Hi, after finding the floating point calculations implementation for fpc[1] I'm asking myself if it is possible to force fpc to use it although the CPU does actually have a FPU built in. How can this be done? [1] /usr/home/marc/fpc-2.2.2/share/src/fpc-2.2.2/rtl/inc/softfpu.pp -- Marc Santhoff <M.Santhoff at web.de>