On Sun, 12 Feb 2006 15:51:02 -0500, Shmuel Metz (Seymour J.) <[EMAIL PROTECTED]> wrote:
> In <[log in to unmask]>, on > 02/11/2006 > at 12:53 PM, Johnny Luo <[log in to unmask]> said: > > >A little suprising for me.With 31-bit addressing,if you have 10G real > >storage,only 2G can be addressable as central storage.Then,how about > >the use of other 8G real storage?Used as a substitiution for paging > >data sets? > > Yes; the instructions that move pages between CSTOR and ESTOR address > the ES by block rather than by byte. > Given that hardware innovation generally outpaces software exploitation, it's easy to conjecture that the continuing distinction between CS and ES arises more from a software than a hardware constraint. It might well have been easier for hardware developers than software developers to add more bits to page and segment tables. I can readily imagine software developers asking hardware developers, "Please don't make us redesign the memory management software; rather give us a very fast virtual disk." > the line. Neither can include space allocated to common and neither is > backed up by aux or real storage until pages are actually allocated. > "Until ... allocated"? Or until accessed? I know the controversy over AIX's "lazy malloc", which allows a programmer to allocate a reserve work area for emergency shutdown, then fails because resources are unavailable when he actually tries to use it. An MVS partisan has told me he considers the AIX behavior quite reasonable, and faithful to the MVS tradition. -- gil -- StorageTek INFORMATION made POWERFUL ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
