Ticket History #40: K5b7 util/pty/update_utmp.c and HPUX 10.10

From b17783@achilles.ctd.anl.gov Fri Sep 27 15:56:25 1996
Received: from MIT.EDU (PACIFIC-CARRIER-ANNEX.MIT.EDU [18.69.0.28]) by rt-11.MIT.EDU (8.7.5/8.7.3) with SMTP id PAA09118 for <bugs@RT-11.MIT.EDU>; Fri, 27 Sep 1996 15:56:25 -0400
Received: from achilles.ctd.anl.gov by MIT.EDU with SMTP
id AA06378; Fri, 27 Sep 96 15:56:24 EDT
Received: from pembroke.ctd.anl.gov (pembroke.ctd.anl.gov [146.137.64.73]) by achilles.ctd.anl.gov (8.6.11/8.6.11) with ESMTP id OAA02253 for <krb5-bugs@mit.edu>; Fri, 27 Sep 1996 14:56:18 -0500
Received: (b17783@localhost) by pembroke.ctd.anl.gov (8.6.11/8.6.11) id OAA32602; Fri, 27 Sep 1996 14:56:17 -0500
Message-Id: <199609271956.OAA32602@pembroke.ctd.anl.gov>
Date: Fri, 27 Sep 1996 14:56:17 -0500
From: Doug Engert <DEEngert@anl.gov>
To: krb5-bugs@MIT.EDU
Subject: K5b7 util/pty/update_utmp.c and HPUX 10.10

HPUX 10.10
The K5 beta 7 krlogind does not link, because the HPUX 10.10 system
does not have the getutmpx routine.

The following fix takes care of this.
--

Douglas E. Engert <DEEngert@anl.gov>
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois 60439
(630) 252-5444 <New Area Code 8/3/96>
PGP Key fingerprint = 20 2B 0C 78 43 8A 9C A6 29 F7 A3 6D 5E 30 A6 7F



*** ,update_utmp.c Tue Apr 16 21:11:19 1996
--- update_utmp.c Fri Sep 27 10:21:29 1996
***************
*** 124,129 ****
--- 124,130 ----
pututline(&ent);
endutent();

+ #ifdef HAVE_UPDWTMPX
#ifdef HAVE_SETUTXENT
setutxent();
getutmpx(&ent, &utx);
***************
*** 134,139 ****
--- 135,141 ----
pututxline(&utx);
endutxent();
#endif /* HAVE_SETUTXENT */
+ #endif /* HAVE_UPDWTMPX */

#else /* HAVE_SETUTENT */
if (flags&PTY_TTYSLOT_USABLE)



Responsible-Changed-From-To: gnats-admin->hartmans
Responsible-Changed-By: hartmans
Responsible-Changed-When: Mon Sep 30 00:49:29 1996
Responsible-Changed-Why:
Assign to category and update PR.


From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Doug Engert <DEEngert@anl.gov>
Cc: krb5-bugs@MIT.EDU
Subject: Re: pty/40: K5b7 util/pty/update_utmp.c and HPUX 10.10
Date: Wed, 30 Oct 1996 18:34:29 -0500

The problem with your proposed fix (to comment out that whole block of
code) is that you no longer update the utmpx file, which is the wrong
thing. I think this patch solves the problem in a cleaner fashion, but
I haven't been able to test it yet. Does anyone with access to HPUX or
DEC Unix want to comment?

Index: update_utmp.c
===================================================================
RCS file: /mit/krb5/.cvsroot/src/util/pty/update_utmp.c,v
retrieving revision 1.12
diff -u -r1.12 update_utmp.c
--- update_utmp.c 1996/04/17 02:11:19 1.12
+++ update_utmp.c 1996/10/30 23:31:06
@@ -126,7 +126,19 @@

#ifdef HAVE_SETUTXENT
setutxent();
+#ifdef HAVE_GETUTMPX
getutmpx(&ent, &utx);
+#else
+ /* For platforms like HPUX and Dec Unix which don't have getutmpx */
+ strncpy(utx.ut_user, ent.ut_user, sizeof(ent.ut_user));
+ strncpy(utx.ut_id, ent.ut_id, sizeof(ent.ut_id));
+ strncpy(utx.ut_line, ent.ut_line, sizeof(ent.ut_line));
+ utx.ut_pid = ent.ut_pid;
+ utx.ut_type = ent.ut_type;
+ utx.ut_exit = ent.ut_exit;
+ utx.ut_tv.tv_sec = ent.ut_time;
+ utx.ut_tv.tv_usec = 0;
+#endif
if (host)
strncpy(utx.ut_host, host, sizeof(utx.ut_host));
else

- Ted

From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Sam Hartman <hartmans@MIT.EDU>
Cc: krb5-bugs@MIT.EDU
Subject: Re: pty/40: K5b7 util/pty/update_utmp.c and HPUX 10.10
Date: Wed, 30 Oct 1996 18:35:37 -0500

`Theodore Y. Ts'o' made changes to this PR.

