A4091 SCSI controller configuration issues

[codewiz]

The GUI to configure the A4091 SCSI controller (and possibly others) has several issues:

• The SCSI ID of the controller defaults to 0 rather than the default value 7 (all DIP switches off).
• The drop-down to change the SCSI ID is hard to find and has no label
• Selecting an alternative firmware ROM with "..." sometimes causes the controller to be deactivated (the * mark on the left of the name disappears).
• Changing any of the additional settings also causes the A4091 board to deactivate (the * disappears)

Minor issues:

• It is not clear what the drop-down with numbers from 1 to 4 does. Is it the Zorro slot number? Can I install multiple A4091 cards?

Context:

• OS: Fedora Linux (Flatpak)
• Version 7.0.5

[codewiz]

Attaching the ROM I'm trying to use with the A4091 controller. Note that it's a 64KB ROM, while the original ROM was 32KB. The real hardware takes both types (with a jumper).

a4091-4231-fixed.zip

[codewiz]

This is what gets written to the config file:

a4091_rom_file=/home/bernie/emu/amiga/kick/a4091-4231-fixed.rom
a4091_rom_options=id=0,fastbus,syncmode

The parameter id=0 should be id=7, and I had to edit a4091_rom_file by hand to use my ROM.

With these changes, the A4091 works and can boot off an attached hard-drive.

[midwan]

This area is basically coming from WinUAE, which I agree, doesn't have the most intuitive approach.
The dropdown next to the controller (1-4) relates to the slot.
There is another dropdown above the ROM one, which is the Controller ID (0-7). Did you try changing that one? It's only enabled after you select a ROM.

[midwan]

After a quick test here, it seems that using that dropdown does the expected thing:

a4091_rom_file=/home/midwan/Amiberry/roms/harddrive/A4091 v40.13.rom
a4091_rom_file_id=54CB9E85,A4091 v40.13
a4091_rom_options=id=7

(I used one of the 32K ROMs)

[codewiz]

Try selecting the .rom file I attached using the "..." file browser: the star on the side of the A4091 entry should disappear.

[midwan]

Hmm, true - but the same thing happens in WinUAE as well. So maybe this is an upstream issue, I'll have to check with Toni.