static matrix<T, N, M> matrix<T, R, C>.dmul<U>( U a, matrix<T, N, M> b) where U : __BuiltinRealType where T : __BuiltinFloatingPointType;