[PATCH 00/12] For 3.4 (3rd take)

[PATCH 00/12] For 3.4 (3rd take)

[Chuck Lever]

Hi-

Please consider these minor fixes and migration pre-requisites for
3.4.  I've ported them to nfs-for-next on top of commit abd96698.

This take addresses yesterday's review comments.  I'm leaving the 
nfs4_verifier and nfs4_deviceid patches for a separate submission,
since the following patches seem to have reached a low level of
controversy and should be ready to include now.

---

Chuck Lever (10):
      NFS: Request fh_expire_type attribute in "server caps" operation
      NFS: Introduce NFS_ATTR_FATTR_V4_LOCATIONS
      NFS: Simplify arguments of encode_renew()
      NFS: Add a client-side function to display NFS file handles
      commit 6f38b4ba433ac6494f83cb73dd07dcbde797e1e0
      SUNRPC: Add API to acquire source address
      NFS: Reduce debugging noise from encode_compound_hdr
      NFS: Add debugging messages to NFSv4's CLOSE procedure
      NFS: Clean up debugging in decode_pathname()
      NFS: Make nfs_cache_array.size a signed integer

Trond Myklebust (2):
      SUNRPC: Move clnt->cl_server into struct rpc_xprt
      SUNRPC: Use RCU to dereference the rpc_clnt.cl_xprt field


 fs/nfs/callback.c              |    3 
 fs/nfs/callback_proc.c         |    9 +
 fs/nfs/client.c                |   31 +++-
 fs/nfs/dir.c                   |    2 
 fs/nfs/inode.c                 |   45 +++++
 fs/nfs/nfs4namespace.c         |    2 
 fs/nfs/nfs4proc.c              |   34 +++-
 fs/nfs/nfs4state.c             |   25 ++-
 fs/nfs/nfs4xdr.c               |   52 ++++--
 fs/nfs/super.c                 |   11 +
 include/linux/nfs_fs.h         |   14 ++
 include/linux/nfs_fs_sb.h      |    3 
 include/linux/nfs_xdr.h        |   12 +
 include/linux/sunrpc/clnt.h    |    6 -
 include/linux/sunrpc/debug.h   |   13 +
 include/linux/sunrpc/xprt.h    |    2 
 net/sunrpc/auth_gss/auth_gss.c |    4 
 net/sunrpc/clnt.c              |  347 ++++++++++++++++++++++++++++++++--------
 net/sunrpc/rpc_pipe.c          |    6 +
 net/sunrpc/rpcb_clnt.c         |   19 +-
 net/sunrpc/stats.c             |    6 +
 net/sunrpc/xprt.c              |   15 ++
 22 files changed, 527 insertions(+), 134 deletions(-)

-- 
Chuck Lever

[PATCH 01/12] NFS: Make nfs_cache_array.size a signed integer

[Chuck Lever]

Eliminate a number of implicit type casts in comparisons, and these
compiler warnings:

fs/nfs/dir.c: In function ‘nfs_readdir_clear_array’:
fs/nfs/dir.c:264:16: warning: comparison between signed and unsigned
		integer expressions [-Wsign-compare]
fs/nfs/dir.c: In function ‘nfs_readdir_search_for_cookie’:
fs/nfs/dir.c:352:16: warning: comparison between signed and unsigned
		integer expressions [-Wsign-compare]
fs/nfs/dir.c: In function ‘nfs_do_filldir’:
fs/nfs/dir.c:769:38: warning: comparison between signed and unsigned
		integer expressions [-Wsign-compare]
fs/nfs/dir.c:780:9: warning: comparison between signed and unsigned
		integer expressions [-Wsign-compare]

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

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

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index bb132a8..9952170 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -207,7 +207,7 @@ struct nfs_cache_array_entry {
 };
 
 struct nfs_cache_array {
-	unsigned int size;
+	int size;
 	int eof_index;
 	u64 last_cookie;
 	struct nfs_cache_array_entry array[0];

[PATCH 02/12] NFS: Clean up debugging in decode_pathname()

[Chuck Lever]

I noticed recently that decode_attr_fs_locations() is not generating
very pretty debugging output.  The pathname components each appear on
a separate line of output, though that does not appear to be the
intended display behavior.  The preferred way to generate continued
lines of output on the console is to use pr_cont().

Note that incoming pathname4 components contain a string that is not
necessarily NUL-terminated.  I did actually see some trailing garbage
on the console.  In addition to correcting the line continuation
problem, add a string precision format specifier to ensure that each
component string is displayed properly, and that vsnprintf() does
not Oops.

