[email protected] (Tom Marchant) writes:
> Bi-modal?  You mean like MVS/XA?  Or do you mean that you 
> could run it in ESA mode?  From the 1.6 announcement:
>
> "z/OS V1.6 must execute in a z/Architecture (64-bit) mode."

above/below line was originally introduced with >16mbyte real storage
for 3033 ... 64mbyte option or "26bit" real addressing.

all instructions were still 24bit (16mbyte) ... but a hack to the
pagetables allowed pages to be located in real storage above 16mbytes.
24bit virtual addresses were run thru address translation ... and the
pagetable hack could result in 14bit real page number.

original 370 (4k page option) was 12bit virtual page number into
pagetables with 12bit real page number out. page table entry was 16bits;
12bit real page number, 2 defined bits, and 2 undefined bits. the 3033
64mbyte hack ... retasked the undefined bits as 2bit prefix to the
standard 12bit real page number ... for 14bit real (4kbyte) page number
(up to 64mbytes real). 370 IDAL was also hacked to support 26bit i/o
address (enabling direct page in/out above 16mbyte line).

however ... all sorts of stuff still had to be in the first 16mbytes
real storage (introducing above/below line issues).

misc. recent posts mentioning >16mbyte support originally for 3033:
http://www.garlic.com/~lynn/2010.html#84 locate mode, was Happy DEC-10 Day
http://www.garlic.com/~lynn/2010g.html#11 Mainframe Executive article on the 
death of tape
http://www.garlic.com/~lynn/2010g.html#36 16:32 far pointers in OpenWatcom C/C++
http://www.garlic.com/~lynn/2010g.html#55 Mainframe Executive article on the 
death of tape

for moving stuff between above & below the 16mbyte line ... there was
originally plan to do page out and then page back in (running page move
thru i/o system courtesy of IDAL). I provided a little example code
where pagetable entries were fiddled for above & below the line
addresses ... with a MVCL for move across the line (lot more efficient
than doing it thru the page i/o system). some old email providing the
MVCL hack (for crossing 16mbyte line, also mentions that 3081 & XA
schedule was slipping so additional storage was extending life/revenue
for 3033):
http://www.garlic.com/~lynn/2006t.html#email800121
in this post
http://www.garlic.com/~lynn/2006t.html#15 more than 16mbyte support for 370

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
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

Reply via email to