[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

[Bug 214479] kdm no answer to xdmcp forward_query

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


--- Comment #9 from Dennis W. Tokarski <dwt poltec com>  2009-11-07 21:34:20 EDT ---

I tried cloning this bug to make a new posting out of it, but
nothing happened, so let me try just add a comment.

This bug occurs identically as described above in Fedora 10
*and* Fedora 11.

Since gdm no long supports xdmcp, I am now using kdm like the
original reporter.

To summarize:
 1) kdm correctly configured on machine A and machine B
 2) At the graphical login screen of either, select menu->remote login,
    and see A and B in the chooser list
 3) Select the other machine
 4) Successful connection occurs, and login to the remote machine is possible

So far, so good. Now the failure scenario:
 2) Login to either machine, from e.g., a gnome-terminal, run
      Xephyr -indirect localhost :1
      (or substitute the local machine's name for localhost)
    and see A and B in the chooser list
 3) Select the other machine
 4) Xephyr screen goes grey with an X cursor and stays that way
    until the connection attempt times out, then the chooser reappears.

During the timeout period, wireshark shows FORWARD_QUERY packets going
to the remote machine, and wireshark running on the remote machine shows
the arrival of the FORWARD_QUERYs. The remote machine generates no response.

When the remote connection is attempted from the login screen menu as in 
1-4) earlier, the local machine is not sending FORWARD_QUERYs, it is sending
a direct query to the remote.

This bug has now been demonstrated using kdm and, previously, gdm, spanning
multiple versions of Fedora over a three year period.

The three conspicuous features of this bug are:
  * recipients of xdmcp FORWARD_QUERYs don't seem to know what to do with them
  * a connection attempt via kdm from the login screen results in a direct
    query while an attempt via kdm from a nested server such as Xnest or
    Xephyr results in a forward query. Whey the different behavior?
  * Xnest or Xephyr both work when a direct query is forced on the
    command line, e.g.,
      Xephyr -query other_machine :1

Please keep this bug alive. Broken xdmcp/chooser in a nested server is
a real sore spot.

Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]