This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug c++/79555] New: Warning 'base class should be explicitly initialized in the copy constructor' issued in wrong case


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79555

            Bug ID: 79555
           Summary: Warning 'base class should be explicitly initialized
                    in the copy constructor' issued in wrong case
           Product: gcc
           Version: 7.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: reichelt at gcc dot gnu.org
  Target Milestone: ---

The C++ front-end (with -Wextra enabled) complains about the following code
snippet:

================================
struct A
{
  A();
  A(const A&);
};

struct B : virtual A
{
  B(const B&) {}
//  B(const B&) : A() {}
  virtual int foo() = 0;
};
================================

bug.cc: In copy constructor 'B::B(const B&)':
bug.cc:9:3: warning: base class 'struct A' should be explicitly initialized in
the copy constructor [-Wextra]
   B(const B&) {}
   ^

However, since the base class A is virtual and the class B itself is abstract,
A isn't initialized in this function anyway (and so the warning is pointless).
Even more annoying is the fact that other major compilers complain if you try
to
please GCC and initialize A (as in the comment line above):

* clang with (clang++ -Wabstract-vbase-init -c):
bug.cc:10:17: warning: initializer for virtual base class 'A' of abstract
     class 'B' will never be used [-Wabstract-vbase-init]

* VisualStudio 2015 (cl -Wall)
bug.cc(10): warning C4589: Constructor of abstract class 'B' ignores
initializer for virtual base class 'A'
bug.cc(10): note: virtual base classes are only initialized by the most-derived
type

So it would be nice to suppress the warning in this specific case.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]