You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Based on the description on GBATEK, it appears to me, that the MSW and LSW for IWRAM actually should be the other way around.
Out of general curiosity I wrote a small (currently very immature) test ROM last night, and its result in mGBA 0.8 seems to match my observation.
(left value is actual result, right value is what it expects, according to real HW)
It is also interesting to note, that with Thumb code in IWRAM it is possible that data reads/writes influence observed open bus values, as each instruction prefetch only trashes a HWORD of the latch. The test ROM demonstrates this. Although I'd guess (or hope?) that no game actually relies on this :p
https://github.com/mgba-emu/mgba/blob/master/src/gba/memory.c#L355
Based on the description on GBATEK, it appears to me, that the MSW and LSW for IWRAM actually should be the other way around.
Out of general curiosity I wrote a small (currently very immature) test ROM last night, and its result in mGBA 0.8 seems to match my observation.
(left value is actual result, right value is what it expects, according to real HW)
It is also interesting to note, that with Thumb code in IWRAM it is possible that data reads/writes influence observed open bus values, as each instruction prefetch only trashes a HWORD of the latch. The test ROM demonstrates this. Although I'd guess (or hope?) that no game actually relies on this :p
The test ROM (the code is also in the repo, but currently very messy):
https://github.com/fleroviux/openbuster/releases/download/0.1/openbuster.gba
The text was updated successfully, but these errors were encountered: