AW: L4Linux on L4Ka::Pistachio

[Christian, Martin, OPEE45]

Thanks a lot Adam!
I'll think about which road to take during Christmas.

Nice holidays,

Martin.

-----Ursprüngliche Nachricht-----
Von: l4-hackers-bounces at os.inf.tu-dresden.de [mailto:l4-hackers-bounces at os.inf.tu-dresden.de] Im Auftrag von Adam Lackorzynski
Gesendet: Samstag, 16. Dezember 2006 00:10
An: l4-hackers at os.inf.tu-dresden.de
Betreff: Re: L4Linux on L4Ka::Pistachio

Hi,

On Fri Dec 15, 2006 at 10:11:14 +0100, Christian, Martin, OPEE45 wrote:
> >> a) Is the L4Env required for running L4Linux?
> > Yes but that's no strict requirement.
> How much time would it probably take, to adjust L4Linux without L4Env to
> the L4 Interface of Pistachio? Did anybody try this?

Something like this is in 2.4. I do not think this is particularly hard.
The concepts are the same. One needs to get some memory from somewhere,
start tasks somewhere, etc. I don't see big obstacles.

> >> e) If Fiasco would be running on PowerPC, would it be possible
> >> to run L4Linux without changes?
> > If one wants to have binary compatibility then some PPC code needs
> > to be there. And then that would not be the way I'm currently doing
> > things.
> As far as I understood Michael Hohmuths Diplomarbeit, all Hardware
> access from Linux is replaced by sycalls to L4. Is that still the case?

I'd rather they that it's adapted to run in user-level, using L4. Some
of the instructions cannot be executes on user-level, they're replaced,
or removed, or whatever.

> And so far only the IA32 part of Linux is changed that way, isn't it? So

L4Linux is also running on ARM.

> I would basically need to rewrite the PowerPC architecture part of
> Linux, basically. Isn't it the way things are supposed to be done?

Basically yes but all the architecture independant code is already
there, and two example architectures.





Adam