logging related segmentation fault

[GoogleCodeExporter]

Working with ngs_backbone we have found an strange segmentation fault. In only 
happens with some bam files and only after initializing the logging module.

We have isolated the problem and created a minimal script and two bam files 
capable of triggering the segmentation fault.

What steps will reproduce the problem?

Run the provided fail_test.py.
If the same script is run like "fail_test.py 2" it won't fail. The fail cases 
are 1 and 3. All these cases use different combinations of loggings and bam 
files.

What is the expected output? What do you see instead?

If shouldn't end with a segmentation fault.

What version of the product are you using? On what operating system?

We're using pysam 0.3 with python2.6 in Debian squeeze.

We hope this bug report helps you.
Best regards

Peio Ziarsolo
Jose Blanca

Original issue reported on code.google.com by pziars...@gmail.com on 28 Sep 2010 at 8:00

Attachments:

• pysam_bug.tar.gz

[GoogleCodeExporter]

I have tested the script with the latest code in mercurial repository 
(30/09/2010) and it works. The bug seems fixed.

Original comment by pziars...@gmail.com on 30 Sep 2010 at 11:00

[GoogleCodeExporter]

Thanks, for reporting this.

It seems to work for me, but there is definitely 
something strange going on: if I ctrl-c the script, I get 
a segfault.

It might be related to me trying to suppress samtools
output on stdout/stderr. 

I will try to understand this behaviour.

Original comment by andreas....@gmail.com on 7 Oct 2010 at 3:36

• Changed state: Accepted

[GoogleCodeExporter]

The bug has appeared again. I have tried with pysam 3.1 and with the latest 
mercurial and in both cases it gives a segmentation fault.
Maybe it could be a good idea to put this tests into pysam tests.
thanks
p.

Original comment by pziars...@gmail.com on 20 Dec 2010 at 10:38

[GoogleCodeExporter]

Hi Peio, 

thanks for the up-date - could you please try the latest development branch, 
dev-0.4?

(hg update dev-0.4)

I changed the stdtout/stderr redirection in dev-0.4. At one point I had set 
dev-0.4 as the head branch and thus you might have inadvertently tried dev-0.4. 
I was told
it would be more intuitive to set rel-0.3 as the head and hence I have set it 
back.

I think the bug will be hard to fix in the 0.3 branch.

Please let me know if this works.

Best wishes,
Andreas

Original comment by andreas....@gmail.com on 22 Dec 2010 at 3:05

[GoogleCodeExporter]

Hi Andreas,
I have just tried with the 0.4 branch and it works fine again.

Thanks for the reply.

p.

Original comment by pziars...@gmail.com on 22 Dec 2010 at 3:15

[GoogleCodeExporter]

Thanks, I close this issue.

Original comment by andreas....@gmail.com on 10 Feb 2011 at 10:56

• Changed state: Fixed