi breaks DT schema validation (dtbs_check) for disabled nodes and improperly delegates fixed SoC routing to board files.
> --
>
> commit e9dc7ad92d5f4a805cf362fdd51c6ae160da1934
> Author: Aurelien Jarno <aurelien at aurel32.net>
>
> riscv: dts: spacemit: k3: move USB3 phy to board level
>
> This commit moves the USB3 phy definitions from the SoC dtsi to the board
> level dts files for the SpacemiT K3 SoC. It aims to prevent PHY mode
> configuration errors when the combo PHY lines are used for PCIe rather than
> USB3.
>
> [ ... ]
> > diff --git a/arch/riscv/boot/dts/spacemit/k3.dtsi b/arch/riscv/boot/dts/spacemit/k3.dtsi
> > index c3f2dce0969c0..6f8d55eae6f6b 100644
> > --- a/arch/riscv/boot/dts/spacemit/k3.dtsi
> > +++ b/arch/riscv/boot/dts/spacemit/k3.dtsi
> > @@ -452,9 +452,6 @@ usb3d: usb at 81a00000 {
> > <&syscon_apmu RESET_APMU_USB3_D_VCC>,
> > <&syscon_apmu RESET_APMU_USB3_D_PHY>;
> > reset-names = "ahb", "vcc", "phy";
> > - phys = <&usb3d_u2phy>,
> > - <&combophy 4 PHY_TYPE_USB3>;
> > - phy-names = "usb2-phy", "usb3-phy";
>
> [Severity: Low]
> Does removing phys and phy-names from the base k3.dtsi break DT schema
> validation (dtbs_check) for disabled nodes?
No it doesn't. dtbs_check seems to ignore disabled nodes.
> To satisfy the schema's minItems: 1 requirement while supporting board-level
> combo PHY configuration, should the dedicated internal SoC PHY (usb3d_u2phy)
> have been retained in the base file?
>
> The dtbs_check validation will fail for any board DT that includes k3.dtsi
> but does not override the usb3d node to provide the phys and phy-names
> properties (e.g., k3-com260-ifx.dts and k3-deepcomputing-fml13v05.dts).
I have just tried, in practice it doesn't. I'll send a new version
mentioning that.
> This happens because the schema spacemit,k1-dwc3.yaml unconditionally
> requires these properties, and dt-schema evaluates all nodes, even disabled
> ones.
>
> > phy_type = "utmi";
> > snps,dis_enblslpm_quirk;
> > snps,dis_u2_susphy_quirk;
>
> --
> Sashiko AI review · https://sashiko.dev/#/patchset/20260831044540.3304350-1-aurelien@aurel32.net?part=1
>
--
Aurelien Jarno GPG: 4096R/1DDD8C9B
aurelien at aurel32.net http://aurel32.net
More information about the linux-riscv
mailing list