My mail contained a proposed fix, and I'm looking for comments....

From: Sam Hartman <hartmans@MIT.EDU>
To: "Theodore Y. Ts'o" <tytso@MIT.EDU>
Cc: krb5-bugs@MIT.EDU, krb5-prs@rt11.MIT.EDU
Subject: Re: pty/40: K5b7 util/pty/update_utmp.c and HPUX 10.10
Date: 31 Oct 1996 16:16:16 -0500

I'll try an HP and Alpha build tonight; thanks for the patch.

--Sam

From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Sam Hartman <hartmans@MIT.EDU>
Cc: "Theodore Y. Ts'o" <tytso@MIT.EDU>, krb5-bugs@MIT.EDU,
krb5-prs@rt11.MIT.EDU
Subject: Re: pty/40: K5b7 util/pty/update_utmp.c and HPUX 10.10
Date: Thu, 31 Oct 1996 19:01:47 -0500

From: Sam Hartman <hartmans@MIT.EDU>
Date: 31 Oct 1996 16:16:16 -0500

I'll try an HP and Alpha build tonight; thanks for the patch.

Don't forget the obvious changes to configure.in, which I didn't
include.... (not that it matters for HP and Alpha, since HAVE_GETUTXENT
will be undefined anyway).

- Ted

State-Changed-From-To: open-feedback
State-Changed-By: hartmans
State-Changed-When: Thu Nov 14 22:09:26 1996
State-Changed-Why:

Doug, could you look at the following patch that Ted checked in? (I am
not including the configure.in hunk). We don't have any HPs to try it
on.


Index: update_utmp.c
===================================================================
RCS file: /mit/krb5/.cvsroot/src/util/pty/update_utmp.c,v
retrieving revision 1.12
diff -u -r1.12 update_utmp.c
--- update_utmp.c 1996/04/17 02:11:19 1.12
+++ update_utmp.c 1996/10/30 23:31:06
@@ -126,7 +126,19 @@

#ifdef HAVE_SETUTXENT
setutxent();
+#ifdef HAVE_GETUTMPX
getutmpx(&ent, &utx);
+#else
+ /* For platforms like HPUX and Dec Unix which don't have getutmpx */
+ strncpy(utx.ut_user, ent.ut_user, sizeof(ent.ut_user));
+ strncpy(utx.ut_id, ent.ut_id, sizeof(ent.ut_id));
+ strncpy(utx.ut_line, ent.ut_line, sizeof(ent.ut_line));
+ utx.ut_pid = ent.ut_pid;
+ utx.ut_type = ent.ut_type;
+ utx.ut_exit = ent.ut_exit;
+ utx.ut_tv.tv_sec = ent.ut_time;
+ utx.ut_tv.tv_usec = 0;
+#endif
if (host)
strncpy(utx.ut_host, host, sizeof(utx.ut_host));
else


From: "Theodore Y. Ts'o" <tytso@MIT.EDU>
To: Doug Engert <DEEngert@anl.gov>
Cc: krb5-bugs@MIT.EDU
Subject: Re: pty/40: K5b7 util/pty/update_utmp.c and HPUX 10.10
Date: Thu, 5 Dec 1996 22:29:48 -0500

(This is just a note to update our gnats database.)

Doug reported that on top of the changes I made, the following patch was
still necessary. It's not a final fix, but it is suitable for
application during the 1.0 code thaw, since it won't cause any problems
for other platforms; in the long run, we want to try to make some
configure tests that test for this cleanly. For now, though, I'll be
applying this suggested patch to the 1.0 release branch.

*** ./util/pty/,update_utmp.c Fri Nov 15 07:40:31 1996
--- ./util/pty/update_utmp.c Sun Dec 1 11:48:56 1996
***************
*** 138,144 ****
--- 138,149 ----
#ifdef UT_EXIT_STRUCTURE_DIFFER
utx.ut_exit.ut_exit = ent.ut_exit.e_exit;
#else
+ #ifdef __hpux
+ utx.ut_exit.__e_termination = ent.ut_exit.e_termination;
+ utx.ut_exit.__e_exit = ent.ut_exit.e_exit;
+ #else
utx.ut_exit = ent.ut_exit;
+ #endif
#endif
utx.ut_tv.tv_sec = ent.ut_time;
utx.ut_tv.tv_usec = 0;