Shouldn't the compiler assume that function `f` always returns
something for all valid inputs `t`?
I believe no, because for some crazy reason I can call `f` in the
following way[1]:
f(static_cast<T>(123));
Therefore, argument `t` is neither `T::A`, nor `T::B`.
Moreover:
> 9.3.4 List-initialization [dcl.init.list]
> 3 List-initialization of an object or reference of type T is defined
> as follows:
> (3.8) Otherwise, if T is an enumeration with a fixed underlying type
> (9.6), the initializer-list has a single element v , and the
> initialization is direct-list-initialization, the object is
> initialized with the value T(v) (7.6.1.3); if a narrowing conversion
> is required to convert v to the underlying type of T , the program is
> ill-formed.
Thus (I believe with `-std=c++17`) we can "end up" with:
T bar{123};
int foo = f(bar);
Once again, the `f` function gets neither `T::A`, nor `T::B`.
[1] "7.6.1.9 Static cast [expr.static.cast]. 10 A value of integral or
enumeration type can be explicitly converted to a complete enumeration
type."
On 9/6/2018 10:58 PM, Juan Cabrera wrote:
Hello,
I'm getting a "control reaches end of non-void function" warning
message on the following code (tested with gcc version 7.3 and above):
enum class T { A, B };
int f(T t) {
switch (t) {
case T::A: return 10;
case T::B: return 20;
}
}
https://godbolt.org/z/IXC4Ff
Shouldn't the compiler assume that function `f` always returns
something for all valid inputs `t`? (Given that the parameter `t` is
of type `T` which is an enum class and the siwtch statement covers all
the enum values).
Regards,
Juan.
