On Tuesday, 04/29/2008 at 07:00 EDT, Mark Post <[EMAIL PROTECTED]> wrote: > >>> On Sat, Apr 26, 2008 at 11:55 PM, in message > <[EMAIL PROTECTED]>, Alan > Altmark <[EMAIL PROTECTED]> wrote: > > On Saturday, 04/26/2008 at 08:59 EDT, Mark Post <[EMAIL PROTECTED]> wrote: > >> Compared to a CP, an IFL has one "less" instruction. > > > > This is not true, and if it were it wouldn't apply to GCS or VM/VTAM. > > Care to explain exactly why this is not true, particularly in the context of > the sentence that followed that one: "If you try to execute it though, you > don't get a S0C1, you get a check stop." ? I put the word "less" in quotes > for reason.
Not in any detail, no, since that would delve into areas where Balrogs live, and I am sorely afraid of them, but all of the instructions on a CP are present on an IFL. If instructions were missing you'd see S0C1 (operation exception), but as you note there is a Check Stop, not an S0C1. Ergo, no missing instruction. Q.E.D. Thou Shalt Not Futz With The Architecture. So sayeth the Master Architects. "All Hail the Master Architects! Huzzzah!" Alan Altmark z/VM Development IBM Endicott
