Claude review: ARM: dts: imx6ul-var-som: add support for EC configuration option (ENET1)

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

**Issue 1 (Critical, same as patch 9)**: The `pinctrl_enet1_gpio` in `imx6ull-var-som.dtsi` uses `MX6ULL_PAD_SNVS_TAMPER0__GPIO5_IO00` under `&iomuxc` — this must be `&iomuxc_snvs`.

**Issue 2 (Significant)**: There is a duplicate definition of `ethphy0: ethernet-phy@1` — it appears in both `imx6ul-var-som-enet2.dtsi` (added by this patch) and `imx6ul-var-som-enet1.dtsi`. When both are included in the `-full` boards, the same node is defined twice with the same label at the same path. While DTS merge semantics allow this (properties get overwritten with identical values), it is confusing and error-prone.

A cleaner approach would be to have `ethphy0` defined only in `enet1.dtsi` (since it's the ENET1 PHY), and use the `mdio_enet2` label from enet2.dtsi to place it. Don't duplicate it in enet2.dtsi.

**Issue 3 (Significant)**: The author acknowledges the design problem in the `---` notes:
> In order for this to work, imx6ul-var-som-enet2.dtsi must be included first, and thus enabled, even if not used. Maybe there is a better way to support both independently, but I'm not sure how.

This forced dependency on enet2.dtsi for enet1 to work is fragile. Consider defining the shared MDIO bus in the common dtsi or in a separate shared file, rather than making enet1 depend on enet2's internal implementation.

**Issue 4**: The `pinctrl_enet1_mdio` group from the original code was dropped. The original had:
```c
pinctrl-0 = <&pinctrl_enet1>, <&pinctrl_enet1_gpio>, <&pinctrl_enet1_mdio>;
```
The new version has:
```c
pinctrl-0 = <&pinctrl_enet1>, <&pinctrl_enet1_gpio>;
```
This seems intentional since fec1 now uses fec2's MDIO bus, but should be noted in the commit message.

---
Generated by Claude Code Patch Reviewer