RE: [cosmacelf] Re: CDP 1802/4/5/6 ICEing on the cake
- oooh, you could do memory reference bp's with the larger-than-byte memory too. The execution BP is just a special case.
Date: Fri, 1 Mar 2013 20:17:44 -0800
Subject: Re: [cosmacelf] Re: CDP 1802/4/5/6 ICEing on the cakeYou can only do ROM breakpoints with an emulator or simulator or a hardware debugger.
One interesting way to do a hardware-software debugger is to have the CPU
interrupt be executed after one instruction is executed. This would work
when interrupts aren't being used, or probably if interrupts are being used
and you do some extra work.
One nice way to do breakpoints in the 1802 is to use a dedicated 16-bit
register as a breakpoint using SEP Rn. Before you run or single-step the code,
you exchange all of the breakpoint addresses with that value and save what
was stored in each one so it can be restored after a breakpoint.
This is only for an execution breakpoint, and won't work for memory reference BP's.
You can also use it to single-step by intelligently looking at the next instruction
and setting breakpoints where needed to stop after the next instruction. This takes
some good intelligence because of what can happen in the next instruction execution.
((conditional) branches, (conditional) long skips, SEP instructions, etc.)
On 3/1/2013 4:25 PM, Kevin wrote:
One should be able to set breakpoints when running code from ROM. This would dictate that an ICE have a list of breakpoint addresses instead of needing to alter the code undergoing testing.
Modern debuggers also allow conditional breakpoints, which would be extremely nice!
--- In email@example.com, "urrossum@..." wrote:
> > > the otherwise unused 68h... Does anyone know what a *real* 1802 does
> > > when it encounters this instruction?
> > That's easy! :-) Notice that 69-6F are the INP instructions? They put
> <...clip of excellent explanation..>
> Thanks, Lee! That makes perfect sense, and also affirms that I can use this instruction with impunity, since it should be of no use to any existing software.
> Mark Moulding
- Dave Jones did a review of the TL866CS on his EEVBLOG.- Tom
--- In firstname.lastname@example.org, "joshbensadon" <joshbensadon@...> wrote:
> --- In email@example.com, "thinkpast" hjohnson@ wrote:
> > http://www.retrotechnology.com/herbs_stuff/gal.html
> > Some of the simplest PROMS can just be poked at and manually programmed. I reference some of that work on another Web page linked to the above page.
> > Herb
> Thanks Herb,
> Your web site lead me back to do another search on Ebay. I searched a few years ago without success, but now there are several to choose from!
> -TL866CS -$49
> -TOP853 $46 (does not list GAL 22V10)
> -TOP3000 $160
> -G540 -$60
> -G840 -$80
> -RT809F -$88 (does not have large ZIF socket, less versatile?)
> Several of these models include RAM testing and other CMOS Chip Testing, very impressive (if the tests are rigorous?).
> For these prices, it's not worth messing about building a home brew programmer. Heck, I've paid more for an adapter on my Andromeda Labs programmer.
> Right now I'm considering the G840, just on a hunch, but if anyone here can offer any advise, it would be much appreciated.