[23052] in Kerberos

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

Re: kadmind - SIGSEGV

daemon@ATHENA.MIT.EDU (Tom Yu)
Sun Dec 12 16:48:43 2004

To: Michael Tautschnig <michael.tautschnig@zt-consulting.com>
From: Tom Yu <tlyu@mit.edu>
Date: Sun, 12 Dec 2004 16:47:49 -0500
In-Reply-To: <Pine.LNX.4.61.0412111847150.28405@l01.thnet> (Michael
 Tautschnig's message of "Sat, 11 Dec 2004 18:51:01 +0100 (CET)")
Message-ID: <ldvr7lvfat6.fsf@cathode-dark-space.mit.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
cc: kerberos@mit.edu
Errors-To: kerberos-bounces@mit.edu

>>>>> "Michael" == Michael Tautschnig <michael.tautschnig@zt-consulting.com> writes:

Michael> ==23442== Invalid read of size 4
Michael> ==23442==    at 0x1B92629A: add_to_history (svr_principal.c:1057)
Michael> ==23442==    by 0x1B9269D8: kadm5_chpass_principal_3 (svr_principal.c:1284)
Michael> ==23442==    by 0x1B92653E: kadm5_chpass_principal (svr_principal.c:1174)
Michael> ==23442==    by 0x1B91FF0D: _kadm5_chpass_principal_util (chpass_util.c:125)
Michael> ==23442==  Address 0x1C4F2000 is not stack'd, malloc'd or (recently) free'd
Michael> ==23442==
Michael> ==23442== Process terminating with default action of signal 11 (SIGSEGV)

This is very interesting.  Purify sees nothing of the sort when run on
a simple test case on recent code, which should be mostly identical in
this area to krb5-1.3.4.

Does this only happen when you attempt to change the password of one
particular principal, or does it do so for all principals, including
brand new ones?  Do you have password policies enabled?  How many old
passwords do you keep, and have you changed that number changed
recently?

Also, what release of krb5 did you upgrade from?

---Tom
________________________________________________
Kerberos mailing list           Kerberos@mit.edu
https://mailman.mit.edu/mailman/listinfo/kerberos

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