[petsc-dev] segfault in __pyx_module_cleanup()

[Matthew Knepley]

On Mon, Sep 10, 2012 at 3:00 AM, Aron Ahmadia <aron at ahmadia.net> wrote:

> Hey guys,
>
> I'm running somebody else's petsc4py application script and I'm seeing a
> segmentation fault in __pyx_module_cleanup ().
>
> I'll spend some time debugging this later, but has anybody run into this
> before and have hints on where to start looking?
>

You should always compile the petsc4py stuff with symbols.

  Matt


> petsc4py: 1.2
> petsc: 3.2-p5
>
> The application script is here:
> https://github.com/cbm755/cp_matrices/blob/refactor/python/examples/petsc/heat_circle_petsc_solve.py
>
> Here's the stack trace:
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000101df589b in __pyx_module_cleanup ()
>    from
> /Users/aron/.virtualenvs/pyclaw/lib/python2.7/site-packages/petsc4py/lib/arch-darwin-c-debug/PETSc.so
> (gdb) where
> #0  0x0000000101df589b in __pyx_module_cleanup ()
>    from
> /Users/aron/.virtualenvs/pyclaw/lib/python2.7/site-packages/petsc4py/lib/arch-darwin-c-debug/PETSc.so
> #1  0x00000001000c4960 in PyEval_EvalFrameEx () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #2  0x00000001000c66f6 in PyEval_EvalCodeEx () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #3  0x000000010003d440 in function_call () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #4  0x000000010000c122 in PyObject_Call () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #5  0x00000001000be0b7 in PyEval_CallObjectWithKeywords () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #6  0x00000001000eb33a in Py_Finalize () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #7  0x0000000100101fff in Py_Main () from
> /usr/local/Cellar/python/2.7.2/lib/libpython2.7.dylib
> #8  0x0000000100000e22 in _start ()
> #9  0x0000000100000d41 in start ()
>
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20120910/d527bfa5/attachment.html>