Ubuntu

SIGSEGV in _nss_wins_gethostbyname_r

Reported by Michael on 2009-04-29
144
This bug affects 25 people
Affects Status Importance Assigned to Milestone
glibc (Ubuntu)
Undecided
Unassigned
Nominated for Jaunty by Nizar Kerkeni
Nominated for Karmic by Nizar Kerkeni
liferea (Ubuntu)
Medium
Unassigned
Nominated for Jaunty by Nizar Kerkeni
Nominated for Karmic by Nizar Kerkeni

Bug Description

Binary package hint: liferea

After starting liferea and press the 'update all feeds' button the program crashes and generates a segmentation fault (signal 11)

I have included the backtrace of the crash

ProblemType: Bug
Architecture: i386
DistroRelease: Ubuntu 9.04
NonfreeKernelModules: nvidia
Package: liferea 1.5.15-1~jjwkt1
ProcEnviron:
 PATH=(custom, user)
 LANG=nl_NL.UTF-8
 SHELL=/bin/bash
SourcePackage: liferea
Uname: Linux 2.6.28-11-generic i686
UnreportableReason: Dit is geen authentiek Ubuntu-pakket.

Michael (m-gruys) wrote :
Michael (m-gruys) wrote :

Maybe also duplicate of: Bug #333645?

Michael wrote:
> Maybe also duplicate of: Bug #333645?

Nope, that one is a Liferea bug, this is looks like a bug in samba

The Status of the liferea package is marked as Invalid in bug #286119
If this is the case then it should not be a duplicate of this bug.
Could you elaborate a bit more on this sir?
Thanks in advance

Michael wrote:
> The Status of the liferea package is marked as Invalid in bug #286119
> If this is the case then it should not be a duplicate of this bug.
> Could you elaborate a bit more on this sir?

The reason is because the bug is not in Liferea itself (even if Liferea crashes)
but in Samba. But now that I look at it again, #286119 is about a double free
whereas this is a crash, so I'm not sure whether this is a duplicate of #286119
or not...

Thierry Carrez (ttx) on 2009-06-08
Changed in liferea (Ubuntu):
status: New → Invalid
summary: - Liferea did receive signal 11 (Segmentatiefout).
+ SIGSEGV in _nss_wins_gethostbyname_r
Nizar Kerkeni (nizarus) wrote :

I got this bug with liferea on 2 PC with Jaunty 64 bits where I installed winbind. When I remove winbind, liferea don't crash any more, but I need winbind to access to other PC on my network :/
I have the same issue too with liferea on karmic alpha1 64 bits.

Nizar Kerkeni (nizarus) wrote :

This bug is still reproducible on jaunty and karmic

Changed in liferea (Ubuntu):
status: Invalid → Confirmed
Thierry Carrez (ttx) wrote :

Nizar, Michael:
If you can reproduce on a Jaunty test system, could you test and install glibc 2.9-4ubuntu6.1~ppa1 from my PPA at https://launchpad.net/~ttx/+archive/ppa and let me know if it fixes the problem or not. You should upgrade the libc6 binary package (which should pull the libc6-i686 or libc6-amd64 and libc6-dev packages). Please reboot before testing to make sure you fully test the right glibc.

Changed in liferea (Ubuntu):
importance: Undecided → Medium
status: Confirmed → Invalid
Thierry Carrez (ttx) wrote :

glibc was updated in karmic removing the problematic patch. Would be good to confirm that the problem is gone on karmic now.

Thierry Carrez (ttx) wrote :

This is believed to be fixed in karmic by the fix for bug 374674. Please reopen if you can reproduce it on a current karmic setup.

affects: samba (Ubuntu) → glibc (Ubuntu)
Changed in glibc (Ubuntu):
status: New → Fix Released

Thierry Carrez wrote:
> This is believed to be fixed in karmic by the fix for bug 374674. Please
> reopen if you can reproduce it on a current karmic setup.

Bug 439666 looks like a duplicate, and happened in Karmic.

Emilio Pozuelo Monfort (pochu) wrote :

Bug #446649and #379371 are also dups, reopening.

Changed in glibc (Ubuntu):
status: Fix Released → Confirmed
Ciso (cisoprogressivo) wrote :

I have this problem in Karmic, and
removing wins from /etc/nsswitch.conf allows liferea to work again.
But it need it to browse my network.

isecore (isecore) wrote :

Ciso: As far as I can tell it works fine as long as the wins entry isn't the final one on that line.

i.e.

hosts: files mdns4_minimal [NOTFOUND=return] dns wins mdns4

and not

hosts: files mdns4_minimal [NOTFOUND=return] dns mdns4 wins

However, I'm running Karmic and have no problems with the latter configuration of nsswitch. Back when I was running Jaunty though, it had to be the former line or Liferea would segfault.

Kay Parker (kayparker) wrote :

Guess liferea crashes when selecting a headline with an embedded (youtube) video like that: http://www.guitarnoize.com/blog/comments/5-for-the-weekend-1995/

Micheal Mand (mikemand) wrote :

Thanks to isecore. I had hosts: files mdns4_minimal [NOTFOUND=return] wins dns mdns4 in /etc/nsswitch.conf - switching it to hosts: files mdns4_minimal [NOTFOUND=return] dns mdns4 wins solved the seg fault problem.

Radu (radubaetica) wrote :

Thank you all; I've used the solution described at #16 above by Micheal Mand and it solved my problem, which was posted as bug #498577.

Adam Porter (alphapapa) wrote :

I'm getting this bug in Maverick using Clementine 0.7 when it tries to lookup names. I think this should be given a priority.

[New Thread 0xaacfdb70 (LWP 25165)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xaacfdb70 (LWP 25165)]
0x04f4f31a in _nss_wins_gethostbyname_r () from /lib/libnss_wins.so.2

(gdb) backtrace

#0 0x04f4f31a in _nss_wins_gethostbyname_r () from /lib/libnss_wins.so.2
#1 0x04f4f61f in _nss_wins_gethostbyname2_r () from /lib/libnss_wins.so.2
#2 0x01ced1a5 in gethostbyname2_r () from /lib/libc.so.6
#3 0x01cac82c in ?? () from /lib/libc.so.6
#4 0x01cae3b5 in getaddrinfo () from /lib/libc.so.6
#5 0x017b9672 in ?? () from /usr/lib/libQtNetwork.so.4
#6 0x017ad4a2 in ?? () from /usr/lib/libQtNetwork.so.4
#7 0x01896523 in ?? () from /usr/lib/libQtCore.so.4
#8 0x018a1df9 in ?? () from /usr/lib/libQtCore.so.4
#9 0x001c7cc9 in start_thread () from /lib/libpthread.so.0
#10 0x01cd569e in clone () from /lib/libc.so.6

Adam Porter (alphapapa) wrote :

Also, I had to remove wins entirely to stop the segfaulting in Clementine.

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers