http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19377
Harald van Dijk <harald at gigawatt dot nl> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |harald at gigawatt dot nl
--- Comment #12 from Harald van Dijk <harald at gigawatt dot nl> ---
Fabien, are you sure that the test case in comment 9 is invalid? C++11 reads
"In a using-declaration used as a member-declaration, the nested-name-specifier
shall name a base class of the class being defined. If such a using-declaration
names a constructor, the nested-name-specifier shall name a direct base class
of the class being defined; otherwise it introduces the set of declarations
found by member name lookup (10.2, 3.4.3.1)."
which makes sense: any inherited member, even from an indirect base class, can
be introduced into the current class, except for an indirect base class's
constructor.
As for "It is invalid for a second reason, 'using Base::i' is declared
(implicitly) in a private section, so inaccessible in DerivedDerived.", the
DerivedDerived using declaration uses a fully qualified ns::Base::i, the
visibility of which should depend on the visibility in Base, not the visibility
in Derived. If the code used ns::Derived::i, then it would make sense to issue
an error, and clang does emit an error in that case.
As for being separate from this bug, a trivial change of that test case to
namespace ns {
class Base {
protected:
int i;
};
class Derived : public Base {
using Base::i;
};
}
class DerivedDerived : public ns::Derived {
using ns::Base::i;
};
makes GCC emit
test.cc:4:9: error: ‘int ns::Base::i’ is protected
int i;
^
test.cc:11:7: error: within this context
class DerivedDerived : public ns::Derived {
^
which is exactly the error message that this bug is about. Similarly, changing
the "protected:" to "public:" in the original test case changes the error
message from "protected" to "inaccessible".
