On 17/07/2024 10.37, Daniel P. Berrangé wrote:
On Wed, Jul 17, 2024 at 10:04:19AM +0200, Thomas Huth wrote:
On 16/07/2024 19.03, Thomas Huth wrote:
On 16/07/2024 18.51, Daniel P. Berrangé wrote:
On Tue, Jul 16, 2024 at 01:26:03PM +0200, Thomas Huth wrote:
...
So instead of trying to update the python-based test suite in QEMU
to a newer version of Avocado, we should maybe try to better integrate
it with the meson test runner instead. Indeed most tests work quite
nicely without the Avocado framework already, as you can see with
this patch series - it does not convert all tests, just a subset so
far, but this already proves that many tests only need small modifi-
cations to work without Avocado.
...
Now if you want to try out these patches: Apply the patches, then
recompile and then run:
make check-functional
You can also run single targets e.g. with:
make check-functional-ppc
You can also run the tests without any test runner now by
setting the PYTHONPATH environment variable to the "python" folder
of your source tree, and by specifying the build directory via
QEMU_BUILD_ROOT (if autodetection fails) and by specifying the
QEMU binary via QEMU_TEST_QEMU_BINARY. For example:
export PYTHONPATH=$HOME/qemu/python
export QEMU_TEST_QEMU_BINARY=qemu-system-x86_64
export PYTHONPATH=$HOME/qemu/build
~/qemu/tests/functional/test_virtio_version.py
For the whole series as is
Tested-by: Daniel P. Berrangé <berra...@redhat.com>
as it does what you claim it does here when I tried it.
Thanks!
The logs of the tests can be found in the build directory under
tests/functional/<testname> - console log and general logs will
be put in separate files there.
As an example, one dir name appears to be:
__main__.MemAddrCheck.test_phybits_ok_pentium_pae
I'd rather prefer it if the dir name matched the test script
file name - in this case test_mem_addr_space.py, as I don't
want to have to lookup which class names were defined inside
each test script. We could drop the "test_" prefix from the
method name too
IOW, could we make this dir name be:
test_mem_addr_space.phybits_ok_pentium_pae
I can try to change that, indeed ... but the boilerplate code will
increase a little bit, I guess, since I cannot simply rely on the
unittest.id() function in that case anymore...
After looking at this for a while, I think it's maybe best to ditch the idea
of making the .py files directly runnable and run the tests via a simple
pycotap runner instead. Then you get proper module names:
I'd really not want to loose that. To me, eliminating the test harness
entirely when debugging is the single biggest improvement of this new
approach, especially when I want to 'strace' the test without
extraneous processes.
$ pyvenv/bin/python3 -m pycotap test_virtio_version
TAP version 13
ok 1 test_virtio_version.VirtioVersionCheck.test_conventional_devs
ok 2 test_virtio_version.VirtioVersionCheck.test_modern_only_devs
1..
With the following change, you get the same output with direct
execution, by making argv look the same as you'd get when
running your pycotap example.
diff --git a/tests/functional/qemu_test/__init__.py
b/tests/functional/qemu_test/__init__.py
index cc49fd4c94..3a3e65252d 100644
--- a/tests/functional/qemu_test/__init__.py
+++ b/tests/functional/qemu_test/__init__.py
@@ -266,7 +266,10 @@ def fetch_asset(self, url, asset_hash):
def main():
tr = pycotap.TAPTestRunner(message_log = pycotap.LogMode.LogToError,
test_output_log =
pycotap.LogMode.LogToError)
- unittest.main(testRunner = tr)
+ import sys
+ import os.path
+ path = os.path.basename(sys.argv[0])[:-3]
+ unittest.main(module = None, testRunner = tr, argv=["__dummy__", path])
Sweet, thank you very much, looks like this will do the job!
Thomas
