Prev: iPad for Sysadmins
Next: Jailbreaking iPad - advantages?
From: Rowland McDonnell on 5 Jun 2010 07:55
David Empson <dempson(a)actrix.gen.nz> wrote:

> Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote:
>
> > Jim <jim(a)magrathea.plus.com> wrote:
> >
> > > <http://reviews.cnet.com/8301-13727_7-20006694-263.html>
> >
> > But /why/ is `wired' memory called that?
> >
> > I don't get it.
>
> It is "wired down" in that the system can't reallocate that memory for
> another purpose.

Hmm - okay, that's a way of looking at it that makes sense. I've always
thought `Stupid term: all the bloody memory's wired up, 'cos if it
wasn't, it wouldn't be accessible, exotica aside. So what am I missing,
eh?'.

But actually, it's more like lock-wiring fasteners on yer racing vehicle
so they don't fall off on the race track (I wish I knew *exactly* what
the deal was there, is it just to keep the engine together so it's
harder to get oil spills? - 'cos I've seen that racing bike fairing
fasteners are allowed to be quick-release.)

> The term does seem a little odd, because "wired memory" was a term used
> to describe read-only memory back in the early days where it really was
> made up of wires.

Hrumpghfff. Ish, I'd say. ENIAC was programmed that way, sortathing.
You wired a program into it using that crude method, from what I've been
able to work out from what I've read.

But you know, there was also the LOL memory (little old lady) used for
the ROM in the Apollo programme's flying computers - magnetic core,
wired to a fixed pattern, knitted that way by a regiment of little old
ladies (honest!). The programs (US spelling had taken over by then,
AFAICT) had to be written and debugged a long time in advance - and
those little old ladies had to knit it perfectly. They succeeded.

<shrug> What else? I dunno. The early Manchester computers didn't
have ROM at all that I know of - even Atlas had to have a `Load tape'
program keyed into it via binary switches on the main console if you
wanted to boot the thing from cold.

They used to have the Manchester Uni. Atlas console out on display in
the CS building - including the paper note with the coding required
still taped in place, underneath the perspex screen fitted to keep
students from nicking bits. And a big row of big fat binary data entry
switches.

Not been back to check for the last N decades, so who knows where it is
now?

They had an Atlas RAM cabinet there too. Woo!

(ISTR the MU 5 console as well, and at least one rack that had been used
for an incarnation of the very first Manchester computer - but who knows
if it had been used on Baby itself? No-one, I'd guess)

> Mac OS X's "wired memory" is certainly not read-only.

Aye indeed.

> I haven't found anything which clearly explains where the Mac OS X usage
> of the term originates.

It's a Unix thing, innit? So it'll go back to the way they thought at
Bell Labs back in the late '60s.

> The best analogy I can think of is a garden where part has been covered
> in mesh wire to protect it. You can't uproot the plants in the wired
> section without first removing the wire netting, but the rest of the
> garden can be uprooted and replanted at will.

<heh> Not an analogy really, but certainly a way of remembering what
the terms mean in a fashion that makes good sense.

Rowland.

--
Remove the animal for email address: rowland.mcdonnell(a)dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
From: Jaimie Vandenbergh on 5 Jun 2010 08:25
On Sat, 5 Jun 2010 23:39:02 +1200, dempson(a)actrix.gen.nz (David
Empson) wrote:

>Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote:
>
>> Jim <jim(a)magrathea.plus.com> wrote:
>>
>> > <http://reviews.cnet.com/8301-13727_7-20006694-263.html>
>>
>> But /why/ is `wired' memory called that?
>>
>> I don't get it.
>
>It is "wired down" in that the system can't reallocate that memory for
>another purpose.
>
>The term does seem a little odd, because "wired memory" was a term used
>to describe read-only memory back in the early days where it really was
>made up of wires.
>
>Mac OS X's "wired memory" is certainly not read-only.
>
>I haven't found anything which clearly explains where the Mac OS X usage
>of the term originates.

