Log. 7th February, 2013.
Probing the 6309
Aaron Wolfe very kindly sent me a couple of 63C09Es, and they arrived on Tuesday. Today I've thrown one into a Dragon and started probing it. The story so far: An invalid register in a memory-to-register-bit operation does not cause an illegal instruction trap.
Most of what I'll be doing is just verifying behaviour already documented by others and checking whether native mode only makes NVMA cycles disappear, but there are one or two ambiguities like this one to clear up!
A happy (?) side-effect of testing invalid registers in TFM (which does cause a trap) showed up a crash bug in XRoar, so that's fixed now too.
Tags: 6309, dragon, xroar