[Bug tree-optimization/66422] -Warray-bounds false positive with -O3

["rguenth at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66422

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2015-06-08
                 CC|                            |hubicka at gcc dot gnu.org
          Component|c                           |tree-optimization
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
Ok, so we peeled the loop like

run_foo ()
{
...
  <bb 10>:
  _33 = myfoo_28->buf[3];
  if (_33 != 1)
    goto <bb 13>;
  else
    goto <bb 11>;

  <bb 11>:
  _34 = (int) _27;
  if (_34 > 4)
    goto <bb 12>;
  else
    goto <bb 13>;

  <bb 12>:
  __builtin_unreachable ();
  _35 = myfoo_28->buf[4];

  <bb 13>:
  # _36 = PHI <0(2), 1(3), 0(4), 1(5), 0(6), 1(7), 0(8), 1(9), 0(10), 1(11),
0(12)>
  if (_36 != 0)
    goto <bb 15>;
  else
    goto <bb 14>;

  <bb 14>:
  i_37 = 1;

  <bb 15>:
  # _2 = PHI <0(13), -1(14)>
  return _2;

and the unreachable () remains in the CFG.

Honza - it seems that remove_exits_and_undefined_stmts inserts these
unreachable calls but fails to split the BBs.

I have a patch.