Weird problem on a couple of Beebs

[maniacminer]

I have a flashing cursor in the top left of the screen on three Beebs (all issue 7), connecting up the Tube Snooper I can see the processor trapped in an endless loop, only interrupts give it a chance to do something else, then back at it, spinning over and over.

It took me a while to find the bit o'code from the disassembly of OS1.2 here https://tobylobster.github.io/mos/mos/S-s11.html unfortunately, despite all the lovely work, I would have preferred the hex address and hex of the instructions to be like a traditional disassembly (maybe I'll fork it into that when I get some spare time)

The lack of BRK handler is the cause of the endless loop

This bit o'code shows the principle (CALL CODE)

I assume somewhere BASIC or something is supposed to take ownership and clear the carry flag?

Has anyone seen this before?

[Ukwebb]

looking at that code, the value its using is

.startupMessageSuppressionAndBootOptions = $0267
bit 7 = 0 means ignore OS startup message
bit 0 = 1 means if !BOOT errors from DISC because no language found, then lock up machine
bit 0 = 0 means if !BOOT errors from *ROM because no language found, then lock up machine
(Read/Write using OSBYTE 215)

and bit 0 will end up in the carryflag due to the ROR instruction

and OSBYTE 215 leads here .. https://beebwiki.mdfs.net/OSBYTE_%26D7

plenty of reading here

[maniacminer]

Thanks for the info. It turns out that IC20 had gone "bad" in all the boards. Removing it and testing it, they passed OK, then replacing it with a 74F139 and I could then see that there was not enough time for the circuit to lock on causing weird problems if I could get the machine to boot that it would show several copies of the DNFS ROMs and then randomly lockup. In the end replacing IC30 with a fresh flip-flop cured the problem and the old 74LS139 worked again. Damn these squirrelly Beebs. Now back onto building the 40th Anniversary boards again. Yay!