Performing arithmetic operations in quadruple precision

So you should pass the symbol T through to the function, and build up an expression that converts things to the type given by the symbol T.