[Bug middle-end/39937] New: [4.5 Regression] Failed SPEC CPU 2006

[Bug middle-end/39937] New: [4.5 Regression] Failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

On Linux/Intel64, revision 146848 failed with -O2 -ffast-math:

Error: 1x416.gamess 1x435.gromacs

Revision 146820 is OK.


-- 
           Summary: [4.5 Regression] Failed SPEC CPU 2006
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: hjl dot tools at gmail dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #1 from hjl dot tools at gmail dot com  2009-04-28 04:27 -------
I got

gfortran -c -o dftint.fppized.o    -O2 -ffast-math       -DSPEC_CPU_LP64 
-ffixed-form       dftint.fppized.f
gfortran -c -o dgeev.o    -O2 -ffast-math       -DSPEC_CPU_LP64  -ffixed-form  
    dgeev.f
dgeev.f: In function 'dtrevc':
dgeev.f:6294: error: type mismatch in binary expression
<unnamed-unsigned:64>

integer(kind=8)

integer(kind=8)

D.9709_2748 = D.9706_2745 - D.9708_2747;

dgeev.f:6294: error: type mismatch in binary expression
<unnamed-unsigned:64>

<unnamed-unsigned:64>

integer(kind=8)

D.9710_2749 = D.9709_2748 * stride.389_98;

dgeev.f:6294: internal compiler error: verify_stmts failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
specmake[3]: *** [dgeev.o] Error 1


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #2 from hjl dot tools at gmail dot com  2009-04-28 06:01 -------
Revision 146831:

http://gcc.gnu.org/ml/gcc-cvs/2009-04/msg01473.html

is the cause.


-- 

hjl dot tools at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rguenther at suse dot de
            Summary|[4.5 Regression] Failed SPEC|[4.5 Regression] Revision
                   |CPU 2006                    |146831 failed SPEC CPU 2006


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #3 from rguenth at gcc dot gnu dot org  2009-04-28 11:55 -------
      SUBROUTINE DTREVC( SIDE, HOWMNY, SELECT, N, T, LDT, VL, LDVL, VR,
     $                   LDVR, MM, M, WORK, INFO )
      DOUBLE PRECISION   T( LDT, * ), VL( LDVL, * ), VR( LDVR, * ),
     $                   WORK( * )
      DOUBLE PRECISION   X( 2, 2 )
      CALL DLALN2( .FALSE., 1, 1, SMIN, ONE, T( J, J ),
     $                            ZERO, X, 2, SCALE, XNORM, IERR )
      CALL DSCAL( KI, SCALE, WORK( 1+N ), 1 )
      DO 90 J = KI - 2, 1, -1
      IF( J.GT.JNXT )
     $               GO TO 90
      JNXT = J - 1
      IF( J.GT.1 ) THEN
          IF( T( J, J-1 ).NE.ZERO ) THEN
              IF( WORK( J ).GT.BIGNUM / XNORM ) THEN
                  X( 1, 1 ) = X( 1, 1 ) / XNORM
              END IF
          END IF
          CALL DLALN2( .FALSE., 2, 2, SMIN, ONE,
     $                            T( J-1, J-1 ), LDT, ONE, ONE,
     $                            XNORM, IERR )
          CALL DAXPY( J-2, -X( 1, 1 ), T( 1, J-1 ), 1,
     $                           WORK( 1+N ), 1 )
          CALL DAXPY( J-2, -X( 2, 2 ), T( 1, J ), 1,
     $                           WORK( 1+N2 ), 1 )
      END IF
   90          CONTINUE
      END


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2009-04-28 11:55:23
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #4 from rguenth at gcc dot gnu dot org  2009-04-28 12:45 -------
I have a patch for the gamess failure.  What was the gromacs failure?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #5 from hjl dot tools at gmail dot com  2009-04-28 13:36 -------
(In reply to comment #4)
> I have a patch for the gamess failure.  What was the gromacs failure?
> 

cc -c -o pdbio.o -DSPEC_CPU -DNDEBUG  -I. -DHAVE_CONFIG_H  -O2 -ffast-math     
 -DSPEC_CPU_LP64         pdbio.c
pdbio.c: In function 'read_pdbfile':
pdbio.c:384: error: type mismatch in address expression
const char * restrict

char *

D.10701_504 = &line[29];

pdbio.c:384: internal compiler error: verify_stmts failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
specmake[3]: *** [pdbio.o] Error 1


