My Vintage Dream PC

[jmfbahciv]

Right. A multiprocessor system would behave similarly. But in a
multiprocessor-oriented system, the processes would be making far
fewer calls to the kernal... a kernal that has nothing to do but
service those infrequent calls.

Take the huge GUI, the Taiwanese device drivers, the interrupt
handlers, all that dangerous stuff out of the OS kernal. A lot of that
can even be removed on uniprocessor systems.

But your idea is to have the Boss CPU have control of the whole
system; if it does, then the other CPUs cannot run the device drivers
without the Boss knowing about it. Having the control of the
system means that the scheduling for I/O and memory management
has to be done by the Boss, not the other CPUs. Thus, when
a slave CPU needs any resources, it has to ask the Boss for
it. This will cause the system to grind down to almost a halt
because the other CPUs will be in a constant wait state waiting
for the Boss to service their requests.

/BAH

 

[jmfbahciv]

OK, define the kernal as the chunk of code that sets up processes,
starts them, checks them for health, and kills them. It probably sets
up memory management and disk quotas, although it could even delegate
that to another processor. Remove the GUI, the drivers, most of the
interrupt handlers, the file systems, the comm stacks. The number of
system calls into this code is tiny compared to what Windows or even
Linux has to handle. Move all that other traditional "OS" stuff into
50 or 60 other processors that can't take down the system and don't
need to be context switched.

That's what I'm trying to get you to think about.

We didn't have to think about it; that was the situation with
our master/slave implementation.

There was a Datamation article written by Alan Wilson about
our SMP implementation. I have no idea how you could find
it online but it should be somewhere out there. My hardcopy
is packed in one of my unpacked boxes. It might be a good
idea to read it if you can find it.

/BAH


/BAH

 

[jmfbahciv]

Thermite is probably the cheapest way of doing that.

I met a device that could do that. Nice piece of equipment.

/BAH

 

[jmfbahciv]

ISTR hearing XP decribed in exactly those terms around the time it
was being launched.

Yep. :-(

/BAH

 

[jmfbahciv]

Written by engineers!

then the biz has improved a tad. However, those engineers still
don't know how to beat up a system like a timesharing monitor
can.

/BAH

 

[jmfbahciv]

It sure acted like one.

Not if you used in the same machine room which also had TOPS-10
and TOPS-20 systems.

The letter "T" stands for "Time Sharing." I
still have some books around that say so.

that was an -11 nomenclature. Substitute task for time and
it would have been more descriptive.

Somebody famously said that, no matter how powerful a timeshared CPU
is, it will now and then piss off its users by being too slow.

Systems will also be too slow because the user will hone usage habits
that will create some waiting. It's the same kind of phenomena w.r.t.
disk usage. The bigger the disk, the less space a user will have
for storage because the user will never clean up the disk areas until
forced to.

Most systems that are not connected to a network are "secure." That's
easy; lock the door when you leave.

Are you trying to be silly?

However,

Gosh, are you suggesting that some day we'll have a new approach to OS
design? Tell us more.

I've been trying to explain why MS' stuff is the way it is. But
I guess I'm not able to find the right word salad to get this
from my head to your TTY screen. This stuff is obvious to me.

/BAH

 

[jmfbahciv]

She said "really." She's being a timeshare snob just because I had a
PDP-11 and she had a VAX.

John

There is no reason to be insulting. You don't know what you're
talking about now.

/BAH

 

[jmfbahciv]

Virtualization gadgets are creeping up on the nanokernal function.
They let you run multiple copies of moderately unreliable OSs, and
kill/restart them when they misbehave.

And those are applications because they don't really run at the
exec level of the computer system.

We're starting to virtualize multiple copies of Windows just so they
can run incompatible apps, like different versions of the Xilinx
stuff, with clean installs.