Someone pointed out that allowing incoming network data to possibly
generate a console line of unbounded length may not be such a good
idea.  Since this output will rarely be enabled, and there is a hard
upper bound (NFS4_PATHNAME_MAXCOMPONENTS) in our implementation, this
is probably not a major concern.

It might be useful to additionally sanity-check the length of each
incoming component, however.  RFC 3530bis15 does not suggest a maximum
number of UTF-8 characters per component for either the pathname4 or
component4 types.  However, we could invent one that is appropriate
for our implementation.

Another possibility is to scrap all of this and print these pathnames
in upper layers after a reasonable amount of sanity checking in the
XDR layer.  This would give us an opportunity to allocate a full
buffer so that the whole pathname would be output via a single
dprintk.

Introduced by commit 7aaa0b3b: "NFSv4: convert fs-locations-components
to conform to RFC3530," (June 9, 2006).

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

 fs/nfs/nfs4xdr.c |   16 ++++++++--------
 1 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index ae78343..4ca87cb 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -3514,16 +3514,17 @@ static int decode_pathname(struct xdr_stream *xdr, struct nfs4_pathname *path)
 	n = be32_to_cpup(p);
 	if (n == 0)
 		goto root_path;
-	dprintk("path ");
+	dprintk("pathname4: ");
 	path->ncomponents = 0;
 	while (path->ncomponents < n) {
 		struct nfs4_string *component = &path->components[path->ncomponents];
 		status = decode_opaque_inline(xdr, &component->len, &component->data);
 		if (unlikely(status != 0))
 			goto out_eio;
-		if (path->ncomponents != n)
-			dprintk("/");
-		dprintk("%s", component->data);
+		if (unlikely(nfs_debug & NFSDBG_XDR))
+			pr_cont("%s%.*s ",
+				(path->ncomponents != n ? "/ " : ""),
+				component->len, component->data);
 		if (path->ncomponents < NFS4_PATHNAME_MAXCOMPONENTS)
 			path->ncomponents++;
 		else {
@@ -3532,14 +3533,13 @@ static int decode_pathname(struct xdr_stream *xdr, struct nfs4_pathname *path)
 		}
 	}
 out:
-	dprintk("\n");
 	return status;
 root_path:
 /* a root pathname is sent as a zero component4 */
 	path->ncomponents = 1;
 	path->components[0].len=0;
 	path->components[0].data=NULL;
-	dprintk("path /\n");
+	dprintk("pathname4: /\n");
 	goto out;
 out_eio:
 	dprintk(" status %d", status);
@@ -3565,7 +3565,7 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st
 	/* Ignore borken servers that return unrequested attrs */
 	if (unlikely(res == NULL))
 		goto out;
-	dprintk("%s: fsroot ", __func__);
+	dprintk("%s: fsroot:\n", __func__);
 	status = decode_pathname(xdr, &res->fs_path);
 	if (unlikely(status != 0))
 		goto out;
@@ -3586,7 +3586,7 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st
 		m = be32_to_cpup(p);
 
 		loc->nservers = 0;
-		dprintk("%s: servers ", __func__);
+		dprintk("%s: servers:\n", __func__);
 		while (loc->nservers < m) {
 			struct nfs4_string *server = &loc->servers[loc->nservers];
 			status = decode_opaque_inline(xdr, &server->len, &server->data);

[PATCH 03/12] NFS: Add debugging messages to NFSv4's CLOSE procedure

[Chuck Lever]

CLOSE is new with NFSv4.  Sometimes it's important to know the timing
of this operation compared to things like lease renewal.

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

 fs/nfs/nfs4proc.c |   11 ++++++++---
 1 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 87c584d..53daab9 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1983,6 +1983,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data)
 	struct nfs4_state *state = calldata->state;
 	struct nfs_server *server = NFS_SERVER(calldata->inode);
 
+	dprintk("%s: begin!\n", __func__);
 	if (!nfs4_sequence_done(task, &calldata->res.seq_res))
 		return;
         /* hmm. we are done with the inode, and in the process of freeing
@@ -2010,6 +2011,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data)
 	}
 	nfs_release_seqid(calldata->arg.seqid);
 	nfs_refresh_inode(calldata->inode, calldata->res.fattr);
+	dprintk("%s: done, ret = %d!\n", __func__, task->tk_status);
 }
 
 static void nfs4_close_prepare(struct rpc_task *task, void *data)
@@ -2018,6 +2020,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 	struct nfs4_state *state = calldata->state;
 	int call_close = 0;
 
+	dprintk("%s: begin!\n", __func__);
 	if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0)
 		return;
 
