On Thu, 9 Dec 2010 13:31:56 -0500, John P. Baker <[email protected]> wrote: >There seems to be some belief here that 2G thru 4G-1 is unavailable for some reason other than that specifically addressed by the undocumented >USE2GTO32G={NO|YES} operand. > >I do not believe that to be case. Am I correct?
USE2GTO32G is very recent, John; perhaps z/OS R10 or R11. And it's much larger than the the gap in storage allocation from 2G to 4G, which has existed since the inception of 64-bit z/OS storage. As far as I know, the reasoning expressed in this thread is correct: we have the 2G to 4G gap to avoid potential problems with AMODE(64) programs that have a 31-bit address with the high-order bit on, perhaps because it signals an addressing mode or (perhaps more likely) because it's an end-of-plist indicator for a 31-bit parameter list. -- Walt Farrell IBM STSM, z/OS Security Design
