[PATCH 00/37] Patches to be added to the 'devel' branch

[PATCH 00/37] Patches to be added to the 'devel' branch

[Trond Myklebust]

The following series of patches will be added to the 'devel' branch
today. There are no major bugfixes or new features. Most of the
patches are just cleanups, and there are a few optimisations.

The only minor new feature to report is Jeff's patch to report an
error if we are unable to accommodate a remount request.

-- 
Trond Myklebust
Linux NFS client maintainer

NetApp
Trond.Myklebust@netapp.com
www.netapp.com

[PATCH 16/37] NFS: Make nfs_fsync methods consistent

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: Report the same debugging info, count function calls the same,
and use similar function naming in nfs_fsync_dir() and nfs_fsync().

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/dir.c  |    3 ++-
 fs/nfs/file.c |   12 +++++++-----
 2 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 58d43da..1bc8bc6 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -629,10 +629,11 @@ out:
  */
 static int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync)
 {
-	dfprintk(VFS, "NFS: fsync_dir(%s/%s) datasync %d\n",
+	dfprintk(VFS, "NFS: fsync dir(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
+	nfs_inc_stats(dentry->d_inode, NFSIOS_VFSFSYNC);
 	return 0;
 }
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 0213c21..1789de2 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -50,7 +50,7 @@ static ssize_t nfs_file_read(struct kiocb *, const struct iovec *iov,
 static ssize_t nfs_file_write(struct kiocb *, const struct iovec *iov,
 				unsigned long nr_segs, loff_t pos);
 static int  nfs_file_flush(struct file *, fl_owner_t id);
-static int  nfs_fsync(struct file *, struct dentry *dentry, int datasync);
+static int  nfs_file_fsync(struct file *, struct dentry *dentry, int datasync);
 static int nfs_check_flags(int flags);
 static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl);
 static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl);
@@ -72,7 +72,7 @@ const struct file_operations nfs_file_operations = {
 	.open		= nfs_file_open,
 	.flush		= nfs_file_flush,
 	.release	= nfs_file_release,
-	.fsync		= nfs_fsync,
+	.fsync		= nfs_file_fsync,
 	.lock		= nfs_lock,
 	.flock		= nfs_flock,
 	.splice_read	= nfs_file_splice_read,
@@ -181,7 +181,7 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 }
 
 /*
- * Helper for nfs_file_flush() and nfs_fsync()
+ * Helper for nfs_file_flush() and nfs_file_fsync()
  *
  * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to
  * disk, but it retrieves and clears ctx->error after synching, despite
@@ -296,12 +296,14 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
  * whether any write errors occurred for this process.
  */
 static int
-nfs_fsync(struct file *file, struct dentry *dentry, int datasync)
+nfs_file_fsync(struct file *file, struct dentry *dentry, int datasync)
 {
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
 	struct inode *inode = dentry->d_inode;
 
-	dfprintk(VFS, "nfs: fsync(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino);
+	dfprintk(VFS, "NFS: fsync file(%s/%s) datasync %d\n",
+			dentry->d_parent->d_name.name, dentry->d_name.name,
+			datasync);
 
 	nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
 	return nfs_do_fsync(ctx, inode);

[PATCH 12/37] SUNRPC: Rename "call_" functions that are no longer FSM states

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

The RPC client uses a finite state machine to move RPC tasks through each
step of an RPC request.  Each state is contained in a function in
net/sunrpc/clnt.c, and named call_foo.

Some of the functions named call_foo have changed over the past few years and
are no longer states in the FSM.  These include: call_encode, call_header,
and call_verify.  As a clean up, rename the functions that have changed.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |   35 ++++++++++++++---------------------
 1 files changed, 14 insertions(+), 21 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 68ea6dd..ab8038d 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -58,7 +58,6 @@ static void	call_start(struct rpc_task *task);
 static void	call_reserve(struct rpc_task *task);
 static void	call_reserveresult(struct rpc_task *task);
 static void	call_allocate(struct rpc_task *task);
-static void	call_encode(struct rpc_task *task);
 static void	call_decode(struct rpc_task *task);
 static void	call_bind(struct rpc_task *task);
 static void	call_bind_status(struct rpc_task *task);
@@ -70,9 +69,9 @@ static void	call_refreshresult(struct rpc_task *task);
 static void	call_timeout(struct rpc_task *task);
 static void	call_connect(struct rpc_task *task);
 static void	call_connect_status(struct rpc_task *task);
-static __be32 *	call_header(struct rpc_task *task);
-static __be32 *	call_verify(struct rpc_task *task);
 
+static __be32	*rpc_encode_header(struct rpc_task *task);
+static __be32	*rpc_verify_header(struct rpc_task *task);
 static int	rpc_ping(struct rpc_clnt *clnt, int flags);
 
 static void rpc_register_client(struct rpc_clnt *clnt)
@@ -876,7 +875,7 @@ rpc_xdr_buf_init(struct xdr_buf *buf, void *start, size_t len)
  * 3.	Encode arguments of an RPC call
  */
 static void
-call_encode(struct rpc_task *task)
+rpc_xdr_encode(struct rpc_task *task)
 {
 	struct rpc_rqst	*req = task->tk_rqstp;
 	kxdrproc_t	encode;
@@ -891,13 +890,14 @@ call_encode(struct rpc_task *task)
 			 (char *)req->rq_buffer + req->rq_callsize,
 			 req->rq_rcvsize);
 
-	/* Encode header and provided arguments */
-	encode = task->tk_msg.rpc_proc->p_encode;
-	if (!(p = call_header(task))) {
-		printk(KERN_INFO "RPC: call_header failed, exit EIO\n");
+	p = rpc_encode_header(task);
+	if (p == NULL) {
+		printk(KERN_INFO "RPC: couldn't encode RPC header, exit EIO\n");
 		rpc_exit(task, -EIO);
 		return;
 	}
+
+	encode = task->tk_msg.rpc_proc->p_encode;
 	if (encode == NULL)
 		return;
 
@@ -1056,7 +1056,7 @@ call_transmit(struct rpc_task *task)
 	/* Encode here so that rpcsec_gss can use correct sequence number. */
 	if (rpc_task_need_encode(task)) {
 		BUG_ON(task->tk_rqstp->rq_bytes_sent != 0);
-		call_encode(task);
+		rpc_xdr_encode(task);
 		/* Did the encode result in an error condition? */
 		if (task->tk_status != 0) {
 			/* Was the error nonfatal? */
@@ -1240,8 +1240,7 @@ call_decode(struct rpc_task *task)
 		goto out_retry;
 	}
 
-	/* Verify the RPC header */
-	p = call_verify(task);
+	p = rpc_verify_header(task);
 	if (IS_ERR(p)) {
 		if (p == ERR_PTR(-EAGAIN))
 			goto out_retry;
@@ -1259,7 +1258,7 @@ call_decode(struct rpc_task *task)
 	return;
 out_retry:
 	task->tk_status = 0;
-	/* Note: call_verify() may have freed the RPC slot */
+	/* Note: rpc_verify_header() may have freed the RPC slot */
 	if (task->tk_rqstp == req) {
 		req->rq_received = req->rq_rcv_buf.len = 0;
 		if (task->tk_client->cl_discrtry)
@@ -1306,11 +1305,8 @@ call_refreshresult(struct rpc_task *task)
 	return;
 }
 
-/*
- * Call header serialization
- */
 static __be32 *
-call_header(struct rpc_task *task)
+rpc_encode_header(struct rpc_task *task)
 {
 	struct rpc_clnt *clnt = task->tk_client;
 	struct rpc_rqst	*req = task->tk_rqstp;
@@ -1330,11 +1326,8 @@ call_header(struct rpc_task *task)
 	return p;
 }
 
-/*
- * Reply header verification
- */
 static __be32 *
-call_verify(struct rpc_task *task)
+rpc_verify_header(struct rpc_task *task)
 {
 	struct kvec *iov = &task->tk_rqstp->rq_rcv_buf.head[0];
 	int len = task->tk_rqstp->rq_rcv_buf.len >> 2;
@@ -1408,7 +1401,7 @@ call_verify(struct rpc_task *task)
 			task->tk_action = call_bind;
 			goto out_retry;
 		case RPC_AUTH_TOOWEAK:
-			printk(KERN_NOTICE "call_verify: server %s requires stronger "
+			printk(KERN_NOTICE "RPC: server %s requires stronger "
 			       "authentication.\n", task->tk_client->cl_server);
 			break;
 		default:

[PATCH 07/37] NFS: Revert commit 44dd151d

[Trond Myklebust]

Revert commit 44dd151d "NFS: Don't mark a written page as uptodate until it
is on disk". While it is true that the write may fail, that is always the
case. There is no reason why we should treat data on pages that are not
already marked as PG_uptodate as being special. The only thing we gain is a
noticeable slowdown when re-reading these pages.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/write.c |   13 ++-----------
 1 files changed, 2 insertions(+), 11 deletions(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index eea2d2b..ee6fcde 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -188,6 +188,7 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page,
 	}
 	/* Update file length */
 	nfs_grow_file(page, offset, count);
+	nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes);
 	nfs_clear_page_tag_locked(req);
 	return 0;
 }
@@ -743,12 +744,7 @@ int nfs_updatepage(struct file *file, struct page *page,
 static void nfs_writepage_release(struct nfs_page *req)
 {
 
-	if (PageError(req->wb_page)) {
-		nfs_end_page_writeback(req->wb_page);
-		nfs_inode_remove_request(req);
-	} else if (!nfs_reschedule_unstable_write(req)) {
-		/* Set the PG_uptodate flag */
-		nfs_mark_uptodate(req->wb_page, req->wb_pgbase, req->wb_bytes);
+	if (PageError(req->wb_page) || !nfs_reschedule_unstable_write(req)) {
 		nfs_end_page_writeback(req->wb_page);
 		nfs_inode_remove_request(req);
 	} else
@@ -1069,8 +1065,6 @@ static void nfs_writeback_release_full(void *calldata)
 			dprintk(" marked for commit\n");
 			goto next;
 		}
-		/* Set the PG_uptodate flag? */
-		nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes);
 		dprintk(" OK\n");
 remove_request:
 		nfs_end_page_writeback(page);
@@ -1309,9 +1303,6 @@ static void nfs_commit_release(void *calldata)
 		 * returned by the server against all stored verfs. */
 		if (!memcmp(req->wb_verf.verifier, data->verf.verifier, sizeof(data->verf.verifier))) {
 			/* We have a match */
-			/* Set the PG_uptodate flag */
-			nfs_mark_uptodate(req->wb_page, req->wb_pgbase,
-					req->wb_bytes);
 			nfs_inode_remove_request(req);
 			dprintk(" OK\n");
 			goto next;

[PATCH 04/37] SUNRPC: Ensure we exit early in case of an encode error

[Trond Myklebust]

All errors from call_encode(), with exception of EAGAIN are fatal, so we
should immediately return instead of proceeding to xprt_transmit().

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 8945307..9503b4c 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -890,7 +890,6 @@ call_encode(struct rpc_task *task)
 			task->tk_msg.rpc_argp);
 	if (task->tk_status == -ENOMEM) {
 		/* XXX: Is this sane? */
-		rpc_delay(task, 3*HZ);
 		task->tk_status = -EAGAIN;
 	}
 }
@@ -1048,8 +1047,14 @@ call_transmit(struct rpc_task *task)
 		BUG_ON(task->tk_rqstp->rq_bytes_sent != 0);
 		call_encode(task);
 		/* Did the encode result in an error condition? */
-		if (task->tk_status != 0)
+		if (task->tk_status != 0) {
+			/* Was the error nonfatal? */
+			if (task->tk_status == -EAGAIN)
+				rpc_delay(task, HZ >> 4);
+			else
+				rpc_exit(task, task->tk_status);
 			return;
+		}
 	}
 	xprt_transmit(task);
 	if (task->tk_status < 0)

[PATCH 11/37] SUNRPC: Add a function to display the name of an RPC procedure

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Improve debugging messages in call_start() and call_verify() by having
them show the RPC procedure name instead of the procedure number.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |   23 +++++++++++++++++++----
 1 files changed, 19 insertions(+), 4 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 1af4f16..68ea6dd 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -690,6 +690,21 @@ rpc_restart_call(struct rpc_task *task)
 }
 EXPORT_SYMBOL_GPL(rpc_restart_call);
 
+#ifdef RPC_DEBUG
+static const char *rpc_proc_name(const struct rpc_task *task)
+{
+	const struct rpc_procinfo *proc = task->tk_msg.rpc_proc;
+
+	if (proc) {
+		if (proc->p_name)
+			return proc->p_name;
+		else
+			return "NULL";
+	} else
+		return "no proc";
+}
+#endif
+
 /*
  * 0.  Initial state
  *
@@ -701,9 +716,9 @@ call_start(struct rpc_task *task)
 {
 	struct rpc_clnt	*clnt = task->tk_client;
 
-	dprintk("RPC: %5u call_start %s%d proc %d (%s)\n", task->tk_pid,
+	dprintk("RPC: %5u call_start %s%d proc %s (%s)\n", task->tk_pid,
 			clnt->cl_protname, clnt->cl_vers,
-			task->tk_msg.rpc_proc->p_proc,
+			rpc_proc_name(task),
 			(RPC_IS_ASYNC(task) ? "async" : "sync"));
 
 	/* Increment call count */
@@ -1432,10 +1447,10 @@ call_verify(struct rpc_task *task)
 		error = -EPROTONOSUPPORT;
 		goto out_err;
 	case RPC_PROC_UNAVAIL:
-		dprintk("RPC: %5u %s: proc %p unsupported by program %u, "
+		dprintk("RPC: %5u %s: proc %s unsupported by program %u, "
 				"version %u on server %s\n",
 				task->tk_pid, __func__,
-				task->tk_msg.rpc_proc,
+				rpc_proc_name(task),
 				task->tk_client->cl_prog,
 				task->tk_client->cl_vers,
 				task->tk_client->cl_server);

[PATCH 15/37] SUNRPC: Display some debugging information as text rather than numbers

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

In rpc_show_tasks(), display the program name, version number, procedure
name and tk_action as human-readable variable-length text fields rather
than columnar numbers.

Doing the symbol lookup here helps in cases where we have actual
debugging output from a kernel log, but don't have access to the kernel
image or RPC module that generated the output.

Sample output:

 -pid- flgs status -client- --rqstp- -timeout ---ops--
  5608 0001    -11 eeb42690 f6d93710        0 f8fa1764 nfsv3 WRITE a:call_transmit_status q:none
  5609 0001    -11 eeb42690 f6d937e0        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5610 0001    -11 eeb42690 f6d93230        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5611 0001    -11 eeb42690 f6d93300        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5612 0001    -11 eeb42690 f6d93090        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5613 0001    -11 eeb42690 f6d933d0        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5614 0001    -11 eeb42690 f6d93cc0        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5615 0001    -11 eeb42690 f6d93a50        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5616 0001    -11 eeb42690 f6d93640        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5617 0001    -11 eeb42690 f6d93b20        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending
  5618 0001    -11 eeb42690 f6d93160        0 f8fa1764 nfsv3 WRITE a:call_status q:xprt_sending

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |   29 ++++++++++++++++-------------
 1 files changed, 16 insertions(+), 13 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 7964a98..0530eea 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -25,6 +25,7 @@
 
 #include <linux/module.h>
 #include <linux/types.h>
+#include <linux/kallsyms.h>
 #include <linux/mm.h>
 #include <linux/slab.h>
 #include <linux/smp_lock.h>
@@ -1528,29 +1529,31 @@ EXPORT_SYMBOL_GPL(rpc_call_null);
 #ifdef RPC_DEBUG
 static void rpc_show_header(void)
 {
-	printk(KERN_INFO "-pid- proc flgs status -client- -prog- --rqstp- "
-		"-timeout -rpcwait -action- ---ops--\n");
+	printk(KERN_INFO "-pid- flgs status -client- --rqstp- "
+		"-timeout ---ops--\n");
 }
 
 static void rpc_show_task(const struct rpc_clnt *clnt,
 			  const struct rpc_task *task)
 {
 	const char *rpc_waitq = "none";
-	int proc = -1;
-
-	if (task->tk_msg.rpc_proc)
-		proc = task->tk_msg.rpc_proc->p_proc;
+	char *p, action[KSYM_SYMBOL_LEN];
 
 	if (RPC_IS_QUEUED(task))
 		rpc_waitq = rpc_qname(task->tk_waitqueue);
 
-	printk(KERN_INFO "%5u %04d %04x %6d %8p %6d %8p %8ld %8s %8p %8p\n",
-		task->tk_pid, proc,
-		task->tk_flags, task->tk_status,
-		clnt, clnt->cl_prog,
-		task->tk_rqstp, task->tk_timeout,
-		rpc_waitq,
-		task->tk_action, task->tk_ops);
+	/* map tk_action pointer to a function name; then trim off
+	 * the "+0x0 [sunrpc]" */
+	sprint_symbol(action, (unsigned long)task->tk_action);
+	p = strchr(action, '+');
+	if (p)
+		*p = '\0';
+
+	printk(KERN_INFO "%5u %04x %6d %8p %8p %8ld %8p %sv%u %s a:%s q:%s\n",
+		task->tk_pid, task->tk_flags, task->tk_status,
+		clnt, task->tk_rqstp, task->tk_timeout, task->tk_ops,
+		clnt->cl_protname, clnt->cl_vers, rpc_proc_name(task),
+		action, rpc_waitq);
 }
 
 void rpc_show_tasks(void)

[PATCH 01/37] NFS: Fix a preemption count leak in nfs_update_request

[Trond Myklebust]

The commit 2785259631697ebb0749a3782cca206e2e542939 (nfs: use GFP_NOFS
preloads for radix-tree insertion) appears to have introduced a bug:
We only want to call radix_tree_preload() once after creating a request.
Calling it every time we loop after we created the request, will cause
preemption count leaks.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Nick Piggin <npiggin@suse.de>
---

 fs/nfs/write.c |   11 ++++-------
 1 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 6d8ace3..c2dd219 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -584,13 +584,6 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx,
 		/* Loop over all inode entries and see if we find
 		 * A request for the page we wish to update
 		 */
-		if (new) {
-			if (radix_tree_preload(GFP_NOFS)) {
-				nfs_release_request(new);
-				return ERR_PTR(-ENOMEM);
-			}
-		}
-
 		spin_lock(&inode->i_lock);
 		req = nfs_page_find_request_locked(page);
 		if (req) {
@@ -630,6 +623,10 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx,
 		new = nfs_create_request(ctx, inode, page, offset, bytes);
 		if (IS_ERR(new))
 			return new;
+		if (radix_tree_preload(GFP_NOFS)) {
+			nfs_release_request(new);
+			return ERR_PTR(-ENOMEM);
+		}
 	}
 
 	/* We have a request for our page.

[PATCH 05/37] SUNRPC: An ENOMEM error from call_encode is always fatal

[Trond Myklebust]

The special 'ENOMEM' case that was previously flagged as non-fatal is
bogus: auth_gss always returns EAGAIN for non-fatal errors, and may in fact
return ENOMEM in the special case where xdr_buf_read_netobj runs out of
preallocated buffer space (invariably a _fatal_ error, since there is no
provision for preallocating larger buffers).

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |    4 ----
 1 files changed, 0 insertions(+), 4 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 9503b4c..1af4f16 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -888,10 +888,6 @@ call_encode(struct rpc_task *task)
 
 	task->tk_status = rpcauth_wrap_req(task, encode, req, p,
 			task->tk_msg.rpc_argp);
-	if (task->tk_status == -ENOMEM) {
-		/* XXX: Is this sane? */
-		task->tk_status = -EAGAIN;
-	}
 }
 
 /*

Re: [PATCH 05/37] SUNRPC: An ENOMEM error from call_encode is always fatal

[J. Bruce Fields]

On Thu, Jun 12, 2008 at 03:22:00PM -0400, Trond Myklebust wrote:
> The special 'ENOMEM' case that was previously flagged as non-fatal is
> bogus: auth_gss always returns EAGAIN for non-fatal errors, and may in fact
> return ENOMEM in the special case where xdr_buf_read_netobj runs out of
> preallocated buffer space (invariably a _fatal_ error, since there is no
> provision for preallocating larger buffers).

For a minute I was worried about the krb5p code that tries to do some
allocations that may fail, but you're right, it takes care to return
EAGAIN in that case (in alloc_enc_pages).  Makes sense to me.

--b.

> 
> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
> ---
> 
>  net/sunrpc/clnt.c |    4 ----
>  1 files changed, 0 insertions(+), 4 deletions(-)
> 
> diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
> index 9503b4c..1af4f16 100644
> --- a/net/sunrpc/clnt.c
> +++ b/net/sunrpc/clnt.c
> @@ -888,10 +888,6 @@ call_encode(struct rpc_task *task)
>  
>  	task->tk_status = rpcauth_wrap_req(task, encode, req, p,
>  			task->tk_msg.rpc_argp);
> -	if (task->tk_status == -ENOMEM) {
> -		/* XXX: Is this sane? */
> -		task->tk_status = -EAGAIN;
> -	}
>  }
>  
>  /*
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 10/37] NFS: Update help text for CONFIG_NFS_FS

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: refresh the help text for Kconfig items related to the NFS
client.  Remove obsolete URLs, and make the language consistent among
the options.

Also move the ROOT_NFS config option next to the options related to the
NFS client.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/Kconfig |  115 ++++++++++++++++++++++++++++++------------------------------
 1 files changed, 57 insertions(+), 58 deletions(-)

diff --git a/fs/Kconfig b/fs/Kconfig
index cf12c40..07a61f5 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -1544,10 +1544,6 @@ config UFS_FS
           The recently released UFS2 variant (used in FreeBSD 5.x) is
           READ-ONLY supported.
 
-	  If you only intend to mount files from some other Unix over the
-	  network using NFS, you don't need the UFS file system support (but
-	  you need NFS file system support obviously).
-
 	  Note that this option is generally not needed for floppies, since a
 	  good portable way to transport files and directories between unixes
 	  (and even other operating systems) is given by the tar program ("man
@@ -1587,6 +1583,7 @@ menuconfig NETWORK_FILESYSTEMS
 	  Say Y here to get to see options for network filesystems and
 	  filesystem-related networking code, such as NFS daemon and
 	  RPCSEC security modules.
+
 	  This option alone does not add any kernel code.
 
 	  If you say N, all options in this submenu will be skipped and
@@ -1595,76 +1592,92 @@ menuconfig NETWORK_FILESYSTEMS
 if NETWORK_FILESYSTEMS
 
 config NFS_FS
-	tristate "NFS file system support"
+	tristate "NFS client support"
 	depends on INET
 	select LOCKD
 	select SUNRPC
 	select NFS_ACL_SUPPORT if NFS_V3_ACL
 	help
-	  If you are connected to some other (usually local) Unix computer
-	  (using SLIP, PLIP, PPP or Ethernet) and want to mount files residing
-	  on that computer (the NFS server) using the Network File Sharing
-	  protocol, say Y. "Mounting files" means that the client can access
-	  the files with usual UNIX commands as if they were sitting on the
-	  client's hard disk. For this to work, the server must run the
-	  programs nfsd and mountd (but does not need to have NFS file system
-	  support enabled in its kernel). NFS is explained in the Network
-	  Administrator's Guide, available from
-	  <http://www.tldp.org/docs.html#guide>, on its man page: "man
-	  nfs", and in the NFS-HOWTO.
-
-	  A superior but less widely used alternative to NFS is provided by
-	  the Coda file system; see "Coda file system support" below.
+	  Choose Y here if you want to access files residing on other
+	  computers using Sun's Network File System protocol.  To compile
+	  this file system support as a module, choose M here: the module
+	  will be called nfs.
 
-	  If you say Y here, you should have said Y to TCP/IP networking also.
-	  This option would enlarge your kernel by about 27 KB.
+	  To mount file systems exported by NFS servers, you also need to
+	  install the user space mount.nfs command which can be found in
+	  the Linux nfs-utils package, available from http://linux-nfs.org/.
+	  Information about using the mount command is available in the
+	  mount(8) man page.  More detail about the Linux NFS client
+	  implementation is available via the nfs(5) man page.
 
-	  To compile this file system support as a module, choose M here: the
-	  module will be called nfs.
+	  Below you can choose which versions of the NFS protocol are
+	  available in the kernel to mount NFS servers.  Support for NFS
+	  version 2 (RFC 1094) is always available when NFS_FS is selected.
 
-	  If you are configuring a diskless machine which will mount its root
-	  file system over NFS at boot time, say Y here and to "Kernel
-	  level IP autoconfiguration" above and to "Root file system on NFS"
-	  below. You cannot compile this driver as a module in this case.
-	  There are two packages designed for booting diskless machines over
-	  the net: netboot, available from
-	  <http://ftp1.sourceforge.net/netboot/>, and Etherboot,
-	  available from <http://ftp1.sourceforge.net/etherboot/>.
+	  To configure a system which mounts its root file system via NFS
+	  at boot time, say Y here, select "Kernel level IP
+	  autoconfiguration" in the NETWORK menu, and select "Root file
+	  system on NFS" below.  You cannot compile this file system as a
+	  module in this case.
 
-	  If you don't know what all this is about, say N.
+	  If unsure, say N.
 
 config NFS_V3
-	bool "Provide NFSv3 client support"
+	bool "NFS client support for NFS version 3"
 	depends on NFS_FS
 	help
-	  Say Y here if you want your NFS client to be able to speak version
-	  3 of the NFS protocol.
+	  This option enables support for version 3 of the NFS protocol
+	  (RFC 1813) in the kernel's NFS client.
 
 	  If unsure, say Y.
 
 config NFS_V3_ACL
-	bool "Provide client support for the NFSv3 ACL protocol extension"
+	bool "NFS client support for the NFSv3 ACL protocol extension"
 	depends on NFS_V3
 	help
-	  Implement the NFSv3 ACL protocol extension for manipulating POSIX
-	  Access Control Lists.  The server should also be compiled with
-	  the NFSv3 ACL protocol extension; see the CONFIG_NFSD_V3_ACL option.
+	  Some NFS servers support an auxiliary NFSv3 ACL protocol that
+	  Sun added to Solaris but never became an official part of the
+	  NFS version 3 protocol.  This protocol extension allows
+	  applications on NFS clients to manipulate POSIX Access Control
+	  Lists on files residing on NFS servers.  NFS servers enforce
+	  ACLs on local files whether this protocol is available or not.
+
+	  Choose Y here if your NFS server supports the Solaris NFSv3 ACL
+	  protocol extension and you want your NFS client to allow
+	  applications to access and modify ACLs on files on the server.
+
+	  Most NFS servers don't support the Solaris NFSv3 ACL protocol
+	  extension.  You can choose N here or specify the "noacl" mount
+	  option to prevent your NFS client from trying to use the NFSv3
+	  ACL protocol.
 
 	  If unsure, say N.
 
 config NFS_V4
-	bool "Provide NFSv4 client support (EXPERIMENTAL)"
+	bool "NFS client support for NFS version 4 (EXPERIMENTAL)"
 	depends on NFS_FS && EXPERIMENTAL
 	select RPCSEC_GSS_KRB5
 	help
-	  Say Y here if you want your NFS client to be able to speak the newer
-	  version 4 of the NFS protocol.
+	  This option enables support for version 4 of the NFS protocol
+	  (RFC 3530) in the kernel's NFS client.
 
-	  Note: Requires auxiliary userspace daemons which may be found on
-		http://www.citi.umich.edu/projects/nfsv4/
+	  To mount NFS servers using NFSv4, you also need to install user
+	  space programs which can be found in the Linux nfs-utils package,
+	  available from http://linux-nfs.org/.
 
 	  If unsure, say N.
 
+config ROOT_NFS
+	bool "Root file system on NFS"
+	depends on NFS_FS=y && IP_PNP
+	help
+	  If you want your system to mount its root file system via NFS,
+	  choose Y here.  This is common practice for managing systems
+	  without local permanent storage.  For details, read
+	  <file:Documentation/filesystems/nfsroot.txt>.
+
+	  Most people say N here.
+
 config NFSD
 	tristate "NFS server support"
 	depends on INET
@@ -1746,20 +1759,6 @@ config NFSD_V4
 
 	  If unsure, say N.
 
-config ROOT_NFS
-	bool "Root file system on NFS"
-	depends on NFS_FS=y && IP_PNP
-	help
-	  If you want your Linux box to mount its whole root file system (the
-	  one containing the directory /) from some other computer over the
-	  net via NFS (presumably because your box doesn't have a hard disk),
-	  say Y. Read <file:Documentation/filesystems/nfsroot.txt> for
-	  details. It is likely that in this case, you also want to say Y to
-	  "Kernel level IP autoconfiguration" so that your box can discover
-	  its network address at boot time.
-
-	  Most people say N here.
-
 config LOCKD
 	tristate
 

[PATCH 03/37] NFS: Add correct bounds checking to NFSv2 locks

[Trond Myklebust]

NFSv2 file locking currently fails the Connectathon tests, because the
calls to the VFS locking code do not return an EINVAL error if the
struct file_lock overflows the 32-bit boundaries.

The problem is due to the fact that we occasionally call helpers from
fs/locks.c in order to avoid RPC calls to the server when we know that a
local process holds the lock. These helpers are, of course, always
64-bit enabled, so EINVAL is not returned in cases when it would if
the call had gone to the NLM code.

For consistency, we therefore add support for a bounds-checking helper.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/file.c           |   20 +++++++++++++++-----
 fs/nfs/proc.c           |   24 ++++++++++++++++++++++++
 include/linux/nfs_xdr.h |    1 +
 3 files changed, 40 insertions(+), 5 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index d84a3d8..7c73f06 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -593,6 +593,7 @@ out:
 static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl)
 {
 	struct inode * inode = filp->f_mapping->host;
+	int ret = -ENOLCK;
 
 	dprintk("NFS: nfs_lock(f=%s/%ld, t=%x, fl=%x, r=%Ld:%Ld)\n",
 			inode->i_sb->s_id, inode->i_ino,
@@ -602,13 +603,22 @@ static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl)
 
 	/* No mandatory locks over NFS */
 	if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
-		return -ENOLCK;
+		goto out_err;
+
+	if (NFS_PROTO(inode)->lock_check_bounds != NULL) {
+		ret = NFS_PROTO(inode)->lock_check_bounds(fl);
+		if (ret < 0)
+			goto out_err;
+	}
 
 	if (IS_GETLK(cmd))
-		return do_getlk(filp, cmd, fl);
-	if (fl->fl_type == F_UNLCK)
-		return do_unlk(filp, cmd, fl);
-	return do_setlk(filp, cmd, fl);
+		ret = do_getlk(filp, cmd, fl);
+	else if (fl->fl_type == F_UNLCK)
+		ret = do_unlk(filp, cmd, fl);
+	else
+		ret = do_setlk(filp, cmd, fl);
+out_err:
+	return ret;
 }
 
 /*
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index 03599bf..5c35b02 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -598,6 +598,29 @@ nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
 	return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl);
 }
 
+/* Helper functions for NFS lock bounds checking */
+#define NFS_LOCK32_OFFSET_MAX ((__s32)0x7fffffffUL)
+static int nfs_lock_check_bounds(const struct file_lock *fl)
+{
+	__s32 start, end;
+
+	start = (__s32)fl->fl_start;
+	if ((loff_t)start != fl->fl_start)
+		goto out_einval;
+
+	if (fl->fl_end != OFFSET_MAX) {
+		end = (__s32)fl->fl_end;
+		if ((loff_t)end != fl->fl_end)
+			goto out_einval;
+	} else
+		end = NFS_LOCK32_OFFSET_MAX;
+
+	if (start < 0 || start > end)
+		goto out_einval;
+	return 0;
+out_einval:
+	return -EINVAL;
+}
 
 const struct nfs_rpc_ops nfs_v2_clientops = {
 	.version	= 2,		       /* protocol version */
@@ -633,4 +656,5 @@ const struct nfs_rpc_ops nfs_v2_clientops = {
 	.file_open	= nfs_open,
 	.file_release	= nfs_release,
 	.lock		= nfs_proc_lock,
+	.lock_check_bounds = nfs_lock_check_bounds,
 };
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 24263bb..8d780de 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -832,6 +832,7 @@ struct nfs_rpc_ops {
 	int	(*file_open)   (struct inode *, struct file *);
 	int	(*file_release) (struct inode *, struct file *);
 	int	(*lock)(struct file *, int, struct file_lock *);
+	int	(*lock_check_bounds)(const struct file_lock *);
 	void	(*clear_acl_cache)(struct inode *);
 };
 

[PATCH 06/37] NFS: Optimise append writes with holes

[Trond Myklebust]

If a file is being extended, and we're creating a hole, we might as well
declare the entire page to be up to date.

