GDB 5.2 vs. Ada (and probably other unknown languages)

[Daniel Jacobowitz]

On Fri, Apr 05, 2002 at 07:18:39AM +0200, Florian Weimer wrote:
> Andrew Cagney <ac131313@cygnus.com> writes:
> 
> >> Is there a workaround?  Even if full Ada support is not available,
> >> getting backtraces would be helpful in some cases.
> >
> > Definitly sounds like a new (and unknown) regression :-(
> > Can you create a bug report to track this
> > (http://sources.redhat.com/gdb/bugs/).
> 
> Okay.
> 
> > For a workaround, what happens if you enter ``set language c''?
> 
> The problem persits.  It seems to be related to DWARF2.  If I compile
> using "-gstabs", GDB 5.1.1 works fine.

That's because stabs doesn't have particularly unique language markers,
I'd imagine.  Most places in GDB default to C, but DWARF2 trusts the
object file.

Perhaps if DWARF2 defaults the language to C instead of unknown...

> 
> > There is also a source code dropping of ACT's Ada support in:
> > ftp://sources.redhat.com/pub/gdb/contrib/gnat/ you may find the
> > relevant change in there.
> 
> Aidan Skinner's patch based on this appears to help, too.
> 

-- 
Daniel Jacobowitz                           Carnegie Mellon University
MontaVista Software                         Debian GNU/Linux Developer