From 160bc1604f8a33202578846c9a63e2a61105a4b7 Mon Sep 17 00:00:00 2001
From: Jesper Juhl <jj@chaosbits.net>
Date: Sun, 10 Apr 2011 17:57:07 +0200
Subject: [PATCH 01/20] NFS: Remove dead code from nfs_fs_mount()

In fs/nfs/super.c::nfs_fs_mount() we test for a NULL 'data':

...
 		if (data == NULL || mntfh == NULL)
 			goto out_free_fh;
...

and then further down in the function we test 'data' again:

...
 			nfs_fscache_get_super_cookie(
 				s, data ? data->fscache_uniq : NULL, NULL);
...

this second check is just dead code since there is no way 'data' could
possibly be NULL here.
We also rely on a non-NULL 'data' in more than one location between these
two tests, further proving the point that the second test is bogus.

This patch removes the dead code.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/super.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 2b8e9a5e366a..685a8a7bd17b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2235,8 +2235,7 @@ static struct dentry *nfs_fs_mount(struct file_system_type *fs_type,
 	if (!s->s_root) {
 		/* initial superblock/root creation */
 		nfs_fill_super(s, data);
-		nfs_fscache_get_super_cookie(
-			s, data ? data->fscache_uniq : NULL, NULL);
+		nfs_fscache_get_super_cookie(s, data->fscache_uniq, NULL);
 	}
 
 	mntroot = nfs_get_root(s, mntfh, dev_name);

From 4b38a6db01b09198f4045661815a0039c3d80660 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Mon, 11 Apr 2011 11:56:23 -0400
Subject: [PATCH 02/20] NFS: Eliminate duplicate call to nfs_mark_request_dirty

We only need to call nfs_mark_request_dirty() once in nfs_writepage_setup().

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/write.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index e4cbc11a74ab..bb608186b099 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -680,7 +680,6 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page,
 	req = nfs_setup_write_request(ctx, page, offset, count);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
-	nfs_mark_request_dirty(req);
 	/* Update file length */
 	nfs_grow_file(page, offset, count);
 	nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes);

From 561f0b0ad073859a87d22845ddfd9df149b22e5f Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Tue, 12 Apr 2011 08:47:15 -0400
Subject: [PATCH 03/20] NFS: Remove unused argument from nfs_find_best_sec()

The inode was used in an earlier version of the code, but it isn't
used anymore.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/namespace.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 89fc160fd5b0..1f063bacd285 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -119,7 +119,7 @@ Elong:
 }
 
 #ifdef CONFIG_NFS_V4
-static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors, struct inode *inode)
+static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors)
 {
 	struct gss_api_mech *mech;
 	struct xdr_netobj oid;
@@ -166,7 +166,7 @@ static int nfs_negotiate_security(const struct dentry *parent,
 		}
 		flavors = page_address(page);
 		ret = secinfo(parent->d_inode, &dentry->d_name, flavors);
-		*flavor = nfs_find_best_sec(flavors, dentry->d_inode);
+		*flavor = nfs_find_best_sec(flavors);
 		put_page(page);
 	}
 

From c0d0e96b840dcc73f9b9d45bf8f405dbce72a079 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Tue, 12 Apr 2011 12:29:15 -0400
Subject: [PATCH 04/20] NFS: Get rid of pointless test in nfs_commit_done

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/write.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index bb608186b099..3bd5d7e80f6c 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1417,8 +1417,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata)
                                 task->tk_pid, task->tk_status);
 
 	/* Call the NFS version-specific code */
-	if (NFS_PROTO(data->inode)->commit_done(task, data) != 0)
-		return;
+	NFS_PROTO(data->inode)->commit_done(task, data);
 }
 
 void nfs_commit_release_pages(struct nfs_write_data *data)

From 79a48a1f5d99ab66cf83d2d5d805e7a0e08452ed Mon Sep 17 00:00:00 2001
From: Weston Andros Adamson <dros@netapp.com>
Date: Wed, 13 Apr 2011 10:53:51 -0400
Subject: [PATCH 05/20] Don't mark_inode_dirty_sync() while holding lock

mark_inode_dirty_sync() grabs the same inode lock!

race conditions between holding the lock in pnfs_set_layoutcommit() and in
mark_inode_dirty_sync() can result in a second call to pnfs_layoutcommit_inode(), but
this will be a noop as NFS_INO_LAYOUTCOMMIT won't be set in the second call

Signed-off-by: Weston Andros Adamson <dros@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/pnfs.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index d9ab97269ce6..ff681ab65d31 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -1004,6 +1004,7 @@ pnfs_set_layoutcommit(struct nfs_write_data *wdata)
 {
 	struct nfs_inode *nfsi = NFS_I(wdata->inode);
 	loff_t end_pos = wdata->args.offset + wdata->res.count;
+	bool mark_as_dirty = false;
 
 	spin_lock(&nfsi->vfs_inode.i_lock);
 	if (!test_and_set_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags)) {
@@ -1011,13 +1012,18 @@ pnfs_set_layoutcommit(struct nfs_write_data *wdata)
 		get_lseg(wdata->lseg);
 		wdata->lseg->pls_lc_cred =
 			get_rpccred(wdata->args.context->state->owner->so_cred);
-		mark_inode_dirty_sync(wdata->inode);
+		mark_as_dirty = true;
 		dprintk("%s: Set layoutcommit for inode %lu ",
 			__func__, wdata->inode->i_ino);
 	}
 	if (end_pos > wdata->lseg->pls_end_pos)
 		wdata->lseg->pls_end_pos = end_pos;
 	spin_unlock(&nfsi->vfs_inode.i_lock);
+
+	/* if pnfs_layoutcommit_inode() runs between inode locks, the next one
+	 * will be a noop because NFS_INO_LAYOUTCOMMIT will not be set */
+	if (mark_as_dirty)
+		mark_inode_dirty_sync(wdata->inode);
 }
 EXPORT_SYMBOL_GPL(pnfs_set_layoutcommit);
 