This patch significantly improves the write performance for sparse files
in the case where lseek(SEEK_END) is used to append several non-contiguous
writes at intervals of < PAGE_SIZE.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/file.c  |   20 ++++++++++++++++++++
 fs/nfs/write.c |   12 +++---------
 2 files changed, 23 insertions(+), 9 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 7c73f06..7ac89a8 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -344,6 +344,26 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,
 	unsigned offset = pos & (PAGE_CACHE_SIZE - 1);
 	int status;
 
+	/*
+	 * Zero any uninitialised parts of the page, and then mark the page
+	 * as up to date if it turns out that we're extending the file.
+	 */
+	if (!PageUptodate(page)) {
+		unsigned pglen = nfs_page_length(page);
+		unsigned end = offset + len;
+
+		if (pglen == 0) {
+			zero_user_segments(page, 0, offset,
+					end, PAGE_CACHE_SIZE);
+			SetPageUptodate(page);
+		} else if (end >= pglen) {
+			zero_user_segment(page, end, PAGE_CACHE_SIZE);
+			if (offset == 0)
+				SetPageUptodate(page);
+		} else
+			zero_user_segment(page, pglen, PAGE_CACHE_SIZE);
+	}
+
 	lock_kernel();
 	status = nfs_updatepage(file, page, offset, copied);
 	unlock_kernel();
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index dc62bc5..eea2d2b 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -616,7 +616,7 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx,
 			spin_unlock(&inode->i_lock);
 			radix_tree_preload_end();
 			req = new;
-			goto zero_page;
+			goto out;
 		}
 		spin_unlock(&inode->i_lock);
 
@@ -649,19 +649,13 @@ static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx,
 		req->wb_offset = offset;
 		req->wb_pgbase = offset;
 		req->wb_bytes = max(end, rqend) - req->wb_offset;
-		goto zero_page;
+		goto out;
 	}
 
 	if (end > rqend)
 		req->wb_bytes = end - req->wb_offset;
 
-	return req;
-zero_page:
-	/* If this page might potentially be marked as up to date,
-	 * then we need to zero any uninitalised data. */
-	if (req->wb_pgbase == 0 && req->wb_bytes != PAGE_CACHE_SIZE
-			&& !PageUptodate(req->wb_page))
-		zero_user_segment(req->wb_page, req->wb_bytes, PAGE_CACHE_SIZE);
+out:
 	return req;
 }
 

[PATCH 14/37] SUNRPC: Refactor rpc_show_tasks

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: move the logic that displays each task to its own function.
This removes indentation and makes future changes easier.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |   46 ++++++++++++++++++++++++----------------------
 1 files changed, 24 insertions(+), 22 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 7dda328..7964a98 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -1532,40 +1532,42 @@ static void rpc_show_header(void)
 		"-timeout -rpcwait -action- ---ops--\n");
 }
 
+static void rpc_show_task(const struct rpc_clnt *clnt,
+			  const struct rpc_task *task)
+{
+	const char *rpc_waitq = "none";
+	int proc = -1;
+
+	if (task->tk_msg.rpc_proc)
+		proc = task->tk_msg.rpc_proc->p_proc;
+
+	if (RPC_IS_QUEUED(task))
+		rpc_waitq = rpc_qname(task->tk_waitqueue);
+
+	printk(KERN_INFO "%5u %04d %04x %6d %8p %6d %8p %8ld %8s %8p %8p\n",
+		task->tk_pid, proc,
+		task->tk_flags, task->tk_status,
+		clnt, clnt->cl_prog,
+		task->tk_rqstp, task->tk_timeout,
+		rpc_waitq,
+		task->tk_action, task->tk_ops);
+}
+
 void rpc_show_tasks(void)
 {
 	struct rpc_clnt *clnt;
-	struct rpc_task *t;
+	struct rpc_task *task;
 	int header = 0;
 
 	spin_lock(&rpc_client_lock);
 	list_for_each_entry(clnt, &all_clients, cl_clients) {
 		spin_lock(&clnt->cl_lock);
-		list_for_each_entry(t, &clnt->cl_tasks, tk_task) {
-			const char *rpc_waitq = "none";
-			int proc;
-
+		list_for_each_entry(task, &clnt->cl_tasks, tk_task) {
 			if (!header) {
 				rpc_show_header();
 				header++;
 			}
-
-			if (t->tk_msg.rpc_proc)
-				proc = t->tk_msg.rpc_proc->p_proc;
-			else
-				proc = -1;
-
-			if (RPC_IS_QUEUED(t))
-				rpc_waitq = rpc_qname(t->tk_waitqueue);
-
-			printk("%5u %04d %04x %6d %8p %6d %8p %8ld %8s %8p %8p\n",
-				t->tk_pid, proc,
-				t->tk_flags, t->tk_status,
-				t->tk_client,
-				(t->tk_client ? t->tk_client->cl_prog : 0),
-				t->tk_rqstp, t->tk_timeout,
-				rpc_waitq,
-				t->tk_action, t->tk_ops);
+			rpc_show_task(clnt, task);
 		}
 		spin_unlock(&clnt->cl_lock);
 	}

[PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[Trond Myklebust]

Since the credentials may be allocated during the call to rpc_new_task(),
which again may be called by a memory allocator...

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/auth_gss/auth_gss.c        |   10 +++++-----
 net/sunrpc/auth_gss/gss_krb5_mech.c   |    4 ++--
 net/sunrpc/auth_gss/gss_spkm3_mech.c  |    4 ++--
 net/sunrpc/auth_gss/gss_spkm3_token.c |    2 +-
 net/sunrpc/auth_unix.c                |    2 +-
 5 files changed, 11 insertions(+), 11 deletions(-)

diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index cc12d5f..bf7585b 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -146,7 +146,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *dest)
 	q = (const void *)((const char *)p + len);
 	if (unlikely(q > end || q < p))
 		return ERR_PTR(-EFAULT);
-	dest->data = kmemdup(p, len, GFP_KERNEL);
+	dest->data = kmemdup(p, len, GFP_NOFS);
 	if (unlikely(dest->data == NULL))
 		return ERR_PTR(-ENOMEM);
 	dest->len = len;
@@ -171,7 +171,7 @@ gss_alloc_context(void)
 {
 	struct gss_cl_ctx *ctx;
 
-	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	ctx = kzalloc(sizeof(*ctx), GFP_NOFS);
 	if (ctx != NULL) {
 		ctx->gc_proc = RPC_GSS_PROC_DATA;
 		ctx->gc_seq = 1;	/* NetApp 6.4R1 doesn't accept seq. no. 0 */
@@ -341,7 +341,7 @@ gss_alloc_msg(struct gss_auth *gss_auth, uid_t uid)
 {
 	struct gss_upcall_msg *gss_msg;
 
-	gss_msg = kzalloc(sizeof(*gss_msg), GFP_KERNEL);
+	gss_msg = kzalloc(sizeof(*gss_msg), GFP_NOFS);
 	if (gss_msg != NULL) {
 		INIT_LIST_HEAD(&gss_msg->list);
 		rpc_init_wait_queue(&gss_msg->rpc_waitqueue, "RPCSEC_GSS upcall waitq");
@@ -503,7 +503,7 @@ gss_pipe_downcall(struct file *filp, const char __user *src, size_t mlen)
 	if (mlen > MSG_BUF_MAXSIZE)
 		goto out;
 	err = -ENOMEM;
-	buf = kmalloc(mlen, GFP_KERNEL);
+	buf = kmalloc(mlen, GFP_NOFS);
 	if (!buf)
 		goto out;
 
@@ -806,7 +806,7 @@ gss_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags)
 	dprintk("RPC:       gss_create_cred for uid %d, flavor %d\n",
 		acred->uid, auth->au_flavor);
 
-	if (!(cred = kzalloc(sizeof(*cred), GFP_KERNEL)))
+	if (!(cred = kzalloc(sizeof(*cred), GFP_NOFS)))
 		goto out_err;
 
 	rpcauth_init_cred(&cred->gc_base, acred, auth, &gss_credops);
diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c
index 60c3dba..ef45eba 100644
--- a/net/sunrpc/auth_gss/gss_krb5_mech.c
+++ b/net/sunrpc/auth_gss/gss_krb5_mech.c
@@ -70,7 +70,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
 	q = (const void *)((const char *)p + len);
 	if (unlikely(q > end || q < p))
 		return ERR_PTR(-EFAULT);
-	res->data = kmemdup(p, len, GFP_KERNEL);
+	res->data = kmemdup(p, len, GFP_NOFS);
 	if (unlikely(res->data == NULL))
 		return ERR_PTR(-ENOMEM);
 	res->len = len;
@@ -131,7 +131,7 @@ gss_import_sec_context_kerberos(const void *p,
 	struct	krb5_ctx *ctx;
 	int tmp;
 
-	if (!(ctx = kzalloc(sizeof(*ctx), GFP_KERNEL)))
+	if (!(ctx = kzalloc(sizeof(*ctx), GFP_NOFS)))
 		goto out_err;
 
 	p = simple_get_bytes(p, end, &ctx->initiate, sizeof(ctx->initiate));
diff --git a/net/sunrpc/auth_gss/gss_spkm3_mech.c b/net/sunrpc/auth_gss/gss_spkm3_mech.c
index 5deb4b6..035e1dd 100644
--- a/net/sunrpc/auth_gss/gss_spkm3_mech.c
+++ b/net/sunrpc/auth_gss/gss_spkm3_mech.c
@@ -76,7 +76,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
 	q = (const void *)((const char *)p + len);
 	if (unlikely(q > end || q < p))
 		return ERR_PTR(-EFAULT);
-	res->data = kmemdup(p, len, GFP_KERNEL);
+	res->data = kmemdup(p, len, GFP_NOFS);
 	if (unlikely(res->data == NULL))
 		return ERR_PTR(-ENOMEM);
 	return q;
@@ -90,7 +90,7 @@ gss_import_sec_context_spkm3(const void *p, size_t len,
 	struct	spkm3_ctx *ctx;
 	int	version;
 
-	if (!(ctx = kzalloc(sizeof(*ctx), GFP_KERNEL)))
+	if (!(ctx = kzalloc(sizeof(*ctx), GFP_NOFS)))
 		goto out_err;
 
 	p = simple_get_bytes(p, end, &version, sizeof(version));
diff --git a/net/sunrpc/auth_gss/gss_spkm3_token.c b/net/sunrpc/auth_gss/gss_spkm3_token.c
index 6cdd241..3308157 100644
--- a/net/sunrpc/auth_gss/gss_spkm3_token.c
+++ b/net/sunrpc/auth_gss/gss_spkm3_token.c
@@ -90,7 +90,7 @@ asn1_bitstring_len(struct xdr_netobj *in, int *enclen, int *zerobits)
 int
 decode_asn1_bitstring(struct xdr_netobj *out, char *in, int enclen, int explen)
 {
-	if (!(out->data = kzalloc(explen,GFP_KERNEL)))
+	if (!(out->data = kzalloc(explen,GFP_NOFS)))
 		return 0;
 	out->len = explen;
 	memcpy(out->data, in, enclen);
diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c
index 44920b9..46b2647 100644
--- a/net/sunrpc/auth_unix.c
+++ b/net/sunrpc/auth_unix.c
@@ -66,7 +66,7 @@ unx_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags)
 	dprintk("RPC:       allocating UNIX cred for uid %d gid %d\n",
 			acred->uid, acred->gid);
 
-	if (!(cred = kmalloc(sizeof(*cred), GFP_KERNEL)))
+	if (!(cred = kmalloc(sizeof(*cred), GFP_NOFS)))
 		return ERR_PTR(-ENOMEM);
 
 	rpcauth_init_cred(&cred->uc_base, acred, auth, &unix_credops);

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[J. Bruce Fields]

On Thu, Jun 12, 2008 at 03:22:00PM -0400, Trond Myklebust wrote:
> Since the credentials may be allocated during the call to rpc_new_task(),
> which again may be called by a memory allocator...

Most of these are only called in gssd  process's context when it
performs the downcall.  Which doesn't change the fact that they're
needed for the rpc_new_task() to succeed, OK, so this isn't an objection
to this patch--but is there any plan for how we'll deal with e.g. memory
allocations that gssd does on its own?

--b.

> 
> Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
> ---
> 
>  net/sunrpc/auth_gss/auth_gss.c        |   10 +++++-----
>  net/sunrpc/auth_gss/gss_krb5_mech.c   |    4 ++--
>  net/sunrpc/auth_gss/gss_spkm3_mech.c  |    4 ++--
>  net/sunrpc/auth_gss/gss_spkm3_token.c |    2 +-
>  net/sunrpc/auth_unix.c                |    2 +-
>  5 files changed, 11 insertions(+), 11 deletions(-)
> 
> diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
> index cc12d5f..bf7585b 100644
> --- a/net/sunrpc/auth_gss/auth_gss.c
> +++ b/net/sunrpc/auth_gss/auth_gss.c
> @@ -146,7 +146,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *dest)
>  	q = (const void *)((const char *)p + len);
>  	if (unlikely(q > end || q < p))
>  		return ERR_PTR(-EFAULT);
> -	dest->data = kmemdup(p, len, GFP_KERNEL);
> +	dest->data = kmemdup(p, len, GFP_NOFS);
>  	if (unlikely(dest->data == NULL))
>  		return ERR_PTR(-ENOMEM);
>  	dest->len = len;
> @@ -171,7 +171,7 @@ gss_alloc_context(void)
>  {
>  	struct gss_cl_ctx *ctx;
>  
> -	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
> +	ctx = kzalloc(sizeof(*ctx), GFP_NOFS);
>  	if (ctx != NULL) {
>  		ctx->gc_proc = RPC_GSS_PROC_DATA;
>  		ctx->gc_seq = 1;	/* NetApp 6.4R1 doesn't accept seq. no. 0 */
> @@ -341,7 +341,7 @@ gss_alloc_msg(struct gss_auth *gss_auth, uid_t uid)
>  {
>  	struct gss_upcall_msg *gss_msg;
>  
> -	gss_msg = kzalloc(sizeof(*gss_msg), GFP_KERNEL);
> +	gss_msg = kzalloc(sizeof(*gss_msg), GFP_NOFS);
>  	if (gss_msg != NULL) {
>  		INIT_LIST_HEAD(&gss_msg->list);
>  		rpc_init_wait_queue(&gss_msg->rpc_waitqueue, "RPCSEC_GSS upcall waitq");
> @@ -503,7 +503,7 @@ gss_pipe_downcall(struct file *filp, const char __user *src, size_t mlen)
>  	if (mlen > MSG_BUF_MAXSIZE)
>  		goto out;
>  	err = -ENOMEM;
> -	buf = kmalloc(mlen, GFP_KERNEL);
> +	buf = kmalloc(mlen, GFP_NOFS);
>  	if (!buf)
>  		goto out;
>  
> @@ -806,7 +806,7 @@ gss_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags)
>  	dprintk("RPC:       gss_create_cred for uid %d, flavor %d\n",
>  		acred->uid, auth->au_flavor);
>  
> -	if (!(cred = kzalloc(sizeof(*cred), GFP_KERNEL)))
> +	if (!(cred = kzalloc(sizeof(*cred), GFP_NOFS)))
>  		goto out_err;
>  
>  	rpcauth_init_cred(&cred->gc_base, acred, auth, &gss_credops);
> diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c
> index 60c3dba..ef45eba 100644
> --- a/net/sunrpc/auth_gss/gss_krb5_mech.c
> +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c
> @@ -70,7 +70,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
>  	q = (const void *)((const char *)p + len);
>  	if (unlikely(q > end || q < p))
>  		return ERR_PTR(-EFAULT);
> -	res->data = kmemdup(p, len, GFP_KERNEL);
> +	res->data = kmemdup(p, len, GFP_NOFS);
>  	if (unlikely(res->data == NULL))
>  		return ERR_PTR(-ENOMEM);
>  	res->len = len;
> @@ -131,7 +131,7 @@ gss_import_sec_context_kerberos(const void *p,
>  	struct	krb5_ctx *ctx;
>  	int tmp;
>  
> -	if (!(ctx = kzalloc(sizeof(*ctx), GFP_KERNEL)))
> +	if (!(ctx = kzalloc(sizeof(*ctx), GFP_NOFS)))
>  		goto out_err;
>  
>  	p = simple_get_bytes(p, end, &ctx->initiate, sizeof(ctx->initiate));
> diff --git a/net/sunrpc/auth_gss/gss_spkm3_mech.c b/net/sunrpc/auth_gss/gss_spkm3_mech.c
> index 5deb4b6..035e1dd 100644
> --- a/net/sunrpc/auth_gss/gss_spkm3_mech.c
> +++ b/net/sunrpc/auth_gss/gss_spkm3_mech.c
> @@ -76,7 +76,7 @@ simple_get_netobj(const void *p, const void *end, struct xdr_netobj *res)
>  	q = (const void *)((const char *)p + len);
>  	if (unlikely(q > end || q < p))
>  		return ERR_PTR(-EFAULT);
> -	res->data = kmemdup(p, len, GFP_KERNEL);
> +	res->data = kmemdup(p, len, GFP_NOFS);
>  	if (unlikely(res->data == NULL))
>  		return ERR_PTR(-ENOMEM);
>  	return q;
> @@ -90,7 +90,7 @@ gss_import_sec_context_spkm3(const void *p, size_t len,
>  	struct	spkm3_ctx *ctx;
>  	int	version;
>  
> -	if (!(ctx = kzalloc(sizeof(*ctx), GFP_KERNEL)))
> +	if (!(ctx = kzalloc(sizeof(*ctx), GFP_NOFS)))
>  		goto out_err;
>  
>  	p = simple_get_bytes(p, end, &version, sizeof(version));
> diff --git a/net/sunrpc/auth_gss/gss_spkm3_token.c b/net/sunrpc/auth_gss/gss_spkm3_token.c
> index 6cdd241..3308157 100644
> --- a/net/sunrpc/auth_gss/gss_spkm3_token.c
> +++ b/net/sunrpc/auth_gss/gss_spkm3_token.c
> @@ -90,7 +90,7 @@ asn1_bitstring_len(struct xdr_netobj *in, int *enclen, int *zerobits)
>  int
>  decode_asn1_bitstring(struct xdr_netobj *out, char *in, int enclen, int explen)
>  {
> -	if (!(out->data = kzalloc(explen,GFP_KERNEL)))
> +	if (!(out->data = kzalloc(explen,GFP_NOFS)))
>  		return 0;
>  	out->len = explen;
>  	memcpy(out->data, in, enclen);
> diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c
> index 44920b9..46b2647 100644
> --- a/net/sunrpc/auth_unix.c
> +++ b/net/sunrpc/auth_unix.c
> @@ -66,7 +66,7 @@ unx_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags)
>  	dprintk("RPC:       allocating UNIX cred for uid %d gid %d\n",
>  			acred->uid, acred->gid);
>  
> -	if (!(cred = kmalloc(sizeof(*cred), GFP_KERNEL)))
> +	if (!(cred = kmalloc(sizeof(*cred), GFP_NOFS)))
>  		return ERR_PTR(-ENOMEM);
>  
>  	rpcauth_init_cred(&cred->uc_base, acred, auth, &unix_credops);
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[Trond Myklebust]

