On Thu, Aug 04, 2005 at 09:45:42AM +0800, Feng Wang wrote:
> __float80 a;
> __float128 b;
> a = 2.0;
> b = 2.0;
> printf ("%f, %f\n", a, b);
Certainly you shouldn't expect this to ever work. You
need to use %Lf for long double, and there is no standard
flag letter for __float128.
r~