vinagre crashed with SIGSEGV in setcontext()

Bug #178101 reported by LEVIS Cyril
This bug report is a duplicate of:  Bug #183169: Crash if hostname is not found. Edit Remove
24
Affects Status Importance Assigned to Milestone
vinagre (Ubuntu)
Incomplete
Medium
Unassigned

Bug Description

Binary package hint: vinagre

I can't enter the password and then computer is freeze

ProblemType: Crash
Architecture: i386
Date: Sat Dec 22 11:08:58 2007
DistroRelease: Ubuntu 8.04
ExecutablePath: /usr/bin/vinagre
NonfreeKernelModules: cdrom
Package: vinagre 0.3-0ubuntu2
PackageArchitecture: i386
ProcCmdline: vinagre
ProcCwd: /home/atlas
ProcEnviron:
 PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games
 LANG=fr_FR.UTF-8
 SHELL=/bin/bash
Signal: 11
SourcePackage: vinagre
Stacktrace:
 #0 0xb78b1e9a in setcontext () from /lib/tls/i686/cmov/libc.so.6
 #1 0xb78b1f3e in makecontext () from /lib/tls/i686/cmov/libc.so.6
 #2 0x084882dc in ?? ()
 #3 0x00000000 in ?? ()
StacktraceTop:
 setcontext () from /lib/tls/i686/cmov/libc.so.6
 makecontext () from /lib/tls/i686/cmov/libc.so.6
 ?? ()
 ?? ()
Title: vinagre crashed with SIGSEGV in setcontext()
Uname: Linux atlas-laptop 2.6.24-2-generic #1 SMP Thu Dec 20 17:36:12 GMT 2007 i686 GNU/Linux
UserGroups: adm admin audio cdrom dialout dip floppy lpadmin netdev plugdev powerdev sambashare scanner video

Tags: apport-crash
Revision history for this message
LEVIS Cyril (atlas95) wrote :
Changed in vinagre:
importance: Undecided → Medium
Revision history for this message
Anthony Liguori (anthony-codemonkey) wrote :

What version of gtk-vnc are you using?

Jonh Wendell (wendell)
Changed in vinagre:
status: New → Incomplete
Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

Anthony, you can view that in Dependencies.txt:

libgtk-vnc-1.0-0 0.3.1-1

Revision history for this message
Anthony Liguori (anthony-codemonkey) wrote :

Can you provide more details on the steps to reproduce this problem? How many servers were you connecting to, what types of servers were they (RealVNC, Vino, etc.), what type of authentication where the servers configured to use?

Revision history for this message
Chris Jones (cmsj) wrote :

I can reliably reproduce this crash while using vinagre to connect to an Adder (a hardware VGA->VNC unit) which works correctly with vncviewer.
The authentication it offers is:
 CConnection: Server offers security type RA2(5)
 CConnection: Server offers security type RA2ne(6)
 CConnection: Server offers security type VncAuth(2)
vncviewer chooses VncAuth(2).

Please let me know if you need more information.

Changed in vinagre:
status: Incomplete → Confirmed
Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

Could you try with vinagre 0.4? There's a deb at http://emilio.pozuelo.org/~deb/vinagre_0.4-1_i386.deb, and the package will be synced soonish to Hardy (once it's newed in Debian). Alternatively the sources are in the same directory, in case you have another arch or want to build the package yourself.

Revision history for this message
Jonh Wendell (wendell) wrote :

Also, could you compile gtk-vnc with --enable-debug flag ? With this option, you can obtain detailed output of the connection, if running vinagre in a terminal.

Please, do that and paste here the output you see in terminal.

Also, it would help if you could use debug information in your build. (something like running: export CFLAGS="-g -O0" before buiding gtk-vnc and vinagre)

Changed in vinagre:
status: Confirmed → Incomplete
Revision history for this message
Hervé Cauwelier (hcauwelier-deactivatedaccount) wrote :

I just reproduced this bug trying to connect to a vino server protected by a password. I viewed the remote desktop, tried to move the mouse but it was "stuck" and selecting text on the remote terminal. I typed Ctrl+Alt to release and... bang!

You also asked for the amount of connections. It was the first (and only) connection I opened.

Revision history for this message
Andrew Conkling (andrewski) wrote :

I just got what appears to be the same bug by closing my only connection via the tab bar. Looking to provide a better backtrace, but why are there no debug packages available for libgtk-vnc or vinagre?

