RE: [stella] Stella 2.0 Alpha 2 available

Subject: RE: [stella] Stella 2.0 Alpha 2 available
From: "Fred Quimby" <c9r@xxxxxxxxxxx>
Date: Wed, 29 Jun 2005 11:13:47 -0400
>BTW: When does the code get executed? During VSYNC? I wonder, what
>happens when you clear return address bits while the code is in a
>subroutine (which is quite unlikly for VSYNC). Since then your are 
>effectively
>clearing bits of the PC.

Yes, but I had the code only reach $FE the way it was written, which will 
have the effect of changing the low byte of the return address only, 
assuming we were in a subroutine and the stack pointer was $FF initially.  
Maybe this is less accurate than real frying, but I think that a crash would 
be much more likely if the high byte were changed,

> > I did try messing with the 6502 registers.  Randomizing the PC or SP 
>often
> > just crashed the system.  Scrambling A, X, Y, and flags by themselves
> > sometimes did interesting things but weren't good at replicating known
> > behaviors.  I didn't try this together with clearing random bits, nor 
>did I
> > mess with the timers.
>
>Maybe you should just clear them too?

The timers probably get affected if the RAM does, since they're all on the 
same chip, but regarding the 6502 regs, maybe I'll run an experiment on a 
real hardware to see if any 6502 regs are ever changed by frying, kind of 
like Erik suggested.  It should be easy enough to check for changes in PC or 
even A, X, Y or flags by creating small conditional loops and filling the 
rest of the ROM with zeros.

Or maybe I can take my 2600 and apply semi-persistent power to the RIOT by 
putting a big capacitor across its power pins, thus ruling out any RAM 
changes, and see if anything weird still happens.

Or maybe even hooking the 2600 to a logic analyzer would reveal something.  
I'll look into this a little more if I get some time.


Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://stella.biglist.com

Current Thread