On 12/15/2024 12:30 AM, Doug Jackson wrote:
I saw that, byt the owner doesn't want the archive opened - Its some sort of proprietary archive format.
Hmm, the author of the files is on list (Royce Taft), and I'm sure that was not his intent. From this list, 11/21/2024: http://lists.bitchin100.com/private.cgi/m100-bitchin100.com/2024-November/090131.html The list archive scrubbed the HTML, but I pasted the original text below The repo is mine, so if there's something Royce or I need to do so make it more public, let one of us know ---------------------- If anyone tried to access the Ghidra disassembly files in Jim’s GitHub repository and found that the project doesn’t load, please disregard those files. The correct file has been uploaded which can be imported (or “restored” as Ghidra puts it) into Ghidra 11.2 or higher to load a snapshot of the disassembly thus far. Even for the code I’ve stepped through and commented, I don’t have everything figured out yet. I’ve got a handle on good portions of it up to the DOS IPL which was loaded to RAM from the first sector. When the ROM copies data to RAM, I used the byte viewer tool in Ghidra and manually copied the data into RAM at the appropriate location. From there, I could continue the disassembly in RAM. I did the same for the first sector of the boot disk. I used a separate hex editor to copy the 256 bytes from sector 1 of the boot disk and pasted that into Ghidra’s bytes viewer at the appropriate location in RAM. As I stepped through code, when writes were made to data in RAM, I manually made those changes using the bytes viewer. When playing around with it, keep in mind that reads and writes to 0x0000-0x07FF are not referencing the ROM and are instead referencing RAM1 once the bank has been switched to RAM1. Ghidra doesn’t know this, so those references must be manually changed. I probably missed some references as I stepped my way through. The amount of code I’ve stepped through and tried to decipher+comment is relatively small in terms of bytes compared to the size of DOS or disk basic, so there’s much left to do. I’m hoping to get back to this soon. I should probably try to capture a bunch of this FYI type info for the Ghidra disassembly project and compile it into a readme.
