I had this running in March of 1987.
One document lists many hardware addresses, and as I no longer have anything but a paper listing, I will transcribe some of them here. I'll note that this describes a "Unistar CTS-300" cpu board which seems to be some alias for the Pacific Microsystems PM68K
I found my binder full of manuals, so I can augment the following now with some authority.
Note that all addresses 0x20_0000 and above are restricted to supervisor access only.
SUNADDR 0x60_0000 // dual uart (7201) "sunio" CLKADDR 0x80_0000 // 9513 clock (5 timers) MBIO 0x1F_F800 // multibus IO base MBIO 0x1F_8000 // multibus buffers PAGE 0xA0_0000 // page map SEG 0xC0_0000 // segment map CONTEXT 0xE0_0000 // select one of 16 contexts CONFIG 0xE0_0000 // read as 16 bit input port Onboard ram (256k) 0_0000 to 03_ffff Multibus ram 10_0000 to 1e_ffff (we had 576K) Multibus io 1f_0000 to 1f_ffff Rom pair 101,103 20_0000 to 23_ffff Rom pair 102,104 40_0000 to 43_ffffThere must have been some monkey business to map the first ROM pair to address 0 for the first few cycles so SP and PC could get fetched from it at reset.
Interestingly, the on-board RAM is not dual ported. This means that multibus DMA cannot take place into on-board ram. This means that the Callan hard drive controller will not work without an additional ram card that supports multibus DMA.
From my notes, apparently the board was shipped with an interactive ROM monitor in the auxiliary rom pair. This has been removed and replaced with a pair of roms that serves only to boot from the Callan winchester or floppy boards.
We had a pair of multibus ram boards. to augment the onboard ram. One with 512k, the other (partially populated with ram chips) gave 64k
The winchester controller had a base address of 0x00A0
The floppy controller had a base address of 0x0100
Our hard drive had 320 cyl, 8 heads, 17 sectors per track
It is a Rodime RO200.
320*8*17 = 43,520 sectors = 21.25 megabytes
I have notes on the actual disk partitioning.
Tom's Computer Info / tom@mmto.org