Re: G++ test failure analysis

[Alexandre Oliva <oliva at dcc dot unicamp dot br>]

On Jul  4, 1999, Robert Lipe <robertl@sco.com> wrote:

> conpr-*.C   
> 	allegedly known failures.

shouldn't they be marked as XFAILs for the release?

> g++-ext/attrib5.C
> 	collect2: ld returned 13 exit status

IMO, this should only be run on targets that support aliases, but we'd 
have to find out which they are.

>         /play/egcs/gcc/testsuite/g++.old-deja/g++.oliva/template3.C:11:
> 	no matching func tion for call to `foo::template <class T> C<T> ()'

Looks like the this bug was fixed.  I'm marking it as such.

> coff:static3.C
> 	A<int>::arr                         /usr/tmp/ccZdLzla.o
> coff:static6.C
> 	A<int>::l                           /usr/tmp/cc4yq9jd.o

These are expected on coff, but I don't know how to arrange that they
don't appear as FAILs.  Is there any #ifdef we could use to provide an
explicit instantiation of the template static data members, as
required on such targets?

> g++.pt/typename8.C
> 	FAIL: g++.pt/typename8.C (test for excess errors)
> 	/play/egcs/gcc/testsuite/g++.old-deja/g++.pt/typename8.C:14: invalid 
> 		use of undefined type `class B'
> 	/play/egcs/gcc/testsuite/g++.old-deja/g++.pt/typename8.C:11: `class 
> 		A<B>' has nomember named `t'

Looks like a few more `// ERROR' marks will be needed here.

The others seem to be platform-specific, but I could swear I've
already met these multiple-definition errors on other platforms
(SunOS4 and IRIX5, if my memory serves), and they have already been
fixed.   Is this a regression or was the fix platform-specific?

-- 
Alexandre Oliva http://www.dcc.unicamp.br/~oliva IC-Unicamp, Bra[sz]il
{oliva,Alexandre.Oliva}@dcc.unicamp.br  aoliva@{acm.org,computer.org}
oliva@{gnu.org,kaffe.org,{egcs,sourceware}.cygnus.com,samba.org}
*** E-mail about software projects will be forwarded to mailing lists