Segfault with psutil.net_connections()

[wowi42]

Hi
I have a little segfault with psutil on FreeBSD 10 :

wowi@wowi.io:/root# uname -a
FreeBSD wowi.io 10.0-RELEASE-p3 FreeBSD 10.0-RELEASE-p3 #0: Tue May 13 18:31:10 UTC 2014     root@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC  amd64
wowi@wowi.io:/root# python
Python 2.7.6 (default, May 15 2014, 18:58:42)
[GCC 4.2.1 Compatible FreeBSD Clang 3.3 (tags/RELEASE_33/final 183502)] on freebsd10
Type "help", "copyright", "credits" or "license" for more information.
>>> import psutil
>>> psutil.net_connections()
Segmentation fault (core dumped)

Python come directly from pkg, and psutil version is 2.1.1
Peace

[giampaolo]

Sorry for replying so late. I've never bumped into this issue. As such the only thing I can suggest is to try to debug the problem yourself and post here in case you collect some useful information.

[lucien2k]

I've run into the same problem myself with python2.7.9 on freebsd 10.

It threw these warnings while installing via pip:
Collecting psutil
Using cached psutil-2.2.1.tar.gz
Installing collected packages: psutil
Running setup.py install for psutil
building '_psutil_bsd' extension
cc -fno-strict-aliasing -O2 -pipe -fstack-protector -fno-strict-aliasing -DNDEBUG -fPIC -DPSUTIL_VERSION=221 -I/usr/local/include/python2.7 -c psutil/_psutil_bsd.c -o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_bsd.o
psutil/_psutil_bsd.c:106:1: warning: control reaches end of non-void function [-Wreturn-type]
}
^
psutil/_psutil_bsd.c:1160:44: warning: field precision should have type 'int', but argument has type 'unsigned long' [-Wformat]
path, sizeof(path), "%._s",
~~^~
psutil/_psutil_bsd.c:2004:41: warning: field precision should have type 'int', but argument has type 'unsigned long' [-Wformat]
snprintf(path, sizeof(path), "%._s",
~~^~
3 warnings generated.
cc -fno-strict-aliasing -O2 -pipe -fstack-protector -fno-strict-aliasing -DNDEBUG -fPIC -DPSUTIL_VERSION=221 -I/usr/local/include/python2.7 -c psutil/_psutil_common.c -o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_common.o
cc -fno-strict-aliasing -O2 -pipe -fstack-protector -fno-strict-aliasing -DNDEBUG -fPIC -DPSUTIL_VERSION=221 -I/usr/local/include/python2.7 -c psutil/arch/bsd/process_info.c -o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/arch/bsd/process_info.o
cc -shared -pthread -Wl,-rpath,/usr/lib:/usr/local/lib -Wl,-rpath=/usr/lib:/usr/local/lib -fstack-protector build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_bsd.o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_common.o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/arch/bsd/process_info.o -L/usr/local/lib -ldevstat -lpython2.7 -o build/lib.freebsd-10.1-RELEASE-p5-amd64-2.7/_psutil_bsd.so
building '_psutil_posix' extension
cc -fno-strict-aliasing -O2 -pipe -fstack-protector -fno-strict-aliasing -DNDEBUG -fPIC -I/usr/local/include/python2.7 -c psutil/_psutil_posix.c -o build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_posix.o
cc -shared -pthread -Wl,-rpath,/usr/lib:/usr/local/lib -Wl,-rpath=/usr/lib:/usr/local/lib -fstack-protector build/temp.freebsd-10.1-RELEASE-p5-amd64-2.7/psutil/_psutil_posix.o -L/usr/local/lib -lpython2.7 -o build/lib.freebsd-10.1-RELEASE-p5-amd64-2.7/_psutil_posix.so
Successfully installed psutil-2.2.1

Python 2.7.9 (default, Jan 28 2015, 05:37:45)
[GCC 4.2.1 Compatible FreeBSD Clang 3.3 (tags/RELEASE_33/final 183502)] on freebsd10
Type "help", "copyright", "credits" or "license" for more information.

import psutil
psutil.net_connections()
Segmentation fault (core dumped)

[lucien2k]

Ok I did some more testing this morning.

The good news seems to be that its fixed in the master version. I tried 2.2.1 from pip and from source and both had the same seg fault issue, but using the master version I was able to use it correctly.

Based on the history of commits to the bsd files (and from glancing at the code), I suspect its this one fix: 5feb936

Thanks Giampaolo, and lets hope its fixed in 2.2.2!

[giampaolo]

Mmmm no, unfortunately I think it's unlikely that commit has anything to do with this.

[lucien2k]

Hmm, well the good news is it definitely seems better. I've played around quite a lot now and no more seg faults with the master version.

[giampaolo]

OK, I'm going to close this out then.

[giampaolo]

I've just bumped into this myself. Reopening.

[giampaolo]

OK I think I've fixed this in f37b720