> > i = detect_MMMX();
> > if !i then usenormalstuff=1;
> > else usemmx=1;
> >
> > then had code for both cases???
>
> no....
>
> If you can assume the MMX functions are there then EACH application
> doesn't have to emulate all of the MMX stuff itself if it doesn't have
> the REAL hardware.
>
> This should be treated EXACTLY like the Floating Point hardware... at
> least in my opinion.
The *ONLY* benefit from having MMX emulation like FPU emulation,
would be that apps that use MMX would still FUNCTION on machines
that don't have it. Such an app would be incredibly slow
however. MMX is about speed. Emulating it is about "slow".
The two conflict greatly. Intel has an MMX microcode update that
is downloadable. This puts REAL MMX into some of the CPU's that
didn't have it. If that doesn't work, then it's time to upgrade
if you need to run MMX IMHO. Or use an app that doesn't have MMX
(since the non-MMX app will be faster than emulated MMX).
How many applications currently exist for Linux that even USE MMX
at ALL?
I would guess very few if any, and I'd also guess that they have
code for running non-MMX as well.
I think that it is an application land issue, however I'm sure
someone will write an MMX emulator for the kernel nonetheless.
It's fine by me, so long as someone gets benefit out of it, but
it certainly isn't going to give any speed to any apps. I
personally wouldn't use the FPU emulation code either, but that
is a different issue...
TTYL
-- Mike A. Harris - Computer Consultant - Linux advoacateEscape from the confines of Microsoft's operating systems and push your PC to it's limits with LINUX - a real OS. http://www.redhat.com
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu