[Bug lto/46525] New: [4.6 Regression] Failed to bootstrap-lto

[Bug lto/46525] New: [4.6 Regression] Failed to bootstrap-lto

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

           Summary: [4.6 Regression] Failed to bootstrap-lto
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hjl.tools@gmail.com


On Linux/x86-64, revision 166867] failed with bootstrap-lto:

http://gcc.gnu.org/ml/gcc-regression/2010-11/msg00229.html

Revision 166855 is OK.

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2010-11-17 18:56:09 UTC ---
Revision 166872 gave:

http://gcc.gnu.org/ml/gcc-regression/2010-11/msg00230.html

xgcc: internal compiler error: Segmentation fault (program lto1)
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
lto-wrapper: /export/gnu/import/svn/gcc-test/bld/./prev-gcc/xgcc returned 4
exit status
lto-wrapper failedcollect2: ld returned 1 exit status
make[6]: *** [jc1] Error 1
make[6]: *** Waiting for unfinished jobs....
xgcc: internal compiler error: Segmentation fault (program lto1)

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.6.0

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

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

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

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2010-11-18 00:31:24 UTC ---
It is caused by revision 166870:

http://gcc.gnu.org/ml/gcc-cvs/2010-11/msg00759.html

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2010.11.18 10:04:31
     Ever Confirmed|0                           |1

--- Comment #3 from Richard Guenther <rguenth at gcc dot gnu.org> 2010-11-18 10:04:31 UTC ---
Mine then.  How do you build bootstrap-lto?

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

--- Comment #4 from Richard Guenther <rguenth at gcc dot gnu.org> 2010-11-18 11:16:35 UTC ---
Hm.  We have

  t (leader)  ->  main-variant (but t is its leader)

so we violated

  /* Always register the main variant first.  This is important so we
     pick up the non-typedef variants as canonical, otherwise we'll end
     up taking typedef ids for structure tags during comparison.  */
  if (TYPE_MAIN_VARIANT (t) != t)
    gimple_register_type (TYPE_MAIN_VARIANT (t));

somehow.  This results in endless recursion between the two fns.

I have a patch.

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

--- Comment #5 from Richard Guenther <rguenth at gcc dot gnu.org> 2010-11-18 11:55:33 UTC ---
I'm past the failure but now see comparison fails of all files.  Ah, I forgot
--frandom-seed=0.

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

--- Comment #6 from Richard Guenther <rguenth at gcc dot gnu.org> 2010-11-18 13:10:45 UTC ---
Author: rguenth
Date: Thu Nov 18 13:10:40 2010
New Revision: 166902

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=166902
Log:
2010-11-18  Richard Guenther  <rguenther@suse.de>

    PR lto/46525
    * gimple.c (gimple_register_type): Update TYPE_MAIN_VARIANT of
    type leaders.
    (gimple_register_canonical_type): Also cache the canoncial type
    for non type leaders.

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

[Bug lto/46525] [4.6 Regression] Failed to bootstrap-lto

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46525

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

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

--- Comment #7 from Richard Guenther <rguenth at gcc dot gnu.org> 2010-11-18 13:12:10 UTC ---
Fixed.