From d1a8016a2d1e75021ecc8715e3c81442d7218eb6 Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 13 Apr 2011 14:31:28 -0400
Subject: [PATCH 06/20] NFS: Fix infinite loop in gss_create_upcall()

There can be an infinite loop if gss_create_upcall() is called without
the userspace program running.  To prevent this, we return -EACCES if
we notice that pipe_version hasn't changed (indicating that the pipe
has not been opened).

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c              | 5 +++--
 net/sunrpc/auth_gss/auth_gss.c | 8 +++++---
 2 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 9bf41eab3e46..8a03ee0689f3 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2224,8 +2224,9 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 
 	for (i = 0; i < len; i++) {
 		status = nfs4_lookup_root_sec(server, fhandle, info, flav_array[i]);
-		if (status != -EPERM)
-			break;
+		if (status == -EPERM || status == -EACCES)
+			continue;
+		break;
 	}
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);
diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index f3914d0c5079..339ba64cce1e 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -520,7 +520,7 @@ gss_refresh_upcall(struct rpc_task *task)
 		warn_gssd();
 		task->tk_timeout = 15*HZ;
 		rpc_sleep_on(&pipe_version_rpc_waitqueue, task, NULL);
-		return 0;
+		return -EAGAIN;
 	}
 	if (IS_ERR(gss_msg)) {
 		err = PTR_ERR(gss_msg);
@@ -563,10 +563,12 @@ retry:
 	if (PTR_ERR(gss_msg) == -EAGAIN) {
 		err = wait_event_interruptible_timeout(pipe_version_waitqueue,
 				pipe_version >= 0, 15*HZ);
+		if (pipe_version < 0) {
+			warn_gssd();
+			err = -EACCES;
+		}
 		if (err)
 			goto out;
-		if (pipe_version < 0)
-			warn_gssd();
 		goto retry;
 	}
 	if (IS_ERR(gss_msg)) {

From 0fabee243a2c6edd66284a4d8948ccbe6727e3bb Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 13 Apr 2011 14:31:29 -0400
Subject: [PATCH 07/20] NFS: flav_array honors NFS_MAX_SECFLAVORS

NFS_MAX_SECFLAVORS should already take into account RPC_AUTH_UNIX
and RPC_AUTH_NULL, so we don't need to set aside extra slots
for them.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 8a03ee0689f3..8833472c09a9 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2215,7 +2215,7 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 			      struct nfs_fsinfo *info)
 {
 	int i, len, status = 0;
-	rpc_authflavor_t flav_array[NFS_MAX_SECFLAVORS + 2];
+	rpc_authflavor_t flav_array[NFS_MAX_SECFLAVORS];
 
 	flav_array[0] = RPC_AUTH_UNIX;
 	len = gss_mech_list_pseudoflavors(&flav_array[1]);

From 801a16dc7b5c146f7980a0c61c30cef3ba93344d Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 13 Apr 2011 14:31:30 -0400
Subject: [PATCH 08/20] NFS: Attempt mount with default sec flavor first

nfs4_lookup_root() is already configured to use either RPC_AUTH_UNIX
or a user specified flavor (through -o sec=<whatever>).  We should
use this flavor first, and only attempt negotiation if it fails
with -EPERM.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 8833472c09a9..7e27ebf5be5d 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2208,19 +2208,15 @@ out:
 	return ret;
 }
 
-/*
- * get the file handle for the "/" directory on the server
- */
-static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
+static int nfs4_find_root_sec(struct nfs_server *server, struct nfs_fh *fhandle,
 			      struct nfs_fsinfo *info)
 {
 	int i, len, status = 0;
 	rpc_authflavor_t flav_array[NFS_MAX_SECFLAVORS];
 
-	flav_array[0] = RPC_AUTH_UNIX;
-	len = gss_mech_list_pseudoflavors(&flav_array[1]);
-	flav_array[1+len] = RPC_AUTH_NULL;
-	len += 2;
+	len = gss_mech_list_pseudoflavors(&flav_array[0]);
+	flav_array[len] = RPC_AUTH_NULL;
+	len += 1;
 
 	for (i = 0; i < len; i++) {
 		status = nfs4_lookup_root_sec(server, fhandle, info, flav_array[i]);
@@ -2228,6 +2224,18 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 			continue;
 		break;
 	}
+	return status;
+}
+
+/*
+ * get the file handle for the "/" directory on the server
+ */
+static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
+			      struct nfs_fsinfo *info)
+{
+	int status = nfs4_lookup_root(server, fhandle, info);
+	if (status == -EPERM)
+		status = nfs4_find_root_sec(server, fhandle, info);
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);
 	if (status == 0)

From 9b7160c55a41dd2fec3d467f979e55782d3f92ad Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 13 Apr 2011 14:31:30 -0400
Subject: [PATCH 09/20] NFS: don't negotiate when user specifies sec flavor