@@ -2042,7 +2045,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 	if (!call_close) {
 		/* Note: exit _without_ calling nfs4_close_done */
 		task->tk_action = NULL;
-		return;
+		goto out;
 	}
 
 	if (calldata->arg.fmode == 0) {
@@ -2051,7 +2054,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 		    pnfs_roc_drain(calldata->inode, &calldata->roc_barrier)) {
 			rpc_sleep_on(&NFS_SERVER(calldata->inode)->roc_rpcwaitq,
 				     task, NULL);
-			return;
+			goto out;
 		}
 	}
 
@@ -2061,8 +2064,10 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 				&calldata->arg.seq_args,
 				&calldata->res.seq_res,
 				task))
-		return;
+		goto out;
 	rpc_call_start(task);
+out:
+	dprintk("%s: done!\n", __func__);
 }
 
 static const struct rpc_call_ops nfs4_close_ops = {

[PATCH 04/12] NFS: Reduce debugging noise from encode_compound_hdr

[Chuck Lever]

Get rid of

  encode_compound: tag=

when XDR debugging is enabled.  The current Linux client never sets
compound tags.

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

 fs/nfs/nfs4xdr.c |    1 -
 1 files changed, 0 insertions(+), 1 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 4ca87cb..ea01e31 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -889,7 +889,6 @@ static void encode_compound_hdr(struct xdr_stream *xdr,
 	 * but this is not required as a MUST for the server to do so. */
 	hdr->replen = RPC_REPHDRSIZE + auth->au_rslack + 3 + hdr->taglen;
 
-	dprintk("encode_compound: tag=%.*s\n", (int)hdr->taglen, hdr->tag);
 	BUG_ON(hdr->taglen > NFS4_MAXTAGLEN);
 	p = reserve_space(xdr, 4 + hdr->taglen + 8);
 	p = xdr_encode_opaque(p, hdr->tag, hdr->taglen);

[PATCH 05/12] SUNRPC: Use RCU to dereference the rpc_clnt.cl_xprt field

[Chuck Lever]

From: Trond Myklebust <Trond.Myklebust@netapp.com>

A migration event will replace the rpc_xprt used by an rpc_clnt.  To
ensure this can be done safely, all references to cl_xprt must now use
a form of rcu_dereference().

Special care is taken with rpc_peeraddr2str(), which returns a pointer
to memory whose lifetime is the same as the rpc_xprt.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
[ cel: fix lockdep splats and layering violations ]
[ cel: forward ported to 3.4 ]
[ cel: remove rpc_max_reqs(), add rpc_net_ns() ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---

 fs/nfs/callback_proc.c         |    9 ++-
 fs/nfs/client.c                |   16 ++++--
 fs/nfs/nfs4namespace.c         |    2 -
 fs/nfs/nfs4proc.c              |   13 ++++-
 fs/nfs/nfs4state.c             |   25 ++++++---
 fs/nfs/super.c                 |    5 ++
 include/linux/sunrpc/clnt.h    |    4 +
 include/linux/sunrpc/debug.h   |   13 +++++
 net/sunrpc/auth_gss/auth_gss.c |    4 +
 net/sunrpc/clnt.c              |  110 ++++++++++++++++++++++++++++++++--------
 net/sunrpc/rpc_pipe.c          |    3 +
 net/sunrpc/rpcb_clnt.c         |   15 ++++-
 net/sunrpc/stats.c             |    6 ++
 13 files changed, 175 insertions(+), 50 deletions(-)

diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index f71978d..2db26be 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -8,6 +8,7 @@
 #include <linux/nfs4.h>
 #include <linux/nfs_fs.h>
 #include <linux/slab.h>
+#include <linux/rcupdate.h>
 #include "nfs4_fs.h"
 #include "callback.h"
 #include "delegation.h"
@@ -33,7 +34,7 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args,
 	res->bitmap[0] = res->bitmap[1] = 0;
 	res->status = htonl(NFS4ERR_BADHANDLE);
 
-	dprintk("NFS: GETATTR callback request from %s\n",
+	dprintk_rcu("NFS: GETATTR callback request from %s\n",
 		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
 
 	inode = nfs_delegation_find_inode(cps->clp, &args->fh);
@@ -73,7 +74,7 @@ __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
 	if (!cps->clp) /* Always set for v4.0. Set in cb_sequence for v4.1 */
 		goto out;
 
-	dprintk("NFS: RECALL callback request from %s\n",
+	dprintk_rcu("NFS: RECALL callback request from %s\n",
 		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
 
 	res = htonl(NFS4ERR_BADHANDLE);
@@ -537,7 +538,7 @@ __be32 nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy,
 	if (!cps->clp) /* set in cb_sequence */
 		goto out;
 