It wasn't a term used in MacOS9 and below?

Cheers - Jaimie
--
Unix was written to play Spacewar and cheat at Scrabble, but Linux was
created merely to prove that it booted. - AdB, ASR
From: Rowland McDonnell on 5 Jun 2010 08:50
Jaimie Vandenbergh <jaimie(a)sometimes.sessile.org> wrote:

> (David Empson) wrote:
[snip]

> >Mac OS X's "wired memory" is certainly not read-only.
> >
> >I haven't found anything which clearly explains where the Mac OS X usage
> >of the term originates.
>
> It wasn't a term used in MacOS9 and below?

Not even slightly. Pre MacOS X, and there was no protected memory at
all. 68000s just plain can't do it IIRC. (68030s can IIRC; could look
it up, can't be arsed).

Pre System 7, and there wasn't even any virtual memory as such. The
original idea created for the original Macintosh System Software was to
rely on the code fragment manager, which in effect provided a virtual
memory function, but on an application-by-application basis, as set up
by the original programmer's decision on what size chunks to write the
app in.

Wired memory in the MacOS X sense is a pure Unix term.

Rowland.

--
Remove the animal for email address: rowland.mcdonnell(a)dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
From: Richard Kettlewell on 5 Jun 2010 09:26
dempson(a)actrix.gen.nz (David Empson) writes:
> It is "wired down" in that the system can't reallocate that memory for
> another purpose.
>
> The term does seem a little odd, because "wired memory" was a term used
> to describe read-only memory back in the early days where it really was
> made up of wires.
>
> Mac OS X's "wired memory" is certainly not read-only.
>
> I haven't found anything which clearly explains where the Mac OS X usage
> of the term originates.

The term goes back _at least_ to the early days of Unix, e.g. _The UNIX
Time-Sharing System_, Ritchie & Thompson, 1974:

Likewise, the process‐control scheme and the command
interface have proved both convenient and efficient.
Because the shell operates as an ordinary, swappable user
program, it consumes no ''wired‐down'' space in the system
proper, and it may be made as powerful as desired at little
cost. In particular, given the framework in which the shell
executes as a process that spawns other processes to perform
commands, the notions of I/O redirection, background pro‐
cesses, command files, and user‐selectable system interfaces
all become essentially trivial to implement.

--
http://www.greenend.org.uk/rjk/
From: Chris Ridd on 5 Jun 2010 09:35
On 2010-06-05 14:26:03 +0100, Richard Kettlewell said:

> dempson(a)actrix.gen.nz (David Empson) writes:
>> It is "wired down" in that the system can't reallocate that memory for
>> another purpose.
>>
>> The term does seem a little odd, because "wired memory" was a term used
>> to describe read-only memory back in the early days where it really was
>> made up of wires.
>>
>> Mac OS X's "wired memory" is certainly not read-only.
>>
>> I haven't found anything which clearly explains where the Mac OS X usage
>> of the term originates.
>
> The term goes back _at least_ to the early days of Unix, e.g. _The UNIX
> Time-Sharing System_, Ritchie & Thompson, 1974:
>
> Likewise, the process‐control scheme and the command
> interface have proved both convenient and efficient.
> Because the shell operates as an ordinary, swappable user
> program, it consumes no ''wired‐down'' space in the system
> proper, and it may be made as powerful as desired at little
> cost. In particular, given the framework in which the shell
> executes as a process that spawns other processes to perform
> commands, the notions of I/O redirection, background pro‐
> cesses, command files, and user‐selectable system interfaces
> all become essentially trivial to implement.

I've always assumed it was "wired-down" in the sense of being
non-relocatable (*not* reallocatable) for the purposes of things like
device drivers that use memory mapped I/O.

Most of the memory used in the kernel is wired memory.
--
Chris

First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: iPad for Sysadmins
Next: Jailbreaking iPad - advantages?