Ticket History #3239: AI_NUMERICSERV conflicts with "discard" getaddrinfo hack (AIX 5)

From krb5-bugs-incoming-bounces@PCH.mit.edu Mon Nov 14 14:29:40 2005
Received: from pch.mit.edu (PCH.MIT.EDU [18.7.21.90]) by krbdev.mit.edu (8.9.3p2) with ESMTP
id OAA09368; Mon, 14 Nov 2005 14:29:40 -0500 (EST)
Received: from pch.mit.edu (pch.mit.edu [127.0.0.1])
by pch.mit.edu (8.12.8p2/8.12.8) with ESMTP id jAEJT0px016335
for <krb5-send-pr@krbdev.mit.edu>; Mon, 14 Nov 2005 14:29:00 -0500
Received: from fort-point-station.mit.edu (FORT-POINT-STATION.MIT.EDU
[18.7.7.76])
by pch.mit.edu (8.12.8p2/8.12.8) with ESMTP id jAEJC3px012423
for <krb5-bugs-incoming@PCH.mit.edu>; Mon, 14 Nov 2005 14:12:03 -0500
Received: from mxout3.cac.washington.edu (mxout3.cac.washington.edu
[140.142.32.166])jAEJC5H9014570
for <krb5-bugs@mit.edu>; Mon, 14 Nov 2005 14:12:06 -0500 (EST)
Received: from herbert.u.washington.edu (herbert.u.washington.edu
[140.142.13.4])ESMTP id jAEJC4EE030907
(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO)
for <krb5-bugs@mit.edu>; Mon, 14 Nov 2005 11:12:05 -0800
Received: (from donn@localhost)jAEJC3Bw162960;
Mon, 14 Nov 2005 11:12:04 -0800
Date: Mon, 14 Nov 2005 11:12:04 -0800
Message-Id: <200511141912.jAEJC3Bw162960@herbert.u.washington.edu>
To: krb5-bugs@mit.edu
From: donn@u.washington.edu
X-send-pr-version: 3.99
X-Uwash-Spam: Gauge=IIIIIII, Probability=7%, Report='NO_REAL_NAME 0,
__HAS_MSGID 0, __MIME_TEXT_ONLY 0, __SANE_MSGID 0, __STOCK_SUBJ_7 0'
X-Spam-Score: 0.178
X-Spam-Flag: NO
X-Scanned-By: MIMEDefang 2.42
X-Mailman-Approved-At: Mon, 14 Nov 2005 14:27:47 -0500
X-BeenThere: krb5-bugs-incoming@mailman.mit.edu
X-Mailman-Version: 2.1
Precedence: list
Reply-To: donn@u.washington.edu
Sender: krb5-bugs-incoming-bounces@PCH.mit.edu
Errors-To: krb5-bugs-incoming-bounces@PCH.mit.edu


University of Washington
AIX 5
System: AIX herbert 2 5 000895634C00


(Same as 3107) Use of AI_NUMERICSERV in library functions breaks
the hack that fixes an alleged NUMERIC_SERVICE_BROKEN problem.
I can't duplicate the NUMERIC_SERVICE_BROKEN problem on AIX 5.1
or AIX 5.2, so I believe it's fixed. That makes for an easy
way to resolve the present problem.
Build on AIX 5, observe that kinit can't find KDC.
*** include/fake-addrinfo.h.dist Wed Jun 1 12:24:32 2005
--- include/fake-addrinfo.h Mon Nov 14 10:58:53 2005
***************
*** 144,151 ****
# define COPY_FIRST_CANONNAME
#endif

! #ifdef _AIX
# define NUMERIC_SERVICE_BROKEN
# define COPY_FIRST_CANONNAME
#endif

--- 144,153 ----
# define COPY_FIRST_CANONNAME
#endif

! #if defined(_AIX) && !defined(AI_NUMERICSERV)
# define NUMERIC_SERVICE_BROKEN
+ #endif
+ #if defined(_AIX)
# define COPY_FIRST_CANONNAME
#endif

Looks like this problem goes back a few versions and I found a patch on
the web the fixed the problem and just wanted to make sure that it got
into the next release:

The symptoms in AIX v5.3 were that when doing an initial kinit on the
server, I get the following message:

kinit(v5): Cannot resolve network address for KDC in requested realm
while getting initial credentials

...and a core dump.

The following patch fixed the problem. Many thanks to Donn Cave. I
searched the bug list, but couldn't find this mentioned in results of
the searches and wanted to make sure this was included.

Thanks!


Fix by Donn Cave:

Donn Cave, donn <at> u.washington.edu

-----------------------------------
*** include/fake-addrinfo.h.dist Wed Jun 1 12:24:32 2005
--- include/fake-addrinfo.h Fri Aug 12 09:10:48 2005
***************
*** 1193,1199 ****
a known service name for tcp or udp (as appropriate), an error
code (for "host not found") is returned. If the port maps to a
known service for both udp and tcp, all is well. */
! if (serv && serv[0] && isdigit(serv[0])) {
unsigned long lport;
char *end;
lport = strtoul(serv, &end, 10);
--- 1193,1208 ----
a known service name for tcp or udp (as appropriate), an error
code (for "host not found") is returned. If the port maps to a
known service for both udp and tcp, all is well. */
! /*
! **
! ** However, where AI_NUNERICSERV is defined (AIX 5) and was
specified,
! ** this is unneeded and and broken - "discard" is not numeric.
! */
! if (serv && serv[0]
! #ifdef AI_NUMERICSERV
! && !(hint->ai_flags & AI_NUMERICSERV)
! #endif
! && isdigit(serv[0])) {
unsigned long lport;
char *end;
lport = strtoul(serv, &end, 10);