[Bug rtl-optimization/68695] [6 Regression] Performance regression related to ssa patch / ifcvt

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68695

--- Comment #21 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
int
foo (int x, int y, int a)
{
  int i = x;
  int j = y;
  if (__builtin_expect (x > y, 0))
    {
      i = a;
      j = i;
    }
  return i * j;
}
shows that even without if IRA does the right job we get the same code as
before.
But the #c20 testcase does (or should have) approx. equal frequencies of the
bbs.