Fun schematic rev-engineering Fire Shark;)

trm

Who loves you, and who do you love?
Feedback
2 (100%)
Credits
2,877CR
Thanks dude! $7036 is written on my scratchpad notes about 15 times but I think fatigue and being super happy I'd been able to trigger the OBJ RAM ERROR meant I brain farted.

Something I was pondering this aft was patching the RAM test around $70D4 to modify the readback part. I'd expect D2 to have the offset from base of RAM where the r/w mismatch occurs so I could push that to the display using $180002 and $180004. I've figured out the tile numbering sequence so I don't even need to decode it before dumping it onscreen.

And with your pointers on patching out the conditional check I should be able to get away with it. The RAM test annoyingly occurs after the ROM test (going from memory here as my notes are insulating a double-layered board atm) so I shouldn't need to worry about later checks as I'll have got my address/value data by then.
 

trm

Who loves you, and who do you love?
Feedback
2 (100%)
Credits
2,877CR
I really need to go to sleep earlier tonight instead of staying up 'til daybreak - forgot to mention I'm using the samesam2.zip file here as that matches my HW.

I've not checked to see where the 2P code relocated other stuff from, so I don't know if the addresses are common but I can use pattern matching to implement any good ideas you guys have :)

Also any tips on a decent 68k assembler for Win would be much appreciated. My old haunt of Devpac3 (but under WinUAE) is looking likely instead of Easy68k now.

cheers
 

cmonkey

Active member
vacBacker
Feedback
4 (100%)
Credits
1,659CR
Don't forget to go to bed!
smiley36.gif
 

trm

Who loves you, and who do you love?
Feedback
2 (100%)
Credits
2,877CR
I stayed up as I found more Redbull. I now have an almost working modified self-test which bangs the 68k data bus with the test iteration and loop offset where the failure is found so I should be able to pull that off with my logic analyser. I might need to spend a bit more time making sure that the code I'm overwriting isn't important in the earlier parts of the boot process cos I'm currently running everything inside a mame debug, patching 'live' ROM so I've not done a hard reset since inserting my patches. Thanks for the ROM check skip too cmonkey.

I'd have thought booting a clean ROM, doing a wpset over the region I'm nicking and letting the game pass POST should be enough; I earlier thought of dumping PC to a trace but realised that wouldn't cover any indirect addressing should there be data in the region, it'd generate a crapton of log that'd need parsing and that mame already had the perfect tool in watchpoints. So then I knew it was bedtime again
smiley1.gif
.

I'm pausing the code hacking for a bit to finish off the schems as this has gone way beyond its original purpose and I want to understand completely how the custom causes sprites to be generated
smiley36.gif
.

cheers chaps
 
Top