[Bug debug/54205] New: recursive .debug_macro inclusions

[Bug debug/54205] New: recursive .debug_macro inclusions

[tromey at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54205

             Bug #: 54205
           Summary: recursive .debug_macro inclusions
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: debug
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: tromey@gcc.gnu.org


Created attachment 27963
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27963
readelf output

I built git master gcc today.
Then I ran the libstdc++ pretty-printer tests.
One of these tests (48362.cc) caused gdb to die, so I investigated.
It turns out that gcc is emitting recursive DW_MACRO_GNU_transparent_include
calls in .debug_macro.  I previously thought I fixed this problem in
http://sourceware.org/bugzilla/show_bug.cgi?id=13568
-- but the fix was insufficient; and in any case I think gcc should not do
this.
I'm attaching the readelf dump of the section so you can see the issue.

[Bug debug/54205] recursive .debug_macro inclusions

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54205

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-08-10 06:30:43 UTC ---
All the offsets in the dump are zero, which sounds like you are running readelf
on a relocatable file and relocations aren't applied.

[Bug debug/54205] recursive .debug_macro inclusions

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54205

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-08-10 14:47:03 UTC ---
Looks to be a --gc-sections bug in some versions of the GNU ld.
2.21.53.0.[12] seems to be buggy, 2.21.51.0.7 and 2.22.52.0.1-2 still resp.
again works just fine.