68060?

From: Mikey <mgrove_at_iwvisp.com>
Date: Thu, 09 Mar 2000 23:06:16 -0800

>All,

> If I recall, Mike (another Mike, not me :) was attempting
>to fit an 060 onto the Afterburner. Mike, would you mind telling
>me how far you had gotten on it? I may get an opportunity to
>borrow a 68060 emulator, and was wondering if that would do any
>good....


>Michael White (michael_at_fastlane.net)


End of Original message

Hi Michael

That would be me. I could go on and on about what I have learned the
last year or so, but I can't type that much:) I will try to hilite
where I am at.

I use an Interconnect Socket adapter that plugs into the 040 socket on
the Afterburner. It provides 3.3V for the 060, a delayed C clock in
that the 060 uses internally for bus timing, is configured to run the
060 at 1/2 bus speed, and conditions some of the pins on the CPU that
are more specific for an 060 designed system and O/S (basically makes
it more 040 compatible).

One of the problems is the 060 does not provide an output signal from
the MI pin as the 040 does (there is no MI pin on the 060). The A/B
utilizes this signal from the 040 for some sort of timing. It's (MI)
function is to hold off assess to memory until the 040 decides if
snooping is enabled, and if so how to snoop the bus. Still, even if
snooping is disabled, MI is still asserted inbetween bus cycles. As
the A/B does not snoop the bus, I assume that it uses the MI signal to
detect the end of a bus cycle. I am not to sure on this point, but am
without a bus analizer. MI is also asserted during a reset. I emulated
the signal using a truth gate. The reset,TS and TS signals were used
to do this. The 040 would boot with the truth gate attached, but was a
little buggy telling me that the timing isn't just right. Usually
problems were seen at startup, but during normal operation, the
machine functions normally. I figure that this is enough to think that
it would work with the 060, and indeed with the 060 installed, the
machine will boot off a Falcon Diagnostic Cartridge and perform
diagnostic test. Motorola instructions say that in systems that
utilize the MI pin tie it HI when installing an 060 into an 040
system, but they assume the system is utilizing it for snooping
functions, and I do not believe that to be the case with the A/B.
Ideally, to emulate the MI signal inbetween bus cycles and during
reset would allow the A/B to function properly. I don't know yet how
to do this. Tieing the MI pin HI on the A/B will not allow it to boot.

I think what is the main problem in keeping the 060 from booting from
TOS ROM is an illegal 060 instruction around line 6 of TOS. MOVEP is an
illegal instruction on the 060, and at startup, it apears in TOS to
initailize the MFP. I talked to the Milan people, and they too had
this problem. Not only in TOS, but HD Driver as well had the
instruction. It is not present in all software, but is produced by
some compilers. They elimated it in their TOS, and use Motorolas MOVEP
emulator in the autofolder to elimate the problem.
If I was a coder (which I am not, more of a hacker) I would write a
boot sequence that would start the Falcon up from the cartridge port,
install the emulator, then start TOS from ROM. I am sure there is
more to it than that, or others on the mailing list would have helped
out, or some of the coders in the newsgroup would have helped out. So
far, no takers. The other option would be to write a "patcher
application" like Doug Little did for the keyboard fix and just let it
find and replace instances of MOVEP in TOS with a compatible routine.
I have an EPROM ready to burn a TOS image into if someone ever is able
to do this (or in a couple of years when my programming skills are a
little more useful than they are now:). I have at least been able to
hack into the boot program that loads TOS 4.92, and have been playing
around booting the 040 off of floppy. I kind of got off into changing
the boot logo, but at least it is experience:).

I have offered to let a couple people borrow one of my adapter sockets
and an 060, but no one is interested at this time. Peter S. said he
was working on an 040 compatible TOS to burn into ROM, but was waiting
for Magnus to supply the EPROM. He said he now has an adapter for his
burner to burn the ROM, but still no EPROM, and is also working on his
house. I haven't heard from him in a while.

As far as the 060 emulator is concerned I am not sure. If you could
execute or step TOS through the emulator and find problems, that would
be great. I just don't have the background for all this, it has been
like wearing a blindfold in a cow pasture, only looking to see what I
have done after I steped on something.

Sure wish Doug was around......

Best Wishes,

Michael Grove
Received on to. april 13 2000 - 15:25:57 CEST