On Mon, Apr 07, 2014 at 04:40:24PM -0600, Eric Blake wrote: > On 04/07/2014 02:02 AM, Guido Günther wrote: > > When building packages in a clean chroot the QEMU_USER and QEMU_GROUP > > don't exist making VirQemuDriverConfigNew fail with privileged=true. > > > > Avoid that by not requiring priviliged mode and skipping tests that need > > s/priviliged/privileged/ > > > it. > > --- > > tests/qemuxml2argvtest.c | 24 ++++++++++++++++-------- > > 1 file changed, 16 insertions(+), 8 deletions(-) > > Seems like this is what avoids the fail pointed out in 1/3. It still > feels fishy that our testsuite is that dependent on the system (ideally, > we'd provide a way to mock things up so that creating the config file > NEVER fails when run from the testsuite, even if the uid doesn't exist - > because we shouldn't be probing the live system, only our mockups). I'd > wait for a second opinion on whether this patch is papering over a > bigger problem of depending on the current system state, or whether it > is an acceptable way to avoid the issue without investing the effort to > tackle at the uid lookup level. IMHO we should be passing privileged == false unconditionally, so that we always skip any magic username lookups. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :| -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list