The X'80000000' that I discovered was neither the end of a parameter list nor a last parameter address. It was one of the five exit routine addresses, not all of which are even contiguous, in an IOSB, the driving I/O for which had been intercepted and mishandled by my code. IIRC, the IOSB was built by that most pathological of all products, DB2. It has always been my belief that even pathological cases have to be tested for and supported. They certainly don't go away because of our opinion of them. I once asked a question in a SHARE session about some arcane "what if" scenario in the (then) new CVAF, and the IBM presenter's answer was that my hypothetical question was pathological, meaning it could be safely ignored in the test and debug phase, I suppose.
But you are right - the pathological case that bit me was not under normal conditions. OTOH, under normal conditions nothing pathological can ever happen by definition of "normal." Bill Fairchild Rocket Software -----Original Message----- From: IBM Mainframe Assembler List [mailto:assembler-l...@listserv.uga.edu] On Behalf Of Gerhard Postpischil Sent: Friday, August 13, 2010 4:29 PM To: ASSEMBLER-LIST@LISTSERV.UGA.EDU Subject: Re: Parameter passing: overly cautious or properly paranoid? X'80000000' under normal conditions is pathological - it's not the zero expected for an end of parameter list, and it's not a valid last parameter address. Gerhard Postpischil Bradford, VT
