Claude review: arm64: dts: qcom: hamoa: Add inline crypto for UFS

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

Patch Review

Adds the ICE node and wires it to the UFS host controller in the shared hamoa.dtsi:

```dts
+		ice: crypto@1d88000 {
+			compatible = "qcom,x1e80100-inline-crypto-engine",
+				     "qcom,inline-crypto-engine";
+			reg = <0x0 0x01d88000 0x0 0x8000>;
+
+			clocks = <&gcc GCC_UFS_PHY_ICE_CORE_CLK>;
+		};
```

The `qcom,ice` property is added to the UFS node:

```dts
+			qcom,ice = <&ice>;
```

The ICE node address `0x1d88000` is correctly placed before `cryptobam` at `0x1dc4000` in ascending address order. The compatible uses the binding from patch 2/7. This affects all hamoa-based boards (x1e80100 family), which is intentional since the ICE hardware is part of the SoC.

**Potential concern:** This change affects the shared `hamoa.dtsi` which is used by all x1e80100/x1p42100 boards. Adding `qcom,ice` to the UFS node and a new ICE device node will affect existing boards (Surface Pro 11 / Denali, Lenovo ThinkPad, ASUS VivoBook, etc.). This should be fine since the ICE hardware exists on all these SoCs — but it would be good to confirm this doesn't break any existing boards. The ICE node has no `status` property so it defaults to "okay", and the `qcom,ice` property in the UFS node is optional in the driver. This should be safe.

---

---
Generated by Claude Code Patch Reviewer