We were always attempting sec flavor negotiation, even if the user
told us a specific sec flavor to use.  If that sec flavor fails,
we should return an error rather than continuing with sec flavor
negotiation.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c | 3 ++-
 fs/nfs/super.c    | 1 +
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 7e27ebf5be5d..b8e1ac69b743 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -46,6 +46,7 @@
 #include <linux/nfs4.h>
 #include <linux/nfs_fs.h>
 #include <linux/nfs_page.h>
+#include <linux/nfs_mount.h>
 #include <linux/namei.h>
 #include <linux/mount.h>
 #include <linux/module.h>
@@ -2234,7 +2235,7 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 			      struct nfs_fsinfo *info)
 {
 	int status = nfs4_lookup_root(server, fhandle, info);
-	if (status == -EPERM)
+	if ((status == -EPERM) && !(server->flags & NFS_MOUNT_SECFLAVOUR))
 		status = nfs4_find_root_sec(server, fhandle, info);
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 685a8a7bd17b..75bcc3f0e0b5 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1004,6 +1004,7 @@ static int nfs_parse_security_flavors(char *value,
 		return 0;
 	}
 
+	mnt->flags |= NFS_MOUNT_SECFLAVOUR;
 	mnt->auth_flavor_len = 1;
 	return 1;
 }

From c3dfc2808ab82b13f8b6db62189da959c2eadeea Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 13 Apr 2011 14:31:31 -0400
Subject: [PATCH 10/20] NFS: Use correct variable for page bounds checking

While decoding a secinfo reply, I store the list of supported sec
flavors on a page accessible through res->flavors.  Before reading
each new flavor, I do some math to determine if there is enough
space left on this page, and I break out of my read look if there
isn't.  In order to perform this check correctly, I need to use the
address of res->flavors, rather than the address of res.

When this loop was broken early I lied to the caller and told them
that the entire list had been decoded.  This could lead to problems
if the caller tries to use any the garbage data claiming to be a
valid sec flavor.  I fixed this by using res->flavors->num_flavors
as a counter, incrementing it every time a sec flavor is
successfully decoded.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4xdr.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index dddfb5795d7b..ba952bdc4d62 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -4838,17 +4838,19 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res)
 	struct nfs4_secinfo_flavor *sec_flavor;
 	int status;
 	__be32 *p;
-	int i;
+	int i, num_flavors;
 
 	status = decode_op_hdr(xdr, OP_SECINFO);
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		goto out_overflow;
-	res->flavors->num_flavors = be32_to_cpup(p);
 
-	for (i = 0; i < res->flavors->num_flavors; i++) {
+	res->flavors->num_flavors = 0;
+	num_flavors = be32_to_cpup(p);
+
+	for (i = 0; i < num_flavors; i++) {
 		sec_flavor = &res->flavors->flavors[i];
-		if ((char *)&sec_flavor[1] - (char *)res > PAGE_SIZE)
+		if ((char *)&sec_flavor[1] - (char *)res->flavors > PAGE_SIZE)
 			break;
 
 		p = xdr_inline_decode(xdr, 4);
@@ -4860,6 +4862,7 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res)
 			if (decode_secinfo_gss(xdr, sec_flavor))
 				break;
 		}
+		res->flavors->num_flavors++;
 	}
 
 	return 0;

From e3b2854faabd10438f5e7e34e078b099c3375577 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Fri, 15 Apr 2011 12:58:56 -0400
Subject: [PATCH 11/20] SUNRPC: Fix the SUNRPC Kerberos V RPCSEC_GSS module
 dependencies

Since kernel 2.6.35, the SUNRPC Kerberos support has had an implicit
dependency on a number of additional crypto modules. The following
patch makes that dependency explicit.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 net/sunrpc/Kconfig | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/net/sunrpc/Kconfig b/net/sunrpc/Kconfig
index 8873fd8ddacd..b2198e65d8bb 100644
--- a/net/sunrpc/Kconfig
+++ b/net/sunrpc/Kconfig
@@ -18,14 +18,13 @@ config SUNRPC_XPRT_RDMA
 	  If unsure, say N.
 
 config RPCSEC_GSS_KRB5
-	tristate
+	tristate "Secure RPC: Kerberos V mechanism"
 	depends on SUNRPC && CRYPTO
-	prompt "Secure RPC: Kerberos V mechanism" if !(NFS_V4 || NFSD_V4)
+	depends on CRYPTO_MD5 && CRYPTO_DES && CRYPTO_CBC && CRYPTO_CTS
+	depends on CRYPTO_ECB && CRYPTO_HMAC && CRYPTO_SHA1 && CRYPTO_AES
+	depends on CRYPTO_ARC4
 	default y
 	select SUNRPC_GSS
-	select CRYPTO_MD5
-	select CRYPTO_DES
-	select CRYPTO_CBC
 	help
 	  Choose Y here to enable Secure RPC using the Kerberos version 5
 	  GSS-API mechanism (RFC 1964).

