[Bug bootstrap/48148] New: [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[Bug bootstrap/48148] New: [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

           Summary: [4.7 Regression] LTO bootstrap failed with
                    bootstrap-profiled
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: bootstrap
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hjl.tools@gmail.com


On Linux/Intel64, revision 171037 failed to bootstrap
with LTO and bootstrap-profiled:

http://gcc.gnu.org/ml/gcc-regression/2011-03/msg00071.html

profiling:/export/gnu/import/svn/gcc-test-profile/bld/gcc/insn-attrtab.gcda:Invocation
mismatch - some data files may have been removed
lto1: internal compiler error: in gen_subprogram_die, at dwarf2out.c:18983
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
make[7]: *** [/tmp/cc0ATx3h.ltrans2.ltrans.o] Error 1
lto-wrapper: make returned 2 exit status
/usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-16 12:52:05 UTC ---
It may be caused by revision 171033:

http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00455.html

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.7.0

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-16 12:54:22 UTC ---
It also may be caused by one of

http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00455.html
http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00457.html
http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00458.html
http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00459.html

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

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

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

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-16 13:37:22 UTC ---
That gcc_assert (!declaration || local_scope_p (context_die));
assert is there for many years, and I don't see anything that might cause it in
my patch.  I'll try to reproduce it, but as it is lto it is hard to guess if I
manage to.

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-16 14:49:05 UTC ---
It is caused by revision 171033:

http://gcc.gnu.org/ml/gcc-cvs/2011-03/msg00455.html

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[4.7 Regression] LTO        |[4.7 Regression] bootstrap
                   |bootstrap failed with       |failed with
                   |bootstrap-profiled          |bootstrap-profiled

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-16 15:21:00 UTC ---
It failed with bootstrap-profiled.  LTO isn't needed.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-16 15:32:01 UTC ---
So, it seems this comes from a DW_TAG_GNU_call_site with
(symbol_ref:DI ("read_string.constprop.0") [flags 0x43] <function_decl
0x7ffff1d47000 read_string.constprop.0>)
as DW_AT_abstract_origin, which during resolve_addr results in force_decl_die.
read_string.constprop.0 is DECL_EXTERNAL FUNCTION_DECL with DECL_CONTEXT being
TRANSLATION_UNIT_DECL and DECL_ABSTRACT_ORIGIN of read_string.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-16 16:19:45 UTC ---
Created attachment 23680
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23680
gcc47-pr48148.patch

I think this patch ought to fix it (or we could just not try to force_decl_die
if DECL_ABSTRACT_ORIGIN on the DECL_EXTERNAL tdecl is non-NULL.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

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

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

--- Comment #9 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-16 16:29:49 UTC ---
That is,

Index: lto-cgraph.c
===================================================================
--- lto-cgraph.c        (revision 171048)
+++ lto-cgraph.c        (working copy)
@@ -975,6 +975,7 @@ input_overwrite_node (struct lto_file_de
          || node->clone_of->decl != node->decl))
     {
       DECL_EXTERNAL (node->decl) = 1;
+      DECL_ABSTRACT_ORIGIN (node->decl) = NULL_TREE;
       TREE_STATIC (node->decl) = 0;
     }
   node->alias = bp_unpack_value (bp, 1);
@@ -1146,6 +1147,7 @@ input_varpool_node (struct lto_file_decl
   if (node->in_other_partition)
     {
       DECL_EXTERNAL (node->decl) = 1;
+      DECL_ABSTRACT_ORIGIN (node->decl) = NULL_TREE;
       TREE_STATIC (node->decl) = 0;
     }
   aliases_p = bp_unpack_value (&bp, 1);

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #8 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-16 16:27:14 UTC ---
Do you think this is a LTO speciality because of partitioning (in which
case we should probably clear DECL_ABSTRACT_ORIGIN on those decls we
promote to DECL_EXTERNAL for other ltrans units)?

Btw, I wonder if using DECL_ABSTRACT_ORIGIN for function clones really makes
sense.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|jakub at redhat dot com     |jan.kratochvil at redhat
                   |                            |dot com

--- Comment #10 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-16 17:03:58 UTC ---
I believe it is a LTO speciality, I can't imagine how non-lto would end up with
DECL_EXTERNAL with DECL_ABSTRACT_ORIGIN set, so I guess your patch is an
alternative.

