Re: [MiNT] Still don't get it - difference between 68020 and 68020-60

[Alan Hourihane <alanh@fairlite.co.uk>]

On Sun, 2010-06-20 at 15:04 -0400, Keith Scroggins wrote:
> > Realistically we have....
> >
> > 68000 (Stock ST)
> > 68000+68881 (Mega STE with FPU)
> > 68020 (Stock ST with accelerator)
> > 68020+68881 (Stock ST with accelerator with FPU)
> > 68030 (Stock Falcon)
> > 68030+68882 (Stock TT)
> > 68040 (Afterburner / Hades 040)
> > 68060 (CT60 / Hades 060)
> 
> If we want all these targets, who is going to do the work to add all this 
> additional support in MiNTLib?  I guess it is not that much, but it is 
> needed before GCC, IMHO.

It's trivial for MiNTlib, but we need GCC. Maybe chicken and egg, but I
think GCC is first up.

> I, also, see no point for 68020 and 68020/68881 if we are doing 68030 and 
> 68030/68881, no stockAtari includes this CPU, and the rare machines that 
> have them, not work the additional 2 targets IMHO.  So, we'd have:

68030/68881 is the TT. 68020 is only on accelerators for the ST/STe's,
so these maybe few and far between, regardless of FPU, but are around.
If GCC support exists, then the package maintainers and developers have
a choice to use it or not. You don't have to build 68020 into RPMs, but
at least the base GCC compiler will be able to out of the box.

> m68kmint (68000)
> m68881 (68000/68881)
> m68030 (68030)
> m68030fpu (68030/68882)
> m68040 (68040)
> m68060 (68060)
> m5475 (v4e)
> 
> And this will definitely change the approach to building RPM packages, 
> which will most likely end up being named like intel then (like above).
> 
> That is 7 targets.  What about 68020-60?  Is there still a point to this?

Again, yes, there probably is. People who want to compile code to run on
a range of CPU's.

For me this discussion is about making GCC do what is capable, not the
fact whether packages, package maintainers, or developers make use of
that support or not. But it gives that flexibility when the time arises.

Alan.