Bug? / Warning: VirtualBox 3.0.12 and kernel 2.6.31.6

[damNageHack]

A warning from me to the community ...

I use a self compiled kernel with kernel.org sources version 2.6.31.6 and VirtualBox version 3.0.12. Found this struggling bug.

Kernel boot failed with "BUG" message of a null pointer reference and some "OOPS" i do not remember of exactly.

It was the first time I had such problems with kernel. And yes I know, do not compile if you do not know what you are doing

http://forums.virtualbox.org/viewtopic.php?f=7&t=23223

My log is this.

Code: Select all

# service start vboxdrv
Starte den vboxdrv Systemdienst
Starting VirtualBox kernel module
Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: Oops: 0002 [#1] SMP 
 ...failed!
  (modprobe vboxnetadp failed. Please use 'dmesg' to find out why)

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: last sysfs file: /sys/module/vboxdrv/initstate

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: Process modprobe (pid: 11068, ti=f3016000 task=f6a572c0 task.ti=f3016000)

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: Stack:

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: Call Trace:

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: Code: 89 7c 24 0c 89 c7 b8 5c 00 00 00 89 5c 24 04 c7 04 24 01 00 00 00 e8 b2 82 a8 c8 89 c3 31 c0 85 db 74 2e 8b 16 8b 83 3c 01 00 00 <89> 10 0f b7 56 04 66 89 50 04 89 d8 e8 41 82 a8 c8 85 c0 89 c6 

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: EIP: [<f89e71fe>] vboxNetAdpOsCreate+0x3e/0xa0 [vboxnetadp] SS:ESP 0068:f3017f1c

Message from syslogd@spritty at Wed Nov 25 22:21:57 2009 ...
spritty kernel: CR2: 0000000000000000

Code: Select all

# dmesg |grep vbox
vboxdrv: Trying to deactivate the NMI watchdog permanently...
vboxdrv: Warning: 2.6.31+ kernel detected. Most likely the hwardware performance
vboxdrv: counter framework which can generate NMIs is active. You have to prevent
vboxdrv: the usage of hardware performance counters by
vboxdrv:   echo 2 > /proc/sys/kernel/perf_counter_paranoid
vboxdrv: Found 2 processor cores.
vboxdrv: fAsync=0 offMin=0x1af offMax=0x216e
vboxdrv: TSC mode is 'synchronous', kernel timer mode is 'normal'.
vboxdrv: Successfully loaded version 3.0.12 (interface 0x00110000).
IP: [<f89e71fe>] vboxNetAdpOsCreate+0x3e/0xa0 [vboxnetadp]
last sysfs file: /sys/module/vboxdrv/initstate
Modules linked in: vboxnetadp(+) vboxnetflt vboxdrv w83627ehf hwmon_vid coretemp snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss fuse nvidia(P) snd_hda_intel rtc_cmos snd_hda_codec rtc_core i2c_nforce2 rtc_lib 3c59x forcedeth snd_pcm processor thermal psmouse agpgart mii thermal_sys snd_timer snd serio_raw button evdev hwmon sg soundcore snd_page_alloc
EIP is at vboxNetAdpOsCreate+0x3e/0xa0 [vboxnetadp]
 [<f89e7403>] ? vboxNetAdpCreate+0x83/0xa0 [vboxnetadp]
 [<f89e7452>] ? vboxNetAdpInit+0x32/0x40 [vboxnetadp]
 [<f804e01f>] ? VBoxNetAdpLinuxInit+0x1f/0x11d [vboxnetadp]
 [<f804e000>] ? VBoxNetAdpLinuxInit+0x0/0x11d [vboxnetadp]
EIP: [<f89e71fe>] vboxNetAdpOsCreate+0x3e/0xa0 [vboxnetadp] SS:ESP 0068:f3017f1c

[gapan]

It usually takes a while until virtualbox catches on with the latest kernel versions. This is probably a case where it just doesn't with with the latest kernel.

[damNageHack]

Code: Select all

vboxdrv: Warning: 2.6.31+ kernel detected. Most likely the hwardware performance
vboxdrv: counter framework which can generate NMIs is active. You have to prevent
vboxdrv: the usage of hardware performance counters by
vboxdrv:   echo 2 > /proc/sys/kernel/perf_counter_paranoid

It does know about the kernel version! ...

[damNageHack]

I am really confused of this.

When the system is running and you do the echo 2 ... at console and then service start vboxdrv, all is going well.

When you enable the service starting with boot process, kernel produces a BUG message with a null pointer dereference.

I do not really understand this. So the user has to let the service without the +x flag to prevent starting at boot time but afterwards - when system is working for some time - it works?

Can somebody please try to reproduce this

I could try to reproduce the behaviour in a virtual machine itself to get the exact text of the kernel stop error. But then I would have to run virtualbox in itself. Or is there a log somewhere where such errors from kernel are kept persistently over reboots to have better debugging functionallity?