I am equally (or perhaps more <g>) ignorant of LGR. I was *assuming* that the way LGR works is that you define an LGR pool of machines to VM, VM looks around and says "aha! Two z14's and a z12" and any guest defined for LGR gets told "you're on a z12" even if at the moment it is actually on a z14.
I was proposing something simpler and more flexible in which the customer just told VM via parm or command "tell guest G this is a z12." Simpler in that there is no pool and no looking around; more flexible in that the customer might not actually own a z12 at all. If LGR does not work the way I assumed (and yes, I had a First Sergeant in the army who continuously reminded us what happens when you assume) then my proposal is indeed possibly redundant with reality. Charles -----Original Message----- From: IBM Mainframe Assembler List [mailto:[email protected]] On Behalf Of Peter Relson Sent: Thursday, March 28, 2019 6:44 AM To: [email protected] Subject: Re: Determing the Presence of an Instruction I'll admit that I have no idea what I'm talking about w/r/t LGR, but it seems to me that "dial down" is exactly what the function does. Tony mentioned that it is "automatic". I don't think it can be fully automatic; the minimum target level has to be defined to z/VM (and I had wondered whether it could be defined to z/VM on a per-guest basis). So imagine a z/VM definition while running on a z14: Guest G is to be able to use LGR to a z13. Now you IPL z/OS as guest G. z/OS would see the facilities available as those for the z13. A program would see the facilities available as those for the z13 (I assume). Isn't that what you would think "dial down" should do?
