On 04/04/2011, at 8:51 PM, Justin Clift wrote:
On 02/04/2011, at 6:34 PM, Matthias Bolte wrote:
<snip>
>> Thanks. It's looking like some kind of problem interacting with
>> VirtualBox (v4.0.4):
>
>> This is the full gdb backtrace from the process:
>>
>> (gdb) thread apply all bt
>>
>> Thread 1 (process 85784):
>> #0 0x00007fff8549b3a6 in swtch_pri ()
>> #1 0x00007fff854d5b51 in _pthread_find_thread ()
<snip>
> Strange point to get stuck.
>
> I assume VirtualBox for OSX comes with it's own GUI tool. Is it also
> affected or only virsh?
>
> Does this happen with VirtualBox 4.0.4 only, or is 4.x or even 3.x affected too?
Just tried with the earlier versions of 4.0.2 and 3.2.12, with full libvirt
autogen.sh and recompile after each installation. Exact same behaviour, hang
point, everything. :(
The VirtualBox GUI seems completely unaffected. While virsh is stuck, the
GUI can be started, VM's started/stopped, GUI shut down, all as per normal.
I guess this is going to take a bunch more debugging, which I personally don't
have the time for atm. (ugh) :(
Are there any other OSX users around that could take a look?
As additional info, connecting to a remote RHEL 6.0 libvirtd using ssh works:
$ tools/virsh -c qemu+ssh://root@servername/system?socket=/var/run/libvirt/libvirt-sock
root@servername's password:
Welcome to virsh, the virtualization interactive terminal.
Type: 'help' for help with commands
'quit' to quit
virsh # ^D
$
Which is nice, as from memory it didn't used to (for ssh specifically).
+ Justin