News from Swissmicros!

[Ángel Martin]

(09-20-2015 08:12 PM)matthiaspaul Wrote:  

(09-20-2015 06:13 PM)Chasfield Wrote:  Seeing more of the stack, via a two line display, would be cool, as Matthias suggests. I guess the original 15C ROM code would have to be broken into and quite heavily re-engineeered to achieve this - a bigger job than just getting the original version to run, as is, on an emulation layer.

While I have no direct insight into their implementation, I don't think it would be very difficult. After all, the Y register of the emulated machine must be stored somewhere in the microcontrollers RAM. I assume that it is quite easy to find out where it resides, even if the emulated code isn't fully understood. Once this location is known, it could be read out periodically (or at certain events) from outside the emulator's domain (that is by the operating system or driver layer underneath the emulator). The numeric contents could then be decoded and translated into whatever representation is necessary to display it on the pixel display. Just reading out the Y register at random times in a concurrent process might cause some temporary garbage to be displayed if the multi-byte register is read out at the same time as it is being updated by the emulated machine, but I'm quite sure that the emulator provides enough state-information to let the read out happen in a mutex and only when appropriate. In fact, it is possible, that a similar technique is used already in order to retrieve and translate the X register value.
Matthias

So here's a completely unqualified, unauthorized, unrequested and unconfirmable (the last two words I just made up) opinion to the above:

as long as there's something akin to DSPCRG in the 15C ROM there should be possible, but even in that case it'll require scalpo MCODE surgery beyond the "dump & go" approach...

BTW, I did *not* place an order for this new machine so I guess you can say this is as impartial an opinion as it gets nowadays.