[Bug tree-optimization/71477] [7 Regression] gcc ICE at -O3 on valid code on x86_64-linux-gnu with “seg fault”

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71477

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2016-06-10
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot 
gnu.org
   Target Milestone|---                         |7.0
            Summary|gcc ICE at -O3 on valid     |[7 Regression] gcc ICE at
                   |code on x86_64-linux-gnu    |-O3 on valid code on
                   |with “seg fault”            |x86_64-linux-gnu with “seg
                   |                            |fault”
     Ever confirmed|0                           |1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.

#4  0x00000000011666b3 in vect_transform_loop (loop_vinfo=0x27db8b0)
    at /space/rguenther/src/svn/trunk/gcc/tree-vect-loop.c:6951
6951        = wi::udiv_floor (loop->nb_iterations_likely_upper_bound + 1,
(gdb) l
6946        }
6947      loop->nb_iterations_upper_bound
6948        = wi::udiv_floor (loop->nb_iterations_upper_bound + 1,
6949                          vectorization_factor) - 1;
6950      loop->nb_iterations_likely_upper_bound
6951        = wi::udiv_floor (loop->nb_iterations_likely_upper_bound + 1,
6952                          vectorization_factor) - 1;

(gdb) p loop->nb_iterations_likely_upper_bound
$2 = {<fixed_wide_int_storage<192>> = {val = {0, 0, 0, 0}, len = 0}, 
  static is_sign_extended = <optimized out>}

so likely_upper_bound is not initialized.  I have an obvious patch.