Reduce some boilerplate in TensorPrimitive's IBinaryOperator (dotnet#92576)

Change a few of the static abstract interface methods to be virtual, as most implementations throw from these methods; we can consolidate that throwing to the base.

Author: stephentoub

src/libraries/System.Numerics.Tensors/src/System/Numerics/Tensors/TensorPrimitives.netcore.cs

@@ -1099,12 +1099,6 @@ private static Vector512<float> FusedMultiplyAdd(Vector512<float> x, Vector512<f
 #if NET8_0_OR_GREATER
             public static Vector512<float> Invoke(Vector512<float> x, Vector512<float> y) => x - y;
 #endif
-
-            public static float Invoke(Vector128<float> x) => throw new NotSupportedException();
-            public static float Invoke(Vector256<float> x) => throw new NotSupportedException();
-#if NET8_0_OR_GREATER
-            public static float Invoke(Vector512<float> x) => throw new NotSupportedException();
-#endif
         }
 
         private readonly struct SubtractSquaredOperator : IBinaryOperator
@@ -1134,12 +1128,6 @@ public static Vector512<float> Invoke(Vector512<float> x, Vector512<float> y)
                 return tmp * tmp;
             }
 #endif
-
-            public static float Invoke(Vector128<float> x) => throw new NotSupportedException();
-            public static float Invoke(Vector256<float> x) => throw new NotSupportedException();
-#if NET8_0_OR_GREATER
-            public static float Invoke(Vector512<float> x) => throw new NotSupportedException();
-#endif
         }
 
         private readonly struct MultiplyOperator : IBinaryOperator
@@ -1195,12 +1183,6 @@ public static float Invoke(Vector512<float> x)
 #if NET8_0_OR_GREATER
             public static Vector512<float> Invoke(Vector512<float> x, Vector512<float> y) => x / y;
 #endif
-
-            public static float Invoke(Vector128<float> x) => throw new NotSupportedException();
-            public static float Invoke(Vector256<float> x) => throw new NotSupportedException();
-#if NET8_0_OR_GREATER
-            public static float Invoke(Vector512<float> x) => throw new NotSupportedException();
-#endif
         }
 
         private readonly struct NegateOperator : IUnaryOperator
@@ -1276,14 +1258,18 @@ private interface IUnaryOperator
         private interface IBinaryOperator
         {
             static abstract float Invoke(float x, float y);
-
             static abstract Vector128<float> Invoke(Vector128<float> x, Vector128<float> y);
-            static abstract float Invoke(Vector128<float> x);
             static abstract Vector256<float> Invoke(Vector256<float> x, Vector256<float> y);
-            static abstract float Invoke(Vector256<float> x);
 #if NET8_0_OR_GREATER
             static abstract Vector512<float> Invoke(Vector512<float> x, Vector512<float> y);
-            static abstract float Invoke(Vector512<float> x);
+#endif
+
+            // Operations for aggregating all lanes in a vector into a single value.
+            // These are not supported on most implementations.
+            static virtual float Invoke(Vector128<float> x) => throw new NotSupportedException();
+            static virtual float Invoke(Vector256<float> x) => throw new NotSupportedException();
+#if NET8_0_OR_GREATER
+            static virtual float Invoke(Vector512<float> x) => throw new NotSupportedException();
 #endif
         }