This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: [PATCH]: testsuite/gdb.base/constvars.exp
On Tue, Sep 25, 2001 at 02:03:30PM -0700, Michael Snyder wrote:
> Fernando Nasser wrote:
> >
> > Michael Snyder wrote:
> > >
> > > Corinna Vinschen wrote:
> > > >
> > > > Hi,
> > > >
> > > > I patched the test constvars.exp to get rid of the $gcc_compiled
> > > > compile dependencies which result in XFAIL behaviour when the
> > > > tests are compiled with GCC. AFAICS, they aren't needed if just
> > > > the tests are more correct.
> > > >
> > > > E.g., the testsuite expects strings like "unsigned long" while
> > > > gdb may also emit "long unsigned" or "long unsigned int".
> > > >
> > > > The below patch cares for that.
> > > >
> > > > Corinna
> > >
> > > The test was originally submitted by HP, and probably
> > > worked only with the HP compiler. I like the idea of
> > > extending itto work with GCC, but I wonder -- will this
> > > work with stabs? Or only with dwarf?
> > >
> > > And if it won't work with stabs -- do we care?
> > >
> >
> > Can someone please help us and try it with stabs?
>
> OK, I've tried it. Alas, it does not work. And native Linux
> still uses stabs, so that means we care. ;-( Sorry, Corinna.
> It's a good change, but we need some kind of test that applies
> the xfails for stabs and not for dwarf.
>
> In the meantime, if you want to just check in the regular expression
> changes, I'm sure that would be OK -- it would preserve your work.
What I don't understand is why it's coupled to `gcc_compiled'
while it's in reality target dependend.
Corinna
--
Corinna Vinschen
Cygwin Developer
Red Hat, Inc.
mailto:vinschen@redhat.com