Re: [f-cpu] IEEE FP exceptions

[Michael Riepe <michael@stud.uni-hannover.de>]

On Sun, Mar 02, 2003 at 06:50:54AM +0100, Yann Guidon wrote:
[...]
> >>hoping to revive this list,
> >>    
> >>
> >I guess that was enough revival for a single mail, wasn't it? ;)
> >
> yup and i already regret it :-)

You asked for it, you got it ;)

> The FP stuff is ok because it's "future work" and the integer part is 
> not even finished.

I've lost track -- what's missing except eu_div?

> However, the redesign of the conditions is less amusing because i am not
> in a mood for digging into F-CPU details :-/
> I am well aware that the "zero" and "msb" conditions contain problems
> but i can live with them and compiler problems are ... compiler problems.

It's not only a `compiler problem', it's a code generation problem
in general:  You'll also have to include extra masking (or shift)
instructions when you write in assembler.

> It's already a big chance that the whole architecture is still alive 
> after almost 4 years.

Alive but not unchanged.  And we're still learning what's good and what's
not so good, which instructions to use and which to avoid -- working on
gcc and fctools was rather enlightening sometimes.  This feedback will
help us to refine the architecture, and close all the open holes in it.

In fact, I wish we would get more feedback from SW people.  That's the
reason why I finished the assembler/disassembler/linker/emulator quartet
before I go hacking VHDL again (it's about time for eu_div, isn't it?).

-- 
 Michael "Tired" Riepe <Michael.Riepe@stud.uni-hannover.de>
 "All I wanna do is have a little fun before I die"
*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/