From 47c2199b6eb5fbe38ddb844db7cdbd914d304f9c Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Fri, 15 Apr 2011 17:34:18 -0400
Subject: [PATCH 12/20] NFSv4.1: Ensure state manager thread dies on last
 umount

Currently, the state manager may continue to try recovering state forever
even after the last filesystem to reference that nfs_client has umounted.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org
---
 fs/nfs/nfs4state.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index a6804f704d9d..4dfb34b43ffb 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1607,7 +1607,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
 	int status = 0;
 
 	/* Ensure exclusive access to NFSv4 state */
-	for(;;) {
+	do {
 		if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) {
 			/* We're going to have to re-establish a clientid */
 			status = nfs4_reclaim_lease(clp);
@@ -1691,7 +1691,7 @@ static void nfs4_state_manager(struct nfs_client *clp)
 			break;
 		if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0)
 			break;
-	}
+	} while (atomic_read(&clp->cl_count) > 1);
 	return;
 out_error:
 	printk(KERN_WARNING "Error: state manager failed on NFSv4 server %s"

From 468f86134ee515234afe5c5b3f39f266c50e61a5 Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Mon, 18 Apr 2011 15:57:32 -0400
Subject: [PATCH 13/20] NFSv4.1: Don't update sequence number if rpc_task is
 not sent

If we fail to contact the gss upcall program, then no message will
be sent to the server.  The client still updated the sequence number,
however, and this lead to NFS4ERR_SEQ_MISMATCH for the next several
RPC calls.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c            | 4 ++--
 include/linux/sunrpc/sched.h | 2 ++
 net/sunrpc/xprt.c            | 1 +
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b8e1ac69b743..e7e2077eebd9 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -444,8 +444,8 @@ static int nfs41_sequence_done(struct rpc_task *task, struct nfs4_sequence_res *
 	if (res->sr_status == 1)
 		res->sr_status = NFS_OK;
 
-	/* -ERESTARTSYS can result in skipping nfs41_sequence_setup */
-	if (!res->sr_slot)
+	/* don't increment the sequence number if the task wasn't sent */
+	if (!RPC_WAS_SENT(task))
 		goto out;
 
 	/* Check the SEQUENCE operation status */
diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h
index d81db8012c63..3b94f804b852 100644
--- a/include/linux/sunrpc/sched.h
+++ b/include/linux/sunrpc/sched.h
@@ -127,6 +127,7 @@ struct rpc_task_setup {
 #define RPC_TASK_KILLED		0x0100		/* task was killed */
 #define RPC_TASK_SOFT		0x0200		/* Use soft timeouts */
 #define RPC_TASK_SOFTCONN	0x0400		/* Fail if can't connect */
+#define RPC_TASK_SENT		0x0800		/* message was sent */
 
 #define RPC_IS_ASYNC(t)		((t)->tk_flags & RPC_TASK_ASYNC)
 #define RPC_IS_SWAPPER(t)	((t)->tk_flags & RPC_TASK_SWAPPER)
@@ -134,6 +135,7 @@ struct rpc_task_setup {
 #define RPC_ASSASSINATED(t)	((t)->tk_flags & RPC_TASK_KILLED)
 #define RPC_IS_SOFT(t)		((t)->tk_flags & RPC_TASK_SOFT)
 #define RPC_IS_SOFTCONN(t)	((t)->tk_flags & RPC_TASK_SOFTCONN)
+#define RPC_WAS_SENT(t)		((t)->tk_flags & RPC_TASK_SENT)
 
 #define RPC_TASK_RUNNING	0
 #define RPC_TASK_QUEUED		1
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 9494c3767356..ce5eb68a9664 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -906,6 +906,7 @@ void xprt_transmit(struct rpc_task *task)
 	}
 
 	dprintk("RPC: %5u xmit complete\n", task->tk_pid);
+	task->tk_flags |= RPC_TASK_SENT;
 	spin_lock_bh(&xprt->transport_lock);
 
 	xprt->ops->set_retrans_timeout(task);

From fb8a5ba8114491467c4067ec0330e1c3dcc81d10 Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Mon, 18 Apr 2011 16:52:25 -0400
Subject: [PATCH 14/20] NFSv4: Handle NFS4ERR_WRONGSEC outside of
 nfs4_handle_exception()

I only want to try other secflavors during an initial mount if
NFS4ERR_WRONGSEC is returned.  nfs4_handle_exception() could
potentially map other errors to EPERM, so we should handle this
error specially for correctness.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c | 28 +++++++++++++++++++++++-----
 1 file changed, 23 insertions(+), 5 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index e7e2077eebd9..628e35f75307 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2186,9 +2186,14 @@ static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	struct nfs4_exception exception = { };
 	int err;
 	do {
-		err = nfs4_handle_exception(server,
-				_nfs4_lookup_root(server, fhandle, info),
-				&exception);
+		err = _nfs4_lookup_root(server, fhandle, info);
+		switch (err) {
+		case 0:
+		case -NFS4ERR_WRONGSEC:
+			break;
+		default:
+			err = nfs4_handle_exception(server, err, &exception);
+		}
 	} while (exception.retry);
 	return err;
 }
@@ -2221,10 +2226,19 @@ static int nfs4_find_root_sec(struct nfs_server *server, struct nfs_fh *fhandle,
 
 	for (i = 0; i < len; i++) {
 		status = nfs4_lookup_root_sec(server, fhandle, info, flav_array[i]);
-		if (status == -EPERM || status == -EACCES)
+		if (status == -NFS4ERR_WRONGSEC || status == -EACCES)
 			continue;
 		break;
 	}
+	/*
+	 * -EACCESS could mean that the user doesn't have correct permissions
+	 * to access the mount.  It could also mean that we tried to mount
+	 * with a gss auth flavor, but rpc.gssd isn't running.  Either way,
+	 * existing mount programs don't handle -EACCES very well so it should
+	 * be mapped to -EPERM instead.
+	 */
+	if (status == -EACCES)
+		status = -EPERM;
 	return status;
 }
 
@@ -2235,7 +2249,11 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 			      struct nfs_fsinfo *info)
 {
 	int status = nfs4_lookup_root(server, fhandle, info);
-	if ((status == -EPERM) && !(server->flags & NFS_MOUNT_SECFLAVOUR))
+	if ((status == -NFS4ERR_WRONGSEC) && !(server->flags & NFS_MOUNT_SECFLAVOUR))
+		/*
+		 * A status of -NFS4ERR_WRONGSEC will be mapped to -EPERM
+		 * by nfs4_map_errors() as this function exits.
+		 */
 		status = nfs4_find_root_sec(server, fhandle, info);
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);

From fd954ae124e8a866e9cc1bc3de9a07be5492f608 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Sun, 24 Apr 2011 14:28:18 -0400
Subject: [PATCH 15/20] NFSv4.1: Don't loop forever in nfs4_proc_create_session

If a server for some reason keeps sending NFS4ERR_DELAY errors, we can end
up looping forever inside nfs4_proc_create_session, and so the usual
mechanisms for detecting if the nfs_client is dead don't work.

