On the buildd, the segfault in the tests happens with both Python 3.13
and 3.14 (so it's not interpreter-specific).
I can't reproduce the segfault locally or on an i386 porter box - the
tests pass.
On both amd64 and i386 and with both Python 3.13 and 3.14, the segfaults
appear to be associated with tests/test_examples.py:test_examples
"widgets/vtk_canvas.enaml". It's likely that this test is the culprit
although not guaranteed. My experience with qtbot tests in other
projects is that the test that provokes the segfault is not necessarily
the actual cause... it can actually be a previous test that has not
disposed of widgets properly and the failure happens later when some
other gc kicks in.
This particular test uses vtk but python-vtk9 only builds for the
default interpreter so doesn't exist for Python 3.14; I'm not sure how
these tests are supposed to work in that situation in any case.
Perhaps skipping that particular test during the transition would be
reasonable. However, without a way of reproducing the test locally, we'd
be using the buildd to test the theory.
--
Stuart Prescott http://www.nanonanonano.net/ [email protected]
Debian Developer http://www.debian.org/ [email protected]
GPG fingerprint 90E2 D2C1 AD14 6A1B 7EBB 891D BBC1 7EBB 1396 F2F7
