I think setting up the XTCF to use LBA is that part that I'm missing. I haven't figured out how to calculate the CHS for a given CF card from Linux (or at least how it will be seen on the XT-IDE), but once I can get it from an actual machine, it's easy enough to go back to my linux box and plug...
Both of mine are stock with whatever shipped, so I'll definitely have to upgrade XUB and see if that helps with something I don't already know the CHS for.
I was able to build a setup in linux to generate CF images I can dd to a real CF card, but it's a PITA. I've have a DOS setup I like to...
YMMV but unless you have the exact CHS geometry, this doesn't work for me. You end up creating an image that an XT-IDE won't boot, but will still be viewable on a real hardware if booted from something else. I had issues with the reverse too: taking a raw CF image and letting PCem or 86box...
It's easiest if you can boot from floppy on the machine in question and at least partition, format and run sys on it there. Then you can bring the CF card over to your (presumably Windows) PC and copy whatever you want on it.
Getting the geometry right can be a PITA without using the actual XT...
Sorry, I can't help much but I have one of these boards (the first version) and it has a BIOS ROM with this same sticker but yellow. Looks original to me.
I took the easy way out and booted from a DOS 6.22 floppy and set it that way :) I was using David_M's clock.com but that didn't seem to work correctly with GLaTICK though I didn't spend enough time to figure out whether it was a real problem or operator error.
Any way to patch PCDOS to address this issue? I have GLaTICK running on my Compaq Portable with an AST MegaPlus II (running PCDOS 2000 with a V20) and was scratching my head on why it didn't quite work!