On Fri, 2008-06-13 at 17:17 -0400, J. Bruce Fields wrote:
> On Thu, Jun 12, 2008 at 03:22:00PM -0400, Trond Myklebust wrote:
> > Since the credentials may be allocated during the call to rpc_new_task(),
> > which again may be called by a memory allocator...
> 
> Most of these are only called in gssd  process's context when it
> performs the downcall.  Which doesn't change the fact that they're
> needed for the rpc_new_task() to succeed, OK, so this isn't an objection
> to this patch--but is there any plan for how we'll deal with e.g. memory
> allocations that gssd does on its own?

You mean alloc_enc_pages()? We should really try to avoid sleeping
there...

-- 
Trond Myklebust
Linux NFS client maintainer

NetApp
Trond.Myklebust@netapp.com
www.netapp.com

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[J. Bruce Fields]

On Fri, Jun 13, 2008 at 05:26:17PM -0400, Trond Myklebust wrote:
> On Fri, 2008-06-13 at 17:17 -0400, J. Bruce Fields wrote:
> > On Thu, Jun 12, 2008 at 03:22:00PM -0400, Trond Myklebust wrote:
> > > Since the credentials may be allocated during the call to rpc_new_task(),
> > > which again may be called by a memory allocator...
> > 
> > Most of these are only called in gssd  process's context when it
> > performs the downcall.  Which doesn't change the fact that they're
> > needed for the rpc_new_task() to succeed, OK, so this isn't an objection
> > to this patch--but is there any plan for how we'll deal with e.g. memory
> > allocations that gssd does on its own?
> 
> You mean alloc_enc_pages()? We should really try to avoid sleeping
> there...

No, that may be a problem too, but I mean: if we're worried about a
deadlock due to some allocations that gssd performs on the downcall at
the end of the context initiation, then shouldn't we also be worried
about all the other allocations that gssd performs itself?

--b.

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[Trond Myklebust]

On Fri, 2008-06-13 at 17:31 -0400, J. Bruce Fields wrote:
> No, that may be a problem too, but I mean: if we're worried about a
> deadlock due to some allocations that gssd performs on the downcall at
> the end of the context initiation, then shouldn't we also be worried
> about all the other allocations that gssd performs itself?

Definitely.

-- 
Trond Myklebust
Linux NFS client maintainer

NetApp
Trond.Myklebust@netapp.com
www.netapp.com

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[J. Bruce Fields]

On Fri, Jun 13, 2008 at 05:34:54PM -0400, Trond Myklebust wrote:
> On Fri, 2008-06-13 at 17:31 -0400, J. Bruce Fields wrote:
> > No, that may be a problem too, but I mean: if we're worried about a
> > deadlock due to some allocations that gssd performs on the downcall at
> > the end of the context initiation, then shouldn't we also be worried
> > about all the other allocations that gssd performs itself?
> 
> Definitely.

Do you have any ideas (however speculative) about how we'll deal with
that?

--b.

Re: [PATCH 08/37] SUNRPC: Use GFP_NOFS when allocating credentials

[Trond Myklebust]

On Fri, 2008-06-13 at 17:58 -0400, J. Bruce Fields wrote:
> On Fri, Jun 13, 2008 at 05:34:54PM -0400, Trond Myklebust wrote:
> > On Fri, 2008-06-13 at 17:31 -0400, J. Bruce Fields wrote:
> > > No, that may be a problem too, but I mean: if we're worried about a
> > > deadlock due to some allocations that gssd performs on the downcall at
> > > the end of the context initiation, then shouldn't we also be worried
> > > about all the other allocations that gssd performs itself?
> > 
> > Definitely.
> 
> Do you have any ideas (however speculative) about how we'll deal with
> that?

Not really: AFAIK there are no mechanisms for fully pre-provisioning the
necessary resources that a userland app may need to ensure it can
function. mlock() won't, for instance, allow you to pre-provision the
socket buffer space that the kerberos stuff needs to communicate...

-- 
Trond Myklebust
Linux NFS client maintainer

NetApp
Trond.Myklebust@netapp.com
www.netapp.com

[PATCH 02/37] NFS: nfs_updatepage(): don't mark page as dirty if an error occurred

[Trond Myklebust]

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/write.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index c2dd219..dc62bc5 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -736,12 +736,13 @@ int nfs_updatepage(struct file *file, struct page *page,
 	}
 
 	status = nfs_writepage_setup(ctx, page, offset, count);
-	__set_page_dirty_nobuffers(page);
+	if (status < 0)
+		nfs_set_pageerror(page);
+	else
+		__set_page_dirty_nobuffers(page);
 
         dprintk("NFS:      nfs_updatepage returns %d (isize %Ld)\n",
 			status, (long long)i_size_read(inode));
-	if (status < 0)
-		nfs_set_pageerror(page);
 	return status;
 }
 

[PATCH 13/37] SUNRPC: Don't display the rpc_show_tasks header if there are no tasks

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: don't display the rpc_show_tasks column header unless there is at
least one task to display.  As far as I can tell, it is safe to let the
list_for_each_entry macro decide that each list is empty.

scripts/checkpatch.pl also wants a KERN_FOO at the start of any newly added
printk() calls, so this and subsequent patches will also add KERN_INFO.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/clnt.c |   19 ++++++++++++-------
 1 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index ab8038d..7dda328 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -1526,24 +1526,30 @@ struct rpc_task *rpc_call_null(struct rpc_clnt *clnt, struct rpc_cred *cred, int
 EXPORT_SYMBOL_GPL(rpc_call_null);
 
 #ifdef RPC_DEBUG
+static void rpc_show_header(void)
+{
+	printk(KERN_INFO "-pid- proc flgs status -client- -prog- --rqstp- "
+		"-timeout -rpcwait -action- ---ops--\n");
+}
+
 void rpc_show_tasks(void)
 {
 	struct rpc_clnt *clnt;
 	struct rpc_task *t;
+	int header = 0;
 
 	spin_lock(&rpc_client_lock);
-	if (list_empty(&all_clients))
-		goto out;
-	printk("-pid- proc flgs status -client- -prog- --rqstp- -timeout "
-		"-rpcwait -action- ---ops--\n");
 	list_for_each_entry(clnt, &all_clients, cl_clients) {
-		if (list_empty(&clnt->cl_tasks))
-			continue;
 		spin_lock(&clnt->cl_lock);
 		list_for_each_entry(t, &clnt->cl_tasks, tk_task) {
 			const char *rpc_waitq = "none";
 			int proc;
 
+			if (!header) {
+				rpc_show_header();
+				header++;
+			}
+
 			if (t->tk_msg.rpc_proc)
 				proc = t->tk_msg.rpc_proc->p_proc;
 			else
@@ -1563,7 +1569,6 @@ void rpc_show_tasks(void)
 		}
 		spin_unlock(&clnt->cl_lock);
 	}
-out:
 	spin_unlock(&rpc_client_lock);
 }
 #endif

[PATCH 17/37] NFS: Make nfs_llseek methods consistent

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: Report the same debugging info in nfs_llseek_dir() and
nfs_llseek_file().

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/dir.c  |   12 ++++++++++--
 fs/nfs/file.c |    5 +++++
 2 files changed, 15 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 1bc8bc6..9f31cc6 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -603,7 +603,15 @@ out:
 
 static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin)
 {
-	mutex_lock(&filp->f_path.dentry->d_inode->i_mutex);
+	struct dentry *dentry = filp->f_path.dentry;
+	struct inode *inode = dentry->d_inode;
+
+	dfprintk(VFS, "NFS: llseek dir(%s/%s, %lld, %d)\n",
+			dentry->d_parent->d_name.name,
+			dentry->d_name.name,
+			offset, origin);
+
+	mutex_lock(&inode->i_mutex);
 	switch (origin) {
 		case 1:
 			offset += filp->f_pos;
@@ -619,7 +627,7 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin)
 		nfs_file_open_context(filp)->dir_cookie = 0;
 	}
 out:
-	mutex_unlock(&filp->f_path.dentry->d_inode->i_mutex);
+	mutex_unlock(&inode->i_mutex);
 	return offset;
 }
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 1789de2..cef3636 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -170,6 +170,11 @@ force_reval:
 
 static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 {
+	dfprintk(VFS, "NFS: llseek file(%s/%s, %lld, %d)\n",
+			filp->f_path.dentry->d_parent->d_name.name,
+			filp->f_path.dentry->d_name.name,
+			offset, origin);
+
 	/* origin == SEEK_END => we must revalidate the cached file length */
 	if (origin == SEEK_END) {
 		struct inode *inode = filp->f_mapping->host;

[PATCH 09/37] NFS: do_setlk(): don't flush caches when we have a delegation

[Trond Myklebust]

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/file.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 7ac89a8..0213c21 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -602,7 +602,8 @@ static int do_setlk(struct file *filp, int cmd, struct file_lock *fl)
 	 * This makes locking act as a cache coherency point.
 	 */
 	nfs_sync_mapping(filp->f_mapping);
-	nfs_zap_caches(inode);
+	if (!nfs_have_delegation(inode, FMODE_READ))
+		nfs_zap_caches(inode);
 out:
 	return status;
 }

[PATCH 26/37] rpc: eliminate unused variable in auth_gss upcall code

[Trond Myklebust]

From: \"J. Bruce Fields\" <bfields@citi.umich.edu>

Also, a minor comment grammar fix in the same file.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/auth_gss/auth_gss.c |    4 +---
 1 files changed, 1 insertions(+), 3 deletions(-)

diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index bf7585b..ebfd630 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -272,7 +272,7 @@ __gss_find_upcall(struct rpc_inode *rpci, uid_t uid)
 	return NULL;
 }
 
-/* Try to add a upcall to the pipefs queue.
+/* Try to add an upcall to the pipefs queue.
  * If an upcall owned by our uid already exists, then we return a reference
  * to that upcall instead of adding the new upcall.
  */
@@ -493,7 +493,6 @@ gss_pipe_downcall(struct file *filp, const char __user *src, size_t mlen)
 {
 	const void *p, *end;
 	void *buf;
-	struct rpc_clnt *clnt;
 	struct gss_upcall_msg *gss_msg;
 	struct inode *inode = filp->f_path.dentry->d_inode;
 	struct gss_cl_ctx *ctx;
@@ -507,7 +506,6 @@ gss_pipe_downcall(struct file *filp, const char __user *src, size_t mlen)
 	if (!buf)
 		goto out;
 
-	clnt = RPC_I(inode)->private;
 	err = -EFAULT;
 	if (copy_from_user(buf, src, mlen))
 		goto err;

[PATCH 22/37] SUNRPC: Remove obsolete messages during transport connect

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Recent changes to the RPC client's transport connect logic make connect
status values ECONNREFUSED and ECONNRESET impossible.

Clean up xprt_connect_status() to account for these changes.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/xprt.c |    8 +-------
 1 files changed, 1 insertions(+), 7 deletions(-)

diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index e1770f7..67996bd 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -690,7 +690,7 @@ static void xprt_connect_status(struct rpc_task *task)
 {
 	struct rpc_xprt	*xprt = task->tk_xprt;
 
-	if (task->tk_status >= 0) {
+	if (task->tk_status == 0) {
 		xprt->stat.connect_count++;
 		xprt->stat.connect_time += (long)jiffies - xprt->stat.connect_start;
 		dprintk("RPC: %5u xprt_connect_status: connection established\n",
@@ -699,12 +699,6 @@ static void xprt_connect_status(struct rpc_task *task)
 	}
 
 	switch (task->tk_status) {
-	case -ECONNREFUSED:
-	case -ECONNRESET:
-		dprintk("RPC: %5u xprt_connect_status: server %s refused "
-				"connection\n", task->tk_pid,
-				task->tk_client->cl_server);
-		break;
 	case -ENOTCONN:
 		dprintk("RPC: %5u xprt_connect_status: connection broken\n",
 				task->tk_pid);

[PATCH 23/37] fs/nfs/nfsroot.c: remove CVS keyword

[Trond Myklebust]

From: Adrian Bunk <bunk@kernel.org>

This patch removes a CVS keyword that wasn't updated for a long time
from a comment.

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/nfsroot.c |    2 --
 1 files changed, 0 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c
index 531379d..15afe3a 100644
--- a/fs/nfs/nfsroot.c
+++ b/fs/nfs/nfsroot.c
@@ -1,6 +1,4 @@
 /*
- *  $Id: nfsroot.c,v 1.45 1998/03/07 10:44:46 mj Exp $
- *
  *  Copyright (C) 1995, 1996  Gero Kuhlmann <gero-TuicA7gkpRym5h6znurzUg@public.gmane.org>
  *
  *  Allow an NFS filesystem to be mounted as root. The way this works is:

[PATCH 28/37] rpc: minor cleanup of scheduler callback code

[Trond Myklebust]

From: \"J. Bruce Fields\" <bfields@citi.umich.edu>

Try to make the comment here a little more clear and concise.

