[461] in Kerberos-V5-bugs

home help back first fref pref prev next nref lref last post

Bug in kpasswd

daemon@ATHENA.MIT.EDU (Lee A. Butler)
Fri Apr 15 08:59:39 1994

Date:     Fri, 15 Apr 94 6:42:26 EDT
From: "Lee A. Butler" <butler@ARL.MIL>
To: kerberos@MIT.EDU
Cc: krb5-bugs@MIT.EDU

There is a bug in kpasswd/libkrb5.a as it exists in Beta 3 of Kerberos V.

Scenario:

	% kpasswd
	Old password for principal@REALM: <old-pwd>
	Enter new password: <CR>
	Re-enter new password for verification: <CR>
	Password changed.
	%

At this point the principal has a null password which cannot be entered or
changed.  It now takes admin priviledges to set a new password for the
principal.  The user should not be able to enter/create a non-reproducible or
non-reusable password.

This can be viewed as a flaw in one (or both) of the following:

read_pwd.c: krb5_read_password()
	This routine performs the fgets() call and then gleefully replaces
	the newline character with a null, thus making (and accepting)
	a zero length password.

kpasswd.c:
	fails to check to see if the new password has length < 1.

In the case of kpasswd.c, I suggest that it exit early if a 0 length "new
password" is provided.

Lee A. Butler
Attn: AMSRL-SL-BV
U.S. Army Research Laboratory			Internet: butler@brl.mil
Aberdeen Proving Ground, MD  21005-5068		Phone: (410) 278-9200

As nightfall does not come at once, neither does oppression.
In both instances, there is a twilight when everything remains seemingly
unchanged.  And it is in such twilight that we all must be most aware of
change in the air--however slight--lest we become unwitting victims of
the darkness.                   --- Justice William O. Douglas

home help back first fref pref prev next nref lref last post