From john@rwsrv05.mby.riverwillow.net.au  Sat Oct 27 09:57:45 2007
Return-Path: <john@rwsrv05.mby.riverwillow.net.au>
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id D6F1316A419
	for <FreeBSD-gnats-submit@freebsd.org>; Sat, 27 Oct 2007 09:57:45 +0000 (UTC)
	(envelope-from john@rwsrv05.mby.riverwillow.net.au)
Received: from omta02ps.mx.bigpond.com (omta02ps.mx.bigpond.com [144.140.83.154])
	by mx1.freebsd.org (Postfix) with ESMTP id 69BDC13C48A
	for <FreeBSD-gnats-submit@freebsd.org>; Sat, 27 Oct 2007 09:57:45 +0000 (UTC)
	(envelope-from john@rwsrv05.mby.riverwillow.net.au)
Received: from oaamta04ps.mx.bigpond.com ([121.210.32.83])
          by omta02ps.mx.bigpond.com with ESMTP
          id <20071027095743.TNCU19085.omta02ps.mx.bigpond.com@oaamta04ps.mx.bigpond.com>
          for <FreeBSD-gnats-submit@freebsd.org>;
          Sat, 27 Oct 2007 09:57:43 +0000
Received: from mail3.riverwillow.net.au ([121.210.32.83])
          by oaamta04ps.mx.bigpond.com with ESMTP
          id <20071027095743.GAVK15825.oaamta04ps.mx.bigpond.com@mail3.riverwillow.net.au>
          for <FreeBSD-gnats-submit@freebsd.org>;
          Sat, 27 Oct 2007 09:57:43 +0000
Received: from rwsrv05.mby.riverwillow.net.au (localhost [127.0.0.1])
	by mail3.riverwillow.net.au (8.14.1/8.14.1) with ESMTP id l9R9vgZ2000787
	for <FreeBSD-gnats-submit@freebsd.org>; Sat, 27 Oct 2007 19:57:42 +1000 (AEST)
Received: (from john@localhost)
	by rwsrv05.mby.riverwillow.net.au (8.14.1/8.14.1/Submit) id l9R9vgMD000786;
	Sat, 27 Oct 2007 19:57:42 +1000 (AEST)
Message-Id: <200710270957.l9R9vgMD000786@rwsrv05.mby.riverwillow.net.au>
Date: Sat, 27 Oct 2007 19:57:42 +1000 (AEST)
From: John Marshall <john@rwsrv05.mby.riverwillow.net.au>
Reply-To: John Marshall <john.marshall@riverwillow.com.au>
To: FreeBSD-gnats-submit@freebsd.org
Cc:
Subject: [patch] NIS-dependent services cannot start if ypbind uses ypset
X-Send-Pr-Version: 3.113
X-GNATS-Notify:

>Number:         117555
>Category:       conf
>Synopsis:       [patch] [nis] NIS-dependent services cannot start if ypbind uses ypset
>Confidential:   no
>Severity:       serious
>Priority:       low
>Responsible:    dougb
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          change-request
>Submitter-Id:   current-users
>Arrival-Date:   Sat Oct 27 10:00:09 UTC 2007
>Closed-Date:    Mon Aug 03 20:28:02 UTC 2009
>Last-Modified:  Mon Aug 03 20:28:02 UTC 2009
>Originator:     John Marshall
>Release:        FreeBSD 7.0-BETA1 i386
>Organization:
Riverwillow Pty Ltd
>Environment:
System: FreeBSD rwsrv05.mby.riverwillow.net.au 7.0-BETA1 FreeBSD 7.0-BETA1 #0: Sat Oct 20 10:13:47 AEST 2007 root@rwsrv05.mby.riverwillow.net.au:/spare/obj/usr/src/sys/RWSRV05 i386

>Description:
/etc/rc.d/ypset needs to be relocated within the rc order so that it comes after /etc/rc.d/ypbind and before anything which requires ypbind. At present, ypset is not REQUIREd by any other rc.d scripts and falls through to way near the end of rcorder. If an NIS client's ypbind does not use ypset for its configuration, then this oversight has no impact at all: ypset does not need to run and ypbind will be properly configured and functional after /etc/rc.d/ypbind. If, however, ypbind is set to be configured by ypset, it will not be functional until after ypset has run.

