Monolithic Vs. Microkernel

Monolithic Vs. Microkernel

[Rakesh Kumar Banka]

Anyone working on the Microkernel implementation
of Linux? Specially in the area of seperating the process
and the file management activities out of the kernel.

Rakesh.

Re: Monolithic Vs. Microkernel

[Larry McVoy]

On Mon, Feb 25, 2002 at 11:11:59PM -0700, Rakesh Kumar Banka wrote:
> Anyone working on the Microkernel implementation
> of Linux? Specially in the area of seperating the process
> and the file management activities out of the kernel.

Not if they learned from history, they aren't.  But the Hurd could use
your help, they're a microkernel.
-- 
---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Re: Monolithic Vs. Microkernel

[Alan Cox]

> On Mon, Feb 25, 2002 at 11:11:59PM -0700, Rakesh Kumar Banka wrote:
> > Anyone working on the Microkernel implementation
> > of Linux? Specially in the area of seperating the process
> > and the file management activities out of the kernel.
> 
> Not if they learned from history, they aren't.  But the Hurd could use
> your help, they're a microkernel.

There are several Linux on microkernel implementations around, thankfully 
using something that can really be called a microkernel. With the "we
want to run 10,000 copies of Linux on a box" market boom it may well prove
to have a practical use one day - as well as the security partitioning one
which some people overlook (and paranoid security people often do not mind
a small performance hit)

Alan

Re: Monolithic Vs. Microkernel

[David Lang]

this is a FAQ

basicly it boils down to the fact that Linus believes that a microkernel
ends up spending to much of it's time formatting messages for other pieces
of itself instead of doing the work you purchased the computer to do.

David Lang

 On
Mon, 25 Feb 2002, Rakesh Kumar Banka wrote:

> Date: Mon, 25 Feb 2002 23:11:59 -0700 (MST)
> From: Rakesh Kumar Banka <Rakesh@asu.edu>
> To: linux-kernel@vger.kernel.org
> Subject: Monolithic Vs. Microkernel
>
>
> Anyone working on the Microkernel implementation
> of Linux? Specially in the area of seperating the process
> and the file management activities out of the kernel.
>
> Rakesh.
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Re: Monolithic Vs. Microkernel

[Eric Krout]

On Tue, 2002-02-26 at 01:11, Rakesh Kumar Banka wrote:
> 
> Anyone working on the Microkernel implementation
> of Linux? Specially in the area of seperating the process
> and the file management activities out of the kernel.
> 
> Rakesh.

AST?  ;-)

(Sorry, that joke is based on 1991 material)

Re: Monolithic Vs. Microkernel

[Rik van Riel]

On Mon, 25 Feb 2002, Rakesh Kumar Banka wrote:

> Anyone working on the Microkernel implementation
> of Linux? Specially in the area of seperating the process
> and the file management activities out of the kernel.

You have to remember that the source code for Linux is available.

This means we can have all the advantages of modularity at the
source level without needing any of the potential disadvantages
of modularity at the binary level.

regards,

Rik
-- 
"Linux holds advantages over the single-vendor commercial OS"
    -- Microsoft's "Competing with Linux" document

http://www.surriel.com/		http://distro.conectiva.com/

Re: Monolithic Vs. Microkernel

[Pavel Machek]

Hi!

> > Anyone working on the Microkernel implementation
> > of Linux? Specially in the area of seperating the process
> > and the file management activities out of the kernel.
> 
> You have to remember that the source code for Linux is available.
> 
> This means we can have all the advantages of modularity at the

Not *all* of them. On vsta, you could do

( killall keyboard; sleep 1; keyboard ) &

to change your keymap. On linux you need special tools for managing
modules and are not protected from module bugs. Try developing filesystem
on production box.... You can do that on u-kernels.
								Pavel
-- 
Philips Velo 1: 1"x4"x8", 300gram, 60, 12MB, 40bogomips, linux, mutt,
details at http://atrey.karlin.mff.cuni.cz/~pavel/velo/index.html.

