I bisected it to
commit 0856119bd7399a416c21e1692855a1077164f21c
Author: Aaron Meurer <[email protected]>
Date: Mon Mar 12 23:05:30 2012 -0600
Factor out setup.py test into run_all_tests() in sympy/utilities/runtests.py
This way, we can easily get the functionality of running all the tests in a
forward compatible way, but still have the ability to pass arguments and
keyword arguments to the various test functions.
This would explain why bin/test does not generate the error. The
problem has something to do with setup.py.
I don't know how this causes a problem. I don't think it's related to
the translation with 2to3. A diff on the Python 2 file and the
translated file reveals that the only changes it made to the contents
of that commit are to add () to the end of the "print" lines. I think
it has something to do with calling a function from the sympy module
inside setup.py somehow.
If necessary, this commit can be reverted. Nothing else depends on this change.
By the way, apparently this failure actually showed up in the pull
request implementing this (https://github.com/sympy/sympy/pull/1115),
but it went unnoticed. I think we need to make it more clear in the
bot test summary if there were failures, as it looks too much like the
passed test case in passing.
Aaron Meurer
On Fri, Apr 6, 2012 at 1:54 AM, Chris Smith <[email protected]> wrote:
>
>
> On Fri, Apr 6, 2012 at 1:04 PM, Aaron Meurer <[email protected]> wrote:
>>
>> Hi.
>>
>> Recently, we started getting "Fatal Python error: Cannot recover from
>> stack overflow." in the Python 3 tests (when running setup.py test),
>> which is immediately followed by "Abort trap: 6", which kills Python.
>> In Mac OS X, this generates a problem report automatically with
>> logging information. I've uploaded the test output and the logging
>> info at https://gist.github.com/2317869. This has shown up in at
>> least two other people's SymPy-Bot tests (see
>> https://github.com/sympy/sympy/pull/1208), so I'm pretty sure it's not
>> a problem on my end.
>>
>> Does anyone have any idea what could be causing this? I don't see how
>> it could be anything but a Python bug. I'm currently in the process
>> of bisecting the history to find the offending commit. It's taking a
>> while, though, as I have to rerun ./bin/use2to3 and setup.py test to
>> reproduce it (just running ./bin/test on the failing test file does
>> not reproduce the error). I'll report back when I get it.
>>
>
> Note that running tests via bin/test does not generate the error.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sympy" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to
> [email protected].
> For more options, visit this group at
> http://groups.google.com/group/sympy?hl=en.
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sympy?hl=en.