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #6 from rguenth at gcc dot gnu dot org  2009-04-28 14:11 -------
int foo (__const char *__restrict __s);
static void 
read_anisou(char line[])
{
  foo (line+1);
}
void
read_pdbfile(void)
{
  char line[4096];
  read_anisou (line);
}


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #7 from rguenth at gcc dot gnu dot org  2009-04-28 15:27 -------
Subject: Bug 39937

Author: rguenth
Date: Tue Apr 28 15:27:09 2009
New Revision: 146901

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=146901
Log:
2009-04-28  Richard Guenther  <rguenther@suse.de>

        PR middle-end/39937
        * fold-const.c (fold_binary): Use distribute_real_division only
        on float types.

        * gfortran.fortran-torture/compile/pr39937.f: New testcase.

Added:
    trunk/gcc/testsuite/gfortran.fortran-torture/compile/pr39937.f
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/fold-const.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #8 from hjl dot tools at gmail dot com  2009-04-29 14:04 -------
On Linux/x86-64, I got

gcc -c -o pp_sort.o -DSPEC_CPU -DNDEBUG -DPERL_CORE   -O3 -ffast-math
-funroll-loops       -DSPEC_CPU_LP64  -DSPEC_CPU_LINUX_X64       pp_sort.c
pp_sort.c: In function 'S_qsortsv':
pp_sort.c:1352: error: type mismatch in pointer plus expression
struct SV * *

struct SV * *

struct SV * *

vect_vec_iv_.1464_184 = vect_vec_iv_.1464_183 + vect_cst_.1463_182;

pp_sort.c:1352: internal compiler error: verify_stmts failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
specmake[3]: *** [pp_sort.o] Error 1


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenther at suse dot de]

------- Comment #10 from rguenther at suse dot de  2009-04-29 14:06 -------
Subject: Re:  [4.5 Regression] Revision 146831 failed
 SPEC CPU 2006

On Wed, 29 Apr 2009, hjl dot tools at gmail dot com wrote:

> ------- Comment #8 from hjl dot tools at gmail dot com  2009-04-29 14:04 -------
> On Linux/x86-64, I got
> 
> gcc -c -o pp_sort.o -DSPEC_CPU -DNDEBUG -DPERL_CORE   -O3 -ffast-math
> -funroll-loops       -DSPEC_CPU_LP64  -DSPEC_CPU_LINUX_X64       pp_sort.c
> pp_sort.c: In function 'S_qsortsv':
> pp_sort.c:1352: error: type mismatch in pointer plus expression
> struct SV * *
> 
> struct SV * *
> 
> struct SV * *
> 
> vect_vec_iv_.1464_184 = vect_vec_iv_.1464_183 + vect_cst_.1463_182;
> 
> pp_sort.c:1352: internal compiler error: verify_stmts failed
> Please submit a full bug report,
> with preprocessed source if appropriate.
> See <http://gcc.gnu.org/bugs.html> for instructions.
> specmake[3]: *** [pp_sort.o] Error 1

Is that after or before

2009-04-28  Richard Guenther  <rguenther@suse.de>

        * tree-vect-loop.c (get_initial_def_for_induction): Use
        correct types for pointer increment.

?

Richard.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #9 from hjl dot tools at gmail dot com  2009-04-29 14:06 -------
(In reply to comment #8)
> On Linux/x86-64, I got
> 
> gcc -c -o pp_sort.o -DSPEC_CPU -DNDEBUG -DPERL_CORE   -O3 -ffast-math
> -funroll-loops       -DSPEC_CPU_LP64  -DSPEC_CPU_LINUX_X64       pp_sort.c
> pp_sort.c: In function 'S_qsortsv':
> pp_sort.c:1352: error: type mismatch in pointer plus expression
> struct SV * *
> 
> struct SV * *
> 
> struct SV * *
> 
> vect_vec_iv_.1464_184 = vect_vec_iv_.1464_183 + vect_cst_.1463_182;
> 
> pp_sort.c:1352: internal compiler error: verify_stmts failed
> Please submit a full bug report,
> with preprocessed source if appropriate.
> See <http://gcc.gnu.org/bugs.html> for instructions.
> specmake[3]: *** [pp_sort.o] Error 1
> 

This is in perlbench.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl dot tools at gmail dot com]

------- Comment #11 from hjl dot tools at gmail dot com  2009-04-29 14:10 -------
(In reply to comment #10)

