28 Jul
2006
28 Jul
'06
7:25 p.m.
On Friday 28 July 2006 21:42, Roy J. Tellason wrote:
On Friday 28 July 2006 09:22 pm, Patrick Finnegan
wrote:
Yeah, BDAL00. From looking at a QBUS pinout, I don't see any other signals to
indicate even/odd.
Pat
--
Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
On Friday 28 July 2006 21:14, Roy J. Tellason
wrote:
It may be that I'm mis-remembering, this _was_ thirty years ago or so I
was last messing with that machine. Wasn't there a line on that bus to
indicate one or the other byte, though? On Thursday 27 July 2006 07:08 pm, Chuck Guzis
wrote:
Nope. 64kB address space - 8kB I/O page (32kW - 4kW = 28kW).
Unlike some other things of the era (like the HP 1000 series), the
address gives you what byte location of memory something is at, not which
word location in memory.
Note that things after the LSI-11/2 which includes an MMU (or any
Unibus/11 that has an MMU) can physically address more ram, but cannot
logically address more than 32kW... Now, back then, it was even more confusing. The
"K" that IBM was
talking about on its S/360 machines was 8-bit bytes; the CDC was
talking about 60-bit words.
I just bumped into that one in an email conversation with somebody, who
said that an LSI-11 would only handle 56K of ram. And my recollection
(?) is that this referred to 56K _words_ as opposed to bytes.