/BAH

 

[jmfbahciv]

John Larkin wrote:

Walter Bushell wrote:
Walter Bushell wrote:
On Mon, 25 May 2009 16:26:50 -0400, Peter Flass

John Larkin wrote:
The ultimate OS should maybe be hardware, fpga probably, or an
entirely separate processor that runs nothing but the os.

CDC-6600.
In a few years, when most any decent CPU has 64 or so cores, I suspect
we'll have one of them run just the OS. But Microsoft will f*** that
up, too.

John
Why only one? Surely the kernel will be multithreaded.
You meant to say reentrant.

/BAH
Well that too.
Not "too" but first.

/BAH
An os nanokernal needs to be neither.
Wrong; then you don't have an OS.


You are playing with words. An OS should be hierarchial, with the
top-most thing (what I call the kernal) being absolutely in charge of
the system.

But it cannot have that kind of control with more than [number picked
out of the air as a guesstimate] 8 threads or processes or services.

It could, in my opinion should, run on a dedicated CPU.

It is impossible to do this without allowing the other CPUs to be
able to make their own decisions about what they're processing.

It's times like this that I regret that "duh" has fallen into disuse.

But you are the one who insisted that the Boss CPU have control
of the whole system and what it does. You cannot have it
both ways.

Of course the kernal has visibility to resources; it manages them. It
just doesn't share its cozy protected CPU with lower-level stuff like
device drivers, file systems, apps. Why is that such a hard thing to
understand?

It isn't hard for me to understand. What you don't know is
the problems of communicating between CPUs. If the Boss CPU
has to have complete control, then the other CPUs cannot
do any resource management; they have to gain permission
with the accompanying data from the Boss CPU before they
can resume execution.

/BAH

 

[jmfbahciv]

re:
http://www.garlic.com/~lynn/2009h.html#68 My Vintage Dream PC

I guess I had also tweaked the mainstream organization when during FS, I
drew parallel with cult movie playing in central sq (something about
the inmates being in charge of the institutions). it didn't help that
FS was a mega-effort that was cancelled w/o even being announced ...
some past posts
http://www.garlic.com/~lynn/submain.html#futuresys

I had sponsored Boyd's briefings at IBM in the early 80s.... so
his line about "be or do" resonated ... referenced in this
recent post
http://www.garlic.com/~lynn/2009h.html#5 mainframe replacement (Z/Journal Does it Again)

quote used at dedication of Boyd Hall, USAF Weapons School, Nellis
Air Force Base, 17Sep1999

"There are two career paths in front of you, and you have to choose
which path you will follow. One path leads to promotions, titles, and
positions of distinction.... The other path leads to doing things that
are truly significant for the Air Force, but the rewards will quite
often be a kick in the stomach because you may have to cross swords with
the party line on occasion. You can't go down both paths, you have to
choose. Do you want to be a man of distinction or do you want to do
things that really influence the shape of the Air Force? To be or to do,
that is the question." Colonel John R. Boyd, USAF 1927-1997

... snip ...

past posts mentioning Boyd
http://www.garlic.com/~lynn/subboyd.html#boyd
and misc. URLs from around the web mentioning Boyd
http://www.garlic.com/~lynn/subboyd.html#boyd2

Yep. That seemed to be true in our neck of the woods, too. The
problem was that a majority of people couldn't understand this.

/BAH

 

[jmfbahciv]

That's because no matter how powerful it is, some PHB will put off
upgrading until the wheels come off.

Not at all. It has to do with the way people do their work.

/BAH

 

[jmfbahciv]

Just a screwy question, how many processors does your monitor support?

DEC politics (and that was a RPITA) demanded that we publish support
for 3. We had a customer site who ran with 5. There was a limit,
due to the width of a field used in the monitor's data base, of 8
because the field was only 8 bits wide...or was it 6?...can't remember.

One of the things on my Todo list is to get a scan of the picture
JMF had of the system configuration of that 5-CPU system.
You need the time stamp for this stuff; early 80s.

/BAH

 

[jmfbahciv]

Not only that, it is still in the idle loop well over 95% of the time.
Who needs more idle cores?

