There's been a bunch of reports of VM startup failures on Fedora 31
and elsewhere:
https://bugzilla.redhat.com/show_bug.cgi?id=1751120
The reproducer I found is running a mock build, which adds a
name=systemd cgroup to /proc/self/cgroup, which messes up libvirt's
logic. The first patch has a bit more detail.
I'm not positive this is entirely correct... maybe we should be doing
something with that cgroup in some way? I'm kinda ignorant here.
I also filed a systemd bug incase that behavior is unintentional. We
probably want to change libvirt regardless though
https://bugzilla.redhat.com/show_bug.cgi?id=1756143
Cole Robinson (2):
vircgroupv2: Fix VM startup when legacy cgroups are defined
vircgroup: Add some VIR_DEBUG statements
src/util/vircgroup.c | 3 ++-
src/util/vircgroupv2.c | 15 ++++++++++++++-
2 files changed, 16 insertions(+), 2 deletions(-)
--
2.23.0