Opened 17 years ago
Closed 17 years ago
#737 closed defect (fixed)
VMs very slow with realtime kernel, CPU maxed out
Reported by: | laga | Owned by: | |
---|---|---|---|
Component: | other | Version: | VirtualBox 1.5.0 |
Keywords: | Cc: | ||
Guest type: | other | Host type: | other |
Description
Hello,
I'm running Ubuntu Gutsy amd64 with their 2.6.22-12-rt kernel and the 100.14.19 nvidia driver.
When I start a VM in Virtualbox 1.5.0, one core of my core 2 duo E6300 gets maxed out and the VM is really slow (or it's indeed hanging, but I seem to recall that it's just slow).
Steps to reproduce:
1) Install Ubuntu gutsy 2) sudo aptitude install linux-rt 3) reboot into the -rt kernel 4) recompile the vboxdrv module 5) start a VM in virtualbox and find out that's it's slow/hanging
According to the package description, the linux-rt package in gutsy contains "Ingo Molnar's full real time preemption patch (2.6.22.1-rt9)".
Regards,
Michael
uname -a: Linux prometheus 2.6.22-12-rt #1 SMP PREEMPT RT Sun Sep 23 21:11:52 GMT 2007 x86_64 GNU/Linux
Change History (3)
comment:1 by , 17 years ago
comment:2 by , 17 years ago
Using 2.6.22-14-rt and Virtualbox 1.5.2, this problem seems to be fixed. VMs are running normally now.
I think this ticket can be closed :)
comment:3 by , 17 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
I've just found out the hard way that this can also cause the host to freeze. I found this in /var/log/messages just prior to the restart:
Oct 3 16:20:26 prometheus kernel: [15575.835450] VirtualBox[719]: segfault at 00000000ffffff9a rip 00002ae5e94be22c rsp 0000000041246090 error
In case it's uncommon for segfaults to show up in that file, it's probably because Ubuntu does some magic to catch them and handle them using apport, their crash reporting tool. If you want me to, I can try to crash this box again and get a backtrace using apport (didn't work this time because some packages were outdated).
--Michael