> 
> 2009-04-28  Richard Guenther  <rguenther@suse.de>
> 
>         * tree-vect-loop.c (get_initial_def_for_induction): Use
>         correct types for pointer increment.
> 

That is before and with revision 146920.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenther at suse dot de]

------- Comment #12 from rguenther at suse dot de  2009-04-29 14:12 -------
Subject: Re:  [4.5 Regression] Revision 146831 failed
 SPEC CPU 2006

On Wed, 29 Apr 2009, hjl dot tools at gmail dot com wrote:

> ------- Comment #11 from hjl dot tools at gmail dot com  2009-04-29 14:10 -------
> (In reply to comment #10)
> 
> > 
> > 2009-04-28  Richard Guenther  <rguenther@suse.de>
> > 
> >         * tree-vect-loop.c (get_initial_def_for_induction): Use
> >         correct types for pointer increment.
> > 
> 
> That is before and with revision 146920.

Ok, it seems to work for me on current trunk.

Richard.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl at gcc dot gnu dot org]

------- Comment #13 from hjl at gcc dot gnu dot org  2009-04-29 14:55 -------
Subject: Bug 39937

Author: hjl
Date: Wed Apr 29 14:54:54 2009
New Revision: 146972

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=146972
Log:
2009-04-29  H.J. Lu  <hongjiu.lu@intel.com>

        Backport from mainline:
        2009-04-29  Richard Guenther  <rguenther@suse.de>

        PR tree-optimization/39941
        * gcc.c-torture/compile/pr39941.c: New testcase.

        2009-04-29  Anmol P. Paralkar  <anmol@freescale.com>

        PR target/39565
        * gcc.dg/pr39565.c: New testcase.

        2009-04-28  Richard Guenther  <rguenther@suse.de>

        PR middle-end/39937
        * gfortran.fortran-torture/compile/pr39937.f: New testcase.

Added:
    branches/gcc-4_4-branch/gcc/testsuite/gcc.c-torture/compile/pr39941.c
      - copied unchanged from r146971,
trunk/gcc/testsuite/gcc.c-torture/compile/pr39941.c
    branches/gcc-4_4-branch/gcc/testsuite/gcc.dg/pr39565.c
      - copied unchanged from r146971, trunk/gcc/testsuite/gcc.dg/pr39565.c
   
branches/gcc-4_4-branch/gcc/testsuite/gfortran.fortran-torture/compile/pr39937.f
      - copied unchanged from r146971,
trunk/gcc/testsuite/gfortran.fortran-torture/compile/pr39937.f
Modified:
    branches/gcc-4_4-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #14 from rguenth at gcc dot gnu dot org  2009-04-29 15:06 -------
Subject: Bug 39937

Author: rguenth
Date: Wed Apr 29 15:05:22 2009
New Revision: 146973

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=146973
Log:
2009-04-29  Richard Guenther  <rguenther@suse.de>

        PR middle-end/39937
        * tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Do not
        loose type conversions.
        (forward_propagate_addr_expr): Fix tuplification bug.  Remove
        stmts only if there are no uses of its definition.

        * gcc.c-torture/compile/pr39937.c: New testcase.

Added:
    trunk/gcc/testsuite/gcc.c-torture/compile/pr39937.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-ssa-forwprop.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[rguenth at gcc dot gnu dot org]

------- Comment #15 from rguenth at gcc dot gnu dot org  2009-04-29 15:06 -------
Fixed.


-- 

rguenth at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937

[Bug middle-end/39937] [4.5 Regression] Revision 146831 failed SPEC CPU 2006

[hjl at gcc dot gnu dot org]

------- Comment #16 from hjl at gcc dot gnu dot org  2009-04-29 15:55 -------
Subject: Bug 39937

Author: hjl
Date: Wed Apr 29 15:54:18 2009
New Revision: 146981

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=146981
Log:
2009-04-29  H.J. Lu  <hongjiu.lu@intel.com>

        Backport from mainline:
        2009-04-29  Richard Guenther  <rguenther@suse.de>

        PR middle-end/39937
        * gcc.c-torture/compile/pr39937.c: New testcase.

Added:
    branches/gcc-4_4-branch/gcc/testsuite/gcc.c-torture/compile/pr39937.c
      - copied unchanged from r146980,
trunk/gcc/testsuite/gcc.c-torture/compile/pr39937.c
Modified:
    branches/gcc-4_4-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39937