Also, this macro definition seems unnecessary.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 include/linux/sunrpc/sched.h |    1 -
 net/sunrpc/sched.c           |   14 +++++---------
 2 files changed, 5 insertions(+), 10 deletions(-)

diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h
index d1a5c8c..64981a2 100644
--- a/include/linux/sunrpc/sched.h
+++ b/include/linux/sunrpc/sched.h
@@ -135,7 +135,6 @@ struct rpc_task_setup {
 #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_DO_CALLBACK(t)	((t)->tk_callback != NULL)
 #define RPC_IS_SOFT(t)		((t)->tk_flags & RPC_TASK_SOFT)
 
 #define RPC_TASK_RUNNING	0
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 6eab9bf..6288af0 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -626,19 +626,15 @@ static void __rpc_execute(struct rpc_task *task)
 		/*
 		 * Execute any pending callback.
 		 */
-		if (RPC_DO_CALLBACK(task)) {
-			/* Define a callback save pointer */
+		if (task->tk_callback) {
 			void (*save_callback)(struct rpc_task *);
 
 			/*
-			 * If a callback exists, save it, reset it,
-			 * call it.
-			 * The save is needed to stop from resetting
-			 * another callback set within the callback handler
-			 * - Dave
+			 * We set tk_callback to NULL before calling it,
+			 * in case it sets the tk_callback field itself:
 			 */
-			save_callback=task->tk_callback;
-			task->tk_callback=NULL;
+			save_callback = task->tk_callback;
+			task->tk_callback = NULL;
 			save_callback(task);
 		}
 

[PATCH 36/37] NFS: Treat "intr" and "nointr" options as deprecated

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up:  the "intr" and "nointr" mount options were recently retired.
Document this in the NFS mount option parser.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/super.c |   10 ++++------
 1 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index dea138b..09e563b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -65,7 +65,6 @@
 enum {
 	/* Mount options that take no arguments */
 	Opt_soft, Opt_hard,
-	Opt_intr, Opt_nointr,
 	Opt_posix, Opt_noposix,
 	Opt_cto, Opt_nocto,
 	Opt_ac, Opt_noac,
@@ -105,8 +104,8 @@ static match_table_t nfs_mount_option_tokens = {
 
 	{ Opt_soft, "soft" },
 	{ Opt_hard, "hard" },
-	{ Opt_intr, "intr" },
-	{ Opt_nointr, "nointr" },
+	{ Opt_deprecated, "intr" },
+	{ Opt_deprecated, "nointr" },
 	{ Opt_posix, "posix" },
 	{ Opt_noposix, "noposix" },
 	{ Opt_cto, "cto" },
@@ -787,9 +786,6 @@ static int nfs_parse_mount_options(char *raw,
 		case Opt_hard:
 			mnt->flags &= ~NFS_MOUNT_SOFT;
 			break;
-		case Opt_intr:
-		case Opt_nointr:
-			break;
 		case Opt_posix:
 			mnt->flags |= NFS_MOUNT_POSIX;
 			break;
@@ -1105,6 +1101,8 @@ static int nfs_parse_mount_options(char *raw,
 
 		case Opt_userspace:
 		case Opt_deprecated:
+			dfprintk(MOUNT, "NFS:   ignoring mount option "
+					"'%s'\n", p);
 			break;
 
 		default:

[PATCH 37/37] NFS: missing newline in NFS mount debugging message

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/super.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 09e563b..226ae6a 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1190,7 +1190,7 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args,
 	if (status == 0)
 		return 0;
 
-	dfprintk(MOUNT, "NFS: unable to mount server %s, error %d",
+	dfprintk(MOUNT, "NFS: unable to mount server %s, error %d\n",
 			hostname, status);
 	return status;
 }

[PATCH 32/37] NFS: Move fs/nfs/iostat.h to include/linux

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

The fs/nfs/iostat.h header has definitions that were designed to be exposed
to user space.  Move these definitions under include/linux so user space can
use the definitions in applications that read /proc/self/mountstats.

Also address a handful of coding style issues called out by checkpatch.pl in
fs/nfs/iostat.h.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/iostat.h            |  119 +++++---------------------------------------
 include/linux/nfs_iostat.h |  119 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 132 insertions(+), 106 deletions(-)

diff --git a/fs/nfs/iostat.h b/fs/nfs/iostat.h
index 6350ecb..2ec65e1 100644
--- a/fs/nfs/iostat.h
+++ b/fs/nfs/iostat.h
@@ -5,135 +5,41 @@
  *
  *  Copyright (C) 2005, 2006 Chuck Lever <cel@netapp.com>
  *
- *  NFS client per-mount statistics provide information about the health of
- *  the NFS client and the health of each NFS mount point.  Generally these
- *  are not for detailed problem diagnosis, but simply to indicate that there
- *  is a problem.
- *
- *  These counters are not meant to be human-readable, but are meant to be
- *  integrated into system monitoring tools such as "sar" and "iostat".  As
- *  such, the counters are sampled by the tools over time, and are never
- *  zeroed after a file system is mounted.  Moving averages can be computed
- *  by the tools by taking the difference between two instantaneous samples
- *  and dividing that by the time between the samples.
  */
 
 #ifndef _NFS_IOSTAT
 #define _NFS_IOSTAT
 
-#define NFS_IOSTAT_VERS		"1.0"
-
-/*
- * NFS byte counters
- *
- * 1.  SERVER - the number of payload bytes read from or written to the
- *     server by the NFS client via an NFS READ or WRITE request.
- *
- * 2.  NORMAL - the number of bytes read or written by applications via
- *     the read(2) and write(2) system call interfaces.
- *
- * 3.  DIRECT - the number of bytes read or written from files opened
- *     with the O_DIRECT flag.
- *
- * These counters give a view of the data throughput into and out of the NFS
- * client.  Comparing the number of bytes requested by an application with the
- * number of bytes the client requests from the server can provide an
- * indication of client efficiency (per-op, cache hits, etc).
- *
- * These counters can also help characterize which access methods are in
- * use.  DIRECT by itself shows whether there is any O_DIRECT traffic.
- * NORMAL + DIRECT shows how much data is going through the system call
- * interface.  A large amount of SERVER traffic without much NORMAL or
- * DIRECT traffic shows that applications are using mapped files.
- *
- * NFS page counters
- *
- * These count the number of pages read or written via nfs_readpage(),
- * nfs_readpages(), or their write equivalents.
- */
-enum nfs_stat_bytecounters {
-	NFSIOS_NORMALREADBYTES = 0,
-	NFSIOS_NORMALWRITTENBYTES,
-	NFSIOS_DIRECTREADBYTES,
-	NFSIOS_DIRECTWRITTENBYTES,
-	NFSIOS_SERVERREADBYTES,
-	NFSIOS_SERVERWRITTENBYTES,
-	NFSIOS_READPAGES,
-	NFSIOS_WRITEPAGES,
-	__NFSIOS_BYTESMAX,
-};
-
-/*
- * NFS event counters
- *
- * These counters provide a low-overhead way of monitoring client activity
- * without enabling NFS trace debugging.  The counters show the rate at
- * which VFS requests are made, and how often the client invalidates its
- * data and attribute caches.  This allows system administrators to monitor
- * such things as how close-to-open is working, and answer questions such
- * as "why are there so many GETATTR requests on the wire?"
- *
- * They also count anamolous events such as short reads and writes, silly
- * renames due to close-after-delete, and operations that change the size
- * of a file (such operations can often be the source of data corruption
- * if applications aren't using file locking properly).
- */
-enum nfs_stat_eventcounters {
-	NFSIOS_INODEREVALIDATE = 0,
-	NFSIOS_DENTRYREVALIDATE,
-	NFSIOS_DATAINVALIDATE,
-	NFSIOS_ATTRINVALIDATE,
-	NFSIOS_VFSOPEN,
-	NFSIOS_VFSLOOKUP,
-	NFSIOS_VFSACCESS,
-	NFSIOS_VFSUPDATEPAGE,
-	NFSIOS_VFSREADPAGE,
-	NFSIOS_VFSREADPAGES,
-	NFSIOS_VFSWRITEPAGE,
-	NFSIOS_VFSWRITEPAGES,
-	NFSIOS_VFSGETDENTS,
-	NFSIOS_VFSSETATTR,
-	NFSIOS_VFSFLUSH,
-	NFSIOS_VFSFSYNC,
-	NFSIOS_VFSLOCK,
-	NFSIOS_VFSRELEASE,
-	NFSIOS_CONGESTIONWAIT,
-	NFSIOS_SETATTRTRUNC,
-	NFSIOS_EXTENDWRITE,
-	NFSIOS_SILLYRENAME,
-	NFSIOS_SHORTREAD,
-	NFSIOS_SHORTWRITE,
-	NFSIOS_DELAY,
-	__NFSIOS_COUNTSMAX,
-};
-
-#ifdef __KERNEL__
-
 #include <linux/percpu.h>
 #include <linux/cache.h>
+#include <linux/nfs_iostat.h>
 
 struct nfs_iostats {
 	unsigned long long	bytes[__NFSIOS_BYTESMAX];
 	unsigned long		events[__NFSIOS_COUNTSMAX];
 } ____cacheline_aligned;
 
-static inline void nfs_inc_server_stats(struct nfs_server *server, enum nfs_stat_eventcounters stat)
+static inline void nfs_inc_server_stats(struct nfs_server *server,
+					enum nfs_stat_eventcounters stat)
 {
 	struct nfs_iostats *iostats;
 	int cpu;
 
 	cpu = get_cpu();
 	iostats = per_cpu_ptr(server->io_stats, cpu);
-	iostats->events[stat] ++;
+	iostats->events[stat]++;
 	put_cpu_no_resched();
 }
 
-static inline void nfs_inc_stats(struct inode *inode, enum nfs_stat_eventcounters stat)
+static inline void nfs_inc_stats(struct inode *inode,
+				 enum nfs_stat_eventcounters stat)
 {
 	nfs_inc_server_stats(NFS_SERVER(inode), stat);
 }
 
-static inline void nfs_add_server_stats(struct nfs_server *server, enum nfs_stat_bytecounters stat, unsigned long addend)
+static inline void nfs_add_server_stats(struct nfs_server *server,
+					enum nfs_stat_bytecounters stat,
+					unsigned long addend)
 {
 	struct nfs_iostats *iostats;
 	int cpu;
@@ -144,7 +50,9 @@ static inline void nfs_add_server_stats(struct nfs_server *server, enum nfs_stat
 	put_cpu_no_resched();
 }
 
-static inline void nfs_add_stats(struct inode *inode, enum nfs_stat_bytecounters stat, unsigned long addend)
+static inline void nfs_add_stats(struct inode *inode,
+				 enum nfs_stat_bytecounters stat,
+				 unsigned long addend)
 {
 	nfs_add_server_stats(NFS_SERVER(inode), stat, addend);
 }
@@ -160,5 +68,4 @@ static inline void nfs_free_iostats(struct nfs_iostats *stats)
 		free_percpu(stats);
 }
 
-#endif
-#endif
+#endif /* _NFS_IOSTAT */
diff --git a/include/linux/nfs_iostat.h b/include/linux/nfs_iostat.h
new file mode 100644
index 0000000..1cb9a3f
--- /dev/null
+++ b/include/linux/nfs_iostat.h
@@ -0,0 +1,119 @@
+/*
+ *  User-space visible declarations for NFS client per-mount
+ *  point statistics
+ *
+ *  Copyright (C) 2005, 2006 Chuck Lever <cel@netapp.com>
+ *
+ *  NFS client per-mount statistics provide information about the
+ *  health of the NFS client and the health of each NFS mount point.
+ *  Generally these are not for detailed problem diagnosis, but
+ *  simply to indicate that there is a problem.
+ *
+ *  These counters are not meant to be human-readable, but are meant
+ *  to be integrated into system monitoring tools such as "sar" and
+ *  "iostat".  As such, the counters are sampled by the tools over
+ *  time, and are never zeroed after a file system is mounted.
+ *  Moving averages can be computed by the tools by taking the
+ *  difference between two instantaneous samples  and dividing that
+ *  by the time between the samples.
+ */
+
+#ifndef _LINUX_NFS_IOSTAT
+#define _LINUX_NFS_IOSTAT
+
+#define NFS_IOSTAT_VERS		"1.0"
+
+/*
+ * NFS byte counters
+ *
+ * 1.  SERVER - the number of payload bytes read from or written
+ *     to the server by the NFS client via an NFS READ or WRITE
+ *     request.
+ *
+ * 2.  NORMAL - the number of bytes read or written by applications
+ *     via the read(2) and write(2) system call interfaces.
+ *
+ * 3.  DIRECT - the number of bytes read or written from files
+ *     opened with the O_DIRECT flag.
+ *
+ * These counters give a view of the data throughput into and out
+ * of the NFS client.  Comparing the number of bytes requested by
+ * an application with the number of bytes the client requests from
+ * the server can provide an indication of client efficiency
+ * (per-op, cache hits, etc).
+ *
+ * These counters can also help characterize which access methods
+ * are in use.  DIRECT by itself shows whether there is any O_DIRECT
+ * traffic.  NORMAL + DIRECT shows how much data is going through
+ * the system call interface.  A large amount of SERVER traffic
+ * without much NORMAL or DIRECT traffic shows that applications
+ * are using mapped files.
+ *
+ * NFS page counters
+ *
+ * These count the number of pages read or written via nfs_readpage(),
+ * nfs_readpages(), or their write equivalents.
+ *
+ * NB: When adding new byte counters, please include the measured
+ * units in the name of each byte counter to help users of this
+ * interface determine what exactly is being counted.
+ */
+enum nfs_stat_bytecounters {
+	NFSIOS_NORMALREADBYTES = 0,
+	NFSIOS_NORMALWRITTENBYTES,
+	NFSIOS_DIRECTREADBYTES,
+	NFSIOS_DIRECTWRITTENBYTES,
+	NFSIOS_SERVERREADBYTES,
+	NFSIOS_SERVERWRITTENBYTES,
+	NFSIOS_READPAGES,
+	NFSIOS_WRITEPAGES,
+	__NFSIOS_BYTESMAX,
+};
+
+/*
+ * NFS event counters
+ *
+ * These counters provide a low-overhead way of monitoring client
+ * activity without enabling NFS trace debugging.  The counters
+ * show the rate at which VFS requests are made, and how often the
+ * client invalidates its data and attribute caches.  This allows
+ * system administrators to monitor such things as how close-to-open
+ * is working, and answer questions such as "why are there so many
+ * GETATTR requests on the wire?"
+ *
+ * They also count anamolous events such as short reads and writes,
+ * silly renames due to close-after-delete, and operations that
+ * change the size of a file (such operations can often be the
+ * source of data corruption if applications aren't using file
+ * locking properly).
+ */
+enum nfs_stat_eventcounters {
+	NFSIOS_INODEREVALIDATE = 0,
+	NFSIOS_DENTRYREVALIDATE,
+	NFSIOS_DATAINVALIDATE,
+	NFSIOS_ATTRINVALIDATE,
+	NFSIOS_VFSOPEN,
+	NFSIOS_VFSLOOKUP,
+	NFSIOS_VFSACCESS,
+	NFSIOS_VFSUPDATEPAGE,
+	NFSIOS_VFSREADPAGE,
+	NFSIOS_VFSREADPAGES,
+	NFSIOS_VFSWRITEPAGE,
+	NFSIOS_VFSWRITEPAGES,
+	NFSIOS_VFSGETDENTS,
+	NFSIOS_VFSSETATTR,
+	NFSIOS_VFSFLUSH,
+	NFSIOS_VFSFSYNC,
+	NFSIOS_VFSLOCK,
+	NFSIOS_VFSRELEASE,
+	NFSIOS_CONGESTIONWAIT,
+	NFSIOS_SETATTRTRUNC,
+	NFSIOS_EXTENDWRITE,
+	NFSIOS_SILLYRENAME,
+	NFSIOS_SHORTREAD,
+	NFSIOS_SHORTWRITE,
+	NFSIOS_DELAY,
+	__NFSIOS_COUNTSMAX,
+};
+
+#endif	/* _LINUX_NFS_IOSTAT */

[PATCH 29/37] NFS: Fix the ftruncate() credential problem

[Trond Myklebust]

ftruncate() access checking is supposed to be performed at open() time,
just like reads and writes.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/inode.c    |    4 ++--
 fs/nfs/nfs3proc.c |    2 ++
 fs/nfs/nfs4proc.c |   47 +++++++++++++++++++++++------------------------
 fs/nfs/proc.c     |    2 ++
 4 files changed, 29 insertions(+), 26 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 596c5d8..2e4ab4a 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -347,7 +347,7 @@ out_no_inode:
 	goto out;
 }
 
