On 6/6/11 16:48 , Tres Seaver wrote:
On 06/05/2011 02:24 PM, Hanno Schlichting wrote:
On Sun, Jun 5, 2011 at 6:31 PM, Tres Seaver <tseaver@palladion.com> wrote:
[1] FAILED Zope Buildbot / zope2.13-py2.7 slave-ubuntu64 https://mail.zope.org/pipermail/zope-tests/2011-June/042535.html
[2] FAILED Zope Buildbot / zope2.14-py2.7 slave-ubuntu64 https://mail.zope.org/pipermail/zope-tests/2011-June/042537.html
These two are both the "segfault-at-exit" issue.
I finally took the time to install me a virtual machine with Ubuntu 64bit on it. And I was able to reproduce the issue!
It's not a segfault at exit, but actually happens inside a test:
test_interfaces (Acquisition.tests) (0.001 s) test_mixed_explicit_and_explicit (Acquisition.tests)Segmentation fault
Or to get the issue quicker:
$ bin/alltests -s Acquisition -t test_mixed_explicit_and_explicit Running zope.testrunner.layer.UnitTests tests: Set up zope.testrunner.layer.UnitTests in 0.000 seconds. Segmentation fault
I haven't yet tried to debug the problem itself.
I cannot reproduce on my "native" Ubuntu x86_64 machine using a built-from-scratch Python 2.7.1::
On the specific buildslave machine I tested with a python-2.7 built from scratch, re-ran bootstrap.py, bin/buildout, bin/alltests and all tests passed. It seems the python-2.7 from the "deadsnakes" PPA that is installed on this buildmachine behaves badly. I do not know why, and I do not really feel like investigating it. I'll remove the python-2.7 PPA install from this machine we'll start using the hand-build python-2.7 instead. regards, jw