-	dprintk("NFS: RECALL_ANY callback request from %s\n",
+	dprintk_rcu("NFS: RECALL_ANY callback request from %s\n",
 		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
 
 	status = cpu_to_be32(NFS4ERR_INVAL);
@@ -572,7 +573,7 @@ __be32 nfs4_callback_recallslot(struct cb_recallslotargs *args, void *dummy,
 	if (!cps->clp) /* set in cb_sequence */
 		goto out;
 
-	dprintk("NFS: CB_RECALL_SLOT request from %s target max slots %d\n",
+	dprintk_rcu("NFS: CB_RECALL_SLOT request from %s target max slots %d\n",
 		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR),
 		args->crsa_target_max_slots);
 
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 8563585..994899d 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1282,16 +1282,18 @@ static int nfs4_init_callback(struct nfs_client *clp)
 	int error;
 
 	if (clp->rpc_ops->version == 4) {
+		struct rpc_xprt *xprt;
+
+		xprt = rcu_dereference_raw(clp->cl_rpcclient->cl_xprt);
+
 		if (nfs4_has_session(clp)) {
-			error = xprt_setup_backchannel(
-						clp->cl_rpcclient->cl_xprt,
+			error = xprt_setup_backchannel(xprt,
 						NFS41_BC_MIN_CALLBACKS);
 			if (error < 0)
 				return error;
 		}
 
-		error = nfs_callback_up(clp->cl_mvops->minor_version,
-					clp->cl_rpcclient->cl_xprt);
+		error = nfs_callback_up(clp->cl_mvops->minor_version, xprt);
 		if (error < 0) {
 			dprintk("%s: failed to start callback. Error = %d\n",
 				__func__, error);
@@ -1676,7 +1678,7 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data,
 				data->addrlen,
 				parent_client->cl_ipaddr,
 				data->authflavor,
-				parent_server->client->cl_xprt->prot,
+				rpc_protocol(parent_server->client),
 				parent_server->client->cl_timeout,
 				parent_client->cl_mvops->minor_version,
 				parent_client->net);
@@ -1903,12 +1905,14 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
 	if (clp->cl_cons_state != NFS_CS_READY)
 		return 0;
 
+	rcu_read_lock();
 	seq_printf(m, "v%u %s %s %3d %s\n",
 		   clp->rpc_ops->version,
 		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
 		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_PORT),
 		   atomic_read(&clp->cl_count),
 		   clp->cl_hostname);
+	rcu_read_unlock();
 
 	return 0;
 }
@@ -1991,6 +1995,7 @@ static int nfs_volume_list_show(struct seq_file *m, void *v)
 		 (unsigned long long) server->fsid.major,
 		 (unsigned long long) server->fsid.minor);
 
+	rcu_read_lock();
 	seq_printf(m, "v%u %s %s %-7s %-17s %s\n",
 		   clp->rpc_ops->version,
 		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
@@ -1998,6 +2003,7 @@ static int nfs_volume_list_show(struct seq_file *m, void *v)
 		   dev,
 		   fsid,
 		   nfs_server_fscache_state(server));
+	rcu_read_unlock();
 
 	return 0;
 }
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index 667ea74..9c8eca3 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -96,8 +96,8 @@ static int nfs4_validate_fspath(struct dentry *dentry,
 static size_t nfs_parse_server_name(char *string, size_t len,
 		struct sockaddr *sa, size_t salen, struct nfs_server *server)
 {
+	struct net *net = rpc_net_ns(server->client);
 	ssize_t ret;
-	struct net *net = server->client->cl_xprt->xprt_net;
 
 	ret = rpc_pton(net, string, len, sa, salen);
 	if (ret == 0) {
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 53daab9..3ec09fb 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3833,6 +3833,7 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 	*p = htonl((u32)clp->cl_boot_time.tv_nsec);
 
 	for(;;) {
+		rcu_read_lock();
 		setclientid.sc_name_len = scnprintf(setclientid.sc_name,
 				sizeof(setclientid.sc_name), "%s/%s %s %s %u",
 				clp->cl_ipaddr,
@@ -3849,6 +3850,7 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 		setclientid.sc_uaddr_len = scnprintf(setclientid.sc_uaddr,
 				sizeof(setclientid.sc_uaddr), "%s.%u.%u",
 				clp->cl_ipaddr, port >> 8, port & 255);
+		rcu_read_unlock();
 
 		status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 		if (status != -NFS4ERR_CLID_INUSE)
@@ -5220,11 +5222,16 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp)
 
