[C++] g++.other/enum5.C

[Ben Elliston]

>>>>> "Nathan" == Nathan Sidwell <nathan@codesourcery.com> writes:

  Nathan> The recently added test case g++.other/enum5.C fails with
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:8: semicolon missing after declaration of `numbers'
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:8: warning: `packed' attribute ignored
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:9: semicolon missing after declaration of `colours'
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:9: warning: `packed' attribute ignored
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:10: parse error before `__attribute__'
  Nathan> /home/nathan/egcs/devel/egcs/gcc/testsuite/g++.old-deja/g++.other/enum5.C:10: ISO C++ forbids declaration of `forecast' with no type

  Nathan> compiler exited with status 1

  Nathan> should the test be xfailed, or ...?

No, the test should not be xfailed.  It is a genuine bug (see PR
c++/80) and this test should serve to remind us of that fact.

Ben