Fix this by ensuring that we loop inside the nfs4_state_manager thread
instead.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4_fs.h          |  1 +
 fs/nfs/nfs4proc.c         | 45 ++++++--------------------------------
 fs/nfs/nfs4state.c        | 46 ++++++++++++++++++++++++++-------------
 include/linux/nfs_fs_sb.h |  1 +
 4 files changed, 40 insertions(+), 53 deletions(-)

diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index e1c261ddd65d..c4a69833dd0d 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -47,6 +47,7 @@ enum nfs4_client_state {
 	NFS4CLNT_LAYOUTRECALL,
 	NFS4CLNT_SESSION_RESET,
 	NFS4CLNT_RECALL_SLOT,
+	NFS4CLNT_LEASE_CONFIRM,
 };
 
 enum nfs4_session_state {
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 628e35f75307..50c814828fcb 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3754,18 +3754,17 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 		status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
 		if (status != -NFS4ERR_CLID_INUSE)
 			break;
-		if (signalled())
+		if (loop != 0) {
+			++clp->cl_id_uniquifier;
 			break;
-		if (loop++ & 1)
-			ssleep(clp->cl_lease_time / HZ + 1);
-		else
-			if (++clp->cl_id_uniquifier == 0)
-				break;
+		}
+		++loop;
+		ssleep(clp->cl_lease_time / HZ + 1);
 	}
 	return status;
 }
 
-static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp,
+int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
 		struct nfs4_setclientid_res *arg,
 		struct rpc_cred *cred)
 {
@@ -3790,26 +3789,6 @@ static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp,
 	return status;
 }
 
-int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
-		struct nfs4_setclientid_res *arg,
-		struct rpc_cred *cred)
-{
-	long timeout = 0;
-	int err;
-	do {
-		err = _nfs4_proc_setclientid_confirm(clp, arg, cred);
-		switch (err) {
-			case 0:
-				return err;
-			case -NFS4ERR_RESOURCE:
-				/* The IBM lawyers misread another document! */
-			case -NFS4ERR_DELAY:
-				err = nfs4_delay(clp->cl_rpcclient, &timeout);
-		}
-	} while (err == 0);
-	return err;
-}
-
 struct nfs4_delegreturndata {
 	struct nfs4_delegreturnargs args;
 	struct nfs4_delegreturnres res;
@@ -5222,20 +5201,10 @@ int nfs4_proc_create_session(struct nfs_client *clp)
 	int status;
 	unsigned *ptr;
 	struct nfs4_session *session = clp->cl_session;
-	long timeout = 0;
-	int err;
 
 	dprintk("--> %s clp=%p session=%p\n", __func__, clp, session);
 
-	do {
-		status = _nfs4_proc_create_session(clp);
-		if (status == -NFS4ERR_DELAY) {
-			err = nfs4_delay(clp->cl_rpcclient, &timeout);
-			if (err)
-				status = err;
-		}
-	} while (status == -NFS4ERR_DELAY);
-
+	status = _nfs4_proc_create_session(clp);
 	if (status)
 		goto out;
 
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 4dfb34b43ffb..4a810c8b0753 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -64,10 +64,15 @@ static LIST_HEAD(nfs4_clientid_list);
 
 int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
 {
-	struct nfs4_setclientid_res clid;
+	struct nfs4_setclientid_res clid = {
+		.clientid = clp->cl_clientid,
+		.confirm = clp->cl_confirm,
+	};
 	unsigned short port;
 	int status;
 
+	if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state))
+		goto do_confirm;
 	port = nfs_callback_tcpport;
 	if (clp->cl_addr.ss_family == AF_INET6)
 		port = nfs_callback_tcpport6;
@@ -75,10 +80,14 @@ int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
 	status = nfs4_proc_setclientid(clp, NFS4_CALLBACK, port, cred, &clid);
 	if (status != 0)
 		goto out;
+	clp->cl_clientid = clid.clientid;
+	clp->cl_confirm = clid.confirm;
+	set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+do_confirm:
 	status = nfs4_proc_setclientid_confirm(clp, &clid, cred);
 	if (status != 0)
 		goto out;
-	clp->cl_clientid = clid.clientid;
+	clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
 	nfs4_schedule_state_renewal(clp);
 out:
 	return status;
@@ -230,13 +239,18 @@ int nfs41_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)
 {
 	int status;
 
