[Bug tree-optimization/42521] New: [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[Bug tree-optimization/42521] New: [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[zsojka at seznam dot cz]

Command line:
gcc -O3 -fgraphite-identity -c testcase.c

Tested versions:
trunk r155434 - crash
trunk r155363 - crash (with and without checking)
trunk r155248 - crash
trunk r154830 - crash
trunk r153685 - different crash (in build_loop_iteration_domains)
4,4 r155365 - OK
4.4 r154975 - OK (with checking)
4.4.2 - OK (gentoo)

Output:
$ /mnt/svn/gcc-trunk/binary-155434-lto/bin/gcc -O3 -fgraphite-identity -c
testcase-manual.c
testcase-manual.c: In function ‘foo’:
testcase-manual.c:4:6: internal compiler error: in graphite_loop_normal_form,
at graphite-sese-to-poly.c:2844
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

valgrind output:
no invalid accesses


-- 
           Summary: [4.5 Regression] ICE: in graphite_loop_normal_form, at
                    graphite-sese-to-poly.c:2844
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: zsojka at seznam dot cz
  GCC host triplet: x86_64-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[zsojka at seznam dot cz]

------- Comment #1 from zsojka at seznam dot cz  2009-12-28 20:10 -------
Created an attachment (id=19400)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19400&action=view)
reduced testcase

Command line:
gcc -O3 -fgraphite-identity -c pr42521.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[zsojka at seznam dot cz]

------- Comment #2 from zsojka at seznam dot cz  2009-12-28 20:17 -------
Compiler flags can be further reduced to:
-O1 -funswitch-loops -fstrict-aliasing -fgraphite-identity


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[hjl dot tools at gmail dot com]

------- Comment #3 from hjl dot tools at gmail dot com  2009-12-28 22:52 -------
It is caused by the new implementation of Graphite:

http://gcc.gnu.org/ml/gcc-cvs/2009-07/msg01187.html


-- 

hjl dot tools at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |spop at gcc dot gnu dot org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2009-12-28 22:52:58
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[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=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[spop at gcc dot gnu dot org]

------- Comment #4 from spop at gcc dot gnu dot org  2010-01-08 04:39 -------
Mine.


-- 

spop at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |spop at gcc dot gnu dot org
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-12-28 22:52:58         |2010-01-08 04:39:23
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[spop at gcc dot gnu dot org]

------- Comment #5 from spop at gcc dot gnu dot org  2010-01-08 06:04 -------
Subject: Bug 42521

Author: spop
Date: Fri Jan  8 06:04:41 2010
New Revision: 155727

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=155727
Log:
Fix PR42521.

2010-01-07  Sebastian Pop  <sebastian.pop@amd.com>

        PR middle-end/42521
        * graphite.c (graphite_finalize): Call scev_reset.
        (graphite_transform_loops): Do not call scev_reset between the code
        generation of scops.

        * testsuite/gcc.dg/graphite/pr42521.c: New.

Added:
    branches/graphite/gcc/testsuite/gcc.dg/graphite/pr42521.c
Modified:
    branches/graphite/gcc/ChangeLog.graphite
    branches/graphite/gcc/graphite.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[spop at gcc dot gnu dot org]

------- Comment #6 from spop at gcc dot gnu dot org  2010-01-08 07:39 -------
Subject: Bug 42521

Author: spop
Date: Fri Jan  8 07:39:01 2010
New Revision: 155728

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=155728
Log:
Fix PR42521.

2010-01-07  Sebastian Pop  <sebastian.pop@amd.com>

        PR middle-end/42521
        * graphite.c (graphite_finalize): Call scev_reset.
        (graphite_transform_loops): Do not call scev_reset between the code
        generation of scops.

        * testsuite/gcc.dg/graphite/pr42521.c: New.

Added:
    trunk/gcc/testsuite/gcc.dg/graphite/pr42521.c
Modified:
    trunk/gcc/ChangeLog.graphite
    trunk/gcc/graphite.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[spop at gcc dot gnu dot org]

------- Comment #7 from spop at gcc dot gnu dot org  2010-01-08 07:40 -------
Fixed.


-- 

spop at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[spop at gcc dot gnu dot org]

------- Comment #8 from spop at gcc dot gnu dot org  2010-01-13 17:00 -------
The fix for this PR produces an ICE in 434.zeusmp:

+ /home/seb/gcc/graphite/usr/bin/gfortran -c -g -O2 -DSPEC_CPU_LP64 -o pdv_d.o
pdv_d.f
pdv_d.f: In function pdv_d:
pdv_d.f:89:0: error: definition in block 40 does not dominate use in block 212
for SSA_NAME: prephitmp.28_439 in statement:
D.2771_606 = D.2770_605 <= prephitmp.28_439;
pdv_d.f:89:0: internal compiler error: verify_ssa failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

This is the reduced testcase:

      parameter(in =           128+5
     &        , jn =           128+5
     &        , kn =           128+5)
      real*8   d (in,jn,kn)
       real*8 dcopy(in,jn,kn)
       call pdv (is, dcopy)
       do k=ks,ke
         do j=je+1,je+2
           do i=is-2,ie+2
             dcopy(i,j,k) = d(i,j,k)
           enddo
         enddo
       enddo
       do k=ks,ke
         do j=js,je
           do i=is-2,is-1
             dcopy(i,j,k) = d(i,j,k)
           enddo
         enddo
       enddo
       end


-- 

spop at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[rguenth at gcc dot gnu dot org]

------- Comment #9 from rguenth at gcc dot gnu dot org  2010-01-13 17:34 -------
Please track this in a separate bug PR42732.


-- 

rguenth at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521

Re: [Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[Sebastian Pop]

> pdv_d.f:89:0: error: definition in block 40 does not dominate use in block 212
> for SSA_NAME: prephitmp.28_439 in statement:
> D.2771_606 = D.2770_605 <= prephitmp.28_439;

The error comes from the fact that we are not clearing the scev
information anymore in between the code generation of two scops.

In this particular case, we have two scops, the second scop contains a
loop for which the number of iterations is a variable computed in the
first scop, and because we do not update the niter/scev info we keep
referring to the old SSA_NAME, prephitmp.28_439.

A solution would be to rename all the scev info based on the
rename_map that is computed by the translation of the first scop.

I am working on a patch to do that.

Sebastian

[Bug tree-optimization/42521] [4.5 Regression] ICE: in graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

[sebpop at gmail dot com]

------- Comment #10 from sebpop at gmail dot com  2010-01-13 18:15 -------
Subject: Re:  [4.5 Regression] ICE: in 
        graphite_loop_normal_form, at graphite-sese-to-poly.c:2844

> pdv_d.f:89:0: error: definition in block 40 does not dominate use in block 212
> for SSA_NAME: prephitmp.28_439 in statement:
> D.2771_606 = D.2770_605 <= prephitmp.28_439;

The error comes from the fact that we are not clearing the scev
information anymore in between the code generation of two scops.

In this particular case, we have two scops, the second scop contains a
loop for which the number of iterations is a variable computed in the
first scop, and because we do not update the niter/scev info we keep
referring to the old SSA_NAME, prephitmp.28_439.

A solution would be to rename all the scev info based on the
rename_map that is computed by the translation of the first scop.

I am working on a patch to do that.

Sebastian


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42521