Wednesday, July 13, 2016

Virtual 2310 drive passes tests, just needs to run more comprehensive diagnostic to get 'signed off'


Restructuring the GUI

-- virtual 2310 --

I ran a quick test this morning before heading to Computer History Museum for my regular Wednesday time with the restoration teams. It showed that two of my FSMs did not return to idle position, as they should. The disk seek machine didn't go to idle, nor did the busy machine, while the operation complete machine remained at idle.

I looked over the logic a bit while I am at CHM as well as designed some new instrumentation for the testing this afternoon, when I get back to the 1130. I wasn't happy with the way I had built the seek timing emulation state machine so I restructured it.

Late afternoon back home, with the air conditioning powered up, I could resume some testing. I found some timing issues with my seek timing simulation, particularly a timer that overflowed before I reached my termination count, but I corrected that to confirm proper operation.

I had success with an operation complete after my seek was executed. Further, it did move the virtual disk cylinder out to the proper location. I subsequently did a read of the sector and everything matched.

If I can get either a DCIP Analyze to run or the disk diagnostic to run properly, I will feel much better about this. As it is, I have to check the disk with hand entered instructions which isn't exhaustive.
Everything I tested matched what I expected. I did seeks in both directions, tried seeking backwards past cylinder zero, read several sectors in different spots on the disk, and that looked good.

My disk diagnostic program is set up as a core load, which I ran on the IBM 1130 simulator then dumped as a file. I just couldn't get it to try a different disk drive, it kept accessing the internal disk drive. The diagnostic listing says that I have to start it with bit switch 15 up, which causes it to pause for customization. I will do a session on the simulator to yield a workable copy of the diagnostic.

My standalone DCIP utility suffers from a problem I have seen before. It can think that it has a 1403 or 1132 printer available and it will go into a loop waiting for the printer to come ready. That won't happen since I am running without printers, but it blocks me from dumping sectors or doing a disk analyze. I need to get a DCIP core image that will run properly without a printer. I think it was caused by the way I set up the core load file.

Thus, it is frustrating to find that my two test programs are both flawed and can't be used now. I do have a third method I could use, entering a boot card image into low core and trying to boot the DMS2 image on the virtual disk cartridge. Testing tomorrow will focus on getting at least one of these methods to satisfy me.

No comments:

Post a Comment