From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4.1)
Gecko/20031114

Description of problem:
When logging in to a fc1 machine using kerberos, the terminal type
is encoded by the client with the terminal speed before being sent
to the klogind server.

klogind decodes this string back into the terminal type and speed.

Versions of krb5 from Red Hat contain a work-around for a bug where
this is not correctly decoded.

krb5-1.3.1 contains a fix for this bug from MIT people.  However,
this package also contains the RH fix (krb5-1.2.1-term.patch) as
well, which results in the bug re-appearing (the two fixes together
reintroduce the bug).

Therefore, I now end up with:

This rlogin session is encrypting all data transmissions.
Last login: Mon Feb  2 09:41:48 on :0
unknown terminal "xterm/38400"
unknown terminal "xterm/38400"

upon login.


Version-Release number of selected component (if applicable):
1.3.1-6

How reproducible:
Always

Steps to Reproduce:
1. krsh somekerberoshost
2.
3.
    

Actual Results:  [rmk@dyn-67 rmk]$ krsh dyn-67
This rlogin session is encrypting all data transmissions.
Last login: Mon Feb  2 09:41:48 on :0
unknown terminal "xterm/38400"
unknown terminal "xterm/38400"
[rmk@dyn-67 rmk]$


Expected Results:  [rmk@dyn-67 rmk]$ krsh dyn-67
This rlogin session is encrypting all data transmissions.
Last login: Mon Feb  2 09:41:48 on :0
[rmk@dyn-67 rmk]$


Additional info:

This bug can be fixed by dropping krb5-1.2.1-term.patch from the
package and rebuilding.  Confirmed by building krb5-1.3.1-6.src.rpm
without this patch for ARM.