Skip Menu |
 

To: rt-krb5@krbdev.mit.edu
Subject: tests fail on alpha/tru64
From: Tom Yu <tlyu@mit.edu>
Date: Wed, 04 Jun 2003 15:41:22 -0400
on rsts-11, with krb5-1.3-current:

334 Using authentication type KERBEROS_V4; ADAT must follow
KERBEROS_V4 accepted as authentication type
ADAT: Kerberos V4 krb_rd_req: Can't decode authenticator (krb_rd_req)
Kerberos V4 authentication failed
Name (rsts-11.mit.edu:krbsnap): FAIL: des-v4: ftp authentication

I suspect this is keytab-related, since it appears that the KDC
successfully reads a service ticket in order to issue the rcmd ticket.
To: rt-comment@krbdev.mit.edu
Subject: Re: [krbdev.mit.edu #1563] tests fail on alpha/tru64
From: Tom Yu <tlyu@mit.edu>
Date: Fri, 06 Jun 2003 10:16:03 -0400
RT-Send-Cc:
My bad. Turns out that the caching introduced to avoid memory leaks
introduces a situation where a free()d pointer may hang around in the
caller of krb__get_srvtabname(). We *heart* Purify. For some reason,
alpha/tru64 was the most sensitive to this sort of things, possibly
due to its malloc() implementation. Working on a fix.
From: tlyu@mit.edu
Subject: CVS Commit
* g_cnffile.c (krb__get_srvtabname): Make retname be a static
array rather than a static pointer, to avoid callers' possible
retention of free()d pointers. Yes, this may cause difficulty
with making this function thread-safe.


To generate a diff of this commit:



cvs diff -r1.184 -r1.185 krb5/src/lib/krb4/ChangeLog
cvs diff -r1.10 -r1.11 krb5/src/lib/krb4/g_cnffile.c
From: tlyu@mit.edu
Subject: CVS Commit
pullup from trunk, for real this time...


To generate a diff of this commit:



cvs diff -r1.174.2.9 -r1.174.2.10 krb5/src/lib/krb4/ChangeLog
cvs diff -r1.10 -r1.10.2.1 krb5/src/lib/krb4/g_cnffile.c