For an installation that does use ypset, /etc/rc.d/ypset starts way too late in rcorder and cannot configure ypbind before ypbind consumers attempt to start. Those NIS-dependent services will not be able to start because they are waiting for NIS services which cannot be provided because ypbind has not yet been configured by ypset. The NIS-dependent services wait indefinately (unless someone is around to hit ctrl-C) and rc does not complete.

I am proposing that those rc.d scripts which REQUIRE "ypbind" be changed to REQUIRE "ypset" instead. /etc/rc.d/ypset already REQUIREs "ypbind", so that keeps the ypbind dependency while relocating ypset to an appropriate position within the rc order.

Scripts which currently REQUIRE "ypbind" are:
- /etc/rc.d/amd
- /etc/rc.d/keyserv
- /etc/rc.d/yppasswdd

Another script which does not currently REQUIRE "ypbind" but has a comment saying that ypbind must already be started is /etc/rc.d/quota. This script just happens to land in the right spot in rcorder so the missing ypbind REQUIRE hasn't been a problem - but I think that omission should be corrected while we're at it.

Testing this solution on this system has shown that the only affect upon rcorder is the re-positioning of ypset. /etc/rc.d/ypset moves from way near the end to immediately after /etc/rc.d/ypbind - and nothing else moves. Please see the following thread in the rc mailing list for details of the symptoms, analysis and test results:
http://lists.freebsd.org/pipermail/freebsd-rc/2007-October/001231.html

My one misgiving about this proposal is that "REQUIRE ypset" does not make the ypbind dependency immediately apparent.

>How-To-Repeat:
nis_client_enable="YES"
nis_client_flags="-ypsetme -s"
nis_ypset_enable="YES"
nis_ypset_flags="NIS.Server.IP.Address"

>Fix:

--- /usr/src/etc/rc.d/amd	2006-10-19 01:56:11.000000000 +1000
+++ /etc/rc.d/amd	2007-10-25 10:40:36.000000000 +1000
@@ -5,7 +5,7 @@
 #
 
 # PROVIDE: amd
-# REQUIRE: rpcbind ypbind nfsclient cleanvar ldconfig
+# REQUIRE: rpcbind ypset nfsclient cleanvar ldconfig
 # BEFORE: DAEMON
 # KEYWORD: nojail


--- /usr/src/etc/rc.d/keyserv	2004-10-07 23:55:26.000000000 +1000
+++ /etc/rc.d/keyserv	2007-10-25 10:40:36.000000000 +1000
@@ -7,7 +7,7 @@
 #
 
 # PROVIDE: keyserv
-# REQUIRE: ypbind
+# REQUIRE: ypset
 # BEFORE: DAEMON
 
 . /etc/rc.subr


--- /usr/src/etc/rc.d/yppasswdd	2007-10-12 12:38:42.000000000 +1000
+++ /etc/rc.d/yppasswdd	2007-10-25 10:40:36.000000000 +1000
@@ -5,7 +5,7 @@
 #
 
 # PROVIDE: yppasswdd
-# REQUIRE: ypserv ypbind
+# REQUIRE: ypserv ypset
 # BEFORE:  LOGIN
 
 . /etc/rc.subr


--- /usr/src/etc/rc.d/quota	2007-10-12 12:38:42.000000000 +1000
+++ /etc/rc.d/quota	2007-10-25 10:55:52.000000000 +1000
@@ -7,7 +7,7 @@
 #
 
 # PROVIDE: quota
-# REQUIRE: mountcritremote
+# REQUIRE: mountcritremote ypset
 # BEFORE: DAEMON
 # KEYWORD: nojail
 
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->analyzed 
State-Changed-By: dougb 
State-Changed-When: Sat Oct 27 22:50:28 UTC 2007 
State-Changed-Why:  

I worked with John to develop this solution and I think it's the right one. 

As for your reservation about the ypbind dependency not being explicit, 
ypset has a REQUIRE for ypbind, which is how rcorder is supposed to 
work. The REQUIRE for ypserv in yppasswdd for example is overkill, and 
arguably should be removed but I don't want to tempt fate too much here. 

