View Single Post
Old 29th Sep 2019, 12:18 am   #237
JohnBHanson
Heptode
 
Join Date: Aug 2009
Location: Worthing, Sussex, UK.
Posts: 661
Default Re: Gemini 80-Bus System

I am getting someway. Looks like the two bits D6 and D5 on port e5 which are specified as indeterminate are critical for the Multiformat Bios to run.. I have also spotted some odd coding
problems. Some subroutines take parameters after the call. That is ok but if a conditional call is used
and the call is not taken then the parameters are executed as instructions. Also at one point it did an inc (ix-2) which incremented the return address on the stack to the middle of an instruction. The second byte of the instruction happened to be 76 which is halt.

I have created two additional scsi controllers (829m and 849m) that have the correct (I hope) behaviour for running MFB. I suspect these were meant to run on a modified card as a way of doing copy protection.

I have even wondered if my emulator is causing the problem in that the emulation is not perfect - but gemini code does some strange things like aborting transfers part way through.

Still getting closer all the time...
JohnBHanson is offline