On Wed, Mar 05, 2025 at 12:51:16PM +0100, Thomas Huth wrote: > On 28/02/2025 11.27, Daniel P. Berrangé wrote: > > A number of fields are set at the class level on QemuBaseTest, even > > though the exact same named field is then set at the object level > > later in most cases. > > > > Reviewed-by: Thomas Huth <th...@redhat.com> > > Signed-off-by: Daniel P. Berrangé <berra...@redhat.com> > > --- > > tests/functional/qemu_test/testcase.py | 6 ------ > > 1 file changed, 6 deletions(-) > > > > diff --git a/tests/functional/qemu_test/testcase.py > > b/tests/functional/qemu_test/testcase.py > > index 058bf270ec..5b18416bed 100644 > > --- a/tests/functional/qemu_test/testcase.py > > +++ b/tests/functional/qemu_test/testcase.py > > @@ -33,12 +33,6 @@ > > class QemuBaseTest(unittest.TestCase): > > - arch = None > > - > > - workdir = None > > - log = None > > - logdir = None > > This seems to break the ACPI test: > > https://gitlab.com/thuth/qemu/-/jobs/9316861224#L164
Our current code initializes self.log in the setUp() method. The ACPI bits code references self.log in the __init__() method. So historically it has been referencing the class level 'log' property which has always been None. This means if it were to ever trigger codepaths calling self._print_log it will crash on referncing None. So this patch has exposed a preexisting flaw that was dormant. I think the fix is to move ACPI bits logic out of __init__ and into its own setUp() method. With regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
