Hello community,

here is the log from the commit of package nfs-utils for openSUSE:Factory 
checked in at 2013-09-05 23:21:41
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/nfs-utils (Old)
 and      /work/SRC/openSUSE:Factory/.nfs-utils.new (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Package is "nfs-utils"

Changes:
--------
--- /work/SRC/openSUSE:Factory/nfs-utils/nfs-utils.changes      2013-08-06 
07:04:06.000000000 +0200
+++ /work/SRC/openSUSE:Factory/.nfs-utils.new/nfs-utils.changes 2013-09-05 
23:21:42.000000000 +0200
@@ -1,0 +2,17 @@
+Thu Sep  5 05:05:13 UTC 2013 - [email protected]
+
+- nfs.init: always start gssd.
+  Recent kernel change means that if gssd isn't running
+  and NFSv4 mount will wait 15 seconds before timing out
+  even if gss security wasn't requested.
+  So just start gssd alway - there is no significant cost
+  in this.
+  (bnc#837262)
+
+-------------------------------------------------------------------
+Mon Sep  2 05:13:25 UTC 2013 - [email protected]
+
+- skip-on-ENOENT.patch - suppress error message
+  that isn't really an error (bnc#800414)
+
+-------------------------------------------------------------------

New:
----
  skip-on-ENOENT.patch

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Other differences:
------------------
++++++ nfs-utils.spec ++++++
--- /var/tmp/diff_new_pack.X7IhSr/_old  2013-09-05 23:21:43.000000000 +0200
+++ /var/tmp/diff_new_pack.X7IhSr/_new  2013-09-05 23:21:43.000000000 +0200
@@ -69,6 +69,8 @@
 Patch5:         0005-gssd-don-t-give-up-on-machine-credential-if-hostname.patch
 # PATCH-FIX-UPSTREAM 0006-gssd-fixed-typo-in-machine-cred-name.patch 
upstream-bugfix [email protected]
 Patch6:         0006-gssd-fixed-typo-in-machine-cred-name.patch
+# PATCH-FIX-UPSTREAM skip-on-ENOENT.patch [email protected]
+Patch7:         skip-on-ENOENT.patch
 Suggests:       python-base
 
 %description
@@ -125,6 +127,7 @@
 %patch4 -p1
 %patch5 -p1
 %patch6 -p1
+%patch7 -p1
 cp %{S:6} .
 
 %build


++++++ nfs.init ++++++
--- /var/tmp/diff_new_pack.X7IhSr/_old  2013-09-05 23:21:43.000000000 +0200
+++ /var/tmp/diff_new_pack.X7IhSr/_new  2013-09-05 23:21:43.000000000 +0200
@@ -37,7 +37,8 @@
 fi
 
 NEED_IDMAPD=no
-NEED_GSSD=no
+#always start gssd, else kernel waits for it.
+NEED_GSSD=yes
 NEED_LDCONFIG=no
 state=0
 usr=""

++++++ skip-on-ENOENT.patch ++++++
>From 8becedab3982d4780dbc010decc1ac7eb9ce914f Mon Sep 17 00:00:00 2001
From: Neil Brown <[email protected]>
Date: Mon, 2 Sep 2013 14:54:16 +1000
Subject: [PATCH] gssd: support error message if rpc_pipefs dir disappears.

It is possible for a race to cause a name to appear when an rpc_pipefs
dir is scanned but to no longer be present when we try to open it.

So if the error is ENOENT, don't complain.

This is similar to

commit 5ac9bcfd820f09af4d3f87f1f7346d896f70bc9a
Author: David Jeffery <[email protected]>
Date:   Wed Jan 16 15:21:55 2013 -0500

    rpc.idmapd: Ignore open failures in dirscancb()

which addressed a similar issue in idmapd.

Signed-off-by: NeilBrown <[email protected]>

diff --git a/utils/gssd/gssd_proc.c b/utils/gssd/gssd_proc.c
index b7e2bbb..2d3dbec 100644
--- a/utils/gssd/gssd_proc.c
+++ b/utils/gssd/gssd_proc.c
@@ -467,8 +467,9 @@ process_clnt_dir(char *dir, char *pdir)
        }
        sprintf(clp->dirname, "%s/%s", pdir, dir);
        if ((clp->dir_fd = open(clp->dirname, O_RDONLY)) == -1) {
-               printerr(0, "ERROR: can't open %s: %s\n",
-                        clp->dirname, strerror(errno));
+               if (errno != ENOENT)
+                       printerr(0, "ERROR: can't open %s: %s\n",
+                                clp->dirname, strerror(errno));
                goto fail_destroy_client;
        }
        fcntl(clp->dir_fd, F_SETSIG, DNOTIFY_SIGNAL);
-- 
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to