Or:
--- dwarf2out.c.jj      2011-03-16 09:37:40.000000000 +0100
+++ dwarf2out.c 2011-03-16 17:28:36.054796163 +0100
@@ -22927,7 +22927,9 @@ resolve_addr (dw_die_ref die)
          {
            tree tdecl = SYMBOL_REF_DECL (a->dw_attr_val.v.val_addr);
            dw_die_ref tdie = lookup_decl_die (tdecl);
-           if (tdie == NULL && DECL_EXTERNAL (tdecl))
+           if (tdie == NULL
+               && DECL_EXTERNAL (tdecl)
+               && DECL_ABSTRACT_ORIGIN (tdecl) == NULL_TREE)
              {
                force_decl_die (tdecl);
                tdie = lookup_decl_die (tdecl);

that I'm testing right now.  DW_TAG_GNU_call_site won't be very useful in that
case anyway, because if the extern is "read_string.constprop.0" with that
DW_AT_name, I bet the debugger won't be able to match it to the definition DIE
anyway.  I think currently for read_string.constprop.0 in the definition CU we
will just have
DW_TAG_subprogram
  DW_AT_abstract_origin <read_string DIE>
and not any DW_AT_name (which is thought to be the same as abstract origin's
name, correct for the clones) and no DW_AT_linkage_name (guess it would be
desirable to provide a linkage name if the clone has a different name from the
origin).

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jan.kratochvil at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #11 from Jan Kratochvil <jan.kratochvil at redhat dot com> 2011-03-16 17:52:08 UTC ---
read_call_site_type in archer-jankratochvil-entryval for
DW_AT_GNU_call_site_target recognizes DW_FORM_ref* pointing to
DW_AT_declaration DIE and stores + later follows the linkage name.

Whether the linkage name is determined correctly depends on the existing GDB
infrastructure and normally it works.  I do not have debugging experience with
the .0 symbols so far - but that is outside of the scope of the
archer-jankratochvil-entryval branch.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #12 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-16 17:56:35 UTC ---
I've been actually talking about direct calls, with
DW_TAG_GNU_call_site
  DW_AT_abstract_origin <declaration DIE of some extern fn>
and how gdb matches that to the actual definition DIE.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jan.kratochvil at redhat dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #13 from Jan Kratochvil <jan.kratochvil at redhat dot com> 2011-03-16 18:02:51 UTC ---
In such case it expects
DW_TAG_GNU_call_site
  DW_AT_low_pc
and it does not look up DW_AT_low_pc through DW_AT_abstract_origin across
linkage names.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #14 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-17 13:25:28 UTC ---
Author: rguenth
Date: Thu Mar 17 13:25:23 2011
New Revision: 171094

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171094
Log:
2011-03-17  Richard Guenther  <rguenther@suse.de>

    PR bootstrap/48148
    * lto-cgraph.c (input_overwrite_node): Clear the abstract
    origin for decls in other ltrans units.
    (input_varpool_node): Likewise.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/lto-cgraph.c

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

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

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

--- Comment #15 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-17 13:25:54 UTC ---
Should be fixed.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
   Last reconfirmed|                            |2011.03.17 13:31:35
         Resolution|FIXED                       |
     Ever Confirmed|0                           |1

--- Comment #16 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-17 13:31:35 UTC ---
Bootstrap also failed with bootstrap-profiled, but without LTO.

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #17 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-17 13:34:17 UTC ---
Sure, but was it the same problem?  Are you sure it wasn't PR48158 ?

[Bug bootstrap/48148] [4.7 Regression] bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED

--- Comment #18 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-17 14:40:54 UTC ---
Fixed.

