Message ID | 20210415215141.1865467-7-crosa@redhat.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | Tests: introduce custom jobs | expand |
On Thu, Apr 15, 2021 at 6:52 PM Cleber Rosa <crosa@redhat.com> wrote: > > When running tests that are not target specific with various target > binaries, some specific behavior appears. For s390x, when there's no > guest code running, it will produce GUEST_PANICKED events as the > firmware will shutdown the machine. > > With this change, no GUEST_PANICKED *event* will be generated. > > For some QMP commands, such as "query-migrate", a proper response > ("guest-panicked" for the s390x target) will still be given. > > Signed-off-by: Cleber Rosa <crosa@redhat.com> > --- > tests/acceptance/migration.py | 4 ++-- > tests/acceptance/version.py | 2 +- > 2 files changed, 3 insertions(+), 3 deletions(-) > Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Hi, On 4/15/21 6:51 PM, Cleber Rosa wrote: > When running tests that are not target specific with various target > binaries, some specific behavior appears. For s390x, when there's no > guest code running, it will produce GUEST_PANICKED events as the > firmware will shutdown the machine. > > With this change, no GUEST_PANICKED *event* will be generated. > > For some QMP commands, such as "query-migrate", a proper response > ("guest-panicked" for the s390x target) will still be given. > > Signed-off-by: Cleber Rosa <crosa@redhat.com> > --- > tests/acceptance/migration.py | 4 ++-- > tests/acceptance/version.py | 2 +- > 2 files changed, 3 insertions(+), 3 deletions(-) > > diff --git a/tests/acceptance/migration.py b/tests/acceptance/migration.py > index 25ee55f36a..b4d46becc6 100644 > --- a/tests/acceptance/migration.py > +++ b/tests/acceptance/migration.py > @@ -46,12 +46,12 @@ def assert_migration(self, src_vm, dst_vm): > > def do_migrate(self, dest_uri, src_uri=None): > dest_vm = self.get_vm('-incoming', dest_uri) > - dest_vm.add_args('-nodefaults') > + dest_vm.add_args('-nodefaults', '-no-shutdown') On the other hand, can't that new argument introduce unwanted behavior on other targets? Maybe the argument should be set only for s390 because we know it prevents the test failure on that target only. - Wainer > dest_vm.launch() > if src_uri is None: > src_uri = dest_uri > source_vm = self.get_vm() > - source_vm.add_args('-nodefaults') > + source_vm.add_args('-nodefaults', '-no-shutdown') > source_vm.launch() > response = source_vm.qmp('migrate', uri=src_uri) > if 'error' in response: > diff --git a/tests/acceptance/version.py b/tests/acceptance/version.py > index 79b923d4fc..3cf18c9878 100644 > --- a/tests/acceptance/version.py > +++ b/tests/acceptance/version.py > @@ -17,7 +17,7 @@ class Version(Test): > :avocado: tags=quick > """ > def test_qmp_human_info_version(self): > - self.vm.add_args('-nodefaults') > + self.vm.add_args('-nodefaults', '-no-shutdown') > self.vm.launch() > res = self.vm.command('human-monitor-command', > command_line='info version')
diff --git a/tests/acceptance/migration.py b/tests/acceptance/migration.py index 25ee55f36a..b4d46becc6 100644 --- a/tests/acceptance/migration.py +++ b/tests/acceptance/migration.py @@ -46,12 +46,12 @@ def assert_migration(self, src_vm, dst_vm): def do_migrate(self, dest_uri, src_uri=None): dest_vm = self.get_vm('-incoming', dest_uri) - dest_vm.add_args('-nodefaults') + dest_vm.add_args('-nodefaults', '-no-shutdown') dest_vm.launch() if src_uri is None: src_uri = dest_uri source_vm = self.get_vm() - source_vm.add_args('-nodefaults') + source_vm.add_args('-nodefaults', '-no-shutdown') source_vm.launch() response = source_vm.qmp('migrate', uri=src_uri) if 'error' in response: diff --git a/tests/acceptance/version.py b/tests/acceptance/version.py index 79b923d4fc..3cf18c9878 100644 --- a/tests/acceptance/version.py +++ b/tests/acceptance/version.py @@ -17,7 +17,7 @@ class Version(Test): :avocado: tags=quick """ def test_qmp_human_info_version(self): - self.vm.add_args('-nodefaults') + self.vm.add_args('-nodefaults', '-no-shutdown') self.vm.launch() res = self.vm.command('human-monitor-command', command_line='info version')
When running tests that are not target specific with various target binaries, some specific behavior appears. For s390x, when there's no guest code running, it will produce GUEST_PANICKED events as the firmware will shutdown the machine. With this change, no GUEST_PANICKED *event* will be generated. For some QMP commands, such as "query-migrate", a proper response ("guest-panicked" for the s390x target) will still be given. Signed-off-by: Cleber Rosa <crosa@redhat.com> --- tests/acceptance/migration.py | 4 ++-- tests/acceptance/version.py | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-)