float_abs() again: Guido pointed out that this could screw up in the

presence of NaNs.  So pass the issue on to the platform libm fabs();
after all, fabs() is a std C function because you can't implement it
correctly in portable C89.

Objects/floatobject.c

@@ -568,12 +568,7 @@ float_pos(PyFloatObject *v)
 static PyObject *
 float_abs(PyFloatObject *v)
 {
-	if (v->ob_fval < 0)
-		return float_neg(v);
-	else if (v->ob_fval > 0)
-		return float_pos(v);
-	else /* ensure abs(-0) is +0 */
-		return PyFloat_FromDouble(+0.0);
+	return PyFloat_FromDouble(fabs(v->ob_fval));
 }
 
 static int