Bug#863507: numba: test collection crashes/errors out [armel, ppc64el]
Rebecca N. Palmer
rebecca_palmer at zoho.com
Sun Jun 14 13:15:52 BST 2020
Control: retitle -1 test collection crashes/errors out [armel, ppc64el]
The tests now do get run on armhf, but still don't on armel and ppc64el.
This appears to be because some test modules crash/abruptly exit
*during test collection*. (pytest-xdist can start a new worker after a
crash, but each worker collects all tests before running any of them, so
if collection crashes no tests ever get run.)
The files affected are
ppc64el and armel: test_typedlist.py
armel only:
test_errorhandling.py
test_inlining.py
test_serialize.py
npyufunc/test_ufuncbuilding.py
The exact errors aren't displayed from pytest, but are if one imports
these modules directly. They suggest that the armel crash might be
related to #863508 (it no longer happens during documentation build, but
might not be totally fixed) and the ppc64el one to
https://github.com/numba/numba/issues/4026
qemu-armel$ python3 -X faulthandler ; echo $?
Python 3.8.3 (default, May 14 2020, 11:03:12)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numba.tests.test_typedlist
LLVM ERROR: Symbol not found: __sync_fetch_and_add_4
1
(If you want to set a breakpoint here, it's probably
https://sources.debian.org/src/llvm-toolchain-8/1:8.0.1-9/lib/Support/ErrorHandling.cpp/?hl=126#L115
)
qemu-ppc64el$ python3 -X faulthandler
Python 3.8.3 (default, May 14 2020, 11:03:12)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numba.tests.test_typedlist
/usr/lib/python3/dist-packages/numpy/core/getlimits.py:53:
RuntimeWarning: divide by zero encountered in log10
self.precision = int(-log10(self.eps))
Fatal Python error: Segmentation fault
Current thread 0x0000004000d5ed20 (most recent call first):
File "/usr/lib/python3/dist-packages/numba/typed/typedlist.py", line
260 in append
File "/usr/lib/python3/dist-packages/numba/tests/test_typedlist.py",
line 25 in <module>
File "<frozen importlib._bootstrap>", line 219 in
_call_with_frames_removed
File "<frozen importlib._bootstrap_external>", line 783 in exec_module
File "<frozen importlib._bootstrap>", line 671 in _load_unlocked
File "<frozen importlib._bootstrap>", line 975 in _find_and_load_unlocked
File "<frozen importlib._bootstrap>", line 991 in _find_and_load
File "<stdin>", line 1 in <module>
qemu: uncaught target signal 11 (Segmentation fault) - core dumped
Segmentation fault
The above testing was with 0.48.0-5 (unstable), but the build logs imply
that this bug still exists in 0.49.1-1exp1 (experimental).
More information about the debian-science-maintainers
mailing list