[Bug lto/48042] lto segfaults while building Qt 4.7.2 with -g -flto -fwhole-program

["rguenth at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48042

Richard Guenther <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW

--- Comment #8 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-09 16:32:51 UTC ---
(In reply to comment #7)
> Created attachment 23605 [details]
> Much smaller test case
> 
> Small self-contained test case, source-only.
> 
> $ tar xzf bug48042.tar.gz
> $ cd bug48042
> $ ./bug48042.sh
> In file included from
> ../../../include/QtCore/../../src/corelib/global/qglobal.h:899:0,
>                  from :10:
> ../../../include/Qt3Support/../../src/qt3support/tools/q3glist.h: In member
> function âcountâ:
> ../../../include/Qt3Support/../../src/qt3support/tools/q3glist.h:161:1:
> internal compiler error: Segmentation fault
> Please submit a full bug report,
> with preprocessed source if appropriate.
> See <http://gcc.gnu.org/bugs.html> for instructions.
> lto-wrapper: g++ returned 1 exit status
> /usr/bin/ld: lto-wrapper failed
> collect2: ld returned 1 exit status
> 
> 
> The problem goes away if -g is added to the first compiler invocation as well
> -- lto seems to fall over some constructs if linking is done with -g but an
> object file was built without -g.

Ah, yeah - that's a known problem (and hard to fix).  What should work
is dropping -g at link time.

Basically to make your way reliably all conditional clearing in
free_lang_data_in_* based on debug_info_level checks have to be dropped.