[Bug middle-end/49545] [4.7 Regression] New C++ test failures

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49545

--- Comment #2 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-06-27 
13:36:03 UTC ---
An anyway useful transform would be to hoist the call in

  iftmp.0_15 = *D.2099_14;

<bb 4>:
  # iftmp.0_1 = PHI <foo(2), iftmp.0_15(3)>
  iftmp.0_1 (&a);

based on the fact that on the edge 2->4 it will be a direct call.