Doug 


Responsible-Changed-From-To: freebsd-bugs->dougb 
Responsible-Changed-By: dougb 
Responsible-Changed-When: Sat Oct 27 22:50:28 UTC 2007 
Responsible-Changed-Why:  

I'll shepherd this one. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=117555 

From: Mike Makonnen <mtm@FreeBSD.Org>
To: bug-followup@FreeBSD.org, john.marshall@riverwillow.com.au
Cc:  
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if
	ypbind uses ypset
Date: Sun, 28 Oct 2007 12:29:08 +0300

 What's wrong with having the scripts REQUIRE both ypbind and ypset?
 
 Cheers.
 -- 
 Mike Makonnen         | GPG-KEY: http://people.freebsd.org/~mtm/mtm.asc
 mmakonnen @ gmail.com | AC7B 5672 2D11 F4D0 EBF8 5279 5359 2B82 7CD4 1F55
 mtm @ FreeBSD.Org     | FreeBSD - http://www.freebsd.org

From: John Marshall <John.Marshall@riverwillow.com.au>
To: Mike Telahun Makonnen <mtm@FreeBSD.Org>, bug-followup@FreeBSD.Org
Cc: Doug Barton <dougb@FreeBSD.Org>
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if ypbind
 uses ypset
Date: Mon, 29 Oct 2007 20:11:18 +1100

 If scripts REQUIRE both ypbind and ypset, then the ypbind entry is 
 superfluous as far as rcorder is concerned; but it *might* be more 
 helpful (or confusing!) from a self-documenting perspective.
 
 Either adding ypset to the scripts' REQUIRE lines or replacing ypbind 
 with ypset in the scripts' REQUIRE lines will achieve my object. I'll 
 leave the pragmatic/aesthetic argument to the experts!
 
 -- 
 John Marshall

From: Mike Makonnen <mtm@FreeBSD.Org>
To: John Marshall <John.Marshall@riverwillow.com.au>
Cc: Mike Telahun Makonnen <mtm@FreeBSD.Org>, bug-followup@FreeBSD.Org, Doug Barton <dougb@FreeBSD.Org>
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if
	ypbind uses ypset
Date: Tue, 30 Oct 2007 09:31:07 +0300

 On Mon, 2007-10-29 at 20:11 +1100, John Marshall wrote:
 > If scripts REQUIRE both ypbind and ypset, then the ypbind entry is 
 > superfluous as far as rcorder is concerned;
 
 Whether or not rcorder(8) finds it superfluous is beside the point.
 The REQUIRE line is there to properly record dependencies among scripts.
 If, in order to work properly, the script requires
 that ypbind and ypset be run before it, then both those scripts should
 be in its REQUIRE line. For example, what would happen if for some
 reason I decided to remove rc.d/ypset as a local hack? 
 
 Cheers 
 -- 
 Mike Makonnen         | GPG-KEY: http://people.freebsd.org/~mtm/mtm.asc
 mmakonnen @ gmail.com | AC7B 5672 2D11 F4D0 EBF8 5279 5359 2B82 7CD4 1F55
 mtm @ FreeBSD.Org     | FreeBSD - http://www.freebsd.org

From: Doug Barton <dougb@FreeBSD.org>
To: Mike Makonnen <mtm@FreeBSD.Org>
Cc: John Marshall <John.Marshall@riverwillow.com.au>, bug-followup@FreeBSD.Org
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if
 ypbind uses ypset
Date: Mon, 29 Oct 2007 23:32:21 -0700 (PDT)

 On Tue, 30 Oct 2007, Mike Makonnen wrote:
 
 > On Mon, 2007-10-29 at 20:11 +1100, John Marshall wrote:
 >> If scripts REQUIRE both ypbind and ypset, then the ypbind entry is
 >> superfluous as far as rcorder is concerned;
 >
 > Whether or not rcorder(8) finds it superfluous is beside the point.
 > The REQUIRE line is there to properly record dependencies among scripts.
 
 ypset has a REQUIRE for ypbind, so IMO it is properly recorded. I think 
 it's also dangerous to specify too many things in REQUIRE because it could 
 easily lead to circular dependencies, and/or unfullfillable prereqs. I'm 
 not willing to waste a lot of energy on opposing this suggestion, but I do 
 hope I can talk you out of it. :)
 
 > For example, what would happen if for some reason I decided to remove 
 > rc.d/ypset as a local hack?
 
 Then things would break, as they should. I really don't think we should 
 emphasize "give users who are inclined to shoot themselves in the foot a 
 belt and suspenders just in case" vs. optimizing for the common case.
 
