[Bug c++/51318] New: [4.7 Regression] segfault on Eigen3

[Bug c++/51318] New: [4.7 Regression] segfault on Eigen3

[marc.glisse at normalesup dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

             Bug #: 51318
           Summary: [4.7 Regression] segfault on Eigen3
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: marc.glisse@normalesup.org


enum { ColMajor = 0, RowMajor = 0x1, AutoAlign = 0, DontAlign = 0x2 };
template<typename _Scalar, int _Rows, int _Cols, int _Options =  (_Rows==1) ?
RowMajor : ColMajor > class Matrix;
template<typename T> class MatrixBase {
        typedef Matrix<double,T::X,0> type;
};

$ g++ -fsyntax-only a.ii
a.ii:4:30: 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.

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[reichelt at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

Volker Reichelt <reichelt at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |ice-on-valid-code
   Last reconfirmed|                            |2011-11-29
                 CC|                            |reichelt at gcc dot gnu.org
     Ever Confirmed|0                           |1
   Target Milestone|---                         |4.7.0

--- Comment #1 from Volker Reichelt <reichelt at gcc dot gnu.org> 2011-11-29 22:07:38 UTC ---
Confirmed. Even simpler testcase:

=================================================
enum { e0, e1 };

template<bool B, int = B ? e0 : e1> struct A {};

template<typename T> struct B
{
  A<T::X> a;
};
=================================================

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

--- Comment #2 from Paolo Carlini <paolo.carlini at oracle dot com> 2011-12-01 21:31:58 UTC ---
Marc, do you have a feeling PR51384 is a duplicate?

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[marc.glisse at normalesup dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

--- Comment #3 from Marc Glisse <marc.glisse at normalesup dot org> 2011-12-01 21:42:35 UTC ---
(In reply to comment #2)
> Marc, do you have a feeling PR51384 is a duplicate?

At least I produced this testcase starting from the one-liner:
#include <Eigen/Dense>
that made g++ segfault and applying delta. So yes, I'd say it's a dup (although
there could in principle be several g++ bugs and delta would highlight one that
may not be the one giving the backtrace in PR51384).

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

--- Comment #4 from Paolo Carlini <paolo.carlini at oracle dot com> 2011-12-01 22:11:11 UTC ---
Thanks. I just had a look to the backtrace and seems the same.

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

Paolo Carlini <paolo.carlini at oracle dot com> changed:

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

--- Comment #5 from Paolo Carlini <paolo.carlini at oracle dot com> 2011-12-01 22:14:51 UTC ---
Let's add Jason in CC, it seems an annoying regression.

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

--- Comment #6 from Paolo Carlini <paolo.carlini at oracle dot com> 2011-12-04 16:32:52 UTC ---
For some reason, in mainline tsubst_template_arg (thus tsubst_expr ->
tsubst_copy_and_build, which crashes because TREE_TYPE (r) is null) handles an
INDIRECT_REF (with a COND_EXPR as arg 0) instead of a COND_EXPR.

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dcb314 at hotmail dot com

--- Comment #7 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-12-06 14:19:48 UTC ---
*** Bug 51384 has been marked as a duplicate of this bug. ***

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

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

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

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-12-07 09:29:29 UTC ---
Started with http://gcc.gnu.org/viewcvs?view=revision&revision=181174, i.e.
PR50835, in particular reverting the lvalue_kind change makes the ICE go away.

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |jason at gcc dot gnu.org
                   |gnu.org                     |

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

--- Comment #9 from Jason Merrill <jason at gcc dot gnu.org> 2011-12-08 22:28:41 UTC ---
Author: jason
Date: Thu Dec  8 22:28:38 2011
New Revision: 182142

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=182142
Log:
    PR c++/51318
    * typeck.c (build_x_conditional_expr): Restrict glvalue games to C++11.

Added:
    trunk/gcc/testsuite/g++.dg/template/cond8.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/typeck.c
    trunk/gcc/testsuite/ChangeLog

[Bug c++/51318] [4.7 Regression] segfault on Eigen3

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51318

Jason Merrill <jason at gcc dot gnu.org> changed:

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

--- Comment #10 from Jason Merrill <jason at gcc dot gnu.org> 2011-12-08 22:28:58 UTC ---
Fixed.