+	if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state))
+		goto do_confirm;
 	nfs4_begin_drain_session(clp);
 	status = nfs4_proc_exchange_id(clp, cred);
 	if (status != 0)
 		goto out;
+	set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+do_confirm:
 	status = nfs4_proc_create_session(clp);
 	if (status != 0)
 		goto out;
+	clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
 	nfs41_setup_state_renewal(clp);
 	nfs_mark_client_ready(clp, NFS_CS_READY);
 out:
@@ -1584,20 +1598,22 @@ static int nfs4_recall_slot(struct nfs_client *clp) { return 0; }
  */
 static void nfs4_set_lease_expired(struct nfs_client *clp, int status)
 {
-	if (nfs4_has_session(clp)) {
-		switch (status) {
-		case -NFS4ERR_DELAY:
-		case -NFS4ERR_CLID_INUSE:
-		case -EAGAIN:
-			break;
+	switch (status) {
+	case -NFS4ERR_CLID_INUSE:
+	case -NFS4ERR_STALE_CLIENTID:
+		clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
+		break;
+	case -NFS4ERR_DELAY:
+	case -EAGAIN:
+		ssleep(1);
+		break;
 
-		case -EKEYEXPIRED:
-			nfs4_warn_keyexpired(clp->cl_hostname);
-		case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
-					 * in nfs4_exchange_id */
-		default:
-			return;
-		}
+	case -EKEYEXPIRED:
+		nfs4_warn_keyexpired(clp->cl_hostname);
+	case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery
+				 * in nfs4_exchange_id */
+	default:
+		return;
 	}
 	set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
 }
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 216cea5db0aa..87694ca86914 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -47,6 +47,7 @@ struct nfs_client {
 
 #ifdef CONFIG_NFS_V4
 	u64			cl_clientid;	/* constant */
+	nfs4_verifier		cl_confirm;	/* Clientid verifier */
 	unsigned long		cl_state;
 
 	spinlock_t		cl_lock;

From 7494d00c7b826b6ceb79ec33892bd0ef59be5614 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Sun, 24 Apr 2011 14:28:45 -0400
Subject: [PATCH 16/20] SUNRPC: Allow RPC calls to return ETIMEDOUT instead of
 EIO

On occasion, it is useful for the NFS layer to distinguish between
soft timeouts and other EIO errors due to (say) encoding errors,
or authentication errors.

The following patch ensures that the default behaviour of the RPC
layer remains to return EIO on soft timeouts (until we have
audited all the callers).

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 include/linux/sunrpc/sched.h | 3 ++-
 net/sunrpc/clnt.c            | 5 ++++-
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h
index 3b94f804b852..f73c482ec9c6 100644
--- a/include/linux/sunrpc/sched.h
+++ b/include/linux/sunrpc/sched.h
@@ -128,12 +128,13 @@ struct rpc_task_setup {
 #define RPC_TASK_SOFT		0x0200		/* Use soft timeouts */
 #define RPC_TASK_SOFTCONN	0x0400		/* Fail if can't connect */
 #define RPC_TASK_SENT		0x0800		/* message was sent */
+#define RPC_TASK_TIMEOUT	0x1000		/* fail with ETIMEDOUT on timeout */
 
 #define RPC_IS_ASYNC(t)		((t)->tk_flags & RPC_TASK_ASYNC)
 #define RPC_IS_SWAPPER(t)	((t)->tk_flags & RPC_TASK_SWAPPER)
 #define RPC_DO_ROOTOVERRIDE(t)	((t)->tk_flags & RPC_TASK_ROOTCREDS)
 #define RPC_ASSASSINATED(t)	((t)->tk_flags & RPC_TASK_KILLED)
-#define RPC_IS_SOFT(t)		((t)->tk_flags & RPC_TASK_SOFT)
+#define RPC_IS_SOFT(t)		((t)->tk_flags & (RPC_TASK_SOFT|RPC_TASK_TIMEOUT))
 #define RPC_IS_SOFTCONN(t)	((t)->tk_flags & RPC_TASK_SOFTCONN)
 #define RPC_WAS_SENT(t)		((t)->tk_flags & RPC_TASK_SENT)
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index e7a96e478f63..8d83f9d48713 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -1508,7 +1508,10 @@ call_timeout(struct rpc_task *task)
 		if (clnt->cl_chatty)
 			printk(KERN_NOTICE "%s: server %s not responding, timed out\n",
 				clnt->cl_protname, clnt->cl_server);
-		rpc_exit(task, -EIO);
+		if (task->tk_flags & RPC_TASK_TIMEOUT)
+			rpc_exit(task, -ETIMEDOUT);
+		else
+			rpc_exit(task, -EIO);
 		return;
 	}
 

From 1bd714f2a14aa4d6a5570956fcec64530b007e4a Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Sun, 24 Apr 2011 14:29:33 -0400
Subject: [PATCH 17/20] NFSv4: Ensure that clientid and session establishment
 can time out

The following patch ensures that we do not get permanently trapped in
the RPC layer when trying to establish a new client id or session.
This again ensures that the state manager can finish in a timely
fashion when the last filesystem to reference the nfs_client exits.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4proc.c  | 13 +++++++------
 fs/nfs/nfs4state.c |  1 +
 2 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 50c814828fcb..69c0f3c5ee7a 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3751,7 +3751,7 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 				sizeof(setclientid.sc_uaddr), "%s.%u.%u",
 				clp->cl_ipaddr, port >> 8, port & 255);
 
-		status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
+		status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 		if (status != -NFS4ERR_CLID_INUSE)
 			break;
 		if (loop != 0) {
@@ -3779,7 +3779,7 @@ int nfs4_proc_setclientid_confirm(struct nfs_client *clp,
 	int status;
 
 	now = jiffies;
-	status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
+	status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 	if (status == 0) {
 		spin_lock(&clp->cl_lock);
 		clp->cl_lease_time = fsinfo.lease_time * HZ;
@@ -4793,7 +4793,7 @@ int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred)
 				init_utsname()->domainname,
 				clp->cl_rpcclient->cl_auth->au_flavor);
 
