Re: TI-H: RE on 68020


[Prev][Next][Index][Thread]

Re: TI-H: RE on 68020



On Thu, 21 Nov 1996, Isaac Salpeter wrote:


> 
> When you get it to run 68020-specific instructions, you will convince me.
> I'm surprised nobody has yet tried this. It's the only way to know for
> sure.
> 


Actually I've tried this. The A68K included in the fargo archive is 
only MC68000 so it can't compile any 68020 instructions. So instead of 
020 instructions I tried this:


move.l d0,$4441


and the TI92 said "address error". This might not prove anything 100% 
but it strongly suggests that the CPU is a 68000, which is not capable 
of addressing words or longs at odd addresses. 68020 on the other 
should have no problems with that one.


Right now I don't have a TI92 to test this with, but if somebody has 
the time you could check the result of this:


prog_code:
dc.l $4C001001           ; mulu.l d0,d1
dc.l $20301A00           ; move.l (a0,d1.l*2),d0
rts


The dc.l's are 68020 instructions. If the CPU is a 020 this should 
work well. If it's a 000 the TI92 should say "illegal instruction".




---------
Toni "Panadol" Kettula
reijok@kontu.bbs.fi


References: