John-
	Thanks for the information.  I hadn't realized that the buggy
LC040s were fine as long as they weren't doing floating-point emulation.
I'd seen the information that a particular instruction was a problem, and
hadn't realized that this instruction was unique to FPE and that
soft-float completely avoided it.
	Out of curiosity, does anybody know how MacOS deals with the
problem?  I'm still wading through my copy of Inside Macintosh, so I
haven't had a chance to look at how exactly SANE works.  I assume that if
there's no FPU, MacOS effectively always runs without using floating
point?

On Thu, 13 Nov 2003, John Klos wrote:

> > Second, assuming that it's still the case, does anybody have information
> > on what exactly fails?  Most importantly, is there a quick way to test if
> > a system has a good or bad mask revision?  Last I knew, the easiest way to
> > test was to run SoftwareFPU or just try NetBSD.  Is there any better
> > method out there now?  Assuming there isn't, does anybody have some leads
> > I could follow to figure out what exactly is failing and a way to easily
> > test it?
>
> Assume it's the buggy one, maybe? I have a stack of 15 LC040s, mostly from
> Quadra 605s and Quadra 630s, and ALL of them are the buggy masks. I have
> never seen the supposedly non-buggy mask revision.

	Really?  Interesting.  I think I have one (maybe my 640?)
with the supposedly safe mask revision, but I'm not certain of that.
I'll let you know if I do find it...  However, I seem to recall some
people being able to run it successfully (other than the FPE bugs) on
LC040 machines.
	Mostly, though, I was curious if anybody is fairly familiar with
the specific instruction that fails.  If I remember correctly, Ken Nakata
was involved with a thread that was discussing it, but I think that was a
fair time ago.  I don't know what he's up to now or if anybody else is
continuing his work, so I was mostly hoping to see if anybody was still
investigating the processor itself rather than working on the soft-float
version to avoid duplicating effort.

> Yes, we need documentation. So, in a nutshell, we have NetBSD 1.6.1
> (release from August) compiled with soft-float patches for mac68k and for
> Amiga. We also have 2,000 soft-float binary packages. And we plan to have
> NetBSD 1.6.2 snapshots when that is ready, too.

I take it that the process with soft-float at this point is, basically, to
compile anything and everything with the -msoft-float option?  Is there
any way I can help?

> Does anyone want to help design a web page with documentation and links?

	Is any of the www.macbsd.com documentation still being updated?
Alternately, is there a newer site where the same sorts of documentation
are being kept?  The current version of the FAQ still has a number of
links to macbsd, and since the install document references the FAQ, that's
probably a sort of bad thing.
	I'm willing to at the very least try to update some of what was on
macbsd.com...  I don't have an ideal site to host it for the long run,
although I've a development server where I can leave pages for people to
check over for errors.


	-Seth