[Fedora-xen] Re: Xen issues, was Re: FC6 is frozen

[Pasi Kärkkäinen]

On Wed, Oct 04, 2006 at 10:55:56PM +0200, Paul Wouters wrote:
> On Tue, 3 Oct 2006, Paul Wouters wrote:
> 
> > On Tue, 3 Oct 2006, Jesse Keating wrote:
> >
> > > We're now into deep freeze state for FC6.  Any new packages brought in will
> > > need two nods from Bill Nottingham, Jeremy Katz, and Myself.  The only things
> > > that should be coming in now will be critical crasher bugs, multilib
> > > conflicts, or installation blockers.
> >
> > hmm. I have an FC 5.92 + yum updates running 2.6.18-1.2724.fc6xen with
> > xen that is completely unusable and gives me non-fatal kernel panics
> > and non-working networking (compared to 2.6.18-1.2724.fc6 not giving
> > panics and a working networking). xenguest-install.py also fails, and
> > the resulting kernels for xenu's hang on boot.
> 
> I upgraded to 2.6.18-1.2726.fc6xen. I found out the two ethernet cards
> got swapped, explaining the "broken" network connectivity.
> 
> I still get the
> 
> Oct  4 15:56:02 echelon kernel: BUG: warning at kernel/lockdep.c:1814/trace_hardirqs_on() (Not tainted)
> Oct  4 15:56:02 echelon kernel:  [<c040567e>] show_trace_log_lvl+0x58/0x177
> Oct  4 15:56:02 echelon kernel:  [<c0405c83>] show_trace+0xd/0x10
> Oct  4 15:56:02 echelon kernel:  [<c0405cc1>] dump_stack+0x19/0x1b
> Oct  4 15:56:02 echelon kernel:  [<c043638f>] trace_hardirqs_on+0xa4/0x120
> Oct  4 15:56:02 echelon kernel:  [<c0404e77>] restore_all+0x37/0x3a
> Oct  4 15:56:02 echelon kernel: DWARF2 unwinder stuck at restore_all+0x37/0x3a
> Oct  4 15:56:02 echelon kernel: Leftover inexact backtrace:
> Oct  4 15:56:02 echelon kernel: Inexact backtrace:
> Oct  4 15:56:02 echelon kernel:  [<c0405c83>] show_trace+0xd/0x10
> Oct  4 15:56:02 echelon kernel:  [<c0405cc1>] dump_stack+0x19/0x1b
> Oct  4 15:56:02 echelon kernel:  [<c043638f>] trace_hardirqs_on+0xa4/0x120
> Oct  4 15:56:02 echelon kernel:  [<c0404e77>] restore_all+0x37/0x3a
> 
> But it seems non-fatal. I also get the warning:
> 

I get this BUG also during system startup:

device-mapper: ioctl: 4.7.0-ioctl (2006-06-24) initialised:
dm-devel at redhat.com
BUG: warning at kernel/lockdep.c:1814/trace_hardirqs_on() (Not tainted)
 [<c040567e>] show_trace_log_lvl+0x58/0x177
 [<c0405c83>] show_trace+0xd/0x10
 [<c0405cc1>] dump_stack+0x19/0x1b
 [<c043638f>] trace_hardirqs_on+0xa4/0x120
 [<c0404e77>] restore_all+0x37/0x3a
DWARF2 unwinder stuck at restore_all+0x37/0x3a
Leftover inexact backtrace:
Inexact backtrace:
 [<c0405c83>] show_trace+0xd/0x10
 [<c0405cc1>] dump_stack+0x19/0x1b
 [<c043638f>] trace_hardirqs_on+0xa4/0x120
 [<c0404e77>] restore_all+0x37/0x3a
kjournald starting.  Commit interval 5 seconds
EXT3-fs: mounted filesystem with ordered data mode.
SELinux:  Disabled at runtime.
SELinux:  Unregistering netfilter hooks


hw: IBM HS20/8843 Blade (Intel Xeon)
kernel: 2.6.18-1.2726.fc6xen

Happened also with earlier kernels. BUG seems to be non-fatal, system seems
to work OK after that.

- Pasi Kärkkäinen