 void nfs4_destroy_session(struct nfs4_session *session)
 {
+	struct rpc_xprt *xprt;
+
 	nfs4_proc_destroy_session(session);
+
+	rcu_read_lock();
+	xprt = rcu_dereference(session->clp->cl_rpcclient->cl_xprt);
+	rcu_read_unlock();
 	dprintk("%s Destroy backchannel for xprt %p\n",
-		__func__, session->clp->cl_rpcclient->cl_xprt);
-	xprt_destroy_backchannel(session->clp->cl_rpcclient->cl_xprt,
-				NFS41_BC_MIN_CALLBACKS);
+		__func__, xprt);
+	xprt_destroy_backchannel(xprt, NFS41_BC_MIN_CALLBACKS);
 	nfs4_destroy_slot_tables(session);
 	kfree(session);
 }
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index c1111a3..bae959e 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1037,19 +1037,28 @@ static void nfs4_clear_state_manager_bit(struct nfs_client *clp)
 void nfs4_schedule_state_manager(struct nfs_client *clp)
 {
 	struct task_struct *task;
+	char buf[INET6_ADDRSTRLEN + sizeof("-manager") + 1];
 
 	if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0)
 		return;
 	__module_get(THIS_MODULE);
 	atomic_inc(&clp->cl_count);
-	task = kthread_run(nfs4_run_state_manager, clp, "%s-manager",
-				rpc_peeraddr2str(clp->cl_rpcclient,
-							RPC_DISPLAY_ADDR));
-	if (!IS_ERR(task))
-		return;
-	nfs4_clear_state_manager_bit(clp);
-	nfs_put_client(clp);
-	module_put(THIS_MODULE);
+
+	/* The rcu_read_lock() is not strictly necessary, as the state
+	 * manager is the only thread that ever changes the rpc_xprt
+	 * after it's initialized.  At this point, we're single threaded. */
+	rcu_read_lock();
+	snprintf(buf, sizeof(buf), "%s-manager",
+			rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
+	rcu_read_unlock();
+	task = kthread_run(nfs4_run_state_manager, clp, buf);
+	if (IS_ERR(task)) {
+		printk(KERN_ERR "%s: kthread_run: %ld\n",
+			__func__, PTR_ERR(task));
+		nfs4_clear_state_manager_bit(clp);
+		nfs_put_client(clp);
+		module_put(THIS_MODULE);
+	}
 }
 
 /*
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 6708f30..72c7f63 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -53,6 +53,7 @@
 #include <linux/magic.h>
 #include <linux/parser.h>
 #include <linux/nsproxy.h>
+#include <linux/rcupdate.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -679,8 +680,10 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
 		else
 			seq_puts(m, nfs_infop->nostr);
 	}
+	rcu_read_lock();
 	seq_printf(m, ",proto=%s",
 		   rpc_peeraddr2str(nfss->client, RPC_DISPLAY_NETID));
+	rcu_read_unlock();
 	if (version == 4) {
 		if (nfss->port != NFS_PORT)
 			seq_printf(m, ",port=%u", nfss->port);
@@ -729,9 +732,11 @@ static int nfs_show_options(struct seq_file *m, struct dentry *root)
 
 	nfs_show_mount_options(m, nfss, 0);
 
+	rcu_read_lock();
 	seq_printf(m, ",addr=%s",
 			rpc_peeraddr2str(nfss->nfs_client->cl_rpcclient,
 							RPC_DISPLAY_ADDR));
+	rcu_read_unlock();
 
 	return 0;
 }
diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index a4c62e9..e3d12b4 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -35,7 +35,7 @@ struct rpc_clnt {
 	struct list_head	cl_clients;	/* Global list of clients */
 	struct list_head	cl_tasks;	/* List of tasks */
 	spinlock_t		cl_lock;	/* spinlock */
-	struct rpc_xprt *	cl_xprt;	/* transport */
+	struct rpc_xprt __rcu *	cl_xprt;	/* transport */
 	struct rpc_procinfo *	cl_procinfo;	/* procedure info */
 	u32			cl_prog,	/* RPC program number */
 				cl_vers,	/* RPC version number */
@@ -156,6 +156,8 @@ struct rpc_task *rpc_call_null(struct rpc_clnt *clnt, struct rpc_cred *cred,
 int		rpc_restart_call_prepare(struct rpc_task *);
 int		rpc_restart_call(struct rpc_task *);
 void		rpc_setbufsize(struct rpc_clnt *, unsigned int, unsigned int);
+int		rpc_protocol(struct rpc_clnt *);
+struct net *	rpc_net_ns(struct rpc_clnt *);
 size_t		rpc_max_payload(struct rpc_clnt *);
 void		rpc_force_rebind(struct rpc_clnt *);
 size_t		rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t);
diff --git a/include/linux/sunrpc/debug.h b/include/linux/sunrpc/debug.h
index b506936..6cb2517 100644
--- a/include/linux/sunrpc/debug.h
+++ b/include/linux/sunrpc/debug.h
@@ -50,19 +50,32 @@ extern unsigned int		nlm_debug;
 #endif
 
 #define dprintk(args...)	dfprintk(FACILITY, ## args)
+#define dprintk_rcu(args...)	dfprintk_rcu(FACILITY, ## args)
 
 #undef ifdebug
 #ifdef RPC_DEBUG			
 # define ifdebug(fac)		if (unlikely(rpc_debug & RPCDBG_##fac))
+
 # define dfprintk(fac, args...)	\
 	do { \
 		ifdebug(fac) \
 			printk(KERN_DEFAULT args); \
 	} while (0)
+
+# define dfprintk_rcu(fac, args...)	\
+	do { \
+		ifdebug(fac) { \
+			rcu_read_lock(); \
+			printk(KERN_DEFAULT args); \
+			rcu_read_unlock(); \
+		} \
+	} while (0)
+
 # define RPC_IFDEBUG(x)		x
 #else
 # define ifdebug(fac)		if (0)
 # define dfprintk(fac, args...)	do ; while (0)
+# define dfprintk_rcu(fac, args...)	do ; while (0)
 # define RPC_IFDEBUG(x)
 #endif
 
diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c
index cb2e564..d3ad81f 100644
--- a/net/sunrpc/auth_gss/auth_gss.c
+++ b/net/sunrpc/auth_gss/auth_gss.c
@@ -799,7 +799,7 @@ err_unlink_pipe_1:
 static void gss_pipes_dentries_destroy_net(struct rpc_clnt *clnt,
 					   struct rpc_auth *auth)
 {
-	struct net *net = clnt->cl_xprt->xprt_net;
+	struct net *net = rpc_net_ns(clnt);
 	struct super_block *sb;
 
 	sb = rpc_get_sb_net(net);
@@ -813,7 +813,7 @@ static void gss_pipes_dentries_destroy_net(struct rpc_clnt *clnt,
 static int gss_pipes_dentries_create_net(struct rpc_clnt *clnt,
 					 struct rpc_auth *auth)
 {
-	struct net *net = clnt->cl_xprt->xprt_net;
+	struct net *net = rpc_net_ns(clnt);
 	struct super_block *sb;
 	int err = 0;
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index bb7ed2f3..f043c7b 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -31,6 +31,7 @@
 #include <linux/in.h>
 #include <linux/in6.h>
 #include <linux/un.h>
+#include <linux/rcupdate.h>
 
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/rpc_pipe_fs.h>
@@ -81,7 +82,8 @@ static int	rpc_ping(struct rpc_clnt *clnt);
 
 static void rpc_register_client(struct rpc_clnt *clnt)
 {
-	struct sunrpc_net *sn = net_generic(clnt->cl_xprt->xprt_net, sunrpc_net_id);
+	struct net *net = rpc_net_ns(clnt);
+	struct sunrpc_net *sn = net_generic(net, sunrpc_net_id);
 
 	spin_lock(&sn->rpc_client_lock);
 	list_add(&clnt->cl_clients, &sn->all_clients);
@@ -90,7 +92,8 @@ static void rpc_register_client(struct rpc_clnt *clnt)
 
 static void rpc_unregister_client(struct rpc_clnt *clnt)
 {
-	struct sunrpc_net *sn = net_generic(clnt->cl_xprt->xprt_net, sunrpc_net_id);
+	struct net *net = rpc_net_ns(clnt);
+	struct sunrpc_net *sn = net_generic(net, sunrpc_net_id);
 
 	spin_lock(&sn->rpc_client_lock);
 	list_del(&clnt->cl_clients);
@@ -109,12 +112,13 @@ static void __rpc_clnt_remove_pipedir(struct rpc_clnt *clnt)
 
 static void rpc_clnt_remove_pipedir(struct rpc_clnt *clnt)
 {
+	struct net *net = rpc_net_ns(clnt);
 	struct super_block *pipefs_sb;
 
-	pipefs_sb = rpc_get_sb_net(clnt->cl_xprt->xprt_net);
+	pipefs_sb = rpc_get_sb_net(net);
 	if (pipefs_sb) {
 		__rpc_clnt_remove_pipedir(clnt);
-		rpc_put_sb_net(clnt->cl_xprt->xprt_net);
+		rpc_put_sb_net(net);
 	}
 }
 
