On Wed, Dec 24, 2008 at 4:20 AM, Alexander Graf <ag...@suse.de> wrote:
>
>
> Ugh. Looks like the emulation part is still broken :-(. Please use the
> attached patch to disable the emulation optimization for now.
>
> Avi, could you please apply that patch for kvm-82 too, so we get something
> working out? I'll take a closer look at what's broken exactly later on.
>
> Alex
>
>
So I am working with the latest git, from today.
The emulation error went away and the nested KVM guest partially works.
The errors that I am seeing late in the normal guest boot (which seem
non-fatal) are:
Dec 29 18:33:31 amdbox kernel: [ 1060.446054] bad partial csum:
csum=5888/5694 len=80
Dec 29 18:33:33 amdbox kernel: [ 1061.934164] bad partial csum:
csum=5888/5694 len=80
Dec 29 18:33:33 amdbox kernel: [ 1062.170127] bad partial csum:
csum=5888/5694 len=60
Dec 29 18:33:34 amdbox kernel: [ 1063.419124] bad partial csum:
csum=5888/5694 len=270
Dec 29 18:33:35 amdbox kernel: [ 1063.667817] bad partial csum:
csum=5888/5694 len=270
Dec 29 18:33:35 amdbox kernel: [ 1063.927839] bad partial csum:
csum=5888/5694 len=270
Dec 29 18:33:35 amdbox kernel: [ 1064.126336] bad partial csum:
csum=5888/5694 len=252
Dec 29 18:33:35 amdbox kernel: [ 1064.274429] bad partial csum:
csum=5888/5694 len=152
Dec 29 18:33:35 amdbox kernel: [ 1064.522702] bad partial csum:
csum=5888/5694 len=152
Dec 29 18:33:36 amdbox kernel: [ 1064.776290] bad partial csum:
csum=5888/5694 len=152
Dec 29 18:33:38 amdbox kernel: [ 1067.309123] __ratelimit: 4 callbacks
suppressed
Dec 29 18:33:38 amdbox kernel: [ 1067.309126] bad partial csum:
csum=5888/5694 len=252
Dec 29 18:33:39 amdbox kernel: [ 1068.160737] bad partial csum:
csum=5888/5694 len=241
Dec 29 18:33:41 amdbox kernel: [ 1070.170049] bad partial csum:
csum=5888/5694 len=60
After that I am able to start the nested guest with:
sudo qemu-system-x86_64 -hda ubuntu-server.img -cdrom
Desktop/ubuntu-8.10-server-amd64.iso
The nested guest also has the latest git checkout
The nested guest shows the Ubuntu install CD welcome and selecting a
language and starting
the boot process starts a very little bit and the screen goes black.
The nested guest doesn't crash, but becomes very unresponsive, can't
ping it, can't ssh, etc.
It seems like it only runs for a short time before it becomes
unresponsive (less than 30
seconds).
I can attach to the qemu-system-x86_64
(gdb) where
#0 0x00007fa8cc4a1482 in select () from /lib/libc.so.6
#1 0x0000000000408bcb in main_loop_wait (timeout=0)
at /backup/src/kvm-src/kvm-userspace/qemu/vl.c:3617
#2 0x00000000005160fa in kvm_main_loop ()
at /backup/src/kvm-src/kvm-userspace/qemu/qemu-kvm.c:599
#3 0x000000000040d106 in main (argc=<value optimized out>,
argv=0x7fffd58e9f48, envp=<value optimized out>)
at /backup/src/kvm-src/kvm-userspace/qemu/vl.c:3779
After some time, the qemu-system-x86_64 process starts to take
between 97 and 100% of the CPU.
The base system is still running OK, but no new messages are printed
in /var/log/syslog
I am sure there are more KVM debugging tricks....
Any suggestions?
Thanks,
Todd
--
Todd Deshane
http://todddeshane.net
http://runningxen.com
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
