Python 3.14.0a5: `test_external_inspection` fails on freethreading-debug build on s390x Linux

[befeleme]

Bug report

Bug description:

The failure happens only on s390x: /builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Objects/call.c:342: _PyObject_Call: Assertion '!_PyErr_Occurred(tstate)' failed.
Built as RPM on Fedora Linux 40-43, all occurrences have the same behavior.
The relevant log excerpt:

0:40:46 load avg: 3.05 [1/2/1] test_external_inspection worker non-zero exit code (Exit code -6 (SIGABRT))
Re-running test_external_inspection in verbose mode (matching: )
test_async_gather_remote_stack_trace (test.test_external_inspection.TestGetStackTrace.test_async_gather_remote_stack_trace) ... python: /builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Objects/call.c:342: _PyObject_Call: Assertion `!_PyErr_Occurred(tstate)' failed.
Fatal Python error: Aborted
<Cannot show all threads while the GIL is disabled>
Stack (most recent call first):
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/test_external_inspection.py", line 271 in test_async_gather_remote_stack_trace
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/case.py", line 606 in _callTestMethod
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/case.py", line 660 in run
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/case.py", line 716 in __call__
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/suite.py", line 122 in run
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/suite.py", line 84 in __call__
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/suite.py", line 122 in run
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/suite.py", line 84 in __call__
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/unittest/runner.py", line 259 in run
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 84 in _run_suite
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 42 in run_unittest
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 162 in test_func
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 118 in regrtest_runner
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 165 in _load_run_test
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 210 in _runtest_env_changed_exc
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 319 in _runtest
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/single.py", line 348 in run_single_test
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/worker.py", line 83 in worker_process
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/worker.py", line 118 in main
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/test/libregrtest/worker.py", line 122 in <module>
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/runpy.py", line 88 in _run_code
  File "/builddir/build/BUILD/python3.14-3.14.0_a5-build/Python-3.14.0a5/Lib/runpy.py", line 198 in _run_module_as_main
Extension modules: _testinternalcapi, _testexternalinspection (total: 2)

CPython versions tested on:

3.14

Operating systems tested on:

Linux

[colesbury]

It seems like some of the error paths don't properly set a Python exception. Does s390x have HAVE_PROCESS_VM_READV?

@pablogsal:

No Python error set here, but it's not free threading specific, so I'm probably missing something else:

cpython/Modules/_testexternalinspection.c

Lines 483 to 484 in 3e222e3

	#else
	return -1;

results leaked on error here:

cpython/Modules/_testexternalinspection.c

Lines 1524 to 1531 in 3e222e3

	PyObject* result = PyList_New(1);
	if (result == NULL) {
	return NULL;
	}
	PyObject* calls = PyList_New(0);
	if (calls == NULL) {
	return NULL;
	}

Nothing else stood out immediately.

[sergey-miryanov]

@colesbury I checked _testexternalinspection.c and found a few other places where exceptions are not set. Do you mind if I send a PR?

[colesbury]

Yes, thanks, please send a PR.

[sergey-miryanov]

gh-130051 fixes a few memory leaks.
gh-130053 fixes exceptions (but not finished yet), jfyi.

[pablogsal]

It seems like some of the error paths don't properly set a Python exception. Does s390x have HAVE_PROCESS_VM_READV?

🤔 It should otherwise the test should not even be attempted so whatever is going on is something else.

[pablogsal]

For example looking at https://buildbot.python.org/#/builders/509

The test runs and runs successfully:

0:00:19 load avg: 3.96 [ 89/484] test_external_inspection passed

[vstinner]

I suggest to mark this issue as a duplicate of #130052.