-#define NFS_VALID_ATTRS (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_SIZE|ATTR_ATIME|ATTR_ATIME_SET|ATTR_MTIME|ATTR_MTIME_SET)
+#define NFS_VALID_ATTRS (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_SIZE|ATTR_ATIME|ATTR_ATIME_SET|ATTR_MTIME|ATTR_MTIME_SET|ATTR_FILE)
 
 int
 nfs_setattr(struct dentry *dentry, struct iattr *attr)
@@ -369,7 +369,7 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	/* Optimization: if the end result is no change, don't RPC */
 	attr->ia_valid &= NFS_VALID_ATTRS;
-	if (attr->ia_valid == 0)
+	if ((attr->ia_valid & ~ATTR_FILE) == 0)
 		return 0;
 
 	lock_kernel();
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index c3523ad..b14b378 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -129,6 +129,8 @@ nfs3_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
 	int	status;
 
 	dprintk("NFS call  setattr\n");
+	if (sattr->ia_valid & ATTR_FILE)
+		msg.rpc_cred = nfs_file_cred(sattr->ia_file);
 	nfs_fattr_init(fattr);
 	status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
 	if (status == 0)
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 1293e0a..15084a0 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1139,8 +1139,9 @@ static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, int
 	return res;
 }
 
-static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
-                struct iattr *sattr, struct nfs4_state *state)
+static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
+			    struct nfs_fattr *fattr, struct iattr *sattr,
+			    struct nfs4_state *state)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
         struct nfs_setattrargs  arg = {
@@ -1154,9 +1155,10 @@ static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
 		.server		= server,
         };
         struct rpc_message msg = {
-                .rpc_proc       = &nfs4_procedures[NFSPROC4_CLNT_SETATTR],
-                .rpc_argp       = &arg,
-                .rpc_resp       = &res,
+		.rpc_proc	= &nfs4_procedures[NFSPROC4_CLNT_SETATTR],
+		.rpc_argp	= &arg,
+		.rpc_resp	= &res,
+		.rpc_cred	= cred,
         };
 	unsigned long timestamp = jiffies;
 	int status;
@@ -1166,7 +1168,6 @@ static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
 	if (nfs4_copy_delegation_stateid(&arg.stateid, inode)) {
 		/* Use that stateid */
 	} else if (state != NULL) {
-		msg.rpc_cred = state->owner->so_cred;
 		nfs4_copy_stateid(&arg.stateid, state, current->files);
 	} else
 		memcpy(&arg.stateid, &zero_stateid, sizeof(arg.stateid));
@@ -1177,15 +1178,16 @@ static int _nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
 	return status;
 }
 
-static int nfs4_do_setattr(struct inode *inode, struct nfs_fattr *fattr,
-                struct iattr *sattr, struct nfs4_state *state)
+static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
+			   struct nfs_fattr *fattr, struct iattr *sattr,
+			   struct nfs4_state *state)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs4_exception exception = { };
 	int err;
 	do {
 		err = nfs4_handle_exception(server,
-				_nfs4_do_setattr(inode, fattr, sattr, state),
+				_nfs4_do_setattr(inode, cred, fattr, sattr, state),
 				&exception);
 	} while (exception.retry);
 	return err;
@@ -1647,29 +1649,25 @@ static int
 nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
 		  struct iattr *sattr)
 {
-	struct rpc_cred *cred;
 	struct inode *inode = dentry->d_inode;
-	struct nfs_open_context *ctx;
+	struct rpc_cred *cred = NULL;
 	struct nfs4_state *state = NULL;
 	int status;
 
 	nfs_fattr_init(fattr);
 	
-	cred = rpc_lookup_cred();
-	if (IS_ERR(cred))
-		return PTR_ERR(cred);
-
 	/* Search for an existing open(O_WRITE) file */
-	ctx = nfs_find_open_context(inode, cred, FMODE_WRITE);
-	if (ctx != NULL)
+	if (sattr->ia_valid & ATTR_FILE) {
+		struct nfs_open_context *ctx;
+
+		ctx = nfs_file_open_context(sattr->ia_file);
+		cred = ctx->cred;
 		state = ctx->state;
+	}
 
-	status = nfs4_do_setattr(inode, fattr, sattr, state);
+	status = nfs4_do_setattr(inode, cred, fattr, sattr, state);
 	if (status == 0)
 		nfs_setattr_update_inode(inode, sattr);
-	if (ctx != NULL)
-		put_nfs_open_context(ctx);
-	put_rpccred(cred);
 	return status;
 }
 
@@ -1897,17 +1895,16 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 		goto out;
 	}
 	state = nfs4_do_open(dir, &path, flags, sattr, cred);
-	put_rpccred(cred);
 	d_drop(dentry);
 	if (IS_ERR(state)) {
 		status = PTR_ERR(state);
-		goto out;
+		goto out_putcred;
 	}
 	d_add(dentry, igrab(state->inode));
 	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 	if (flags & O_EXCL) {
 		struct nfs_fattr fattr;
-		status = nfs4_do_setattr(state->inode, &fattr, sattr, state);
+		status = nfs4_do_setattr(state->inode, cred, &fattr, sattr, state);
 		if (status == 0)
 			nfs_setattr_update_inode(state->inode, sattr);
 		nfs_post_op_update_inode(state->inode, &fattr);
@@ -1916,6 +1913,8 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 		status = nfs4_intent_set_file(nd, &path, state);
 	else
 		nfs4_close_sync(&path, state, flags);
+out_putcred:
+	put_rpccred(cred);
 out:
 	return status;
 }
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index 5c35b02..c760558 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -129,6 +129,8 @@ nfs_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
 	sattr->ia_mode &= S_IALLUGO;
 
 	dprintk("NFS call  setattr\n");
+	if (sattr->ia_valid & ATTR_FILE)
+		msg.rpc_cred = nfs_file_cred(sattr->ia_file);
 	nfs_fattr_init(fattr);
 	status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
 	if (status == 0)

[PATCH 19/37] NFS: Add debugging facility for NFS aops

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Recent work in fs/nfs/file.c neglected to add appropriate trace debugging
for the NFS client's address space operations.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/file.c |   29 +++++++++++++++++++++++++++--
 1 files changed, 27 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 202408d..6c447a3 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -335,6 +335,11 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,
 	struct page *page;
 	index = pos >> PAGE_CACHE_SHIFT;
 
+	dfprintk(PAGECACHE, "NFS: write_begin(%s/%s(%ld), %u@%lld)\n",
+		file->f_path.dentry->d_parent->d_name.name,
+		file->f_path.dentry->d_name.name,
+		mapping->host->i_ino, len, (long long) pos);
+
 	page = __grab_cache_page(mapping, index);
 	if (!page)
 		return -ENOMEM;
@@ -355,6 +360,11 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,
 	unsigned offset = pos & (PAGE_CACHE_SIZE - 1);
 	int status;
 
+	dfprintk(PAGECACHE, "NFS: write_end(%s/%s(%ld), %u@%lld)\n",
+		file->f_path.dentry->d_parent->d_name.name,
+		file->f_path.dentry->d_name.name,
+		mapping->host->i_ino, len, (long long) pos);
+
 	/*
 	 * Zero any uninitialised parts of the page, and then mark the page
 	 * as up to date if it turns out that we're extending the file.
@@ -389,6 +399,8 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,
 
 static void nfs_invalidate_page(struct page *page, unsigned long offset)
 {
+	dfprintk(PAGECACHE, "NFS: invalidate_page(%p, %lu)\n", page, offset);
+
 	if (offset != 0)
 		return;
 	/* Cancel any unstarted writes on this page */
@@ -397,13 +409,20 @@ static void nfs_invalidate_page(struct page *page, unsigned long offset)
 
 static int nfs_release_page(struct page *page, gfp_t gfp)
 {
+	dfprintk(PAGECACHE, "NFS: release_page(%p)\n", page);
+
 	/* If PagePrivate() is set, then the page is not freeable */
 	return 0;
 }
 
 static int nfs_launder_page(struct page *page)
 {
-	return nfs_wb_page(page->mapping->host, page);
+	struct inode *inode = page->mapping->host;
+
+	dfprintk(PAGECACHE, "NFS: launder_page(%ld, %llu)\n",
+		inode->i_ino, (long long)page_offset(page));
+
+	return nfs_wb_page(inode, page);
 }
 
 const struct address_space_operations nfs_file_aops = {
@@ -423,13 +442,19 @@ const struct address_space_operations nfs_file_aops = {
 static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct page *page)
 {
 	struct file *filp = vma->vm_file;
+	struct dentry *dentry = filp->f_path.dentry;
 	unsigned pagelen;
 	int ret = -EINVAL;
 	struct address_space *mapping;
 
+	dfprintk(PAGECACHE, "NFS: vm_page_mkwrite(%s/%s(%ld), offset %lld)\n",
+		dentry->d_parent->d_name.name, dentry->d_name.name,
+		filp->f_mapping->host->i_ino,
+		(long long)page_offset(page));
+
 	lock_page(page);
 	mapping = page->mapping;
-	if (mapping != vma->vm_file->f_path.dentry->d_inode->i_mapping)
+	if (mapping != dentry->d_inode->i_mapping)
 		goto out_unlock;
 
 	ret = 0;

[PATCH 34/37] NFS: Ensure we zap only the access and acl caches when setting new acls

[Trond Myklebust]

...and ensure that we obey the NFS_INO_INVALID_ACL flag when retrieving the
acls.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/inode.c    |    4 +---
 fs/nfs/internal.h |    1 +
 fs/nfs/nfs3acl.c  |    9 ++++++---
 fs/nfs/nfs4proc.c |    5 ++++-
 4 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 2e4ab4a..2c23d06 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -57,8 +57,6 @@ static int enable_ino64 = NFS_64_BIT_INODE_NUMBERS_ENABLED;
 static void nfs_invalidate_inode(struct inode *);
 static int nfs_update_inode(struct inode *, struct nfs_fattr *);
 
-static void nfs_zap_acl_cache(struct inode *);
-
 static struct kmem_cache * nfs_inode_cachep;
 
 static inline unsigned long
@@ -167,7 +165,7 @@ void nfs_zap_mapping(struct inode *inode, struct address_space *mapping)
 	}
 }
 
-static void nfs_zap_acl_cache(struct inode *inode)
+void nfs_zap_acl_cache(struct inode *inode)
 {
 	void (*clear_acl_cache)(struct inode *);
 
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 04ae867..24241fc 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -150,6 +150,7 @@ extern void nfs_clear_inode(struct inode *);
 #ifdef CONFIG_NFS_V4
 extern void nfs4_clear_inode(struct inode *);
 #endif
+void nfs_zap_acl_cache(struct inode *inode);
 
 /* super.c */
 extern struct file_system_type nfs_xdev_fs_type;
diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c
index 9b73625..423842f 100644
--- a/fs/nfs/nfs3acl.c
+++ b/fs/nfs/nfs3acl.c
@@ -5,6 +5,8 @@
 #include <linux/posix_acl_xattr.h>
 #include <linux/nfsacl.h>
 
+#include "internal.h"
+
 #define NFSDBG_FACILITY	NFSDBG_PROC
 
 ssize_t nfs3_listxattr(struct dentry *dentry, char *buffer, size_t size)
@@ -205,6 +207,8 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type)
 	status = nfs_revalidate_inode(server, inode);
 	if (status < 0)
 		return ERR_PTR(status);
+	if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL)
+		nfs_zap_acl_cache(inode);
 	acl = nfs3_get_cached_acl(inode, type);
 	if (acl != ERR_PTR(-EAGAIN))
 		return acl;
@@ -319,9 +323,8 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl,
 	dprintk("NFS call setacl\n");
 	msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_SETACL];
 	status = rpc_call_sync(server->client_acl, &msg, 0);
-	spin_lock(&inode->i_lock);
-	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS;
-	spin_unlock(&inode->i_lock);
+	nfs_access_zap_cache(inode);
+	nfs_zap_acl_cache(inode);
 	dprintk("NFS reply setacl: %d\n", status);
 
 	/* pages may have been allocated at the xdr layer. */
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 0fc1068..66f13dc 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2705,6 +2705,8 @@ static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen)
 	ret = nfs_revalidate_inode(server, inode);
 	if (ret < 0)
 		return ret;
+	if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL)
+		nfs_zap_acl_cache(inode);
 	ret = nfs4_read_cached_acl(inode, buf, buflen);
 	if (ret != -ENOENT)
 		return ret;
@@ -2732,7 +2734,8 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
 	nfs_inode_return_delegation(inode);
 	buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase);
 	ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
-	nfs_zap_caches(inode);
+	nfs_access_zap_cache(inode);
+	nfs_zap_acl_cache(inode);
 	return ret;
 }
 

[PATCH 31/37] NFS: Remove the redundant file_open entry from struct nfs_rpc_ops

[Trond Myklebust]

All instances are set to nfs_open(), so we should just remove the redundant
indirection. Ditto for the file_release op

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/file.c           |    4 ++--
 fs/nfs/nfs3proc.c       |    2 --
 fs/nfs/nfs4proc.c       |    2 --
 fs/nfs/proc.c           |    2 --
 include/linux/nfs_xdr.h |    2 --
 5 files changed, 2 insertions(+), 10 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 78b26f8..509dcb5 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -129,7 +129,7 @@ nfs_file_open(struct inode *inode, struct file *filp)
 
 	nfs_inc_stats(inode, NFSIOS_VFSOPEN);
 	lock_kernel();
-	res = NFS_PROTO(inode)->file_open(inode, filp);
+	res = nfs_open(inode, filp);
 	unlock_kernel();
 	return res;
 }
@@ -147,7 +147,7 @@ nfs_file_release(struct inode *inode, struct file *filp)
 	if (filp->f_mode & FMODE_WRITE)
 		nfs_wb_all(dentry->d_inode);
 	nfs_inc_stats(inode, NFSIOS_VFSRELEASE);
-	return NFS_PROTO(inode)->file_release(inode, filp);
+	return nfs_release(inode, filp);
 }
 
 /**
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index b14b378..5c5fc8a 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -803,8 +803,6 @@ const struct nfs_rpc_ops nfs_v3_clientops = {
 	.write_done	= nfs3_write_done,
 	.commit_setup	= nfs3_proc_commit_setup,
 	.commit_done	= nfs3_commit_done,
-	.file_open	= nfs_open,
-	.file_release	= nfs_release,
 	.lock		= nfs3_proc_lock,
 	.clear_acl_cache = nfs3_forget_cached_acls,
 };
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 15084a0..75cdec4 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3724,8 +3724,6 @@ const struct nfs_rpc_ops nfs_v4_clientops = {
 	.write_done	= nfs4_write_done,
 	.commit_setup	= nfs4_proc_commit_setup,
 	.commit_done	= nfs4_commit_done,
-	.file_open      = nfs_open,
-	.file_release   = nfs_release,
 	.lock		= nfs4_proc_lock,
 	.clear_acl_cache = nfs4_zap_acl_attr,
 };
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index c760558..4dbb84d 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -655,8 +655,6 @@ const struct nfs_rpc_ops nfs_v2_clientops = {
 	.write_setup	= nfs_proc_write_setup,
 	.write_done	= nfs_write_done,
 	.commit_setup	= nfs_proc_commit_setup,
-	.file_open	= nfs_open,
-	.file_release	= nfs_release,
 	.lock		= nfs_proc_lock,
 	.lock_check_bounds = nfs_lock_check_bounds,
 };
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 8d780de..8c77c11 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -829,8 +829,6 @@ struct nfs_rpc_ops {
 	int	(*write_done)  (struct rpc_task *, struct nfs_write_data *);
 	void	(*commit_setup) (struct nfs_write_data *, struct rpc_message *);
 	int	(*commit_done) (struct rpc_task *, struct nfs_write_data *);
-	int	(*file_open)   (struct inode *, struct file *);
-	int	(*file_release) (struct inode *, struct file *);
 	int	(*lock)(struct file *, int, struct file_lock *);
 	int	(*lock_check_bounds)(const struct file_lock *);
 	void	(*clear_acl_cache)(struct inode *);

[PATCH 18/37] NFS: Make nfs_open methods consistent

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: Report the same debugging info and count function calls the
same for files and directories in nfs_opendir() and nfs_file_open().

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/dir.c  |    7 +++++--
 fs/nfs/file.c |    4 ++++
 2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 9f31cc6..9a4fb51 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -133,8 +133,11 @@ nfs_opendir(struct inode *inode, struct file *filp)
 {
 	int res;
 
-	dfprintk(VFS, "NFS: opendir(%s/%ld)\n",
-			inode->i_sb->s_id, inode->i_ino);
+	dfprintk(VFS, "NFS: open dir(%s/%s)\n",
+			filp->f_path.dentry->d_parent->d_name.name,
+			filp->f_path.dentry->d_name.name);
+
+	nfs_inc_stats(inode, NFSIOS_VFSOPEN);
 
 	lock_kernel();
 	/* Call generic open code in order to cache credentials */
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index cef3636..202408d 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -119,6 +119,10 @@ nfs_file_open(struct inode *inode, struct file *filp)
 {
 	int res;
 
+	dfprintk(VFS, "NFS: open file(%s/%s)\n",
+			filp->f_path.dentry->d_parent->d_name.name,
+			filp->f_path.dentry->d_name.name);
+
 	res = nfs_check_flags(filp->f_flags);
 	if (res)
 		return res;

[PATCH 33/37] NFS: Fix a warning in nfs4_async_handle_error

[Trond Myklebust]

We're not modifying the nfs_server when we call nfs_inc_server_stats and
friends, so allow the compiler to pass 'const' pointers too.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/iostat.h   |    8 ++++----
 fs/nfs/nfs4proc.c |    3 +--
 2 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/fs/nfs/iostat.h b/fs/nfs/iostat.h
index 2ec65e1..a369528 100644
--- a/fs/nfs/iostat.h
+++ b/fs/nfs/iostat.h
@@ -19,7 +19,7 @@ struct nfs_iostats {
 	unsigned long		events[__NFSIOS_COUNTSMAX];
 } ____cacheline_aligned;
 
-static inline void nfs_inc_server_stats(struct nfs_server *server,
+static inline void nfs_inc_server_stats(const struct nfs_server *server,
 					enum nfs_stat_eventcounters stat)
 {
 	struct nfs_iostats *iostats;
@@ -31,13 +31,13 @@ static inline void nfs_inc_server_stats(struct nfs_server *server,
 	put_cpu_no_resched();
 }
 
-static inline void nfs_inc_stats(struct inode *inode,
+static inline void nfs_inc_stats(const struct inode *inode,
 				 enum nfs_stat_eventcounters stat)
 {
 	nfs_inc_server_stats(NFS_SERVER(inode), stat);
 }
 
-static inline void nfs_add_server_stats(struct nfs_server *server,
+static inline void nfs_add_server_stats(const struct nfs_server *server,
 					enum nfs_stat_bytecounters stat,
 					unsigned long addend)
 {
@@ -50,7 +50,7 @@ static inline void nfs_add_server_stats(struct nfs_server *server,
 	put_cpu_no_resched();
 }
 
-static inline void nfs_add_stats(struct inode *inode,
+static inline void nfs_add_stats(const struct inode *inode,
 				 enum nfs_stat_bytecounters stat,
 				 unsigned long addend)
 {
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 75cdec4..0fc1068 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2766,8 +2766,7 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server)
 			task->tk_status = 0;
 			return -EAGAIN;
 		case -NFS4ERR_DELAY:
-			nfs_inc_server_stats((struct nfs_server *) server,
-						NFSIOS_DELAY);
+			nfs_inc_server_stats(server, NFSIOS_DELAY);
 		case -NFS4ERR_GRACE:
 			rpc_delay(task, NFS4_POLL_RETRY_MAX);
 			task->tk_status = 0;

[PATCH 35/37] NFS: Allow any value for the "retry" option

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

The kernel NFS mount option parser should ignore the retry= mount option
since it is meaningful only in user space.  Today it expects a number
rather than arbitrary text, so it ignores the option if the value is
numeric, but chokes if there are other characters in the value.

Change it to allow any text (except ",") as its value.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/super.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 4461c00..dea138b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -101,6 +101,8 @@ enum {
 static match_table_t nfs_mount_option_tokens = {
 	{ Opt_userspace, "bg" },
 	{ Opt_userspace, "fg" },
+	{ Opt_userspace, "retry=%s" },
+
 	{ Opt_soft, "soft" },
 	{ Opt_hard, "hard" },
 	{ Opt_intr, "intr" },
@@ -136,7 +138,6 @@ static match_table_t nfs_mount_option_tokens = {
 	{ Opt_acdirmin, "acdirmin=%u" },
 	{ Opt_acdirmax, "acdirmax=%u" },
 	{ Opt_actimeo, "actimeo=%u" },
-	{ Opt_userspace, "retry=%u" },
 	{ Opt_namelen, "namlen=%u" },
 	{ Opt_mountport, "mountport=%u" },
 	{ Opt_mountvers, "mountvers=%u" },

[PATCH 20/37] NFS: Use NFSDBG_FILE for all fops

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: some fops use NFSDBG_FILE, some use NFSDBG_VFS.  Let's use
NFSDBG_FILE for all fops, and consistently report file names instead
of inode numbers.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/dir.c    |   10 +++++----
 fs/nfs/direct.c |    4 ++--
 fs/nfs/file.c   |   59 ++++++++++++++++++++++++++++++++++---------------------
 3 files changed, 43 insertions(+), 30 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 9a4fb51..ff737bf 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -133,7 +133,7 @@ nfs_opendir(struct inode *inode, struct file *filp)
 {
 	int res;
 
-	dfprintk(VFS, "NFS: open dir(%s/%s)\n",
+	dfprintk(FILE, "NFS: open dir(%s/%s)\n",
 			filp->f_path.dentry->d_parent->d_name.name,
 			filp->f_path.dentry->d_name.name);
 
@@ -531,7 +531,7 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir)
 	struct nfs_fattr fattr;
 	long		res;
 
-	dfprintk(VFS, "NFS: readdir(%s/%s) starting at cookie %Lu\n",
+	dfprintk(FILE, "NFS: readdir(%s/%s) starting at cookie %llu\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			(long long)filp->f_pos);
 	nfs_inc_stats(inode, NFSIOS_VFSGETDENTS);
@@ -598,7 +598,7 @@ out:
 	unlock_kernel();
 	if (res > 0)
 		res = 0;
-	dfprintk(VFS, "NFS: readdir(%s/%s) returns %ld\n",
+	dfprintk(FILE, "NFS: readdir(%s/%s) returns %ld\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			res);
 	return res;
@@ -609,7 +609,7 @@ static loff_t nfs_llseek_dir(struct file *filp, loff_t offset, int origin)
 	struct dentry *dentry = filp->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
 
-	dfprintk(VFS, "NFS: llseek dir(%s/%s, %lld, %d)\n",
+	dfprintk(FILE, "NFS: llseek dir(%s/%s, %lld, %d)\n",
 			dentry->d_parent->d_name.name,
 			dentry->d_name.name,
 			offset, origin);
@@ -640,7 +640,7 @@ out:
  */
 static int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync)
 {
-	dfprintk(VFS, "NFS: fsync dir(%s/%s) datasync %d\n",
+	dfprintk(FILE, "NFS: fsync dir(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 4757a2b..08f6b04 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -890,7 +890,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, const struct iovec *iov,
 	count = iov_length(iov, nr_segs);
 	nfs_add_stats(mapping->host, NFSIOS_DIRECTREADBYTES, count);
 
-	dprintk("nfs: direct read(%s/%s, %zd@%Ld)\n",
+	dfprintk(FILE, "NFS: direct read(%s/%s, %zd@%Ld)\n",
 		file->f_path.dentry->d_parent->d_name.name,
 		file->f_path.dentry->d_name.name,
 		count, (long long) pos);
@@ -947,7 +947,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, const struct iovec *iov,
 	count = iov_length(iov, nr_segs);
 	nfs_add_stats(mapping->host, NFSIOS_DIRECTWRITTENBYTES, count);
 
-	dfprintk(VFS, "nfs: direct write(%s/%s, %zd@%Ld)\n",
+	dfprintk(FILE, "NFS: direct write(%s/%s, %zd@%Ld)\n",
 		file->f_path.dentry->d_parent->d_name.name,
 		file->f_path.dentry->d_name.name,
 		count, (long long) pos);
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 6c447a3..78b26f8 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -119,7 +119,7 @@ nfs_file_open(struct inode *inode, struct file *filp)
 {
 	int res;
 
-	dfprintk(VFS, "NFS: open file(%s/%s)\n",
+	dprintk("NFS: open file(%s/%s)\n",
 			filp->f_path.dentry->d_parent->d_name.name,
 			filp->f_path.dentry->d_name.name);
 
@@ -137,9 +137,15 @@ nfs_file_open(struct inode *inode, struct file *filp)
 static int
 nfs_file_release(struct inode *inode, struct file *filp)
 {
+	struct dentry *dentry = filp->f_path.dentry;
+
+	dprintk("NFS: release(%s/%s)\n",
+			dentry->d_parent->d_name.name,
+			dentry->d_name.name);
+
 	/* Ensure that dirty pages are flushed out with the right creds */
 	if (filp->f_mode & FMODE_WRITE)
-		nfs_wb_all(filp->f_path.dentry->d_inode);
+		nfs_wb_all(dentry->d_inode);
 	nfs_inc_stats(inode, NFSIOS_VFSRELEASE);
 	return NFS_PROTO(inode)->file_release(inode, filp);
 }
@@ -174,7 +180,7 @@ force_reval:
 
 static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 {
-	dfprintk(VFS, "NFS: llseek file(%s/%s, %lld, %d)\n",
+	dprintk("NFS: llseek file(%s/%s, %lld, %d)\n",
 			filp->f_path.dentry->d_parent->d_name.name,
 			filp->f_path.dentry->d_name.name,
 			offset, origin);
@@ -216,16 +222,18 @@ static int nfs_do_fsync(struct nfs_open_context *ctx, struct inode *inode)
 
 /*
  * Flush all dirty pages, and check for write errors.
- *
  */
 static int
 nfs_file_flush(struct file *file, fl_owner_t id)
 {
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
-	struct inode	*inode = file->f_path.dentry->d_inode;
+	struct dentry	*dentry = file->f_path.dentry;
+	struct inode	*inode = dentry->d_inode;
 	int		status;
 
-	dfprintk(VFS, "nfs: flush(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino);
+	dprintk("NFS: flush(%s/%s)\n",
+			dentry->d_parent->d_name.name,
+			dentry->d_name.name);
 
 	if ((file->f_mode & FMODE_WRITE) == 0)
 		return 0;
@@ -250,7 +258,7 @@ nfs_file_read(struct kiocb *iocb, const struct iovec *iov,
 	if (iocb->ki_filp->f_flags & O_DIRECT)
 		return nfs_file_direct_read(iocb, iov, nr_segs, pos);
 
-	dfprintk(VFS, "nfs: read(%s/%s, %lu@%lu)\n",
+	dprintk("NFS: read(%s/%s, %lu@%lu)\n",
 		dentry->d_parent->d_name.name, dentry->d_name.name,
 		(unsigned long) count, (unsigned long) pos);
 
@@ -270,7 +278,7 @@ nfs_file_splice_read(struct file *filp, loff_t *ppos,
 	struct inode *inode = dentry->d_inode;
 	ssize_t res;
 
-	dfprintk(VFS, "nfs: splice_read(%s/%s, %lu@%Lu)\n",
+	dprintk("NFS: splice_read(%s/%s, %lu@%Lu)\n",
 		dentry->d_parent->d_name.name, dentry->d_name.name,
 		(unsigned long) count, (unsigned long long) *ppos);
 
@@ -287,7 +295,7 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
 	struct inode *inode = dentry->d_inode;
 	int	status;
 
-	dfprintk(VFS, "nfs: mmap(%s/%s)\n",
+	dprintk("NFS: mmap(%s/%s)\n",
 		dentry->d_parent->d_name.name, dentry->d_name.name);
 
 	status = nfs_revalidate_mapping(inode, file->f_mapping);
@@ -310,7 +318,7 @@ nfs_file_fsync(struct file *file, struct dentry *dentry, int datasync)
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
 	struct inode *inode = dentry->d_inode;
 
-	dfprintk(VFS, "NFS: fsync file(%s/%s) datasync %d\n",
+	dprintk("NFS: fsync file(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
@@ -502,9 +510,9 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov,
 	if (iocb->ki_filp->f_flags & O_DIRECT)
 		return nfs_file_direct_write(iocb, iov, nr_segs, pos);
 
-	dfprintk(VFS, "nfs: write(%s/%s(%ld), %lu@%Ld)\n",
+	dprintk("NFS: write(%s/%s, %lu@%Ld)\n",
 		dentry->d_parent->d_name.name, dentry->d_name.name,
-		inode->i_ino, (unsigned long) count, (long long) pos);
+		(unsigned long) count, (long long) pos);
 
 	result = -EBUSY;
 	if (IS_SWAPFILE(inode))
@@ -649,13 +657,15 @@ out:
  */
 static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl)
 {
-	struct inode * inode = filp->f_mapping->host;
+	struct inode *inode = filp->f_mapping->host;
 	int ret = -ENOLCK;
 
-	dprintk("NFS: nfs_lock(f=%s/%ld, t=%x, fl=%x, r=%Ld:%Ld)\n",
-			inode->i_sb->s_id, inode->i_ino,
+	dprintk("NFS: lock(%s/%s, t=%x, fl=%x, r=%lld:%lld)\n",
+			filp->f_path.dentry->d_parent->d_name.name,
+			filp->f_path.dentry->d_name.name,
 			fl->fl_type, fl->fl_flags,
 			(long long)fl->fl_start, (long long)fl->fl_end);
+
 	nfs_inc_stats(inode, NFSIOS_VFSLOCK);
 
 	/* No mandatory locks over NFS */
@@ -683,9 +693,9 @@ out_err:
  */
 static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl)
 {
-	dprintk("NFS: nfs_flock(f=%s/%ld, t=%x, fl=%x)\n",
-			filp->f_path.dentry->d_inode->i_sb->s_id,
-			filp->f_path.dentry->d_inode->i_ino,
+	dprintk("NFS: flock(%s/%s, t=%x, fl=%x)\n",
+			filp->f_path.dentry->d_parent->d_name.name,
+			filp->f_path.dentry->d_name.name,
 			fl->fl_type, fl->fl_flags);
 
 	/*
@@ -708,12 +718,15 @@ static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl)
 	return do_setlk(filp, cmd, fl);
 }
 
+/*
+ * There is no protocol support for leases, so we have no way to implement
+ * them correctly in the face of opens by other clients.
+ */
 static int nfs_setlease(struct file *file, long arg, struct file_lock **fl)
 {
-	/*
-	 * There is no protocol support for leases, so we have no way
-	 * to implement them correctly in the face of opens by other
-	 * clients.
-	 */
+	dprintk("NFS: setlease(%s/%s, arg=%ld)\n",
+			file->f_path.dentry->d_parent->d_name.name,
+			file->f_path.dentry->d_name.name, arg);
+
 	return -EINVAL;
 }

[PATCH 30/37] SUNRPC: Ensure all transports set rq_xtime consistently

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

The RPC client uses the rq_xtime field in each RPC request to determine the
round-trip time of the request.  Currently, the rq_xtime field is
initialized by each transport just before it starts enqueing a request to
be sent.  However, transports do not handle initializing this value
consistently; sometimes they don't initialize it at all.

To make the measurement of request round-trip time consistent for all
RPC client transport capabilities, pull rq_xtime initialization into the
RPC client's generic transport logic.  Now all transports will get a
standardized RTT measure automatically, from:

  xprt_transmit()

to

  xprt_complete_rqst()

This makes round-trip time calculation more accurate for the TCP transport.
The socket ->sendmsg() method can return "-EAGAIN" if the socket's output
buffer is full, so the TCP transport's ->send_request() method may call
the ->sendmsg() method repeatedly until it gets all of the request's bytes
queued in the socket's buffer.

Currently, the TCP transport sets the rq_xtime field every time through
that loop so the final value is the timestamp just before the *last* call
to the underlying socket's ->sendmsg() method.  After this patch, the
rq_xtime field contains a timestamp that reflects the time just before the
*first* call to ->sendmsg().

This is consequential under heavy workloads because large requests often
take multiple ->sendmsg() calls to get all the bytes of a request queued.
The TCP transport causes the request to sleep until the remote end of the
socket has received enough bytes to clear space in the socket's local
output buffer.  This delay can be quite significant.

