NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/comp/sys/atari/st/19712

Internet Message Format | 1993-01-12 | 1.9 KB

Path: sparky!uunet!spool.mu.edu!olivea!mintaka.lcs.mit.edu!kropotkin.gnu.ai.mit.edu!entropy
From: [e-mail] (maximum entropy)
Newsgroups: comp.sys.atari.st
Subject: Re: on the need for 'virtual memory'
Message-ID: <[e-mail]>
Date: 13 Jan 93 00:51:56 GMT
References: <[e-mail]>
Sender: [e-mail]
Organization: Not Proud Programmers
Lines: 29
In-Reply-To: [e-mail]'s message of 12 Jan 93 19:02:11 GMT

In article <[e-mail]> [e-mail] writes:
>Can someone who knows the history of UNIX please comment on whether the
>early versions of UNIX 'required' virtual memory in order to work?  My
>own impressions are that the PDP computers didn't have any virtual
>memory, and that most programs that were larger than memory consisted of
>a series of intelligent overlays.  (My impression is partially formed by
>experience with UNIX 7.0 on a PDP-11/34, requiring a grand total of
>10 megabytes of hard disk space, and the fact that one of the selling
>points of the VAX series of computers was the fact that they had virtual
>address space.)

Your impressions are accurate; it was even a big deal when the 11/34A
came out because it could split instruction and data space for a grand
total of 128K of per-process memory.  My DEC Pro-350 (11/34 class)
running VENIX 1.0 (V7-like UNIX clone) lets you use more than 64K per
process only by mapping pages of memory on a user-controlled basis
(The C library will do it for you if you use some compilation
switches) with a special OS call.

Why's this in comp.sys.atari.st?  Getting frustrated with trying to
compile 600 meg source files with GCC under MiNT? (1/2 :-) )

Cheers,
entropy
--
[e-mail]
entropy. . .it's not just a good idea, it's the second law.
Boycott AT&T, Lotus, Apple, and Xerox.  Join the League for
Programming Freedom!  Write to [e-mail] for more information.