@@ -155,17 +159,18 @@ static struct dentry *rpc_setup_pipedir_sb(struct super_block *sb,
 static int
 rpc_setup_pipedir(struct rpc_clnt *clnt, const char *dir_name)
 {
+	struct net *net = rpc_net_ns(clnt);
 	struct super_block *pipefs_sb;
 	struct dentry *dentry;
 
 	clnt->cl_dentry = NULL;
 	if (dir_name == NULL)
 		return 0;
-	pipefs_sb = rpc_get_sb_net(clnt->cl_xprt->xprt_net);
+	pipefs_sb = rpc_get_sb_net(net);
 	if (!pipefs_sb)
 		return 0;
 	dentry = rpc_setup_pipedir_sb(pipefs_sb, clnt, dir_name);
-	rpc_put_sb_net(clnt->cl_xprt->xprt_net);
+	rpc_put_sb_net(net);
 	if (IS_ERR(dentry))
 		return PTR_ERR(dentry);
 	clnt->cl_dentry = dentry;
@@ -279,7 +284,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru
 	if (clnt->cl_server == NULL)
 		goto out_no_server;
 
-	clnt->cl_xprt     = xprt;
+	rcu_assign_pointer(clnt->cl_xprt, xprt);
 	clnt->cl_procinfo = version->procs;
 	clnt->cl_maxproc  = version->nrprocs;
 	clnt->cl_protname = program->name;
@@ -294,7 +299,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru
 	INIT_LIST_HEAD(&clnt->cl_tasks);
 	spin_lock_init(&clnt->cl_lock);
 
-	if (!xprt_bound(clnt->cl_xprt))
+	if (!xprt_bound(xprt))
 		clnt->cl_autobind = 1;
 
 	clnt->cl_timeout = xprt->timeout;
@@ -461,6 +466,7 @@ struct rpc_clnt *
 rpc_clone_client(struct rpc_clnt *clnt)
 {
 	struct rpc_clnt *new;
+	struct rpc_xprt *xprt;
 	int err = -ENOMEM;
 
 	new = kmemdup(clnt, sizeof(*new), GFP_KERNEL);
@@ -483,18 +489,25 @@ rpc_clone_client(struct rpc_clnt *clnt)
 		if (new->cl_principal == NULL)
 			goto out_no_principal;
 	}
+	rcu_read_lock();
+	xprt = xprt_get(rcu_dereference(clnt->cl_xprt));
+	rcu_read_unlock();
+	if (xprt == NULL)
+		goto out_no_transport;
+	rcu_assign_pointer(new->cl_xprt, xprt);
 	atomic_set(&new->cl_count, 1);
 	err = rpc_setup_pipedir(new, clnt->cl_program->pipe_dir_name);
 	if (err != 0)
 		goto out_no_path;
 	if (new->cl_auth)
 		atomic_inc(&new->cl_auth->au_count);
-	xprt_get(clnt->cl_xprt);
 	atomic_inc(&clnt->cl_count);
 	rpc_register_client(new);
 	rpciod_up();
 	return new;
 out_no_path:
+	xprt_put(xprt);
+out_no_transport:
 	kfree(new->cl_principal);
 out_no_principal:
 	rpc_free_iostats(new->cl_metrics);
@@ -574,7 +587,7 @@ rpc_free_client(struct rpc_clnt *clnt)
 	rpc_free_iostats(clnt->cl_metrics);
 	kfree(clnt->cl_principal);
 	clnt->cl_metrics = NULL;
-	xprt_put(clnt->cl_xprt);
+	xprt_put(rcu_dereference_raw(clnt->cl_xprt));
 	rpciod_down();
 	kfree(clnt);
 }
@@ -863,13 +876,18 @@ EXPORT_SYMBOL_GPL(rpc_call_start);
 size_t rpc_peeraddr(struct rpc_clnt *clnt, struct sockaddr *buf, size_t bufsize)
 {
 	size_t bytes;
-	struct rpc_xprt *xprt = clnt->cl_xprt;
+	struct rpc_xprt *xprt;
+
+	rcu_read_lock();
+	xprt = rcu_dereference(clnt->cl_xprt);
 
-	bytes = sizeof(xprt->addr);
+	bytes = xprt->addrlen;
 	if (bytes > bufsize)
 		bytes = bufsize;
-	memcpy(buf, &clnt->cl_xprt->addr, bytes);
-	return xprt->addrlen;
+	memcpy(buf, &xprt->addr, bytes);
+	rcu_read_unlock();
+
+	return bytes;
 }
 EXPORT_SYMBOL_GPL(rpc_peeraddr);
 
