[Bug c++/34691] [4.2/4.3/4.4 Regression] Default argument checking not performed after overload resolution with C linkage

["jason at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org>]

------- Comment #7 from jason at gcc dot gnu dot org  2009-01-16 20:03 -------
3.2.3 also fails the first testcase.  This seems to have been broken by

2001-03-01  Nathan Sidwell  <nathan@codesourcery.com>

        * decl2.c (do_nonmember_using_decl): Don't complain if we find
        same function. Do complain about ambiguating extern "C"
        declarations.

which stopped using duplicate_decls

The testcase in comment #2 (rejects-valid) passes with 3.2, but fails with 3.4.

The testcase in comment #3 (accepts-invalid) has never worked.

The first and last tests have to do with us not having a framework for checking
compatibility of extern "C" declarations; the function that complains about
incompatible declarations is duplicate_decls, and that assumes that both decls
are in the same scope and therefore we can clobber one of them.

But Dodji just added a place to check compatibility of extern "C" declarations
in his fix for 13699.  Dodji, want to extend that to handle #1 and #3?  #2
should be a separate PR I think.


-- 

jason at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dodji at gcc dot gnu dot org
      Known to fail|3.3.6                       |3.2.3 3.3.6


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34691