<html>
    <head>
      <base href="https://bugs.llvm.org/">
    </head>
    <body><table border="1" cellspacing="0" cellpadding="8">
        <tr>
          <th>Bug ID</th>
          <td><a class="bz_bug_link 
          bz_status_NEW "
   title="NEW - Failure to optimize integer to float conversions that can never fault under -ftrapping-math"
   href="https://bugs.llvm.org/show_bug.cgi?id=51921">51921</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>Failure to optimize integer to float conversions that can never fault under -ftrapping-math
          </td>
        </tr>

        <tr>
          <th>Product</th>
          <td>libraries
          </td>
        </tr>

        <tr>
          <th>Version</th>
          <td>trunk
          </td>
        </tr>

        <tr>
          <th>Hardware</th>
          <td>PC
          </td>
        </tr>

        <tr>
          <th>OS</th>
          <td>Linux
          </td>
        </tr>

        <tr>
          <th>Status</th>
          <td>NEW
          </td>
        </tr>

        <tr>
          <th>Severity</th>
          <td>enhancement
          </td>
        </tr>

        <tr>
          <th>Priority</th>
          <td>P
          </td>
        </tr>

        <tr>
          <th>Component</th>
          <td>Scalar Optimizations
          </td>
        </tr>

        <tr>
          <th>Assignee</th>
          <td>unassignedbugs@nondot.org
          </td>
        </tr>

        <tr>
          <th>Reporter</th>
          <td>gabravier@gmail.com
          </td>
        </tr>

        <tr>
          <th>CC</th>
          <td>llvm-bugs@lists.llvm.org
          </td>
        </tr></table>
      <p>
        <div>
        <pre>bool f(float x)
{
    return x == 0;
}

With -O3 -ftrapping-math, this gives:

f(float):                                  # @f(float)
        xor     eax, eax ; eax = 0
        cvtsi2ss        xmm1, eax ; xmm1 = float(eax -> 0)
        ucomiss xmm0, xmm1
        setnp   cl
        sete    al
        and     al, cl
        ret

(though this is not unique to x86 and I can reproduce it on platforms like
powerpc64le)

Conversion of 0 to a float can never be inexact, so it seems to me as though
the conversion is unnecessary here. This should also be possible for all
constants that exactly convert to a float/double/long double in equivalent
situations (though from my reading of the IEEE-754 standard it seems the
conversion should be kept under `-ftrapping-math` if the conversion is
inexact).

As such, it should be possible to optimize this to `return x == 0.f;`. This is
done by GCC, but not by LLVM.

Godbolt comparison here: <a href="https://godbolt.org/z/6YPMEGfex">https://godbolt.org/z/6YPMEGfex</a>
(if someone knows how to do alive2 checks for stuff that involves trapping
floating point operations, I would be glad to know)</pre>
        </div>
      </p>


      <hr>
      <span>You are receiving this mail because:</span>

      <ul>
          <li>You are on the CC list for the bug.</li>
      </ul>
    </body>
</html>