100% CPU load of events kernel thread triggered by virtualbox

[Chris Hemsing]

After starting virtualbox on a dual core machine, one of the events kernel threads (nicelevel: -5 !) gets 100% CPU all the time, i.e. occupying one core completely.

Further observations:

1. This happens as soon as the guest os starts.
2. It is also the case, when the guest os is idling.
3. It does NOT depend on the guest operating system. Same with Windows XP as with some Linux flavour.

This is on Ubuntu Intrepid, version: 2.1.4-42893_Ubuntu_intrepid .

[kanatacrude]

I am experiencing the same thing with VirtualBox 2.1.4. XP partition running as guest in Kubuntu 8.10 and on another PC an XP guest running on Ubuntu 8.10. As soon as the VM starts, one of my two processors is running flat out for hours until I stop the VM. My searchs for help show that so many people are experiencing this problem. This seems to be quite widespread. Wide variety of guests running on a variety of host OSes. MacOS, Ubuntu, CentOS.

This product would really kick bass if it just didn't have this glaring bug right now.

[Chris Hemsing]

Some additional information (even though nobody seems to care...):
It seems to be a bug around vboxnetflt. See the following kernel message:

[26054.680083] BUG: unable to handle kernel NULL pointer dereference at 00000004
[26054.680090] IP: [<c011af98>] __ticket_spin_lock+0x8/0x20
[26054.680097] *pde = 00000000 
[26054.680101] Oops: 0002 [#1] SMP 
[26054.680105] Modules linked in: af_packet binfmt_misc radeon drm rfcomm sco bnep l2cap bluetooth vboxnetflt vboxdrv ppdev lirc_atiusb lirc_dev powernow_k8 cpufreq_stats cpufreq_userspace cpufreq_powersave cpufreq_ondemand freq_table cpufreq_conservative pci_slot container sbs sbshc video output wmi battery bridge stp iptable_filter ip_tables x_tables ac it87 hwmon_vid sbp2 lp loop snd_hda_intel dvb_usb_dib0700 snd_pcm_oss dib7000p snd_mixer_oss dib7000m snd_pcm dvb_usb dvb_core snd_seq_dummy dib3000mc dibx000_common dib0070 snd_seq_oss snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq snd_timer snd_seq_device snd k8temp i2c_piix4 soundcore snd_page_alloc pcspkr i2c_core evdev parport_pc parport button shpchp pci_hotplug ati_agp agpgart ipv6 ext3 jbd mbcache sr_mod cdrom sd_mod crc_t10dif ata_generic sg pata_acpi usbhid hid ohci1394 ieee1394 e1000 uhci_hcd pata_atiixp ahci r8169 mii libata ehci_hcd scsi_mod ohci_hcd dock usbcore raid10 raid456 async_xor async_memcpy async_tx xor raid1 raid0 multipath linear md_mod dm_mirror dm_log dm_snapshot dm_mod thermal processor fan fbcon tileblit font bitblit softcursor fuse
[26054.680172] 
[26054.680175] Pid: 10, comm: events/1 Not tainted (2.6.27-11-generic #1)
[26054.680177] EIP: 0060:[<c011af98>] EFLAGS: 00010246 CPU: 1
[26054.680181] EIP is at __ticket_spin_lock+0x8/0x20
[26054.680183] EAX: 00000004 EBX: f7063f5c ECX: f14588b8 EDX: 00000100
[26054.680185] ESI: f14588b4 EDI: f915c1f0 EBP: f7063f34 ESP: f7063f34
[26054.680187]  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[26054.680190] Process events/1 (pid: 10, ti=f7062000 task=f7068000 task.ti=f7062000)
[26054.680191] Stack: f7063f3c c037f45d f7063f44 f91616e3 f7063f6c f915c222 00000000 f7063f5c 
[26054.680197]        00000000 f1458810 00000000 f14588b4 f700e900 f915c1f0 f7063fa4 c01436f5 
[26054.680203]        005ce478 c20f9d00 f7063f90 00000246 f14588b8 e07cf500 f700e904 ffffffff 
[26054.680208] Call Trace:
[26054.680210]  [<c037f45d>] ? _spin_lock+0xd/0x10
[26054.680215]  [<f91616e3>] ? RTSpinlockAcquire+0x13/0x20 [vboxnetflt]
[26054.680225]  [<f915c222>] ? vboxNetFltLinuxXmitTask+0x32/0x180 [vboxnetflt]
[26054.680231]  [<f915c1f0>] ? vboxNetFltLinuxXmitTask+0x0/0x180 [vboxnetflt]
[26054.680236]  [<c01436f5>] ? run_workqueue+0x95/0x160
[26054.680241]  [<c0147626>] ? finish_wait+0x16/0x70
[26054.680244]  [<c0143998>] ? worker_thread+0x88/0xf0
[26054.680247]  [<c0147560>] ? autoremove_wake_function+0x0/0x50
[26054.680250]  [<c0143910>] ? worker_thread+0x0/0xf0
[26054.680254]  [<c01471f1>] ? kthread+0x41/0x80
[26054.680257]  [<c01471b0>] ? kthread+0x0/0x80
[26054.680260]  [<c0105297>] ? kernel_thread_helper+0x7/0x10
[26054.680264]  =======================
[26054.680266] Code: 00 00 00 55 8b 10 89 e5 5d 89 d0 c1 f8 08 29 d0 25 ff 00 00 00 83 e8 01 0f 9f c0 0f b6 c0 c3 8d 74 26 00 55 ba 00 01 00 00 89 e5 <f0> 66 0f c1 10 38 f2 74 06 f3 90 8a 10 eb f6 5d c3 8d b4 26 00 
[26054.680293] EIP: [<c011af98>] __ticket_spin_lock+0x8/0x20 SS:ESP 0068:f7063f34
[26054.680300] ---[ end trace c976274508e23837 ]---

[Frank Mehnert]

Actually we do care. Perhaps better late than never:

• Is this kernel fault reproducible (I assume so reading your original remarks).
• When did this fault occur, that is, can you show a relation between this fault and loading the module, starting a VM, running the VM? Perhaps you could do

  tail -f /var/log/kern.log
  

  to get immediately notified when this problem occurs.
• The chances are low but could you check if VirtualBox 2.2.2 fixes the problem by any chance?