[Bug rtl-optimization/51938] missed optimization: 2 comparisons

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51938

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |missed-optimization
   Last reconfirmed|                            |2012-01-23
          Component|tree-optimization           |rtl-optimization
     Ever Confirmed|0                           |1
           Severity|normal                      |enhancement

--- Comment #1 from Richard Guenther <rguenth at gcc dot gnu.org> 2012-01-23 10:23:31 UTC ---
Confirmed.

Something for PHI-OPT, recognize

<bb 2>:
  if (x_2(D) > 0.0)
    goto <bb 5>;
  else
    goto <bb 3>;

<bb 3>:
  if (x_2(D) < 0.0)
    goto <bb 4>;
  else
    goto <bb 5>;

<bb 4>:

<bb 5>:
  # D.1719_1 = PHI <1(2), -1(4), 0(3)>
  return D.1719_1;

I suspect also worthwhile for integral types.  Note that for real types
you need -ffinite-math-only - I bet the clang result is wrong for NaNs.

Btw, what's the optimal assembly you expect?

Not sure what's the best way to represent this on the tree level either,
so eventually we have to resort to RTL if-conversion (doesn't work
there for integral types either).