Re: Monolithic Vs. Microkernel

[Alexander Viro]

On Mon, 4 Mar 2002, Pavel Machek wrote:

> Not *all* of them. On vsta, you could do
> 
> ( killall keyboard; sleep 1; keyboard ) &
> 
> to change your keymap. On linux you need special tools for managing
> modules and are not protected from module bugs. Try developing filesystem
> on production box.... You can do that on u-kernels.

Userland filesystems != microkernel.

Re: Monolithic Vs. Microkernel

[Pavel Machek]

Hi!

> > Not *all* of them. On vsta, you could do
> > 
> > ( killall keyboard; sleep 1; keyboard ) &
> > 
> > to change your keymap. On linux you need special tools for managing
> > modules and are not protected from module bugs. Try developing filesystem
> > on production box.... You can do that on u-kernels.
> 
> Userland filesystems != microkernel.

Yep, but microkernel => userland filesystems ;-). Anyway, they *can*
do things linux can't do (or linux has hard time with), like
partitioning physical machine into few logical ones, filesystems in
userland, ability to debug drivers on production machines, etc.

I like those features, but I'm not sure if costs introduced by
u-kernels are worth it.
									Pavel
-- 
(about SSSCA) "I don't say this lightly.  However, I really think that the U.S.
no longer is classifiable as a democracy, but rather as a plutocracy." --hpa

Re: Monolithic Vs. Microkernel

[Rik van Riel]

On Mon, 4 Mar 2002, Pavel Machek wrote:

> > This means we can have all the advantages of modularity at the
>
> Not *all* of them. On vsta, you could do
>
> ( killall keyboard; sleep 1; keyboard ) &

How is that different from the following ?

(rmmod keyboard ; sleep 1 ; modprobe keyboard)

[no, no need to talk about hardware access ... vsta's keyboard
driver also has hardware access]

regards,

Rik
-- 
Will hack the VM for food.

http://www.surriel.com/		http://distro.conectiva.com/

Re: Monolithic Vs. Microkernel

[Xavier Bestel]

le mar 05-03-2002 à 22:04, Rik van Riel a écrit :
> On Mon, 4 Mar 2002, Pavel Machek wrote:
> 
> > > This means we can have all the advantages of modularity at the
> >
> > Not *all* of them. On vsta, you could do
> >
> > ( killall keyboard; sleep 1; keyboard ) &
> 
> How is that different from the following ?
> 
> (rmmod keyboard ; sleep 1 ; modprobe keyboard)
> 
> [no, no need to talk about hardware access ... vsta's keyboard
> driver also has hardware access]

killall works when keyboard is deadlocked, I suppose.

Re: Monolithic Vs. Microkernel

[Pavel Machek]

Hi!

> > > This means we can have all the advantages of modularity at the
> >
> > Not *all* of them. On vsta, you could do
> >
> > ( killall keyboard; sleep 1; keyboard ) &
> 
> How is that different from the following ?
> 
> (rmmod keyboard ; sleep 1 ; modprobe keyboard)
> 
> [no, no need to talk about hardware access ... vsta's keyboard
> driver also has hardware access]

Standart tools for standard tasks. And if there's while(1); in VSTa's
keyboard handler will not prevent killall.
								Pavel
-- 
Casualities in World Trade Center: ~3k dead inside the building,
cryptography in U.S.A. and free speech in Czech Republic.

Re: Monolithic Vs. Microkernel

[Bernd Eckenfels]

In article <20020304144923.A96@toy.ucw.cz> you wrote:
> modules and are not protected from module bugs. Try developing filesystem
> on production box.... You can do that on u-kernels.

Well, I agree that debugging get's easier if you do not crash your box every
now and then. But on the other hand... who is developing on production boxes?!
Is that usual suse policy? :)

Greetings
Bernd

Re: Monolithic Vs. Microkernel

[Hans Adams]

Hello,

at least two teams work upon it here in Germany

references:

http://os.inf.tu-dresden.de/index.xml.de
http://i30www.ira.uka.de

A branch of stable kernels upon L4/Fiasco is maintained at TU Dresden
and available at:
http://os.inf.tu-dresden.de/L4/LinuxOnL4/. 

Though Mr. Torvalds seems not to be amused about micro kernel aproach,
this work is not done in vein due to the inherent technical superiority
in certain applications mostly regarding distributed systems including
NUMA architectures.

best, Hans Adams