 Doug
 
 -- 
 
      This .signature sanitized for your protection
 

From: Mike Makonnen <mtm@FreeBSD.Org>
To: Doug Barton <dougb@FreeBSD.org>
Cc: John Marshall <John.Marshall@riverwillow.com.au>,  bug-followup@FreeBSD.Org
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if
	ypbind uses ypset
Date: Tue, 30 Oct 2007 10:42:21 +0300

 On Mon, 2007-10-29 at 23:32 -0700, Doug Barton wrote:
 > On Tue, 30 Oct 2007, Mike Makonnen wrote:
 > 
 > > On Mon, 2007-10-29 at 20:11 +1100, John Marshall wrote:
 > >> If scripts REQUIRE both ypbind and ypset, then the ypbind entry is
 > >> superfluous as far as rcorder is concerned;
 > >
 > > Whether or not rcorder(8) finds it superfluous is beside the point.
 > > The REQUIRE line is there to properly record dependencies among scripts.
 > 
 > ypset has a REQUIRE for ypbind, so IMO it is properly recorded. I think 
 > it's also dangerous to specify too many things in REQUIRE because it could 
 > easily lead to circular dependencies, and/or unfullfillable prereqs.
 
 I would argue that relying on "indirect" dependencies is also dangerous.
 In fact, relying on this sort of "indirect" dependency is likely to make
 it harder to debug circular dependencies when they do happen. Regardless
 of whether you make the dependency explicit or not, it is still there,
 so it is neither more nor less likely to lead to circular dependencies.
 
 I'm not sure what you mean by "unfullfillable prereqs."
 
 > 
 > > For example, what would happen if for some reason I decided to remove 
 > > rc.d/ypset as a local hack?
 > 
 > Then things would break, as they should.
 
 No, they should *not* break. We should be striving to make the system
 more reliable and resilient in the face of change, not less.
 
 >  I really don't think we should 
 > emphasize "give users who are inclined to shoot themselves in the foot a 
 > belt and suspenders just in case" vs. optimizing for the common case.
 
 But one of the selling points of rc.d *was* indeed that it would make it
 easier for users to introduce local hacks *without* breaking their
 system.
 
 Cheers.
 -- 
 Mike Makonnen         | GPG-KEY: http://people.freebsd.org/~mtm/mtm.asc
 mmakonnen @ gmail.com | AC7B 5672 2D11 F4D0 EBF8 5279 5359 2B82 7CD4 1F55
 mtm @ FreeBSD.Org     | FreeBSD - http://www.freebsd.org

From: John Marshall <john.marshall@riverwillow.com.au>
To: bug-followup@FreeBSD.org, john.marshall@riverwillow.com.au,
        dougb@FreeBSD.org
Cc:  
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if ypbind uses ypset
Date: Thu, 12 Jun 2008 15:57:58 +1000

 Is now a good time to do this?
 
 -- 
 John Marshall

From: John Marshall <john.marshall@riverwillow.com.au>
To: bug-followup@FreeBSD.org
Cc: dougb@FreeBSD.org
Subject: Re: conf/117555: [patch] NIS-dependent services cannot start if ypbind uses ypset
Date: Wed, 17 Sep 2008 17:01:15 +1000

 --UugvWAfsgieZRqgk
 Content-Type: text/plain; charset=us-ascii
 Content-Disposition: inline
 
 This morning I upgraded one of our servers to 7.1-PRERELEASE to do our
 bit in helping with testing for this release cycle. The upgrade source
 was RELENG_7 obtained this morning via CVSup.
 
 rwsrv05> uname -a
 FreeBSD rwsrv05.mby.riverwillow.net.au 7.1-PRERELEASE FreeBSD 7.1-PRERELEASE #0: Wed Sep 17 10:41:36 AEST 2008     root@rwsrv05.mby.riverwillow.net.au:/spare/obj/usr/src/sys/RWSRV05  i386
 
 I let mergemaster clobber my patched versions of the NIS rc.d files. No
 surprises for guessing that the original problem persists. The startup
 "hung" at mountd until I hit ctrl-C and similarly for 4 other network
 services (ircd, squid, dkim-milter and cvsupd). Hitting ctrl-T before
 ctrl-C showed "su" in [nanslp]. I have attached patches against the
 7.1-PRERELEASE versions of those files.
 
