Re: UML troubles
Hmmm, I also tested my uml intallations and made some similar expierience.
If I use my 2.6.6 host kernel with skas3 patch I have no problems with booting
and running my uml guest sessions. But without the skas3 patch I have
sometimes (not always) exactly the same error. Sometime after a "killall -s 9
linux" and a restart it worked nicely and sometimes after a reboot it worked
again.
I have no clue whats the problem here. But because it is always at the point
where the clock should be set to the hardware clock I guess it has something
to do with this.
Hmmm... and because I didn't said it in my last mail:
after the installation of the uml package, you have to copy the modules to
your rootfs of the uml-guests and after that you have to run "depmod -a"
after the first reboot of your uml session, because otherwise you cannot use
the modules... Normaly the modules.dep should written correctly by the
package, but something goes wrong there since my upgrade from 2.4 to 2.6
kernels...
Give me hint if you find out something interessting about this problem. I
would like to have a stable package and not something that sometimes (I guess
only without skas patch) crashes at boot time.
--Ralph
Am Donnerstag 19 August 2004 20:04 schrieb Paul Dwerryhouse:
> On Wed, Aug 18, 2004 at 08:39:18PM +0200, Ralph Paßgang wrote:
> > If you are interessted in my user-mode-linux package, then add these line
> > to your sources.list
> >
> > deb http://packages.debianbase.de/<sid/sarge>/i386/uml-exp ./
> > deb-src http://packages.debianbase.de/sources/uml-exp
>
> I've been having terrible trouble getting UML running recently, too.
>
> I just tried out your 2.6.6 uml package, on a host running 2.6.7, and
> it died shortly after starting init:
>
> hwclock is unable to get I/O port access: the iopl(3) call failed.
> System time was Thu Aug 19 17:56:09 UTC 2004.
> Setting the System Clock using the Hardware Clock as reference...
> hwclock is unable to get I/O port access: the iopl(3) call failed.
> Kernel panic: fix_range fixing wrong address space, current = 0xa7e6a0a0
> Kernel panic: fix_range fixing wrong address space, current = 0xa7e6a0a0
> Kernel panic: fix_range fixing wrong address space, current = 0xa7e6a0a0
>
> Anyone have any idea what's causing that?
>
> This happened with a newly installed sid guest filesystem, under both
> ext2 and reiserfs (the problem also occurs with all of my old sid guest
> fs's, which were last updated about three months ago). I've moved
> /lib/tls to /lib/tls.off, which I believe is a fix for some glibc problem
> with 2.6 uml.
>
> Interestingly, the problem does not occur with an old Redhat guest fs
> that I have. It boots fine.
>
> Cheers,
>
> Paul
>
> --
> Paul Dwerryhouse | PGP Key ID:
> Amsterdam, The Netherlands (X) <-> Melbourne, Australia ( ) | 0x6B91B584
Reply to: