[Bug middle-end/64394] New: ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[Bug middle-end/64394] New: ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64394

            Bug ID: 64394
           Summary: ICE: in build_linearized_memory_access, at
                    graphite-interchange.c:121 (isl_constraint.c:558:
                    expecting integer value) with -floop-interchange
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zsojka at seznam dot cz

Created attachment 34327
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34327&action=edit
reduced testcase

Compiler output:
$ gcc -O2 -floop-interchange testcase.c         
/var/tmp/portage/dev-libs/isl-0.12.2/work/isl-0.12.2/isl_constraint.c:558:
expecting integer value
testcase.c: In function 'foo':
testcase.c:10:1: internal compiler error: Aborted
 foo (void)
 ^
0xc5819f crash_signal
        /mnt/svn/gcc-trunk/gcc/toplev.c:359
0x1757f91 build_linearized_memory_access
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:121
0x1757f91 pdr_stride_in_loop
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:207
0x1757f91 memory_strides_in_loop_1
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:293
0x17627d4 lst_interchange_profitable_p
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:410
0x17627d4 lst_try_interchange_loops
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:547
0x17627d4 lst_interchange_select_inner
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:610
0x17655bf lst_interchange_select_outer
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:636
0x176560d lst_interchange_select_outer
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:646
0x176567b scop_do_interchange(scop*)
        /mnt/svn/gcc-trunk/gcc/graphite-interchange.c:658
0x168c65a apply_poly_transforms(scop*)
        /mnt/svn/gcc-trunk/gcc/graphite-poly.c:272
0x1687783 graphite_transform_loops()
        /mnt/svn/gcc-trunk/gcc/graphite.c:302
0x1687db0 graphite_transforms
        /mnt/svn/gcc-trunk/gcc/graphite.c:331
0x1687db0 execute
        /mnt/svn/gcc-trunk/gcc/graphite.c:412
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.

Tested revisions:
r219043 - ICE

[Bug middle-end/64394] ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[jana at saout dot de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64394

Jana Saout <jana at saout dot de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jana at saout dot de

--- Comment #1 from Jana Saout <jana at saout dot de> ---
I'm presumably seeing this bug being hit when trying to build gcc 5 with
lto-bootstrapping. My backtrace is garbled, but the "isl_constraint.cc":
"expecting integer value" is also at the core of the ICE.

[Bug middle-end/64394] ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[asolokha at gmx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64394

Arseny Solokha <asolokha at gmx dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |asolokha at gmx dot com

--- Comment #2 from Arseny Solokha <asolokha at gmx dot com> ---
Created attachment 34877
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34877&action=edit
Another testcase

The following testcase reduced from pr.c from GNU coreutils 8.23 fails for me
the same way when building for powerpc-e500v2-linux-gnuspe at -O[2s]
-floop-block. However I cannot reproduce this w/ different GCC versions or on
different targets at hands.

% powerpc-e500v2-linux-gnuspe-gcc-5.0.0-alpha20150222 -c -O2 -floop-block
pgyx7arv.i

[Bug middle-end/64394] ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[hiraditya at msn dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64394

AK <hiraditya at msn dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hiraditya at msn dot com

--- Comment #3 from AK <hiraditya at msn dot com> ---
This bug appears to be a duplicate of bug61929. The fix (r225942) has been
pushed today. Please verify if this fixes your problem.

[Bug middle-end/64394] ICE: in build_linearized_memory_access, at graphite-interchange.c:121 (isl_constraint.c:558: expecting integer value) with -floop-interchange

[spop at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64394

Sebastian Pop <spop at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |spop at gcc dot gnu.org
         Resolution|---                         |FIXED

--- Comment #5 from Sebastian Pop <spop at gcc dot gnu.org> ---
We removed the code causing the issue.