From: Jeff Layton <jlayton@redhat.com>
Date: Mon, 31 Mar 2008 15:15:48 -0400
Subject: [nfs] fix transposed deltas in nfs v3
Message-id: 1206990948-28285-1-git-send-email-jlayton@redhat.com
O-Subject: [RHEL5.3 PATCH] BZ#437544: fix transposed deltas in nfs3_write_done and nfs3_commit_done
Bugzilla: 437544
RH-Acked-by: Peter Staubach <staubach@redhat.com>

There appear to be transposed deltas in nfs3_write_done and
nfs3_commit_done in the patch for BZ 321111. COMMIT calls shouldn't
change the mtimes and so we shouldn't need to use the force_wcc call
after one. WRITE's on the other hand generally do change the mtime, so we
want to make sure we force_wcc after one so that we don't invalidate the
cache due to our own writes.

The effect of this problem isn't particularly dire, but it can make the
client invalidate the cache more frequently than it should when the
server doesn't send a pre_op_attr struct in a write reply.

 fs/nfs/nfs3proc.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 86e1529..0246904 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -822,7 +822,8 @@ static int nfs3_write_done(struct rpc_task *task, struct nfs_write_data *data)
 {
 	if (nfs3_async_handle_jukebox(task, data->inode))
 		return -EAGAIN;
-	nfs_refresh_inode(data->inode, data->res.fattr);
+	if (task->tk_status >= 0)
+		nfs_post_op_update_inode_force_wcc(data->inode, data->res.fattr);
 	return 0;
 }
 
@@ -850,8 +851,7 @@ static int nfs3_commit_done(struct rpc_task *task, struct nfs_write_data *data)
 {
 	if (nfs3_async_handle_jukebox(task, data->inode))
 		return -EAGAIN;
-	if (task->tk_status >= 0)
-		nfs_post_op_update_inode_force_wcc(data->inode, data->res.fattr);
+	nfs_refresh_inode(data->inode, data->res.fattr);
 	return 0;
 }