-	status = rpc_call_sync(clp->cl_rpcclient, &msg, 0);
+	status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 	if (!status)
 		status = nfs4_check_cl_exchange_flags(clp->cl_exchange_flags);
 	dprintk("<-- %s status= %d\n", __func__, status);
@@ -4876,7 +4876,8 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo)
 		.rpc_client = clp->cl_rpcclient,
 		.rpc_message = &msg,
 		.callback_ops = &nfs4_get_lease_time_ops,
-		.callback_data = &data
+		.callback_data = &data,
+		.flags = RPC_TASK_TIMEOUT,
 	};
 	int status;
 
@@ -5178,7 +5179,7 @@ static int _nfs4_proc_create_session(struct nfs_client *clp)
 	nfs4_init_channel_attrs(&args);
 	args.flags = (SESSION4_PERSIST | SESSION4_BACK_CHAN);
 
-	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0);
+	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 
 	if (!status)
 		/* Verify the session's negotiated channel_attrs values */
@@ -5245,7 +5246,7 @@ int nfs4_proc_destroy_session(struct nfs4_session *session)
 	msg.rpc_argp = session;
 	msg.rpc_resp = NULL;
 	msg.rpc_cred = NULL;
-	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0);
+	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 
 	if (status)
 		printk(KERN_WARNING
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 4a810c8b0753..036f5adc9e1f 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1604,6 +1604,7 @@ static void nfs4_set_lease_expired(struct nfs_client *clp, int status)
 		clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state);
 		break;
 	case -NFS4ERR_DELAY:
+	case -ETIMEDOUT:
 	case -EAGAIN:
 		ssleep(1);
 		break;

From 28331a46d88459788c8fca72dbb0415cd7f514c9 Mon Sep 17 00:00:00 2001
From: Trond Myklebust <Trond.Myklebust@netapp.com>
Date: Wed, 27 Apr 2011 13:47:52 -0400
Subject: [PATCH 18/20] NFSv4: Ensure we request the ordinary fileid when doing
 readdirplus

When readdir() returns a directory entry for the root of a mounted
filesystem, Linux follows the old convention of returning the inode
number of the covered directory (despite newer versions of POSIX declaring
that this is a bug).
To ensure this continues to work, the NFSv4 readdir implementation requests
the 'mounted-on-fileid' from the server.

However, readdirplus also needs to instantiate an inode for this entry, and
for that, we also need to request the real fileid as per this patch.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4xdr.c        | 31 ++++++++++++++-----------------
 include/linux/nfs_xdr.h |  2 ++
 2 files changed, 16 insertions(+), 17 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index ba952bdc4d62..7310d2ec5de8 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1452,26 +1452,25 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args,
 
 static void encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req, struct compound_hdr *hdr)
 {
-	uint32_t attrs[2] = {0, 0};
+	uint32_t attrs[2] = {
+		FATTR4_WORD0_RDATTR_ERROR,
+		FATTR4_WORD1_MOUNTED_ON_FILEID,
+	};
 	uint32_t dircount = readdir->count >> 1;
 	__be32 *p;
 
 	if (readdir->plus) {
 		attrs[0] |= FATTR4_WORD0_TYPE|FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE|
-			FATTR4_WORD0_FSID|FATTR4_WORD0_FILEHANDLE;
+			FATTR4_WORD0_FSID|FATTR4_WORD0_FILEHANDLE|FATTR4_WORD0_FILEID;
 		attrs[1] |= FATTR4_WORD1_MODE|FATTR4_WORD1_NUMLINKS|FATTR4_WORD1_OWNER|
 			FATTR4_WORD1_OWNER_GROUP|FATTR4_WORD1_RAWDEV|
 			FATTR4_WORD1_SPACE_USED|FATTR4_WORD1_TIME_ACCESS|
 			FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY;
 		dircount >>= 1;
 	}
-	attrs[0] |= FATTR4_WORD0_RDATTR_ERROR|FATTR4_WORD0_FILEID;
-	attrs[1] |= FATTR4_WORD1_MOUNTED_ON_FILEID;
-	/* Switch to mounted_on_fileid if the server supports it */
-	if (readdir->bitmask[1] & FATTR4_WORD1_MOUNTED_ON_FILEID)
-		attrs[0] &= ~FATTR4_WORD0_FILEID;
-	else
-		attrs[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID;
+	/* Use mounted_on_fileid only if the server supports it */
+	if (!(readdir->bitmask[1] & FATTR4_WORD1_MOUNTED_ON_FILEID))
+		attrs[0] |= FATTR4_WORD0_FILEID;
 
 	p = reserve_space(xdr, 12+NFS4_VERIFIER_SIZE+20);
 	*p++ = cpu_to_be32(OP_READDIR);
@@ -3140,7 +3139,7 @@ static int decode_attr_mounted_on_fileid(struct xdr_stream *xdr, uint32_t *bitma
 			goto out_overflow;
 		xdr_decode_hyper(p, fileid);
 		bitmap[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID;
-		ret = NFS_ATTR_FATTR_FILEID;
+		ret = NFS_ATTR_FATTR_MOUNTED_ON_FILEID;
 	}
 	dprintk("%s: fileid=%Lu\n", __func__, (unsigned long long)*fileid);
 	return ret;
@@ -4002,7 +4001,6 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
 {
 	int status;
 	umode_t fmode = 0;
-	uint64_t fileid;
 	uint32_t type;
 
 	status = decode_attr_type(xdr, bitmap, &type);
@@ -4101,13 +4099,10 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
 		goto xdr_error;
 	fattr->valid |= status;
 