 I ran a "before and after" sweep to check rcorder. The only change due
 to my patches was that ypset moved back 73 places and landed immediately
 after ypbind - where it needs to be.
 
 --- rc.late.7-1-PRERELEASE	2008-09-17 14:43:18.000000000 +1000
 +++ rc.late.7-1-PRERELEASE_patched	2008-09-17 14:49:18.000000000 +1000
 @@ -54,6 +54,7 @@
  /etc/rc.d/nisdomain
  /etc/rc.d/ypserv
  /etc/rc.d/ypbind
 +/etc/rc.d/ypset
  /etc/rc.d/amd
  /etc/rc.d/atm3
  /etc/rc.d/auditd
 @@ -127,7 +128,6 @@
  /usr/local/etc/rc.d/apache22
  /etc/rc.d/ypxfrd
  /etc/rc.d/ypupdated
 -/etc/rc.d/ypset
  /etc/rc.d/watchdogd
  /etc/rc.d/syscons
  /etc/rc.d/sshd
 
 
 -- 
 John Marshall
 
 --UugvWAfsgieZRqgk
 Content-Type: text/x-diff; charset=us-ascii
 Content-Disposition: attachment; filename="amd.diff"
 
 --- /usr/src/etc/rc.d/amd	2008-01-28 18:55:44.000000000 +1100
 +++ /etc/rc.d/amd	2008-09-17 14:46:57.000000000 +1000
 @@ -4,7 +4,7 @@
  #
  
  # PROVIDE: amd
 -# REQUIRE: rpcbind ypbind nfsclient cleanvar ldconfig
 +# REQUIRE: rpcbind ypset nfsclient cleanvar ldconfig
  # BEFORE: DAEMON
  # KEYWORD: nojail
  
 
 --UugvWAfsgieZRqgk
 Content-Type: text/x-diff; charset=us-ascii
 Content-Disposition: attachment; filename="keyserv.diff"
 
 --- /usr/src/etc/rc.d/keyserv	2008-01-28 18:55:44.000000000 +1100
 +++ /etc/rc.d/keyserv	2008-09-17 14:47:14.000000000 +1000
 @@ -6,7 +6,7 @@
  # Start keyserv if we are running Secure RPC
  
  # PROVIDE: keyserv
 -# REQUIRE: ypbind
 +# REQUIRE: ypset
  # BEFORE: DAEMON
  
  . /etc/rc.subr
 
 --UugvWAfsgieZRqgk
 Content-Type: text/x-diff; charset=us-ascii
 Content-Disposition: attachment; filename="yppasswdd.diff"
 
 --- /usr/src/etc/rc.d/yppasswdd	2008-01-28 18:55:44.000000000 +1100
 +++ /etc/rc.d/yppasswdd	2008-09-17 14:47:58.000000000 +1000
 @@ -4,7 +4,7 @@
  #
  
  # PROVIDE: yppasswdd
 -# REQUIRE: ypserv ypbind
 +# REQUIRE: ypserv ypset
  # BEFORE:  LOGIN
  
  . /etc/rc.subr
 
 --UugvWAfsgieZRqgk
 Content-Type: text/x-diff; charset=us-ascii
 Content-Disposition: attachment; filename="quota.diff"
 
 --- /usr/src/etc/rc.d/quota	2008-01-28 18:55:44.000000000 +1100
 +++ /etc/rc.d/quota	2008-09-17 14:47:39.000000000 +1000
 @@ -6,7 +6,7 @@
  # Enable/Check the quotas (must be after ypbind if using NIS)
  
  # PROVIDE: quota
 -# REQUIRE: mountcritremote
 +# REQUIRE: mountcritremote ypset
  # BEFORE: DAEMON
  # KEYWORD: nojail
  
 
 --UugvWAfsgieZRqgk--
State-Changed-From-To: analyzed->patched 
State-Changed-By: dougb 
State-Changed-When: Mon Jun 1 04:56:20 UTC 2009 
State-Changed-Why:  

This is done in HEAD, I'll MFC it after a week or so. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=117555 

From: dfilter@FreeBSD.ORG (dfilter service)
To: bug-followup@FreeBSD.org
Cc:  
Subject: Re: conf/117555: commit references a PR
Date: Mon,  1 Jun 2009 04:55:23 +0000 (UTC)

 Author: dougb
 Date: Mon Jun  1 04:55:13 2009
 New Revision: 193197
 URL: http://svn.freebsd.org/changeset/base/193197
 
 Log:
   Substitute ypset for ypbind in REQUIRE lines. If you use ypset it has to
   happen right after ypbind, and before anything that uses NIS. The only
   change in rcorder accomplished by this patch is make that happen.
   
