CiscOS and a real 68K?

[Coeus]

In discussion of a possible b-em issue it has come to light that CiscOS misidentifies the CPU it is running on when running in b-em. See https://github.com/stardot/b-em/issues/ ... 1774507835 for more details but code in CiscOS to identify the CPU works by forcing an address exception and then checking the size of the stack frame generated. B-Em uses Musashi (https://github.com/kstenerud/Musashi) as its 68K emulation and that agrees with the 68000 documentation that the stack frame on a plain 68000 should be 14 bytes but CiscOS requires 18 bytes to identify the CPU as a 68000.

Does anyone have a real 68000 processor to try this on? Either CiscOS or an isolated test case?

[atsampson]

D7 is 14 decimal - as expected - on a real 68000 in my Amiga 500 (output by flashing the background colour!).

[Coeus]

Thanks.