[Qemu-devel] Guest OS kernel panic when using '-kernel-kqemu'

[Qemu-devel] Guest OS kernel panic when using '-kernel-kqemu'

[Paul Moore]

I'm currently running into a problem when running recent Linux kernels as a 
guest under QEMU with KQEMU kernel acceleration.  My problem is that early in 
the boot process the guest kernel panics, the host OS appears to be 
unaffected.  The panic can be different depending on the guest Linux 
distribution/kernel but the panic is repeatable and consistent.  In all 
cases, removing the '-kernel-kqemu' flag and only making use of userspace 
acceleration appears to solve the problem.  Using both the QEMU monitor and 
the linux lsmod reference count I have verified that KQEMU is being used by 
QEMU.

For reference, the host OS is a Gentoo 2007.0 x86 system with the Gentoo Linux 
kernel 2.6.23-r1, QEMU 0.9.0 and KQEMU 1.3.0_pre11; greater detail can be 
found in the Gentoo bug report[1].  I have also noticed this on other Gentoo 
based systems as well.

I am able to reproduce this on a regular basis using the Fedora 8 install DVD 
and the following command line:

 # qemu -m 256 -kernel-kqemu -cdrom Fedora-8-i386-DVD.iso -boot d f8.img

However, using a sample image[2] and the following command line I do not see 
any problems:

 # qemu -m 256 -kernel-kqemu linux-0.2.img

Any assistance or pointers to a solution would be greatly appreciated.  If 
there is anything I can do to help track down the problem further please let 
me know.

Thanks,
-Paul

[1] http://bugs.gentoo.org/show_bug.cgi?id=199151
[2] http://fabrice.bellard.free.fr/qemu/linux-0.2.img.bz2

[Qemu-devel] Guest OS kernel panic when using '-kernel-kqemu'

[Paul Moore]

I'm currently running into a problem when running recent Linux kernels as a 
guest under QEMU with KQEMU kernel acceleration.  My problem is that early in 
the boot process the guest kernel panics, the host OS appears to be 
unaffected.  The panic can be different depending on the guest Linux 
distribution/kernel but the panic is repeatable and consistent.  In all 
cases, removing the '-kernel-kqemu' flag and only making use of userspace 
acceleration appears to solve the problem.  Using both the QEMU monitor and 
the linux lsmod reference count I have verified that KQEMU is being used by 
QEMU.

For reference, the host OS is a Gentoo 2007.0 x86 system with the Gentoo Linux 
kernel 2.6.23-r1, QEMU 0.9.0 and KQEMU 1.3.0_pre11; greater detail can be 
found in the Gentoo bug report[1].  I have also noticed this on other Gentoo 
based systems as well.

I am able to reproduce this on a regular basis using the Fedora 8 install DVD 
and the following command line:

 # qemu -m 256 -kernel-kqemu -cdrom Fedora-8-i386-DVD.iso -boot d f8.img

However, using a sample image[2] and the following command line I do not see 
any problems:

 # qemu -m 256 -kernel-kqemu linux-0.2.img

Any assistance or pointers to a solution would be greatly appreciated.  If 
there is anything I can do to help track down the problem further please let 
me know.

Thanks,
-Paul

[1] http://bugs.gentoo.org/show_bug.cgi?id=199151
[2] http://fabrice.bellard.free.fr/qemu/linux-0.2.img.bz2