Bug#607709: XEN kernel crash on DNS process

[Jonathan Nieder <jrnieder@gmail.com>]

reassign 607709 src:linux-2.6 2.6.32-27
found 607709 linux-2.6/2.6.32-29
quit

Hi Giuseppe,

Giuseppe Sacco wrote:

> I confirm the problem is really XEN related. We made more tests with non
> XEN kernel and we never got the error reported here.
> Now we are going to run again a XEN dom0 kernel, updating to latest
> kernel (from 2.6.32-27 to 2.6.32-29). I will report here more
> information whenever I collect them.
[...]
> kernel BUG at [...]/arch/x86/xen/enlighten.c:309!
> invalid opcode: 0000 [#1] SMP
> last sysfs file: /sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/bInterfaceClass
> Modules linked in: bridge stp xen_evtchn xenfs vboxnetadp vboxnetflt vboxdrv lp parport i5100_edac shpchp psmouse edac_core processor pci_hotplug dcdbas serio_raw acpi_processor button evdev ext3 jbd mbcache sd_mod crc_t10dif sg osst sr_mod dm_mod st cdrom usbhid hid mptsas ata_generic ehci_hcd mptspi mptscsih scsi_transport_sas uhci_hcd ata_piix usbcore libata nls_base tg3 libphy mptbase scsi_transport_spi scsi_mod thermal thermal_sys radeonfb fb_ddc i2c_algo_bit i2c_core
> 
> Pid: 2153, comm: server Not tainted (2.6.32-5-xen-686 #1) PowerEdge T300
> EIP: 0061:[<c1003b51>] EFLAGS: 00010282 CPU: 3
> EIP is at set_aliased_prot+0x9e/0x10d
[...]
> Process server (pid: 2153, ti=f3ec8000 task=f13ddd80 task.ti=f3ec8000)
[...]
> Call Trace:
>  [<c1003be2>] ? xen_free_ldt+0x22/0x30
>  [<c100b2ab>] ? destroy_context+0x38/0x79
>  [<c103564a>] ? __mmdrop+0x1d/0x3a
>  [<c1032456>] ? finish_task_switch+0x76/0x95
>  [<c128d759>] ? schedule+0x762/0x7dc
>  [<c1006048>] ? xen_force_evtchn_callback+0xc/0x10
>  [<c128e080>] ? schedule_hrtimeout_range+0xa1/0xda
>  [<c104d7e2>] ? hrtimer_wakeup+0x0/0x18
>  [<c104e13d>] ? hrtimer_start_range_ns+0xf/0x13
>  [<c10c6dda>] ? poll_schedule_timeout+0x26/0x3e
>  [<c10c772e>] ? do_select+0x3e7/0x42d
>  [<c10d60a8>] ? __getblk+0x3c/0x2f3
>  [<c10c7b1a>] ? __pollwait+0x0/0xa5
>  [<c10d8581>] ? __bread+0xa/0x5d
>  [<f865ff47>] ? ext3_get_branch+0x5d/0xc4 [ext3]
[...]

Thanks and sorry for the slow response.  No ideas come to mind, so some
basic questions instead:

 - do you still use this system?  If so, how are you coping?
 - what kernel and hypervisor do you use these days?
 - is it still reproducible with current squeeze or sid kernels and
   hypervisors?
 - any other weird observations?

Hope that helps,
Jonathan