linkonce sections, DWARF2 EH, and the ppc failures

[Geoff Keating]

> Date: Tue, 19 Jun 2001 13:19:56 +0930
> From: Alan Modra <amodra@bigpond.net.au>
> Cc: bkoz@cygnus.com, gcc-bugs@gcc.gnu.org, rth@cygnus.com,
>         binutils@sources.redhat.com
> Mail-Followup-To: Geoff Keating <geoffk@redhat.com>, bkoz@cygnus.com,
> 	gcc-bugs@gcc.gnu.org, rth@cygnus.com, binutils@sources.redhat.com
> Content-Disposition: inline
> User-Agent: Mutt/1.3.17i
> 
> On Sun, Jun 17, 2001 at 08:54:10PM -0700, Geoff Keating wrote:
> > 
> > The `at' method, aka `_ZNKSs2atEj', is coming from the executable, in
> > a linkonce section, but the EH data that is used is coming from
> > string-inst.o, because the code for specifying the offsets is:
> > 
> >         .4byte  .LFB70   # FDE initial location
> >         .4byte  .LFE70-.LFB70    # FDE address range
> 
> This is from .debug_aranges, right?  Why not emit the debug address
> info for linkonce sections in .gnu.linkonce.da.*, say.  That way the
> linker should select the appropriate debug info for the kept
> linkonce section as they'll be brought in from the same file.

No, this is from .eh_frame.  I don't want to break the eh info into
lots of little sections because it'll become about 75% larger due to
duplicating the header info.

-- 
- Geoffrey Keating <geoffk@geoffk.org>