[clang] [HLSL] Implement the `lit` intrinsic (PR #134171)

[Farzon Lotfi via cfe-commits]

================
@@ -253,6 +253,37 @@ const inline float length(__detail::HLSL_FIXED_VECTOR<float, N> X) {
   return __detail::length_vec_impl(X);
 }
 
+//===----------------------------------------------------------------------===//
+// lit builtins
+//===----------------------------------------------------------------------===//
+
+/// \fn vector<T, 4> lit(T NDotL, T NDotH, T M)
+/// \brief Returns a lighting coefficient vector.
+/// \param NDotL The dot product of the normalized surface normal and the
+/// light vector.
+/// \param NDotH The dot product of the half-angle vector and the surface
+/// normal.
+/// \param M A specular exponent.
+///
+/// This function returns a lighting coefficient vector (ambient, diffuse,
+/// specular, 1).
+
+template <typename T>
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+const inline __detail::enable_if_t<__detail::is_arithmetic<T>::Value &&
----------------
farzonl wrote:

> @farzonl The template is needed for the `__detail::is_arithmetic<T>` check since `lit` is only supposed to accept scalars as inputs. Without the check, it would allow types like float4 as input as it would be implicitly truncated to a float.

I understand that logic makes sense. That said i think the decision should fall to what DXC does. If DXC doesn’t implicit cast then i think we should aswell.

https://github.com/llvm/llvm-project/pull/134171