[Pkg-ayatana-devel] Bug#1126327: user-session-migration: FTBFS on arm64: FAIL: test_subsequent_runs_with_new_script (__main__.MigrationTests.test_subsequent_runs_with_new_script)
Emanuele Rocca
ema at debian.org
Mon Jan 26 16:24:53 GMT 2026
Hi,
On 2026-01-24 10:14, Sebastian Ramacher wrote:
> user-session-migration failed to build on the arm64 buildds:
>
> 1/3 Test #3: build_user_session_migration ..... Passed 0.20 sec
> test 2
> Start 2: migration
>
> 2: Test command: /usr/bin/python3 "/build/reproducible-path/user-session-migration-0.5.1+b2/obj-aarch64-linux-gnu/tests/test_migration.py"
> 2: Working Directory: /build/reproducible-path/user-session-migration-0.5.1+b2/obj-aarch64-linux-gnu
> 2: Test timeout computed to be: 10000000
> 1: ..
> 1: ----------------------------------------------------------------------
> 1: Ran 2 tests in 3.557s
> 1:
> 1: OK
> 2/3 Test #1: debhelper ........................ Passed 3.67 sec
> 2: .......F
> 2: ======================================================================
> 2: FAIL: test_subsequent_runs_with_new_script (__main__.MigrationTests.test_subsequent_runs_with_new_script)
> 2: Test subsequent runs with a new script
> 2: ----------------------------------------------------------------------
> 2: Traceback (most recent call last):
> 2: File "/build/reproducible-path/user-session-migration-0.5.1+b2/obj-aarch64-linux-gnu/tests/test_migration.py", line 248, in test_subsequent_runs_with_new_script
> 2: self.assertTrue(re.match(f"""Using state file '{self.migration_home_file}'
> 2: ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 2: Using '{self.script_path}' directory
> 2: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 2: ...<4 lines>...
> 2: Executing: {re.escape(os.path.join(self.script_path, '08_testexecute.sh'))}
> 2: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 2: """, stdout))
> 2: ^^^^^^^^^^^^^
> 2: AssertionError: None is not true
For some reason I cannot reproduce this issue locally. I have tried both
on amdahl as well as a local arm64 system via sbuild. Here the logs of
the latter attempt:
https://people.debian.org/~ema/user-session-migration_arm64.build
The sbuild environment should match pretty closly what we have on the
buildds: sbuild 0.89.3+deb13u4 on trixie using the unshare backend.
Perhaps someone more familiar with the package can suggest how to try
and reproduce?
ema
More information about the Pkg-ayatana-devel
mailing list