[Bug fortran/30207] New: ICE in gfc_dep_resolver

[Bug fortran/30207] New: ICE in gfc_dep_resolver

[deji_aking at yahoo dot ca]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1799 bytes --]

A trimmed down version of the code which produce the ice is below; The full
code once compile with version 4.1.0, I believe

>>
!
    SUBROUTINE POSL(LAT,LON,JUL,HEURE,MIN,PAZ)

     IMPLICIT NONE
!
     REAL,SAVE :: XPI                ! Pi
     REAL, DIMENSION(:), INTENT(IN):: LAT, LON, JUL, HEURE, MIN 
     REAL, DIMENSION(:),INTENT(OUT):: PAZ
     REAL                                :: ZPI2, ZCDR
!
     REAL, DIMENSION(SIZE(LAT))         :: ZXLON, ZXLON2

      ZPI2 = 2*XPI
      ZCDR = ZPI2/360 

! Position du soleil avec Heure = TU
! Azenit,Azimut en radian
 !
      ZXLON(:)  = 0.
      ZXLON2(:)  = LON(:)
      WHERE(ZXLON < 0.0)
         ZXLON(:) = 360. + ZXLON2(:)         
      END WHERE
!
      PAZ(:)= 0.5*XPI - 0.0
!
    END SUBROUTINE POSL 
>>
[armnlib@agape test1]$ gfortran --version
GNU Fortran 95 (GCC) 4.2.0 20061211 (prerelease)
Copyright (C) 2006 Free Software Foundation, Inc.

GNU Fortran comes with NO WARRANTY, to the extent permitted by law.
You may redistribute copies of GNU Fortran
under the terms of the GNU General Public License.
For more information about these matters, see the file named COPYING

[armnlib@agape test1]$ gfortran -c posl.f90
posl.f90: In function ‘posl’:
posl.f90:20: internal compiler error: in gfc_dep_resolver, at
fortran/dependency.c:1167
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.


-- 
           Summary: ICE in gfc_dep_resolver
           Product: gcc
           Version: 4.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: deji_aking at yahoo dot ca


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver

[pinskia at gcc dot gnu dot org]

-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-valid-code
            Summary|ICE in gfc_dep_resolver     |[4.2/4.3 Regression] ICE in
                   |                            |gfc_dep_resolver
   Target Milestone|---                         |4.2.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[pinskia at gcc dot gnu dot org]

------- Comment #1 from pinskia at gcc dot gnu dot org  2006-12-14 03:10 -------
Confirmed, reduced testcase:
     IMPLICIT NONE
     REAL, DIMENSION(2)         :: ZXLON
      WHERE(ZXLON < 0.0)
         ZXLON(:) = 1
      END WHERE
    END PROGRAM


-- 

pinskia 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         |2006-12-14 03:10:50
               date|                            |
            Summary|[4.2/4.3 Regression] ICE in |[4.2/4.3 Regression] ICE in
                   |gfc_dep_resolver            |gfc_dep_resolver with where
                   |                            |(a < 0) a(:) = 1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[kargl at gcc dot gnu dot org]

------- Comment #2 from kargl at gcc dot gnu dot org  2006-12-14 03:28 -------
(In reply to comment #1)
> Confirmed, reduced testcase:
>      IMPLICIT NONE
>      REAL, DIMENSION(2)         :: ZXLON
>       WHERE(ZXLON < 0.0)
>          ZXLON(:) = 1
>       END WHERE
>     END PROGRAM
> 

A work around is to not use the array section.  That is, change
"ZXLON(:) = 1" to "ZXLON = 1"


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[kargl at gcc dot gnu dot org]

------- Comment #3 from kargl at gcc dot gnu dot org  2006-12-14 05:03 -------
I have a patch/hack to fix this.


-- 

kargl at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kargl at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[roger at eyesopen dot com]

------- Comment #4 from roger at eyesopen dot com  2006-12-16 19:28 -------
Steve posted his fix at http://gcc.gnu.org/ml/gcc-patches/2006-12/msg01012.html
and I came up with an improved version that correctly identifies the
equivalence
of "z" and "z(:)" here http://gcc.gnu.org/ml/gcc-patches/2006-12/msg01094.html

Sorry for the inconvenience.  It looks like I introduced this regression whilst
restructuring gfc_dep_resolver.  As penance, I'm currently investigating a
related gfortran optimization, that can use the proposed gfc_full_array_ref_p,
to improve the code generated for your POSL function even further.


-- 

roger at eyesopen dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |roger at eyesopen dot com
           Keywords|                            |patch


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[jvdelisle at gcc dot gnu dot org]

------- Comment #5 from jvdelisle at gcc dot gnu dot org  2006-12-16 21:21 -------
I have tested Rogers patch with the gfortran testsuite and cp2k and no
regressions.  


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[sayle at gcc dot gnu dot org]

------- Comment #6 from sayle at gcc dot gnu dot org  2006-12-17 18:28 -------
Subject: Bug 30207

Author: sayle
Date: Sun Dec 17 18:28:07 2006
New Revision: 119990

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=119990
Log:
2006-12-17  Roger Sayle  <roger@eyesopen.com>
            Steven G. Kargl  <kargl@gcc.gnu.org>

        PR fortran/30207
        * dependency.c (gfc_full_array_ref_p): New function to test whether
        the given array ref specifies the entire array.
        (gfc_dep_resolver): Use gfc_full_array_ref_p to analyze AR_FULL
        array refs against AR_SECTION array refs, and vice versa.
        * dependency.h (gfc_full_array_ref_p): Prototype here.
        * trans-array.c (gfc_conv_expr_descriptor): Use gfc_full_array_ref_p.

        * gfortran.fortran-torture/execute/where21.f90: New test.


Added:
    trunk/gcc/testsuite/gfortran.fortran-torture/execute/where21.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/dependency.c
    trunk/gcc/fortran/dependency.h
    trunk/gcc/fortran/trans-array.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[deji_aking at yahoo dot ca]

------- Comment #7 from deji_aking at yahoo dot ca  2006-12-17 21:10 -------
Thanks Rogers and Steve for the resolution of this bug. I hope you'll not
forget to commit it to the gcc-4.2 branch.

PS: I've applied the patch cleanly to the 4.2 branch but have not got the
opportunity to build and test it.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[deji_aking at yahoo dot ca]

------- Comment #8 from deji_aking at yahoo dot ca  2006-12-21 03:52 -------
I've applied the fix to 4.2 branch, built and tested it, and it works fine. I
guess the bug can be closed now. Thanks all.


-- 

deji_aking at yahoo dot ca changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2/4.3 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[jvdelisle at gcc dot gnu dot org]

------- Comment #9 from jvdelisle at gcc dot gnu dot org  2006-12-21 04:37 -------
Keep open until we remember to commit to 4.2  Thanks for confirming the fix.


-- 

jvdelisle at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[sayle at gcc dot gnu dot org]

------- Comment #10 from sayle at gcc dot gnu dot org  2006-12-21 18:54 -------
Subject: Bug 30207

Author: sayle
Date: Thu Dec 21 18:54:37 2006
New Revision: 120116

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=120116
Log:
2006-12-21  Roger Sayle  <roger@eyesopen.com>
            Steven G. Kargl  <kargl@gcc.gnu.org>

        PR fortran/30207
        * dependency.c (gfc_full_array_ref_p): New function to test whether
        the given array ref specifies the entire array.
        (gfc_dep_resolver): Use gfc_full_array_ref_p to analyze AR_FULL
        array refs against AR_SECTION array refs, and vice versa.
        * dependency.h (gfc_full_array_ref_p): Prototype here.
        * trans-array.c (gfc_conv_expr_descriptor): Use gfc_full_array_ref_p.

        * gfortran.fortran-torture/execute/where21.f90: New test.


Added:
   
branches/gcc-4_2-branch/gcc/testsuite/gfortran.fortran-torture/execute/where21.f90
Modified:
    branches/gcc-4_2-branch/gcc/fortran/ChangeLog
    branches/gcc-4_2-branch/gcc/fortran/dependency.c
    branches/gcc-4_2-branch/gcc/fortran/dependency.h
    branches/gcc-4_2-branch/gcc/fortran/trans-array.c
    branches/gcc-4_2-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207

[Bug fortran/30207] [4.2 Regression] ICE in gfc_dep_resolver with where (a < 0) a(:) = 1

[pinskia at gcc dot gnu dot org]

------- Comment #11 from pinskia at gcc dot gnu dot org  2006-12-24 05:03 -------
Fixed.


-- 

pinskia at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30207