Mail Archives: djgpp-workers/2002/05/17/07:40:24
On 17 May 2002 at 14:00, Eli Zaretskii wrote:
> > From: pavenis AT lanet DOT lv
> > Date: Fri, 17 May 2002 12:58:06 +0300
> >
> > Unlikely. Anyway it's not a fatal problem, as it happens only then when
> > one uses DWARF-2 debugging info and already have unresolved
> > references from linker (so nothing is additionally broken, not counting
> > already broken user program ... )
>
> Is the fix known? If so, perhaps it could be applied to the DJGPP
> port. I mean, previous versions of GCC did allow DWARF-2 without any
> problems; could we just get back to that?
>
> It is IMHO not good to have a GCC port which effectively disallows
> using DWARF-2 debugging. (Although it's not a fatal problem, given
> that COFF and stabs still work.)
It's even more non fatal problem as DWARF2 debugging info can be used.
This problem does not disallow DRAWF2 debugging.
We'll only have additional error message when we already have unresolved
reference from linker, so when there is some error in user program
It will go away when unresolved reference will be fixed.
> > Only we should be prepared for storm of questions about these messages.
>
> Why ``a storm''? -ggdb is not the default, is it? Does RHIDE use it
> by default, perhaps? If not, only users who know about DWARF-2 will
> bump into this problem.
If we look at current troubles with debugging, -ggdb could be one of most
stable
Andris
- Raw text -