[ CC'ed to port-mac68k for -maybe- general interest ]

At 20:04 Uhr +0200 03.05.1997, Michael R Zucca wrote:
>> From looking at the sources, I can say the SWIM III doesn't have anything
>> to do with the IWM/SWIM. The SWIM III is DMA and interrupt enabled and
>> obviously {de,en}codes GCR as well as MFM internally, i.e. its programming
>> model looks pretty similar to that of any vanilla PC controller. The IWM &
>> SWIM don't do DMA nor interrupts, and they lend no hand in decoding the raw
>> disk data stream, either. Ah, well...
>
>Don't the AV's use a SWIM-III?

As they are (afaik...) doing floppy I/O by DMA, I'd assume they either use
some IIfx IOP alike design or - more likely - an "IWM III" alike design.
Either way, floppy access from an AV machine has not much to do with the
"classic" IWM/SWIM (/Apple II) way.

>Maybe you can support those machines?

I don't have access to an AV Mac. Anyhow, you'd have to have a second
driver (like in the ncrscsi/sbc versus esp case) for the IWM III as it's an
entirely different chip. And with the IWM III code in the Linux kernel
sources (and, if it is there, even better in the MKLinux sources), this
should be a far easier task than the original IWM access.

>Oh well. I suppose the only info is hidden in the assembly of the .Sony
>driver.

And deep down in the archives of Ma Apple...

>I've often wondered if MkLinux could be ported to the AV Macs because
>they share so much hardware with the x100 PowerMacs.

Sounds like the sun3x vs. sparc ports of NetBSD; apparently Sun tried out
most of the SPARC/Sun4 hardware design before they switched processor
architectures.

	hauke


--
"It's never straight up and down"     (DEVO)