[Bug tree-optimization/66592] New: [Regression] ICE in vect_get_def_for_operand

[Bug tree-optimization/66592] New: [Regression] ICE in vect_get_def_for_operand

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

            Bug ID: 66592
           Summary: [Regression] ICE in vect_get_def_for_operand
           Product: gcc
           Version: 6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: kargl at gcc dot gnu.org
  Target Milestone: ---

[Bug tree-optimization/66592] [Regression] ICE in vect_get_def_for_operand

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

--- Comment #1 from kargl at gcc dot gnu.org ---
This bug appears to be regression and it requires the -O2 and
-ftree-vectorize options.

gfortran -c -O2 -ftree-vectorize z.f
z.f:1:0:

       subroutine calsvp(nkr,nkz)
^
internal compiler error: in vect_get_vec_def_for_operand, at
tree-vect-stmts.c:1478
0xd6a2a9 vect_get_vec_def_for_operand(tree_node*, gimple_statement_base*,
tree_node**)
        ../../gcc6/gcc/tree-vect-stmts.c:1478
0xd72a0d vectorizable_store
        ../../gcc6/gcc/tree-vect-stmts.c:5330
0xd7d6a6 vect_transform_stmt(gimple_statement_base*, gimple_stmt_iterator*,
bool*, _slp_tree*, _slp_instance*)
        ../../gcc6/gcc/tree-vect-stmts.c:7496
0xd9a541 vect_schedule_slp_instance
        ../../gcc6/gcc/tree-vect-slp.c:3481
0xd9ab16 vect_schedule_slp(_loop_vec_info*, _bb_vec_info*)
        ../../gcc6/gcc/tree-vect-slp.c:3551
0xd82c17 vect_transform_loop(_loop_vec_info*)
        ../../gcc6/gcc/tree-vect-loop.c:6218
0xda0792 vectorize_loops()
        ../../gcc6/gcc/tree-vectorizer.c:495
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.

[Bug tree-optimization/66592] [Regression] ICE in vect_get_def_for_operand

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

--- Comment #2 from kargl at gcc dot gnu.org ---
Created attachment 35805
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35805&action=edit
Fortran code that exhibits the problem

[Bug tree-optimization/66592] [Regression] ICE in vect_get_def_for_operand

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

--- Comment #3 from kargl at gcc dot gnu.org ---
I forgot to mention in comment 1 that the COMMON statement
in the Fortran is also required.

[Bug tree-optimization/66592] [6 Regression] ICE in vect_get_def_for_operand

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |matz at gcc dot gnu.org,
                   |                            |rguenth at gcc dot gnu.org
   Target Milestone|---                         |6.0
            Summary|[Regression] ICE in         |[6 Regression] ICE in
                   |vect_get_def_for_operand    |vect_get_def_for_operand

[Bug tree-optimization/66592] [6 Regression] ICE in vect_get_def_for_operand

[matz at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

--- Comment #4 from Michael Matz <matz at gcc dot gnu.org> ---
Can't reproduce with r224605 and r224647.  Can you update and retry?

[Bug tree-optimization/66592] [6 Regression] ICE in vect_get_def_for_operand

[sgk at troutmask dot apl.washington.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

--- Comment #5 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Fri, Jun 19, 2015 at 11:49:55AM +0000, matz at gcc dot gnu.org wrote:
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592
> 
> --- Comment #4 from Michael Matz <matz at gcc dot gnu.org> ---
> Can't reproduce with r224605 and r224647.  Can you update and retry?
> 

Must be a hessianbug.  With top of tree, I don't see
the issue.  Sorry about the noise.

[Bug tree-optimization/66592] [6 Regression] ICE in vect_get_def_for_operand

[kargl at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66592

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID

--- Comment #6 from kargl at gcc dot gnu.org ---
Closing this as invalid.  With an unpatched and up-to-date tree,
I cannot reproduce the error. :-/