On Tue, 2025-04-15 at 15:12 +0200, Andrew Lunn wrote:
>
> > My Perl-fu isn't good enough for me to review this properly... I think
> > though that Andrew mentioned something along the lines of 'Comment
> > should include PCB somewhere', but I don't know if this is easily
> > doable with checkpatch though.
>
> Maybe make it into a patchset, and change a few DTS files to match the
> condition. e.g.
>
> arm/boot/dts/nxp/ls/ls1021a-tsn.dts:/* RGMII delays added via PCB traces */
> arm/boot/dts/nxp/ls/ls1021a-tsn.dts-&enet2 {
> arm/boot/dts/nxp/ls/ls1021a-tsn.dts- phy-mode = "rgmii";
> arm/boot/dts/nxp/ls/ls1021a-tsn.dts- status = "okay";
>
> This example is not great, but...
> arm64/boot/dts/freescale/imx8mp-skov-reva.dtsi- phy-mode = "rgmii";
> arm64/boot/dts/freescale/imx8mp-skov-reva.dtsi: /* 2ns RX delay is implemented on PCB */
> arm64/boot/dts/freescale/imx8mp-skov-reva.dtsi- tx-internal-delay-ps = <2000>;
> arm64/boot/dts/freescale/imx8mp-skov-reva.dtsi- rx-internal-delay-ps = <0>;
>
> There is one more i know of somewhere which i cannot find at the
> moment which uses rgmii-rxid or rgmii-txid, an has a comment about
> delay.
>
> Andrew
FWIW, I've decided against including the comment changes in this series for now
(going to send a v2 in a moment), as non-functional changes just to improve
style or make checkpatch happy are often frowned upon as unnecessary churn.
Best,
Matthias
--
TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
Amtsgericht München, HRB 105018
Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
https://www.tq-group.com/
