Re: [PATCH 0/3] tests/acceptance: Test virtio-rng and -balloon on s390x

[Cornelia Huck]

On Sat, 12 Dec 2020 08:15:13 +0100
Thomas Huth <thuth@redhat.com> wrote:

> On 11/12/2020 21.10, Willian Rampazzo wrote:
> > On 12/11/20 2:31 PM, Thomas Huth wrote:  
> >> Add two more simple tests to check that virtio-rng and virtio-balloon
> >> are at least (very) basically working on s390x.
> >>
> >> Based-on: 20201204121450.120730-1-cohuck@redhat.com
> >>
> >> Thomas Huth (3):
> >>    tests/acceptance: Extract the code to clear dmesg and wait for CRW
> >>      reports
> >>    tests/acceptance/machine_s390_ccw_virtio: Test virtio-rng via
> >>      /dev/hwrng
> >>    tests/acceptance/machine_s390_ccw_virtio: Test the virtio-balloon
> >>      device
> >>
> >>   tests/acceptance/machine_s390_ccw_virtio.py | 59 +++++++++++++++------
> >>   1 file changed, 43 insertions(+), 16 deletions(-)
> >>  
> > 
> > One observation, test_s390x_devices tends to get longer and difficult to
> > debug in case of problems. If a test covers one specific device type, It
> > will improve readability, flexibility, and debugging. In case you don't want
> > to spend time breaking this into multiple tests, I'll be glad to do that
> > after the whole series is merged.  
> 
> Theoretically yes, but practically we also want to run the tests as fast as
> possible. Quite a bit of time is used to boot the kernel, so if we add a new
> test for each and every device, that would increase the test time quite a
> bit. Thus I'd rather prefer to keep everything in one single test instead
> for now.

I agree.

Is there any other way to make this more debuggable?