-	status = decode_attr_mounted_on_fileid(xdr, bitmap, &fileid);
+	status = decode_attr_mounted_on_fileid(xdr, bitmap, &fattr->mounted_on_fileid);
 	if (status < 0)
 		goto xdr_error;
-	if (status != 0 && !(fattr->valid & status)) {
-		fattr->fileid = fileid;
-		fattr->valid |= status;
-	}
+	fattr->valid |= status;
 
 xdr_error:
 	dprintk("%s: xdr returned %d\n", __func__, -status);
@@ -6411,7 +6406,9 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry,
 	if (decode_getfattr_attrs(xdr, bitmap, entry->fattr, entry->fh,
 					entry->server, 1) < 0)
 		goto out_overflow;
-	if (entry->fattr->valid & NFS_ATTR_FATTR_FILEID)
+	if (entry->fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID)
+		entry->ino = entry->fattr->mounted_on_fileid;
+	else if (entry->fattr->valid & NFS_ATTR_FATTR_FILEID)
 		entry->ino = entry->fattr->fileid;
 
 	entry->d_type = DT_UNKNOWN;
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 78b101e487ea..890dce242639 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -50,6 +50,7 @@ struct nfs_fattr {
 	} du;
 	struct nfs_fsid		fsid;
 	__u64			fileid;
+	__u64			mounted_on_fileid;
 	struct timespec		atime;
 	struct timespec		mtime;
 	struct timespec		ctime;
@@ -83,6 +84,7 @@ struct nfs_fattr {
 #define NFS_ATTR_FATTR_PRECHANGE	(1U << 18)
 #define NFS_ATTR_FATTR_V4_REFERRAL	(1U << 19)	/* NFSv4 referral */
 #define NFS_ATTR_FATTR_MOUNTPOINT	(1U << 20)	/* Treat as mountpoint */
+#define NFS_ATTR_FATTR_MOUNTED_ON_FILEID		(1U << 21)
 
 #define NFS_ATTR_FATTR (NFS_ATTR_FATTR_TYPE \
 		| NFS_ATTR_FATTR_MODE \

From 613e901e1ee0e1096663b649eee8e5d6697919f3 Mon Sep 17 00:00:00 2001
From: Bryan Schumaker <bjschuma@netapp.com>
Date: Wed, 27 Apr 2011 15:28:44 -0400
Subject: [PATCH 19/20] NFS: Return meaningful status from decode_secinfo()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When compiling, I was getting this warning:
fs/nfs/nfs4xdr.c: In function ‘decode_secinfo’:
fs/nfs/nfs4xdr.c:4839:6: warning: variable ‘status’ set but not used
[-Wunused-but-set-variable]

We were unconditionally returning 0 as long as there wasn't an error
coming out of xdr_inline_decode().  We probably want to check the error
status coming out of decode_op_hdr() and decode_secinfo_gss(), rather
than assuming that everything is OK all the time.

Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/nfs4xdr.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 7310d2ec5de8..c3ccd2c46834 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -4836,6 +4836,8 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res)
 	int i, num_flavors;
 
 	status = decode_op_hdr(xdr, OP_SECINFO);
+	if (status)
+		goto out;
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		goto out_overflow;
@@ -4854,14 +4856,15 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res)
 		sec_flavor->flavor = be32_to_cpup(p);
 
 		if (sec_flavor->flavor == RPC_AUTH_GSS) {
-			if (decode_secinfo_gss(xdr, sec_flavor))
-				break;
+			status = decode_secinfo_gss(xdr, sec_flavor);
+			if (status)
+				goto out;
 		}
 		res->flavors->num_flavors++;
 	}
 
-	return 0;
-
+out:
+	return status;
 out_overflow:
 	print_overflow_msg(__func__, xdr);
 	return -EIO;

From 26c4c170731f00008f4317a2888a0a07ac99d90d Mon Sep 17 00:00:00 2001
From: Jeff Layton <jlayton@redhat.com>
Date: Wed, 27 Apr 2011 11:49:09 -0400
Subject: [PATCH 20/20] nfs: don't lose MS_SYNCHRONOUS on remount of noac mount

On a remount, the VFS layer will clear the MS_SYNCHRONOUS bit on the
assumption that the flags on the mount syscall will have it set if the
remounted fs is supposed to keep it.

In the case of "noac" though, MS_SYNCHRONOUS is implied. A remount of
such a mount will lose the MS_SYNCHRONOUS flag since "sync" isn't part
of the mount options.

Reported-by: Max Matveev <makc@redhat.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---
 fs/nfs/super.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 75bcc3f0e0b5..e288f06d3fa7 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1977,6 +1977,15 @@ nfs_remount(struct super_block *sb, int *flags, char *raw_data)
 	if (error < 0)
 		goto out;
 
+	/*
+	 * noac is a special case. It implies -o sync, but that's not
+	 * necessarily reflected in the mtab options. do_remount_sb
+	 * will clear MS_SYNCHRONOUS if -o sync wasn't specified in the
+	 * remount options, so we have to explicitly reset it.
+	 */
+	if (data->flags & NFS_MOUNT_NOAC)
+		*flags |= MS_SYNCHRONOUS;
+
 	/* compare new mount options with old ones */
 	error = nfs_compare_remount_data(nfss, data);
 out: