> Further more interleaving soft-float and hard-float code was promblematic
> due to the way in which floating point values where returned from
> functions. (with 2.8.x I believe a new APCS variation is supported that
> will allow the floating point value to be returned in both integer and fp
> registers to allow hard-float and soft-float code mixing).

Does gcc 2.8.x support NetBSD/arm32?  I've heard rumours that it
doesn't...  (or was that binutils 2.8.x?)

> If one wished to utilise hard-float on the ARM7500FE there are two
> choices. Rebuild everything with -mhard-float or modify libc so that the

What is everything?  Does this mean all the libraries in /usr/lib or just
some key ones?  If I just wanted to compile a particular program (say
RLaB?) with hard-float would that work?  Or am I going to get problems
when it comes to linking with the standard libraries?  

For the time being, I'm experimenting:  I've written a script called gcc
which forces -mhard-float.

> soft-float functions i.e. the basic FP arithmetic is actually done via the
> FP instructions. Whilst this would not be quite as fast as the first

At 48MHz, I don't have much processing to spare!! ;-)

> method it would be a lot simpler to implement if someone really wanted to
> try and use heavy FP.

Which I do...

If I recompiled "everything" (which I'm more than prepared to do), is
there anyone else out there with hardware FP that could/would use them, or
should I just keep them to myself?

> As a matter of interest has anyone benchmarked an ARM7500FE doing hard-fp
> against an SA110 doing soft-fp ?

Well, before I bought my machine I had it on good authority that (under
RISCOS) the ARM7500FE kicked the SA110 by a factor of six when computing
fractals.  I've no idea what it's like with RBSD's soft-float.  If
anyone's got an FP benchmark prog, I'm more than willing to try it out.

Thanks,
David.