[Bug tree-optimization/15777] Fold does not always fold sub trees

[kazu at cs dot umass dot edu]

------- Additional Comments From kazu at cs dot umass dot edu  2004-07-04 08:05 -------
As of today, t54.tailc looks like so:

f (x)
{
  int T.4;
  unsigned int T.3;
  unsigned int T.2;
  unsigned int x.1;
  int T.0;

<bb 0>:
  x.1_2 = (unsigned int)x_1;
  T.4_5 = (int)x.1_2;
  T.0_6 = T.4_5 & 1;
  return T.0_6;

}

Except the redundant conversions, the generated code at tree level
is optimal.  Probably the patch to have exactly one return statement
fixed this problem.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15777