Re: [Xen-users] Kernel BUG at drivers/xen/netfront/netfront.c:783

[Danilo Godec <danilo.godec@xxxxxxxxx>]

Pasi Kärkkäinen pravi:
> On Wed, Jul 22, 2009 at 11:42:57AM +0200, Danilo Godec wrote:
>   
>> Hi,
>>
>> I have a RHEL5 based Xen server. Today I tried to install a new SLES10
>> SP2 DomU and had many troubles. First I tried to install in fully
>> virtualized mode and at first it seemed fine - I got the VNC windows, I
>> could select the installation options, but when it should start
>> installing, it would simply freeze.
>>
>> So I installed SLES 10 DomU on my local OpenSuSE 11.1 based Xen servers
>> in paravirtualized mode (this is how I'd like to run the VM) and
>> transferred the files to the remote Xen server.
>>
>> Again this seems to work initially - the DomU boots up, I can log in vie
>> the Xen console, but when I try to activate the network, I get this:
>>
>>     
>>> ----------- [cut here ] --------- [please bite here ] ---------
>>> Kernel BUG at drivers/xen/netfront/netfront.c:783
>>> invalid opcode: 0000 [1] SMP
>>> last sysfs file: /class/net/eth0/type
>>> CPU 1
>>> Modules linked in: loop dm_mod ext3 jbd xennet edd processor xenblk
>>> Pid: 2340, comm: ip Not tainted 2.6.16.60-0.21-xen #1
>>> RIP: e030:[<ffffffff88017452>]
>>> <ffffffff88017452>{:xennet:network_alloc_rx_buffers+519}RSP:
>>> e02b:ffff880056551d18  EFLAGS: 00010286
>>> RAX: 0000000000000000 RBX: ffff88005d701980 RCX: 0000000000000240
>>> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000001200
>>> RBP: 0000000056f1e000 R08: 0000000000000000 R09: 00000000000536e1
>>> R10: 0000000000000000 R11: 0000000000000001 R12: ffff880002d28500
>>> R13: 0000000000056f1e R14: ffff880002d28620 R15: ffff880002d2a6d0
>>> FS:  00002ad033dafae0(0000) GS:ffffffff803b9080(0000)
>>> knlGS:0000000000000000
>>> CS:  e033 DS: 0000 ES: 0000
>>> Process ip (pid: 2340, threadinfo ffff880056550000, task ffff880002de87d0)
>>> Stack: ffff88005d4b3978 ffff880002d28000 ffff8800011fe000 ffffffff00005ed0
>>>        0000000100000240 ffff88005cada838 0000000000000001 00000000560e9140
>>>        ffff8800560e9140 ffffffff8015cd12
>>> Call Trace: <ffffffff8015cd12>{filemap_nopage+406}
>>> <ffffffff8801a042>{:xennet:network_open+68}
>>>        <ffffffff80284fe0>{dev_open+56}
>>> <ffffffff8028304e>{dev_change_flags+89}
>>>        <ffffffff802c76db>{devinet_ioctl+569}
>>> <ffffffff8027b6da>{sock_ioctl+485}
>>>        <ffffffff801905dd>{do_ioctl+33} <ffffffff8019086f>{vfs_ioctl+584}
>>>        <ffffffff801908e9>{sys_ioctl+100}
>>> <ffffffff8010a1f4>{system_call+104}
>>>        <ffffffff8010a18c>{system_call+0}
>>>
>>> Code: 0f 0b 68 15 b7 01 88 c2 0f 03 48 63 ea 49 8d bc 24 78 15 00
>>> RIP <ffffffff88017452>{:xennet:network_alloc_rx_buffers+519} RSP
>>> <ffff880056551d18>
>>>  <0>Kernel panic - not syncing: Aiee, killing interrupt handler!
>>>       
>> Now I'm not sure - is this a problem of the DomU or could it be the Dom0
>> or even Xen hypervisor?
>>
>>     
>
> Sounds like a bug in SLES10 domU kernel..
>
> 1) Are you running RHEL 5.3 on the host? 
>   
Dom0 is Red Hat Enterprise Linux Server release 5.3 (Tikanga). Xen that
comes with it is 3.0.3 ...

> 2) Did you install SLES10 SP2 on the guest? 
>   
Yes, the DomU guest is SLES10 SP2. This one comes with Xen 3.2.0 (I
guess the 'kernel part' of Xen has the same version).

I'm also running one OpenSuSE 10.3 DomU - it has Xen 3.1.0. This one
works well for over a year now.

> Both have a lot of bugfixes over the .0 releases.
>   

I guess I could update Dom0 (RHEL) as there are some recent updates that
haven't been applied.

But I don't think I'll be able to update SLES without a working network.


 Danilo


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users