fmax(+0,-0)

[Andreas Jaeger]

What is the goal of our fmax implementation?  Should it return +0 for
fmax(+0,-0)?  I think the current C implementation :

double
__fmax (double x, double y)
{
  return (isgreaterequal (x, y) || isnan (y)) ? x : y;
}

returns the first argument for both fmax(+0, -0) and fmax(-0,+0)
instead of +0 in both cases.

Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj