I'm looking into options for speeding up floating point calculations (currently using the default libgcc routines), and I ran accross a mention of using libfp to save code space -- but I can't find any documentation on it.
1) I gather libfp is smaller, but is it also faster?
2) Do NaN and Inf propogate?
3) Are NaN and Inf generated properly for things like 0.0/0.0
and 1.0/0.0?
--
Grant Edwards grante Yow! I have many CHARTS
at and DIAGRAMS...
visi.com
