Thread Rating:
  • 2 Vote(s) - 5 Average
  • 1
  • 2
  • 3
  • 4
  • 5
flashmasta savegame fix
#71
(05-18-2019, 10:18 AM)Loïc Wrote:
(05-18-2019, 06:13 AM)fbfiso Wrote: Sorry for bump your old reply.
As kickerofelves said, I reproduced his crash in option on both unmodified and modified ROM. After set the message speed to low and pressed B, the door close animation played and code crashed. And if the battery is replaced, and boot the NGPC again, the boot system setting menu shows up.

I mention this because it seems the code erased the boot system setting. So is it possible that the block erase/write operation require more timing cycles on flashmasta(which the FPGA implementation caused) and the code in game software write too quick before the actual hardware is ready and start waiting write end signal(which will never happen since the write command happened too quick, no actual write operation started or write to the wrong place like system setting area)? So will some no-op code inserting before actual write command issue help solve this problem.

I have no idea on NGPC programming, I just once dealed with similar problem on a genesis unlicensed cartridge that require several nop instruction after save data bank switch or it will not be able to save. But NGPC uses flash chip. I have no idea if there will be similar things happen on such hardware.

What's odd is that it's the only known issue with savegame on flashmasta, and it occurs only in this specific case. I should try this one on neoSD too...

Either way, tracking it would require to disassemble the code and figure out when exactly the bug occurs. Without a hardware debugger, I don't know if it's worth the pain ^^

Yes, it's really a pain to dig that without a proper hardware debugger. As the game seems works ok in main progress now. I don't think it deserve to do that research. Just write up my guess on this problem. =P
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)