Re: Alternate FP lib for GCC?

["Jurgis Armanavichius" <jurgis at medelkom dot com>]

Hello,

>   > William Gatliff writes:
>   >  > Question: what is gcc/floatlib.c for, if FP emulation is really in
>   >  > config/fp-bit.c?
>   >
>   > It's an older implementation.
>   > Maybe it's kept around for reference sake.  Dunno.
> It was an old implementation and isn't used by any port that I'm aware of.
>
> jeff

May be it is older implementation but it REALLY works faster :-)
You can see some testing results (for Hitachi's H8 target) at URL:

http://www.nettilinja.fi/~ankosken/releases/h83relnot.htm

I'm using this library for H8/300H instead of standard "fp-bit.c", and
my FP arithmetic (and, of course, other mathematic functions) works
2 - 2.5 times faster! Really! It is very important for me. I'm using single
precision arithmetic, and know nothing about double precision, but
IMHO double FP will be 1.5 - 2 times faster also.

Regards,

Jurgis Armanavichius
Firma MEDELKOM, http://www.medelkom.com



------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sourceware.cygnus.com