The method introduced by this patch is a more accurate measure of RTT
for stream transports, since the server can cause enough back pressure
to delay (ie increase the latency of) requests from the client.

Additionally, this patch corrects the behavior of the RDMA transport, which
entirely neglected to initialize the rq_xtime field.  RPC performance
metrics for RDMA transports now display correct RPC request round trip
times.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Acked-by: Tom Talpey <thomas.talpey@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/xprt.c     |    1 +
 net/sunrpc/xprtsock.c |    2 --
 2 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 67996bd..99a52aa 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -872,6 +872,7 @@ void xprt_transmit(struct rpc_task *task)
 		return;
 
 	req->rq_connect_cookie = xprt->connect_cookie;
+	req->rq_xtime = jiffies;
 	status = xprt->ops->send_request(task);
 	if (status == 0) {
 		dprintk("RPC: %5u xmit complete\n", task->tk_pid);
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index ddbe981..4486c59 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -579,7 +579,6 @@ static int xs_udp_send_request(struct rpc_task *task)
 				req->rq_svec->iov_base,
 				req->rq_svec->iov_len);
 
-	req->rq_xtime = jiffies;
 	status = xs_sendpages(transport->sock,
 			      xs_addr(xprt),
 			      xprt->addrlen, xdr,
@@ -671,7 +670,6 @@ static int xs_tcp_send_request(struct rpc_task *task)
 	 * to cope with writespace callbacks arriving _after_ we have
 	 * called sendmsg(). */
 	while (1) {
-		req->rq_xtime = jiffies;
 		status = xs_sendpages(transport->sock,
 					NULL, 0, xdr, req->rq_bytes_sent);
 

[PATCH 21/37] NFS: Fix trace debugging nits in write.c

[Trond Myklebust]

From: Chuck Lever <chuck.lever@oracle.com>

Clean up: fix a few dprintk messages that still need to show the RPC task ID
correctly, and be sure we use the preferred %lld or %llu instead of %Ld or
%Lu.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/write.c |   27 ++++++++++++++-------------
 1 files changed, 14 insertions(+), 13 deletions(-)

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index ee6fcde..21d8a48 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -713,10 +713,10 @@ int nfs_updatepage(struct file *file, struct page *page,
 
 	nfs_inc_stats(inode, NFSIOS_VFSUPDATEPAGE);
 
-	dprintk("NFS:      nfs_updatepage(%s/%s %d@%Ld)\n",
+	dprintk("NFS:       nfs_updatepage(%s/%s %d@%lld)\n",
 		file->f_path.dentry->d_parent->d_name.name,
 		file->f_path.dentry->d_name.name, count,
-		(long long)(page_offset(page) +offset));
+		(long long)(page_offset(page) + offset));
 
 	/* If we're not using byte range locks, and we know the page
 	 * is up to date, it may be more efficient to extend the write
@@ -736,7 +736,7 @@ int nfs_updatepage(struct file *file, struct page *page,
 	else
 		__set_page_dirty_nobuffers(page);
 
-        dprintk("NFS:      nfs_updatepage returns %d (isize %Ld)\n",
+	dprintk("NFS:       nfs_updatepage returns %d (isize %lld)\n",
 			status, (long long)i_size_read(inode));
 	return status;
 }
@@ -821,7 +821,7 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 	NFS_PROTO(inode)->write_setup(data, &msg);
 
 	dprintk("NFS: %5u initiated write call "
-		"(req %s/%Ld, %u bytes @ offset %Lu)\n",
+		"(req %s/%lld, %u bytes @ offset %llu)\n",
 		data->task.tk_pid,
 		inode->i_sb->s_id,
 		(long long)NFS_FILEID(inode),
@@ -965,13 +965,13 @@ static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
 static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata)
 {
 	struct nfs_write_data	*data = calldata;
-	struct nfs_page		*req = data->req;
 
-	dprintk("NFS: write (%s/%Ld %d@%Ld)",
-		req->wb_context->path.dentry->d_inode->i_sb->s_id,
-		(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
-		req->wb_bytes,
-		(long long)req_offset(req));
+	dprintk("NFS: %5u write(%s/%lld %d@%lld)",
+		task->tk_pid,
+		data->req->wb_context->path.dentry->d_inode->i_sb->s_id,
+		(long long)
+		  NFS_FILEID(data->req->wb_context->path.dentry->d_inode),
+		data->req->wb_bytes, (long long)req_offset(data->req));
 
 	nfs_writeback_done(task, data);
 }
@@ -1045,7 +1045,8 @@ static void nfs_writeback_release_full(void *calldata)
 
 		nfs_list_remove_request(req);
 
-		dprintk("NFS: write (%s/%Ld %d@%Ld)",
+		dprintk("NFS: %5u write (%s/%lld %d@%lld)",
+			data->task.tk_pid,
 			req->wb_context->path.dentry->d_inode->i_sb->s_id,
 			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
 			req->wb_bytes,
@@ -1118,7 +1119,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
 		static unsigned long    complain;
 
 		if (time_before(complain, jiffies)) {
-			dprintk("NFS: faulty NFS server %s:"
+			dprintk("NFS:       faulty NFS server %s:"
 				" (committed = %d) != (stable = %d)\n",
 				NFS_SERVER(data->inode)->nfs_client->cl_hostname,
 				resp->verf->committed, argp->stable);
@@ -1287,7 +1288,7 @@ static void nfs_commit_release(void *calldata)
 		dec_bdi_stat(req->wb_page->mapping->backing_dev_info,
 				BDI_RECLAIMABLE);
 
-		dprintk("NFS: commit (%s/%Ld %d@%Ld)",
+		dprintk("NFS:       commit (%s/%lld %d@%lld)",
 			req->wb_context->path.dentry->d_inode->i_sb->s_id,
 			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
 			req->wb_bytes,

[PATCH 25/37] rpc: bring back cl_chatty

[Trond Myklebust]

From: Olga Kornievskaia <aglo@citi.umich.edu>

The cl_chatty flag alows us to control whether a given rpc client leaves

	"server X not responding, timed out"

messages in the syslog.  Such messages make sense for ordinary nfs
clients (where an unresponsive server means applications on the
mountpoint are probably hanging), but not for the callback client (which
can fail more commonly, with the only result just of disabling some
optimizations).

Previously cl_chatty was removed, do to lack of users; reinstate it, and
use it for the nfsd's callback client.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfsd/nfs4callback.c      |    2 +-
 include/linux/sunrpc/clnt.h |    4 +++-
 net/sunrpc/clnt.c           |   16 +++++++++++-----
 3 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 4d4760e..702fa57 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -381,7 +381,7 @@ static int do_probe_callback(void *data)
 		.program	= &cb_program,
 		.version	= nfs_cb_version[1]->number,
 		.authflavor	= RPC_AUTH_UNIX, /* XXX: need AUTH_GSS... */
-		.flags		= (RPC_CLNT_CREATE_NOPING),
+		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
 	};
 	struct rpc_message msg = {
 		.rpc_proc       = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 6fff7f8..764fd4c 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -42,7 +42,8 @@ struct rpc_clnt {
 
 	unsigned int		cl_softrtry : 1,/* soft timeouts */
 				cl_discrtry : 1,/* disconnect before retry */
-				cl_autobind : 1;/* use getport() */
+				cl_autobind : 1,/* use getport() */
+				cl_chatty   : 1;/* be verbose */
 
 	struct rpc_rtt *	cl_rtt;		/* RTO estimator data */
 	const struct rpc_timeout *cl_timeout;	/* Timeout strategy */
@@ -114,6 +115,7 @@ struct rpc_create_args {
 #define RPC_CLNT_CREATE_NONPRIVPORT	(1UL << 3)
 #define RPC_CLNT_CREATE_NOPING		(1UL << 4)
 #define RPC_CLNT_CREATE_DISCRTRY	(1UL << 5)
+#define RPC_CLNT_CREATE_QUIET		(1UL << 6)
 
 struct rpc_clnt *rpc_create(struct rpc_create_args *args);
 struct rpc_clnt	*rpc_bind_new_program(struct rpc_clnt *,
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 0530eea..09631f6 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -324,6 +324,8 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 		clnt->cl_autobind = 1;
 	if (args->flags & RPC_CLNT_CREATE_DISCRTRY)
 		clnt->cl_discrtry = 1;
+	if (!(args->flags & RPC_CLNT_CREATE_QUIET))
+		clnt->cl_chatty = 1;
 
 	return clnt;
 }
@@ -1149,7 +1151,8 @@ call_status(struct rpc_task *task)
 		rpc_exit(task, status);
 		break;
 	default:
-		printk("%s: RPC call returned error %d\n",
+		if (clnt->cl_chatty)
+			printk("%s: RPC call returned error %d\n",
 			       clnt->cl_protname, -status);
 		rpc_exit(task, status);
 	}
@@ -1174,7 +1177,8 @@ call_timeout(struct rpc_task *task)
 	task->tk_timeouts++;
 
 	if (RPC_IS_SOFT(task)) {
-		printk(KERN_NOTICE "%s: server %s not responding, timed out\n",
+		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);
 		return;
@@ -1182,7 +1186,8 @@ call_timeout(struct rpc_task *task)
 
 	if (!(task->tk_flags & RPC_CALL_MAJORSEEN)) {
 		task->tk_flags |= RPC_CALL_MAJORSEEN;
-		printk(KERN_NOTICE "%s: server %s not responding, still trying\n",
+		if (clnt->cl_chatty)
+			printk(KERN_NOTICE "%s: server %s not responding, still trying\n",
 			clnt->cl_protname, clnt->cl_server);
 	}
 	rpc_force_rebind(clnt);
@@ -1213,8 +1218,9 @@ call_decode(struct rpc_task *task)
 			task->tk_pid, task->tk_status);
 
 	if (task->tk_flags & RPC_CALL_MAJORSEEN) {
-		printk(KERN_NOTICE "%s: server %s OK\n",
-			clnt->cl_protname, clnt->cl_server);
+		if (clnt->cl_chatty)
+			printk(KERN_NOTICE "%s: server %s OK\n",
+				clnt->cl_protname, clnt->cl_server);
 		task->tk_flags &= ~RPC_CALL_MAJORSEEN;
 	}
 

[PATCH 27/37] rpc: remove some unused macros

[Trond Myklebust]

From: \"J. Bruce Fields\" <bfields@citi.umich.edu>

There used to be a print_hexl() function that used isprint(), now gone.
I don't know why NFS_NGROUPS and CA_RUN_AS_MACHINE were here.

I also don't know why another #define that's actually used was marked
"unused".

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 net/sunrpc/auth_gss/auth_gss.c |   13 +------------
 1 files changed, 1 insertions(+), 12 deletions(-)

diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index ebfd630..834a831 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -63,22 +63,11 @@ static const struct rpc_credops gss_nullops;
 # define RPCDBG_FACILITY	RPCDBG_AUTH
 #endif
 
-#define NFS_NGROUPS	16
-
-#define GSS_CRED_SLACK		1024		/* XXX: unused */
+#define GSS_CRED_SLACK		1024
 /* length of a krb5 verifier (48), plus data added before arguments when
  * using integrity (two 4-byte integers): */
 #define GSS_VERF_SLACK		100
 
-/* XXX this define must match the gssd define
-* as it is passed to gssd to signal the use of
-* machine creds should be part of the shared rpc interface */
-
-#define CA_RUN_AS_MACHINE  0x00000200
-
-/* dump the buffer in `emacs-hexl' style */
-#define isprint(c)      ((c > 0x1f) && (c < 0x7f))
-
 struct gss_auth {
 	struct kref kref;
 	struct rpc_auth rpc_auth;

[PATCH 24/37] NFS: implement option checking when remounting NFS filesystems (resend)

[Trond Myklebust]

From: Jeff Layton <jlayton@redhat.com>

When remounting an NFS or NFS4 filesystem, the new NFS options are not
respected, yet the remount will still return success. This patch adds
a remount_fs sb op for NFS that checks any new nfs mount options against
the existing ones and fails the mount if any have changed.

This is only implemented for string-based mount options since doing
this with binary options isn't really feasible.

This is essentially the same as the original patch I sent out, but
adds a check to see if the addr= option has changed.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
---

 fs/nfs/super.c |   76 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 76 insertions(+), 0 deletions(-)

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 2a4a024..4461c00 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -207,6 +207,7 @@ static int nfs_xdev_get_sb(struct file_system_type *fs_type,
 		int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
 static void nfs_kill_super(struct super_block *);
 static void nfs_put_super(struct super_block *);
+static int nfs_remount(struct super_block *sb, int *flags, char *raw_data);
 
 static struct file_system_type nfs_fs_type = {
 	.owner		= THIS_MODULE,
@@ -234,6 +235,7 @@ static const struct super_operations nfs_sops = {
 	.umount_begin	= nfs_umount_begin,
 	.show_options	= nfs_show_options,
 	.show_stats	= nfs_show_stats,
+	.remount_fs	= nfs_remount,
 };
 
 #ifdef CONFIG_NFS_V4
@@ -278,6 +280,7 @@ static const struct super_operations nfs4_sops = {
 	.umount_begin	= nfs_umount_begin,
 	.show_options	= nfs_show_options,
 	.show_stats	= nfs_show_stats,
+	.remount_fs	= nfs_remount,
 };
 #endif
 
@@ -1398,6 +1401,79 @@ out_invalid_fh:
 	return -EINVAL;
 }
 
+static int
+nfs_compare_remount_data(struct nfs_server *nfss,
+			 struct nfs_parsed_mount_data *data)
+{
+	if (data->flags != nfss->flags ||
+	    data->rsize != nfss->rsize ||
+	    data->wsize != nfss->wsize ||
+	    data->retrans != nfss->client->cl_timeout->to_retries ||
+	    data->auth_flavors[0] != nfss->client->cl_auth->au_flavor ||
+	    data->acregmin != nfss->acregmin / HZ ||
+	    data->acregmax != nfss->acregmax / HZ ||
+	    data->acdirmin != nfss->acdirmin / HZ ||
+	    data->acdirmax != nfss->acdirmax / HZ ||
+	    data->timeo != (10U * nfss->client->cl_timeout->to_initval / HZ) ||
+	    data->nfs_server.addrlen != nfss->nfs_client->cl_addrlen ||
+	    memcmp(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
+		   data->nfs_server.addrlen) != 0)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int
+nfs_remount(struct super_block *sb, int *flags, char *raw_data)
+{
+	int error;
+	struct nfs_server *nfss = sb->s_fs_info;
+	struct nfs_parsed_mount_data *data;
+	struct nfs_mount_data *options = (struct nfs_mount_data *)raw_data;
+	struct nfs4_mount_data *options4 = (struct nfs4_mount_data *)raw_data;
+
+	/*
+	 * Userspace mount programs that send binary options generally send
+	 * them populated with default values. We have no way to know which
+	 * ones were explicitly specified. Fall back to legacy behavior and
+	 * just return success.
+	 */
+	if ((sb->s_type == &nfs4_fs_type && options4->version == 1) ||
+	    (sb->s_type == &nfs_fs_type && options->version >= 1 &&
+	     options->version <= 6))
+		return 0;
+
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
+	if (data == NULL)
+		return -ENOMEM;
+
+	/* fill out struct with values from existing mount */
+	data->flags = nfss->flags;
+	data->rsize = nfss->rsize;
+	data->wsize = nfss->wsize;
+	data->retrans = nfss->client->cl_timeout->to_retries;
+	data->auth_flavors[0] = nfss->client->cl_auth->au_flavor;
+	data->acregmin = nfss->acregmin / HZ;
+	data->acregmax = nfss->acregmax / HZ;
+	data->acdirmin = nfss->acdirmin / HZ;
+	data->acdirmax = nfss->acdirmax / HZ;
+	data->timeo = 10U * nfss->client->cl_timeout->to_initval / HZ;
+	data->nfs_server.addrlen = nfss->nfs_client->cl_addrlen;
+	memcpy(&data->nfs_server.address, &nfss->nfs_client->cl_addr,
+		data->nfs_server.addrlen);
+
+	/* overwrite those values with any that were specified */
+	error = nfs_parse_mount_options((char *)options, data);
+	if (error < 0)
+		goto out;
+
+	/* compare new mount options with old ones */
+	error = nfs_compare_remount_data(nfss, data);
+out:
+	kfree(data);
+	return error;
+}
+
 /*
  * Initialise the common bits of the superblock
  */