360 DIAGNOSE 1308
i got a new toy that i'm just learning to use (especially trying to cleanup the OCR bit) ... an...
If one were *very* careful there were a few specialized situations in which a customer could make legitimate use of the DIAGNOSE opcode in a non-VM environment, but they were few and far between. As has been noted in most of the a.f.c threads involving that opcode (and as explicitly documented in the S-360 PoO) the DIAGNOSE opcode was completely model-dependent, so any program that used it was thoroughly nonportable unless you could ensure that it was used only on a single S-360 model.
real" diagnose tends to invoke machine specific microcode whose operations aren't defined in the POP (aka the POP says something about diagnose functions are defined as being model specific). I had done a modification...
AFAIK there was no reliable way for a program to determine just which S-360 model it happened to be executing on (the Store CPUID instruction didn't show up until the S-370).
In my case I used DIAGNOSE on a 360-40 to read the console data and address toggles (simulating the sense switches for the port of our 1401 SYSIN-SYSOUT spooler). On the40 the DIAGNOSE forced its effective address into the ROAR, causing the microcode to jump to that address -- with the obvious opportunity for mischief if you sent it to the wrong address.
But...the decision to use DIAGNOSE was purely in the hands of the customer. For shops that were (to be polite) clueless about the consequences of using DIAGNOSE on bare metal I would consider it quite appropriate for the sales rep to strongly discourage its use -- but a formal "you are not permitted to use DIAGNOSE" atbreastude would have had consequences for the branch office marketing goals the next time the customer wanted to upgrade.
Joe Morris
Alt Folklore Computers from Newsgroups