This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: RFA: GCC stabs don't contain prototype info
- From: Daniel Berlin <dan at cgsoftware dot com>
- To: Michael Snyder <msnyder at cygnus dot com>
- Cc: Jim Blandy <jimb at cygnus dot com>, Daniel Jacobowitz <drow at mvista dot com>, <gdb-patches at sources dot redhat dot com>
- Date: Tue, 27 Nov 2001 17:27:17 -0500 (EST)
- Subject: Re: RFA: GCC stabs don't contain prototype info
On Tue, 27 Nov 2001, Michael Snyder wrote:
> Jim Blandy wrote:
> >
> > Daniel Jacobowitz <drow@mvista.com> writes:
> > > Is there a standard stabs convention for this? If so, it would be a
> > > good idea to make development GCC emit it.
> >
> > Yes, there's the encoding developed by Sun, described in the GNU stabs
> > manual. It looks to me like GDB is ready to consume this info, so
> > getting GCC to emit it would be great.
> >
> > Until that's done, however, this test should XFAIL, so other engineers
> > won't waste their time re-analyzing the failure. (I don't mean to
> > imply that you disagree; I just want to make things clear to the test
> > suite maintainers.)
>
> FYI, it is not only stabs that fails this test. I've got a recent
> elf/dwarf target that also fails it.
Err, it shouldn't.
Unless you really mean dwarf and not dwarf2.
GCC adds DW_AT_prototyped to prototyped functions.
And gdb consumes it.
> The problem wherein GDB cannot
> reliably make this distinction is long standing, and much discussed
> in the past. Unfortunately most of the people who have taken part
> in this discussion aren't around any more.
>