Emilio: I tried to download the DEB of 0.4, but that link didn't work.

Revision history for this message
Glenn (glenn-v) wrote :

I also could not find debug packages for libgtk-vnc or vinagre. Also, I tried the 0.4 deb provided above (I downloaded it, but didn't test it until now) and the program still crashes. In my case, all I have to do is connect to a server on the same computer (see duplicate bug).

Revision history for this message
Shawn vega (svega85-gmail) wrote :

i can confirm this bug, because I'm having the same problem

Revision history for this message
Chris Jones (cmsj) wrote : Re: [Bug 178101] Re: vinagre crashed with SIGSEGV in setcontext()

Hi

Emilio Pozuelo Monfort wrote:
> http://emilio.pozuelo.org/~deb/vinagre_0.4-1_i386.deb, and the package

that link is 404

--
Chris Jones
  <email address hidden>
   www.canonical.com

Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

0.4-1 is now in the archive, should be built soonish. Let's wait for it.

Revision history for this message
Emilio Pozuelo Monfort (pochu) wrote :

The package is already available, please test with it.

Thanks.

Revision history for this message
Chris Jones (cmsj) wrote :

Hi

Emilio Pozuelo Monfort wrote:
> The package is already available, please test with it.

I'm not 100% sure it's the same segfault, but I am still unable to
connect to the passworded VNC box.

Cheers,
--
Chris Jones
  <email address hidden>
   www.canonical.com

Revision history for this message
Hervé Cauwelier (hcauwelier-deactivatedaccount) wrote :

- I start vinagre 0.4
- I click on the "connect" button
- I type the name of a remote machine
- I click on the "connect" button

On the first time I saw it consume "nice" CPU time. On the second time it crashed right away.

In both cases, apport showed me the same traceback on setcontext()/makecontext().

I can reproduce it when I type an IP of my network. None of them have a VNC server listening. If I type a class C IP that can't be resolved in my network, it doesn't crash.

Revision history for this message
Glenn (glenn-v) wrote :

Still crashes with 0.4-1 when trying to connect to a local machine.

Revision history for this message
Glenn (glenn-v) wrote :

...with the same stacktrace. Also, there still isn't any -dbgsym pakages for vinagre.

Revision history for this message
Jonh Wendell (wendell) wrote :

I'd ask you guys to test vinagre 0.4 against development version of gtk-vnc. If you can't do that, please wait until next week, when gtk-vnc 0.3.3 will be released.

This sounds like dupe of bug #183169

Revision history for this message
whitelightme (inspheration-deactivatedaccount) wrote :

i just had this bug, too.
goodbye,
ilh700.

Revision history for this message
Sergey Zelenev (zelenev) wrote :

just had the same crash

Revision history for this message
mattismyname (mattismyname) wrote :

Same here with vinagre 0.4-1 in Hardy. All I have to do is try to connect to a non-listening port on localhost. With vanilla vncviewer, the error is:

Mon Jan 28 23:24:59 2008
 main: unable to connect to host: Connection refused (111)
matt@Aluminumy:~$

Revision history for this message
Andrew Conkling (andrewski) wrote :

On Jan 29, 2008 3:28 AM, Jonh Wendell <email address hidden> wrote:

> *** This bug is a duplicate of bug 183169 ***
> https://bugs.launchpad.net/bugs/183169
>
> ** This bug has been marked a duplicate of bug 183169
> Crash if hostname is not found
>

Jonh, for what it's worth, I think these might be two separate issues. There
are a few other issues tracked here other than the "missing hostname" one; I
myself found this bug because vinagre crashed when trying to close the tab
bar.

I upgraded to the latest gtk-vnc and can no longer reproduce my issue or any
of the others here, so perhaps that fixes it as well. Can anyone reproduce
the issues tracked here other than the "missing hostname" one?

Revision history for this message
Chris Jones (cmsj) wrote :

Hi

Andrew Conkling wrote:
> of the others here, so perhaps that fixes it as well. Can anyone reproduce
> the issues tracked here other than the "missing hostname" one?

I'm not able to make it segfault anymore, but I still can't connect to a
 passworded VNC server (a hardware box) I'm putting on localhost via an
ssh tunnel.

Cheers,
--
Chris Jones
  <email address hidden>
   www.canonical.com

Revision history for this message
Jonh Wendell (wendell) wrote :

Chris, what's happening? Tell me more details please.

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

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.