The following gets me to non boxy libvirt:
$ virsh -c qemu+unix:///session?socket=/run/user/1000/libvirt/libvirt-sock list 
--all

For now I'll assume that it is NOT depending on the guest, but lets
modify the working Ubuntu guest one by one to become more like the F29
guest and we will see.

1. different disks/iso's/MAC (obviously)
2. F29 has gl enabled on the spice graphics
3. video F29: virtio Ubuntu: qxl
4. video has <acceleration accel3d='yes'/> set

That is all the difference, so it seems 3d'ish to me.

First change
<model type='qxl' ram='65536' vram='65536' vgamem='16384' heads='1' 
primary='yes'/>
to
<model type='virtio' heads='1' primary='yes'>
=> still working

Second change enable gl
<gl enable='no'/>
to
<gl enable='yes'/>

=> Broken

Lets take back the First change but keep only the second.
=> still broken.

So it is the enablement of gl which I work on anyway recently (some
apparmor changes to make it work in my former setup).

Thanks for sharing this bug, but I need to analyze more in depth what is
wrong here, but that might take a while.

Note: Since your guest crashed on start the crash has no private data -
marking the bug public ...


For the time being as a workaround:
 virsh -c qemu+unix:///session?socket=/run/user/1000/libvirt/libvirt-sock edit 
fedora29-wor
(assuming that is your guest name as well)
and switch off the gl enablement.
Gives me a perfectly working guest, hope that helps you for now until a real 
fix is found.

** Changed in: qemu (Ubuntu)
       Status: Confirmed => Triaged

** Information type changed from Private to Public

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1815889

Title:
  qemu-system-x86_64 crashed with signal 31 in
  __pthread_setaffinity_new()

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1815889/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to