Bug#1040057: #1040057: slixmpp: autopkgtest regression on arm64: Segmentation fault

2023-10-12 Thread Paul Gevers

Hi Martin,

On 02-10-2023 08:42, Martin wrote:

Finally, I had some time to run the autopkgtest suite on my own arm64
machine (MNT reform). It did not segfault, so I assume a specific
problem with the C/I system. I have no idea how to debug this, help is
appreciated!


If I rename tests/test_stanza_error.py to something else such that it's 
not automatically run as test, the segfault doesn't happen.


If I run only that test $(./run_tests.py ./tests/test_stanza_error.py) 
or $(./run_tests.py -d ./tests/test_stanza_error.py) the segfault happens.


Because I couldn't spot what would be interesting in the debugging 
output, I decided to run the test with -d on armhf as well... segfault 
when run alone, but only when the test runs isolated.


Obviously then I tried on amd64: when run in isolation also there the 
segfault occurs.


I hope you now have enough leads to pick it up again.

Paul


OpenPGP_signature.asc
Description: OpenPGP digital signature


Bug#1040057: #1040057: slixmpp: autopkgtest regression on arm64: Segmentation fault

2023-10-02 Thread Martin
Control: severity -1 normal
Control: tag -1 + help
Control: tag -1 + moreinfo

Finally, I had some time to run the autopkgtest suite on my own arm64
machine (MNT reform). It did not segfault, so I assume a specific
problem with the C/I system. I have no idea how to debug this, help is
appreciated!