I've greatly enjoyed the performance and versatility of a QBone in my PDP-11/83 but have always had trouble getting it to boot the DU (MSCP) drives consistently. After powering up, it seemed to take a number of pwr commands on the Qbone or pushes of the reboot switch etc to finally boot up. Once it booted ok, then it worked fine until I powered the 11/83 down.
Also, it didn't seem to make a difference whether I used the "m ll du.lst" commands from the Qbone console or used the "B DU0" commands on the PDP-11/83 console.
In talking to Lee Gleason who has down a good bit of hacking on RQDX3s from bare metal programs, he told me he had had seen some of the steps of bringing a mscp device online fail with hangs, waiting for a mscp change of state. When that happened he would go branch his code to the very beginning and start the process over.
So I modified Joerg's du.lst by adding a counter to the spin loops that wait for a response from the DU CSRs. If the loop count exceeds 10000, then it branches back to the start and tries again. To follow what is happening, I also am writing the number of retries to the 11/83 two digit LED display. It appears that it takes retries at two different steps of initialization to boot successfully and the total retries is 3 for my system to boot.
I've attached the modified du.lst (named duboot.lst) and also the macro-11 source for it if you want to experiment.
I do love the way Joerg has the "m ll" command utilize macro-11 listing files. One minor thing I ran into is that "m ll" will reload the last .lst referenced, but it appears that if I Try "m ll new_du.lst" it reloaded the original file even though I specified a different one.
It was fun to experiment with the boot loader and thanks to Taylor Hush who got me interested in working on this issue. Also, great thanks to Joerg who makes all this retrocomputing fun easier for so many!