GSU revision comparison test

[qwertymodo]

Maybe solder in a socket so you don't have to desolder the memory? o.o

It's a TSOP-48 chip, mounted to a SOIC-32 footprint via an adapter board. Not only would it be a huge pain to wire up, I don't have another TSOP test socket just lying around.

[FitzRoy]

Ouch, that's quite an arduous process. I'm making a test suite so you don't have to reprogram it for each test. Currently got mult, cache, plot timing tests and you can toggle between 10/21mhz, going to add rom/sram buffer tests. Any suggestions are welcome./

I don't know if what you listed covers this, but some of the only remaining game issues in higan-snes are special chip games running slightly faster or slower due to "cache delays" or whatever being unemulated. These issues are very minor and would be hard to detect in a blind comparison, but they're still worth fixing.

One noticeable example is SMW2's long intro that sort of desyncs over time. There's a scratching sound that's supposed to play when Yoshi looks at the map, but in higan it plays much later.
There is a similar issue in MMX2's attract mode (cx4 chip) where Mega Man dies in the attract mode because of unknown cache behavior.

[qwertymodo]

It's been a few weeks with no replies, so I figured I'd check back. Any new tests to run?