Advertisement
Zoinkity

4567 disk writup

Sep 7th, 2015
331
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 2.29 KB | None | 0 0
  1. The disk's actual contents are a build of SPgame from the SDK, with DD and debug options.
  2.  
  3. However, other content on the disk was not erased and reveals they had also tried the tile_rect3D demo at some point.
  4. The tile_rect3D demo uses a ROM-area MFS table, named "SimpleMfs", found at end of ROM segment - 2 LBAs.
  5. The entries listed:
  6. /cfb/cfb.bin LBA 20 + 0 0 bytes
  7. /static/static.bin LBA 20 + 0 0 bytes
  8. /rsp_cfb/rsp_cfb.bin LBA 20 + 0x130 0 bytes
  9. /texture/cafe.bin LBA 20 + 0x130 0xA0040 bytes
  10. /texture/road.bin LBA 53 + 0x1368 0xA0040 bytes
  11. /texture/world.bin LBA 86 + 0x25A0 0xC0040 bytes
  12.  
  13. Please note the offsets given in the table do not correspond to the actual locations of the images!
  14.  
  15. Current disk contents end at LBA 0x27. The "cafe" images are found at this point, with the first getting cut off. They would have started at 0xB0730, with a possible header starting at 0xB06F0. This image set ends within LBA 0x3F.
  16. The "road" textures start at LBA 0x40, or 0x134200 in the dump. "World" textures are at LBA 0x62, 0x1D7D10 in the dump. They're in a MWo2 wrapper, generated by the SDK tools. This is somewhat ironic since several retail games also use these wrappers on their data, such as "Batman Beyond", "Jikkyou Powerful Pro Yakyuu 2000", "Jikkyou Powerful Pro Yakyuu Basic Ban 2001", "Daikatana", "Ridge Racer 64", and the "Wild Waters" prototype.
  17.  
  18. Most likely explanation is that the demo was compiled with MFS for an initial test, then redone without MFS a second time. Images wound up in different spots. This was then overwritten with the SPgame test.
  19.  
  20. There are three unique binaries in this game. The first, and most obvious, is SPgame at the normal boot LBA 0. A second unique one, what looks like part of the build of a version of tile_rect3D, is found in the RAM area, LBA 0x5A3. A third unique one is written five times in the RAM area at periodic LBAs. This appears to be a different build of tile_rect3D, based on the similarity in code but differences in addressing.
  21. It's not uncommon for a program or data to be written periodically in unused space. It's incredibly common with carts, when filler isn't declared and previously written data in the buffer is pushed. Since the RAM area isn't used by any of the demos it isn't much of a stretch to say something similar occurred here.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement