2019/06/27

09:26

The new memory stuff is much cleaner which is nice, although I believe there is an issue with some initialization of the BootROM or similar. PowerPC and x86 crash the same, so at least this one is consistent.

09:28

Very likely is an initialization error.

14:22

Okay so maybe I am writing to invalid memory? Because the debug stuff and everything after the registers is zero.

14:30

No, that is just the first set error break I had that was from the fprintf code.

14:33

Okay so I see 0xe7e5e7e4 in one of the registers so definitely dealing with a bad magic. Okay so it seems this is attempting to allocate an object which has the bad magic size. This is the initialization of config reader. Init ram links is done and then finished. And jvmNew

15:22

Just going to guess that it is the bootstrap loader is messed up since that is the most complex part of the rewrite. Maybe it is a confusion of a RAM or ROM address.