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |
            Summary|[4.7 Regression] bootstrap  |[4.7 Regression] LTO
                   |failed with                 |bootstrap failed with
                   |bootstrap-profiled          |bootstrap-profiled

--- Comment #19 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-17 17:19:16 UTC ---
Revision 171098 still failed with LTO + bootstrap-profiled:

http://gcc.gnu.org/ml/gcc-regression/2011-03/msg00146.html

to1: internal compiler error: in gen_subprogram_die, at dwarf2out.c:18996
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
make[7]: *** [/tmp/ccQwgD1K.ltrans2.ltrans.o] Error 1
make[7]: *** Waiting for unfinished jobs....
lto-wrapper: make returned 2 exit status
/usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
collect2: ld returned 1 exit status

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #20 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-03-17 20:30:19 UTC ---
Confirmed.  Apparently DECL_ABSTRACT_ORIGIN is still NULL at the point where
Richard changed it and DECL_EXTERNAL is being set.  It is set during:
#0  tree_function_versioning (old_decl=0x7ffff1d40300, new_decl=0x7ffff1d47000,
tree_map=0x7ffff1d524b0, update_clones=1 '\001', 
    args_to_skip=0x7ffff1d1be60, blocks_to_copy=0x0, new_entry=0x0) at
../../gcc/tree-inline.c:5058
#1  0x0000000002ab2fb1 in cgraph_materialize_clone (node=0x7ffff1d61000) at
../../gcc/cgraphunit.c:2175
#2  0x0000000002ab3901 in cgraph_materialize_all_clones () at
../../gcc/cgraphunit.c:2372
#3  0x0000000002ab2356 in cgraph_optimize () at ../../gcc/cgraphunit.c:1889
#4  0x00000000004ee63b in lto_main () at ../../gcc/lto/lto.c:2469

So, either tree_function_versioning needs to be changed not to set
DECL_ABSTRACT_ORIGIN for DECL_EXTERNAL new_decls, or lto needs to fix it up
somewhere else, or we need to go with one of the proposed dwarf2out.c changes.

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |NEW

--- Comment #21 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-24 04:09:42 UTC ---
Bootstrap is still broken.

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #22 from H.J. Lu <hjl.tools at gmail dot com> 2011-03-31 14:14:08 UTC ---
Can we apply Jaub's patch to fix bootstrap?

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #23 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-04-01 21:14:39 UTC ---
Author: jakub
Date: Fri Apr  1 21:14:36 2011
New Revision: 171856

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=171856
Log:
    PR bootstrap/48148
    * dwarf2out.c (resolve_addr): Don't call force_decl_die
    if DECL_EXTERNAL has non-NULL DECL_ABSTRACT_ORIGIN.

    Revert:
    2011-03-17  Richard Guenther  <rguenther@suse.de>

    PR bootstrap/48148
    * lto-cgraph.c (input_overwrite_node): Clear the abstract
    origin for decls in other ltrans units.
    (input_varpool_node): Likewise.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/dwarf2out.c
    trunk/gcc/lto-cgraph.c

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #24 from H.J. Lu <hjl.tools at gmail dot com> 2011-04-04 19:24:03 UTC ---
Still doesn't work.  Revision 171942 gave:

http://gcc.gnu.org/ml/gcc-regression/2011-04/msg00253.html

/tmp/ccuKub9z.s: Assembler messages:
/tmp/ccuKub9z.s:180352: Error: symbol `.Ldebug_info0' is already defined
make[7]: *** [/tmp/ccxeOYQm.ltrans21.ltrans.o] Error 1
lto-wrapper: make returned 2 exit status
/usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
collect2: ld returned 1 exit status
make[6]: *** [cc1] Error 1
make[6]: *** Waiting for unfinished jobs....
/tmp/ccxlW3CV.s: Assembler messages:
/tmp/ccxlW3CV.s:164584: Error: symbol `.Ldebug_info0' is already defined
make[7]: *** [/tmp/ccGWmRAf.ltrans21.ltrans.o] Error 1
make[7]: *** Waiting for unfinished jobs....
lto-wrapper: make returned 2 exit status
/usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
collect2: ld returned 1 exit status
make[6]: *** [cc1obj] Error 1

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #25 from H.J. Lu <hjl.tools at gmail dot com> 2011-04-04 20:01:29 UTC ---
(In reply to comment #24)
> Still doesn't work.  Revision 171942 gave:
> 
> http://gcc.gnu.org/ml/gcc-regression/2011-04/msg00253.html
> 
> /tmp/ccuKub9z.s: Assembler messages:
> /tmp/ccuKub9z.s:180352: Error: symbol `.Ldebug_info0' is already defined
> make[7]: *** [/tmp/ccxeOYQm.ltrans21.ltrans.o] Error 1
> lto-wrapper: make returned 2 exit status
> /usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
> collect2: ld returned 1 exit status
> make[6]: *** [cc1] Error 1
> make[6]: *** Waiting for unfinished jobs....
> /tmp/ccxlW3CV.s: Assembler messages:
> /tmp/ccxlW3CV.s:164584: Error: symbol `.Ldebug_info0' is already defined
> make[7]: *** [/tmp/ccGWmRAf.ltrans21.ltrans.o] Error 1
> make[7]: *** Waiting for unfinished jobs....
> lto-wrapper: make returned 2 exit status
> /usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
> collect2: ld returned 1 exit status
> make[6]: *** [cc1obj] Error 1

This started with revision 171856:

http://gcc.gnu.org/ml/gcc-cvs/2011-04/msg00048.html

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #26 from H.J. Lu <hjl.tools at gmail dot com> 2011-04-05 02:52:15 UTC ---
(In reply to comment #25)
> (In reply to comment #24)
> > Still doesn't work.  Revision 171942 gave:
> > 
> > http://gcc.gnu.org/ml/gcc-regression/2011-04/msg00253.html
> > 
> > /tmp/ccuKub9z.s: Assembler messages:
> > /tmp/ccuKub9z.s:180352: Error: symbol `.Ldebug_info0' is already defined
> > make[7]: *** [/tmp/ccxeOYQm.ltrans21.ltrans.o] Error 1
> > lto-wrapper: make returned 2 exit status
> > /usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
> > collect2: ld returned 1 exit status
> > make[6]: *** [cc1] Error 1
> > make[6]: *** Waiting for unfinished jobs....
> > /tmp/ccxlW3CV.s: Assembler messages:
> > /tmp/ccxlW3CV.s:164584: Error: symbol `.Ldebug_info0' is already defined
> > make[7]: *** [/tmp/ccGWmRAf.ltrans21.ltrans.o] Error 1
> > make[7]: *** Waiting for unfinished jobs....
> > lto-wrapper: make returned 2 exit status
> > /usr/local/x86_64-unknown-linux-gnu/bin/ld: lto-wrapper failed
> > collect2: ld returned 1 exit status
> > make[6]: *** [cc1obj] Error 1

I can reproduce it with revision 171033 + patch in comment #7.
That means revision 171033 has 2 problems:

1. 

lto1: internal compiler error: in gen_subprogram_die, at dwarf2out.c:18983
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
make[7]: *** [/tmp/cc0ATx3h.ltrans2.ltrans.o] Error 1

2.

/tmp/ccuKub9z.s: Assembler messages:
/tmp/ccuKub9z.s:180352: Error: symbol `.Ldebug_info0' is already defined
make[7]: *** [/tmp/ccxeOYQm.ltrans21.ltrans.o] Error 1

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #27 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2011-04-19 14:46:39 UTC ---
Author: ebotcazou
Date: Tue Apr 19 14:46:37 2011
New Revision: 172712

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=172712
Log:
    PR lto/48148
    * gimple.c (gimple_types_compatible_p_1) <ENUMERAL_TYPE>: Do not merge
    the types if they have different enumeration identifiers.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/gimple.c

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |ebotcazou at gcc dot
                   |                            |gnu.org
         Resolution|                            |FIXED

--- Comment #28 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2011-04-19 14:48:26 UTC ---
Presumably.

[Bug bootstrap/48148] [4.7 Regression] LTO bootstrap failed with bootstrap-profiled

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48148

--- Comment #29 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2011-04-20 11:18:57 UTC ---
Author: ebotcazou
Date: Wed Apr 20 11:18:50 2011
New Revision: 172767

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=172767
Log:
    Backport from mainline
    2011-04-19  Eric Botcazou  <ebotcazou@adacore.com>

    PR lto/48148
    * gimple.c (gimple_types_compatible_p_1) <ENUMERAL_TYPE>: Do not merge
    the types if they have different enumeration identifiers.

    2011-04-18  Eric Botcazou  <ebotcazou@adacore.com>

    PR lto/48492
    * cfgexpand.c (expand_debug_expr) <VAR_DECL>: Return NULL for a
    DECL_IN_CONSTANT_POOL without RTL.

Modified:
    branches/gcc-4_6-branch/gcc/ChangeLog
    branches/gcc-4_6-branch/gcc/cfgexpand.c
    branches/gcc-4_6-branch/gcc/gimple.c