Right. When I crashed it, I was trying to get both running something.
Didn't work well at all.

/BAH

 

[jmfbahciv]

What do you admire about Windows?

[emoticon stops to think of something]

[20 minutes later, emoticon gives up]

There has to be something but I can't think of one.

It has taught the average layperson to appreciate that computers are neither
infallible nor inherently reliable.

IMO, that is one of the main BAD things about MS crap. They will never
learn that they shouldn't have to put up with all the crap they have
to do to get their work done. anything that takes 20 minutes to get
1 second's worth of work done sucks big time.

/BAH

 

[jmfbahciv]

I think you missed my irony.

Yep. Completely and utterly missed it.

/BAH

 

[jmfbahciv]

I would have if i had been allowed to. I wasn't. I had to make a
system that met managements choices. Many of which were driven by
"grandfathered" existing systems. Also it was required to coexist
with, without disrupting the powerful (within the organization)
CODASYL / IDMS mainframe group. Generation datasets were fun.

And we had to provide a system that could meet your choices
and every other customer's choice.

/BAH

 

[jmfbahciv]

Speed scaling has slowed to a crawl; the GHz race is mostly over.
Lithography has a ways to go yet, so higher-end chips will be
multicore. Keeping them all busy is a separate issue from how an OS is
partitioned.

I'm guessing that you couldn't run all 64 or 256 or whatever cores
compute-bound without toasting the silicon anyhow. Given that, if we
do get many-core chips, the only use for all those cores is to
restructure the OS.

Geez, doesn't anybody have ideas?

Yes but you don't seem to want to investigate. For a real life
implementation, read all of Morten's descriptions about his
experiences and work ...I'd say over the last 8 years or so.
Pay particular attention to the ones that appear to be about
global politics, economies and work getting done in Asia.

/BAH

 

[jmfbahciv]

I make the opposite argument. Applications needn't be parallelized
across multiple processors, they each need to own a single processor.

And that is a bad assumption. Think about networks or apps that
exchange data.

Of course, some apps might use all the cores they can get - games,
Spice, 3D modeling, FFTs - but lots of apps would be happy to have a 2
GHz CPU all to themselves.

Isn't parallelism already one of the darlings of CS?

Parallelism is a term that seems fall in the same category
as "distributed processing" did in the early 80s.

http://en.wikipedia.org/wiki/Parallel_computing

http://en.wikipedia.org/wiki/Task_parallelism

Instead of relying on Wiki learn from those who actually
did the work, not the ones who are entering descriptions
in that place.

/BAH

 

[jmfbahciv]

Accusing Barb of being involved with VAXes is a good way to get hurt.

ROTFLMAO. I'm being kind today and giving him a break.

/BAH

 

[jmfbahciv]

That is way far from the usual definition of daemon. Check your
dictionaries.

Since we implemented a few of them, I know what the functionality
of our daemons were. You asked me what I would have called them.
I told you.

You are missing the boat here, in the current world there are several
cases of things like virtualbox, which run things like BSD, Solaris,
MSwin XP, Freedos, (as applications) and all their (sub)applications
"simultaneously" (time sharing, and supporting multiple CPU cores).
This would place it at the monitor level you have referenced.

No. Those are running as apps w.r.t. the computer system they are
executing on. Those apps will never (or should never) be running
at the exec level (what the hell does Unix call "exec level"?)
of the computer system. That is exclusively the address space
and instruction execution of the monitor (or kernal) running
on that system.

MSwin never was much of a proper OS. Just remember that there are
more things claiming to be an OS besides Multics, RSTS, TOPS-10, VMS,
MVS, VM-CMS, Unix(es), and MSwin.

MS got Cutler's flavor of VMS and called NT. They started out with
a somewhat [emoticon's bias alert here] proper monitor but spoiled
it when Windows' developers had to have direct access to the
nether parts of the monitor.

/BAH