Hi,

During a test rebuild of HDF5's reverse dependcies against HDF5 2.1.0
currently in experimental, silx FTBFS with 1 failed test (twice):

silx/io/test/test_h5py_utils.py:140:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <silx.io.test.test_h5py_utils.TestH5pyUtils testMethod=test_modes_multi_process_swmr>

    @subtests
    @unittest.skipIf(not h5py_utils.HAS_SWMR, "SWMR not supported")
    def test_modes_multi_process_swmr(self):
        filename = self._new_filename()

        with self._open_context(filename, mode="w", libver="latest") as f:
            pass

        # File open by SWMR writer
        with self._open_context_subprocess(filename, mode="a", swmr=True):
            with self._open_context(filename, mode="r") as f:

silx/io/test/test_h5py_utils.py:343: AssertionError
=============================== warnings summary ===============================
silx/io/test/test_h5py_utils.py: 10 warnings
  /usr/lib/python3.13/multiprocessing/popen_fork.py:67: DeprecationWarning: This process (pid=14626) is multi-threaded, use of fork() may lead to deadlocks in the child.
    self.pid = os.fork()

- -- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
=========================== short test summary info ============================
SUBFAILED[{'mode': 'w'}] silx/io/test/test_h5py_utils.py::TestH5pyUtils::test_modes_multi_process_swmr
========== 1 failed, 1153 passed, 1371 skipped, 10 warnings in 25.53s ==========


Don't hesitate to reach out if you need help from HF5 devs which I'm
in touch with and who are willing to help making HDF5 2.1.0 land into
testing.

Best,
_g.

- -- System Information:
Debian Release: 12.0
Architecture: amd64 (x86_64)

Kernel: Linux 6.1.0-44-amd64 (SMP w/12 CPU threads; PREEMPT)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
-----BEGIN PGP SIGNATURE-----

iQEzBAEBCgAdFiEEoJObzArDE05WtIyR7+hsbH/+z4MFAmooaucACgkQ7+hsbH/+
z4P8eAf+PI1p9WEWzmCJ760xM9pUWTCMmFMQFvcApIjLbyYzMFW5vR/gZ1OZnG8t
4GoAI8oBSVUgpasvKgD0bcewUC2DI/NB22+4lMhdiA2fbvaaN93/ChXE3ovJbVdt
o7P5JO+bbrkkamLTydjW2yterHuZexyjmXT8K9TJRl29ulFZ5fZ5YXa6plYl4YFj
mwXmC6h01nNvGMrMUAN1jc/jKxKmIp9Xpmukm7bdxLuesbpGGXWQJNweIVXsN6Z5
aMhdtAlRpldLLu97mT5xZYc3xOrngmFG9cHiTxO5MeBsqvQpaZ5V0qB+sHnZU3K1
aVnqzNlqRuh0pOjWJpoMh08PuC0qYg==
=cdmA
-----END PGP SIGNATURE-----

Hi,

Gilles Filippini a écrit le 09/06/2026 à 21:35 :
The atteched patch fixes the problem.

Best,
_g.