   PR:		conf/117555
   Submitted by:	John Marshall <john@rwsrv05.mby.riverwillow.net.au>
 
 Modified:
   head/etc/rc.d/amd
   head/etc/rc.d/keyserv
   head/etc/rc.d/nisdomain
   head/etc/rc.d/quota
   head/etc/rc.d/yppasswdd
 
 Modified: head/etc/rc.d/amd
 ==============================================================================
 --- head/etc/rc.d/amd	Mon Jun  1 04:44:43 2009	(r193196)
 +++ head/etc/rc.d/amd	Mon Jun  1 04:55:13 2009	(r193197)
 @@ -4,7 +4,7 @@
  #
  
  # PROVIDE: amd
 -# REQUIRE: rpcbind ypbind nfsclient cleanvar ldconfig
 +# REQUIRE: rpcbind ypset nfsclient cleanvar ldconfig
  # BEFORE: DAEMON
  # KEYWORD: nojail shutdown
  
 
 Modified: head/etc/rc.d/keyserv
 ==============================================================================
 --- head/etc/rc.d/keyserv	Mon Jun  1 04:44:43 2009	(r193196)
 +++ head/etc/rc.d/keyserv	Mon Jun  1 04:55:13 2009	(r193197)
 @@ -6,7 +6,7 @@
  # Start keyserv if we are running Secure RPC
  
  # PROVIDE: keyserv
 -# REQUIRE: ypbind
 +# REQUIRE: ypset
  # BEFORE: DAEMON
  # KEYWORD: shutdown
  
 
 Modified: head/etc/rc.d/nisdomain
 ==============================================================================
 --- head/etc/rc.d/nisdomain	Mon Jun  1 04:44:43 2009	(r193196)
 +++ head/etc/rc.d/nisdomain	Mon Jun  1 04:55:13 2009	(r193197)
 @@ -28,7 +28,7 @@
  
  # PROVIDE: nisdomain
  # REQUIRE: SERVERS rpcbind
 -# BEFORE:  ypbind ypserv ypxfrd
 +# BEFORE:  ypset ypbind ypserv ypxfrd
  
  . /etc/rc.subr
  
 
 Modified: head/etc/rc.d/quota
 ==============================================================================
 --- head/etc/rc.d/quota	Mon Jun  1 04:44:43 2009	(r193196)
 +++ head/etc/rc.d/quota	Mon Jun  1 04:55:13 2009	(r193197)
 @@ -6,7 +6,7 @@
  # Enable/Check the quotas (must be after ypbind if using NIS)
  
  # PROVIDE: quota
 -# REQUIRE: mountcritremote
 +# REQUIRE: mountcritremote ypset
  # BEFORE: DAEMON
  # KEYWORD: nojail
  
 
 Modified: head/etc/rc.d/yppasswdd
 ==============================================================================
 --- head/etc/rc.d/yppasswdd	Mon Jun  1 04:44:43 2009	(r193196)
 +++ head/etc/rc.d/yppasswdd	Mon Jun  1 04:55:13 2009	(r193197)
 @@ -4,7 +4,7 @@
  #
  
  # PROVIDE: yppasswdd
 -# REQUIRE: ypserv ypbind
 +# REQUIRE: ypserv ypset
  # BEFORE:  LOGIN
  # KEYWORD: shutdown
  
 _______________________________________________
 svn-src-all@freebsd.org mailing list
 http://lists.freebsd.org/mailman/listinfo/svn-src-all
 To unsubscribe, send any mail to "svn-src-all-unsubscribe@freebsd.org"
 
State-Changed-From-To: patched->closed 
State-Changed-By: dougb 
State-Changed-When: Mon Aug 3 20:27:29 UTC 2009 
State-Changed-Why:  

This fix has now been MFCed. 

http://www.freebsd.org/cgi/query-pr.cgi?pr=117555 
>Unformatted:
