Re: [Freemint-discuss] Disassembly of TOS 1.00 with Ghidra

[Thorsten O. <ad...@th...>]

VDI now also finished. Also some minor tweaks to BIOS (protobt writes 0xf8 
media byte for SS, 80 tracks instead of 0xf9).

The "make check" target still does not work, because bios,gemdos+vdi have to 
be combined into one part, and aes+desktop into another to get the initialized 
data where it is located in the ROMs. But you can use dd to extract those 
parts (up to address 0xfd91d0 of the french ROM, the function gemstart of the 
AES), and compare that to the original.

Only works for the french version so far (still have to figure out where the 
german version differs), but already includes the resource data of the 
desktop.

The sources are a strange mix: the assembler sources gsxasm1.S and gsxasm2.S 
are now mostly identical to the lisa sources, while the rest is mostly 
identical to the 1.04 sources (expect in places where blitter support was 
added).

Apropos lisa sources: in those, there are two settings used, vme and rev_vid. 
The readme for this (https://github.com/emutos/emutos/blob/
f29cf59f826ab0a3957be1a12387ac809d89dba5/vdi/screen.hlp) mentions:

The first, VME10, should always be equal to zero.  This causes the code
generated to work for the contiguous plane memory layout.  The code switched
out by these switches is for other memory layouts,

However, most of the code is compiled as if the vme setting has been set to 
one. I removed them from the sources though (just like emutos did long ago), 
to make the code better readable.