On Fri, Sep 10, 2010 at 5:56 AM, Andreas Henriksson <and...@fatal.se> wrote:
> On Thu, Sep 09, 2010 at 08:20:52PM -0400, Andres Cimmarusti wrote:
>> I've been using vino as vnc server on this computer. Lately I haven't been able to connect to the server using vinagre while using remmina works fine.
>>
>> I get a pop-up message saying connection to server has been closed. It doesn't even prompt me for password.
>
> Please attach the console debug output you get when running
> "vinagre --gtk-vnc-debug" and attempting to connect.
>
> Regards,
> Andreas Henriksson
>
--
To UNSUBSCRIBE, email to debian-bugs...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
It seems you're not the first to find this problem:
http://mail.gnome.org/archives/gtk-vnc-list/2010-September/msg00000.html
The workaround there doesn't seem like a good solution in general to me.
I've tried debugging this a bit and I think the problem lies in the
signal handling of gtk-vnc, rather then vinagre.
If I've understood gtk-vnc correctly, it has a vnc connection and a vnc
display object. The connection signals "vnc-auth-credential" which the
display object subscribes to (on_auth_cred). The display object then
sends the same signal, to anyone (vinagre) that subscribes to
"vnc-auth-credential" on the display object.
When running this in a debugger I think "on_auth_cred" is never called
and thus the signal on the vnc display object that vinagre subscribes to
is never emitted.
Reading symbols from /opt/vinagre/bin/vinagre...done.
(gdb) run
Starting program: /opt/vinagre/bin/vinagre
[Thread debugging using libthread_db enabled]
^C
Program received signal SIGINT, Interrupt.
0x00007ffff2ca10d8 in __poll (fds=0x7af210, nfds=11,
timeout=<value optimized out>) at ../sysdeps/unix/sysv/linux/poll.c:83
83 ../sysdeps/unix/sysv/linux/poll.c: No such file or directory.
in ../sysdeps/unix/sysv/linux/poll.c
(gdb) break on_auth_cred
Breakpoint 1 at 0x7fffedf1114a: file vncdisplay.c, line 1111.
(gdb) c
Continuing.
The above scenario:
gdb vinagre
run
break the execution with ctrl-c
add breakpoint
continue
click in vinagre and attempt connection.
breakpoint never hit and connection failed dialog shown.
Breaking on do_vnc_connection_emit_main_context works and the breakpoint
gets hit..... no idea why the signals never surface.
Another test case is to eliminate vinagre by using
examples/gvncviewer.py and observing that it also never calls the
vnc_auth_cred, which is the function subscribed to the credentials
signal on the vnc display object in the example.
--
> reassign 596287 libgtk-vnc-1.0-0
Bug #596287 [vinagre] vinagre fails to connect to VNC server
Bug reassigned from package 'vinagre' to 'libgtk-vnc-1.0-0'.
Bug No longer marked as found in versions vinagre/2.30.2-1.
> affects 596287 vinagre
Bug #596287 [libgtk-vnc-1.0-0] vinagre fails to connect to VNC server
Added indication that 596287 affects vinagre
> thanks
Stopping processing here.
Please contact me if you need assistance.
--
596287: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=596287
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
This is a regression compared to gtk-vnc 0.3.10 which I've confirmed
working (after rebuilding it from
http://snapshot.debian.org/package/gtk-vnc/0.3.10-6/ and rebuilding
vinagre with 0.3.10 installed).
Here's the vinagre debug output when running with gtk-vnc 0.3.10 where
apparently TLS+password authentication works...
$ vinagre --gtk-vnc-debug
gtk-vnc: Expose 0x0 @ 440,390
gtk-vnc: Started background coroutine
gtk-vnc: Resolving host 127.0.0.1 5900
gtk-vnc: Trying socket 18
gtk-vnc: Protocol initialization
gtk-vnc: Server version: 3.7
gtk-vnc: Using version: 3.7
gtk-vnc: Possible auth 18
gtk-vnc: Possible auth 2
gtk-vnc: Thinking about auth type 18
gtk-vnc: Decided on auth type 18
gtk-vnc: Waiting for auth type
gtk-vnc: Choose auth 18
gtk-vnc: Do TLS handshake
gtk-vnc: Handshake was blocking
gtk-vnc: Handshake was blocking
gtk-vnc: Handshake was blocking
gtk-vnc: Handshake done
gtk-vnc: Completed TLS setup
gtk-vnc: Got 1 subauths
gtk-vnc: Got 1 subauths
gtk-vnc: Possible sub-auth 2
gtk-vnc: Requested auth subtype 2
gtk-vnc: Waiting for auth subtype
gtk-vnc: Choose auth 2
gtk-vnc: Do Challenge
gtk-vnc: Requesting missing credentials
gtk-vnc: Set password credential seahorse
gtk-vnc: Waiting for missing credentials
gtk-vnc: Got all credentials
gtk-vnc: Checking auth result
gtk-vnc: Expose 0x77 @ 268,199
gtk-vnc: Success
gtk-vnc: Pixel format BPP: 32, Depth: 24, Byte order: 1234, True color: 1
Mask red: 255, green: 255, blue: 255
Shift red: 16, green: 8, blue: 0
gtk-vnc: Display name 'gem@amd64'
gtk-vnc: Setting depth color to 24 (32 bpp)
gtk-vnc: Visual mask: 16711680 65280 255
shift: 16 8 0
gtk-vnc: Mask local: 255 255 255
remote: 255 255 255
merged: 255 255 255
gtk-vnc: Pixel shifts
right: 16 8 0
left: 16 8 0
[...]
I can't spot anything obviously wrong there......
There's a small typo in
"sigdata.params.authCred = authTypes;"
which should probably be
"sigdata.params.authTypes = authTypes;
but it doesn't really matter since both of them are in the same union
anyway.
For background on the problem see: http://bugs.debian.org/596287
As a testcase "./examples/gvncviewer localhost" (where localhost is
running vino) was used and a password prompt = success, connection
closed = failure.
Hopefully Daniel P. Berrange can help shed some light on what might be
the cause?
> retitle 596287 Fails to connect to vino
Bug #596287 [libgtk-vnc-1.0-0] vinagre fails to connect to VNC server
Changed Bug title to 'Fails to connect to vino' from 'vinagre fails to connect to VNC server'
> forwarded 596287 gtk-vnc-list gnome org
Bug #596287 [libgtk-vnc-1.0-0] Fails to connect to vino
Set Bug forwarded-to-address to 'gtk-vnc-list gnome org'.
> thanks
Stopping processing here.
Please contact me if you need assistance.
--
596287: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=596287
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)