Which microkernel is used if not pistachio?

Hi list,

when I wrote about a problem I encountered with pistachio about two
weeks ago, I was told that L4Ka::Pistachio is not (mainly) used among
the developers due to security issues. Since I wanted to play around
with Hurd on L4 a little and (if happy) try to contribute to the project
myself (if I can...), I would really like to use the right kernel.

Cheers,
Christian Huetter

--

-- 
   ,= ,-_-. =.                        ,= ,-_-. =.
  ((_/)o o(\_))  GNU Privacy Guard   ((_/)o o(\_))
   `-'(. .)`-'  http://www.gnupg.org  `-'(. .)`-'
       \_/                                \_/

_______________________________________________
L4-hurd mailing list
L4-hurd <at> gnu.org
http://lists.gnu.org/mailman/listinfo/l4-hurd

Re: Which microkernel is used if not pistachio?

Hi,

> when I wrote about a problem I encountered with pistachio about two
> weeks ago, I was told that L4Ka::Pistachio is not (mainly) used among
> the developers due to security issues.

The Hurd3 is based on Coyotos micro-kernel,
                    the successor to EROS (see: http://coyotos.org)
The Hurd2 is based on L4 microkernel
and,
The Hurd1(The GNU/Hurd) is based on Mach microkernel..

Hope, after all this research, Hurd-kernel can define an interface/model,

1. which makes it independent of underlying micro kernel,

2. and also hope, it is almost independent of
RT-Hurd(embedded) and the Hurd(desktop),

3a. and also hope, _user_ has the freedom to do all the things
within his capabilities, in a simple, easy and a straightforward manner.

3b. and also hope, the interface/model is _not_ restricting some silly
things, which
forces an user to use qemu to debug his program, etc.

( I used both #3a and #3b to express third requirement. Can someone clarify?)

> Since I wanted to play around
> with Hurd on L4 a little and (if happy) try to contribute to the project

Re: Which microkernel is used if not pistachio?

At Mon, 03 Jul 2006 09:20:59 +0200,
Christian Hütter <C_Huetter <at> web.de> wrote:
> 
> [1  <multipart/signed (7bit)>]
> [1.1  <text/plain; us-ascii (7bit)>]
> Hi list,
> 
> when I wrote about a problem I encountered with pistachio about two
> weeks ago, I was told that L4Ka::Pistachio is not (mainly) used among
> the developers due to security issues. Since I wanted to play around
> with Hurd on L4 a little and (if happy) try to contribute to the project
> myself (if I can...), I would really like to use the right kernel.

Right now, the main developers are doing something else entirely in
real life, so currently the project is mostly stalled.

We are looking at recent developments in the EROS/Coyotos project
(www.coyotos.org), as well as recent developments in all of the L4
groups (Karlsruhe, Dresden, Sydney) with keen interest.

I should say however, that there are many questions that are more
important than the choice of a particular microkernel implementation.
The order of progress should be: Get certainty about the Hurd design,
then figure out what the requirements are to implement such a design,
then look for a microkernel that fits best.

I think it is fair to say at this point that this project has, in its
history, been heavily influences by both of the above microkernel
projects, and that our current design puts us somewhere in the middle
between these projects, mostly on common ground, actually.

seL4, L4.sec and coyotos mess

Hi,

There seems to be several L4 successors  with security in mind. Is there 
a comparison between them in any place? Coyotos web said there is some 
similarities with l4.sec. Again... is there any place with a comparison? 
Can somebody help see the light? :)

This may be not the right list to make these questions but... I think 
some of you  had done this comparison between  kernels to  be able to 
choose one of them as a basis for the hurd.

Cheers,
Gustavo Romero

--

-- 
Gustavo Romero López
Arquitectura y Tecnología de Computadores
ETSI. Informática. Universidad de Granada
C\Periodista Daniel Saucedo Aranda. 18071
Tel: (+34) 958240589 Fax: (+34) 958248993

Re: seL4, L4.sec and coyotos mess

Gustavo Romero wrote:
> Hi,
> 
> There seems to be several L4 successors  with security in mind.

Yes.

> Is there
> a comparison between them in any place?

I'm not aware of any.

> Coyotos web said there is some
> similarities with l4.sec. Again... is there any place with a comparison?
> Can somebody help see the light? :)
> 

You could read what is available of information (sadly, the L4
successors aren't developed in such an open fashion like coyotos):
http://www.coyotos.org/docs/ukernel/spec.html
http://l4hq.org/docs/manuals/l4_sec_20051019.pdf and
http://os.inf.tu-dresden.de/papers_ps/kauer-diplom.pdf
http://i30www.ira.uka.de/teaching/coursedocuments/105/l4ng-apr28.pdf

> This may be not the right list to make these questions but... I think
> some of you  had done this comparison between  kernels to  be able to
> choose one of them as a basis for the hurd.
> 

l4 hurd on powerpc

_______________________________________________
L4-hurd mailing list
L4-hurd <at> gnu.org
http://lists.gnu.org/mailman/listinfo/l4-hurd

Re: l4 hurd on powerpc

Christophe Ple wrote:
> Is there any status on porting of l4-hurd on powerpc ?

hurd/pistachio is dead, so is the ppc port.
--
-ness-

Re: Re : upgrade

Hello!

On Wed, Jul 12, 2006 at 06:41:43PM +0530, doxa <at> sancharnet.in wrote:
> I had just been a week at IIIT, Hyderabad - http://iiit.ac.in. The
> faculty there has agreed to pool in research students and scholars there
> to assist in the development of Hurd.

That's very interesting to hear.  

But first we should clarify a few issues.  Are you aware that the current
Hurd's implementation (and design) are currently not as actively pushed
forward as they were is in the past?

A number of people instead began to reconsider issues in the Hurd's
design (see the l4-hurd mailing list's archives for example,
<http://lists.gnu.org/mailman/listinfo/l4-hurd>).  This includes both the
actual Hurd's design, as well as the underlying micro kernel.  A very
early, incomplete draft of this ``new Hurd'''s design is available at
<http://www.marcus-brinkmann.org/hurd-ng.pdf> and the beginning of
documenting some issues at
<http://hurd.gnufans.org/bin/view/Hurd/NextHurd>.  Unfortunately its main
pioneers, Marcus Brinkmann and Neal Walfield (and a number of other
people, mostly on the l4-hurd mailing list), can't -- at the moment --
dedicate as much time to this project as it would need.  So, there is
currently no possibility to directly contribute to this effort, as far as
I can see.

If you instead think it is feasible to invest time into the current Hurd
on Mach implementation, there are tons of loose ends where work would be
possible.

> It has been over 20 years of development. I have undertaken to do a
> qualitative study of the development during this period.
> Q2) Where can I find the valid archives ?

You mean like mailing list archives?  They're accessible from
<http://savannah.gnu.org/mail/?group=hurd> and
<http://lists.debian.org/debian-hurd/>.  Very old stuff, the archives of
the hurd-folks mailing list which is documenting parts of the Hurd's
genesis, are available at
<http://web.cecs.pdx.edu/~trent/gnu/hurd/hurd-talk.tar.gz>.

There is no written history, as far as I know.  Only bits here and there.

Regards,
 Thomas