Re: [PATCH 12/12] riscv: dts: ultrarisc: Add Milk-V Titan board device tree
Jia Wang <[email protected]>
| Newsgroups | org.infradead.lists.linux-riscv,org.kernel.vger.dmaengine,org.kernel.vger.linux-devicetree,org.kernel.vger.linux-i2c,org.kernel.vger.linux-kernel,org.kernel.vger.linux-spi,org.kernel.vger.netdev |
|---|---|
| Message-ID | <178779744933.1423015.12629307530730828329.b4-reply@b4> |
On 2026-08-24 15:06 +0200, Andrew Lunn wrote:
> > +ðernet {
> > + phy-handle = <&phy0>;
> > + /*
> > + * RTL8211F: board timing uses PHY strap delays; keep plain "rgmii".
> > + * Enabling PHY internal delays via "rgmii-id" breaks Ethernet traffic.
>
> This actually indicates something is wrong.
>
> When passed PHY_INTERFACE_MODE_RGMII, the PHY driver should change the
> PHY to not add delays, overriding the strapping.
>
> Given my previous comment, you need to spend some time understanding
> what is going on here. Something is not correct.
>
> https://elixir.bootlin.com/linux/v6.15/source/Documentation/devicetree/bindings/net/ethernet-controller.yaml#L287
>
Thanks for pointing this out. You are right that the RTL8211F driver
disables its internal delays for "rgmii", so the strap-delay explanation
was incorrect.
The same fixed DP1000 TX and RX RGMII clock skew described in my M0 reply
applies here. I will correct the comment in the next version.
> Andrew
>
Best regards,
Jia Wang
_______________________________________________
linux-riscv mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-riscv