[clr-ios] Enable SIMD and hardware intrinsics on arm64 Apple devices

[kotlarmilos]

Description

Hardware intrinsics are currently disabled on the CoreCLR interpreter. However, since Apple devices provide advanced SIMD capabilities (e.g., Vector128 and ARM64 NEON), these intrinsics should be enabled by default when running on supported hardware.

The interpreter should be used only as a fallback path when hardware support for specific intrinsics is unavailable.

Tasks

• Review current hardware intrinsic handling in the interpreter and JIT paths
• Enable hardware intrinsics (e.g., Vector128, SIMD operations) on supported Apple architectures
• Implement interpreter-based fallback logic for unsupported hardware or when intrinsics are unavailable
• Add validation tests covering both hardware and fallback execution paths
• Run performance and functional tests on iOS and tvOS devices to confirm correct behavior and expected speedups

[jkotas]

Related to #120047 . We should treat all hardware intrinsics as must-expand and generate R2R code for them so that the interpreter has an implementation to fallback to.

[dotnet-policy-service]

Tagging subscribers to 'os-ios': @vitek-karas, @kotlarmilos, @steveisok, @akoeplinger
See info in area-owners.md if you want to be subscribed.

[dotnet-policy-service]

Tagging subscribers to this area: @BrzVlad, @janvorli, @kg
See info in area-owners.md if you want to be subscribed.