@@ -878,11 +896,16 @@ EXPORT_SYMBOL_GPL(rpc_peeraddr);
  * @clnt: RPC client structure
  * @format: address format
  *
+ * NB: the lifetime of the memory referenced by the returned pointer is
+ * the same as the rpc_xprt itself.  As long as the caller uses this
+ * pointer, it must hold the RCU read lock.
  */
 const char *rpc_peeraddr2str(struct rpc_clnt *clnt,
 			     enum rpc_display_format_t format)
 {
-	struct rpc_xprt *xprt = clnt->cl_xprt;
+	struct rpc_xprt *xprt;
+
+	xprt = rcu_dereference(clnt->cl_xprt);
 
 	if (xprt->address_strings[format] != NULL)
 		return xprt->address_strings[format];
@@ -894,14 +917,51 @@ EXPORT_SYMBOL_GPL(rpc_peeraddr2str);
 void
 rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize)
 {
-	struct rpc_xprt *xprt = clnt->cl_xprt;
+	struct rpc_xprt *xprt;
+
+	rcu_read_lock();
+	xprt = rcu_dereference(clnt->cl_xprt);
 	if (xprt->ops->set_buffer_size)
 		xprt->ops->set_buffer_size(xprt, sndsize, rcvsize);
+	rcu_read_unlock();
 }
 EXPORT_SYMBOL_GPL(rpc_setbufsize);
 
-/*
- * Return size of largest payload RPC client can support, in bytes
+/**
+ * rpc_protocol - Get transport protocol number for an RPC client
+ * @clnt: RPC client to query
+ *
+ */
+int rpc_protocol(struct rpc_clnt *clnt)
+{
+	int protocol;
+
+	rcu_read_lock();
+	protocol = rcu_dereference(clnt->cl_xprt)->prot;
+	rcu_read_unlock();
+	return protocol;
+}
+EXPORT_SYMBOL_GPL(rpc_protocol);
+
+/**
+ * rpc_net_ns - Get the network namespace for this RPC client
+ * @clnt: RPC client to query
+ *
+ */
+struct net *rpc_net_ns(struct rpc_clnt *clnt)
+{
+	struct net *ret;
+
+	rcu_read_lock();
+	ret = rcu_dereference(clnt->cl_xprt)->xprt_net;
+	rcu_read_unlock();
+	return ret;
+}
+EXPORT_SYMBOL_GPL(rpc_net_ns);
+
+/**
+ * rpc_max_payload - Get maximum payload size for a transport, in bytes
+ * @clnt: RPC client to query
  *
  * For stream transports, this is one RPC record fragment (see RFC
  * 1831), as we don't support multi-record requests yet.  For datagram
@@ -910,7 +970,12 @@ EXPORT_SYMBOL_GPL(rpc_setbufsize);
  */
 size_t rpc_max_payload(struct rpc_clnt *clnt)
 {
-	return clnt->cl_xprt->max_payload;
+	size_t ret;
+
+	rcu_read_lock();
+	ret = rcu_dereference(clnt->cl_xprt)->max_payload;
+	rcu_read_unlock();
+	return ret;
 }
 EXPORT_SYMBOL_GPL(rpc_max_payload);
 
@@ -921,8 +986,11 @@ EXPORT_SYMBOL_GPL(rpc_max_payload);
  */
 void rpc_force_rebind(struct rpc_clnt *clnt)
 {
-	if (clnt->cl_autobind)
-		xprt_clear_bound(clnt->cl_xprt);
+	if (clnt->cl_autobind) {
+		rcu_read_lock();
+		xprt_clear_bound(rcu_dereference(clnt->cl_xprt));
+		rcu_read_unlock();
+	}
 }
 EXPORT_SYMBOL_GPL(rpc_force_rebind);
 
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 6873c9b..bbd636b 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -16,6 +16,7 @@
 #include <linux/namei.h>
 #include <linux/fsnotify.h>
 #include <linux/kernel.h>
+#include <linux/rcupdate.h>
 
 #include <asm/ioctls.h>
 #include <linux/poll.h>
@@ -383,12 +384,14 @@ rpc_show_info(struct seq_file *m, void *v)
 {
 	struct rpc_clnt *clnt = m->private;
 
+	rcu_read_lock();
 	seq_printf(m, "RPC server: %s\n", clnt->cl_server);
 	seq_printf(m, "service: %s (%d) version %d\n", clnt->cl_protname,
 			clnt->cl_prog, clnt->cl_vers);
 	seq_printf(m, "address: %s\n", rpc_peeraddr2str(clnt, RPC_DISPLAY_ADDR));
 	seq_printf(m, "protocol: %s\n", rpc_peeraddr2str(clnt, RPC_DISPLAY_PROTO));
 	seq_printf(m, "port: %s\n", rpc_peeraddr2str(clnt, RPC_DISPLAY_PORT));
+	rcu_read_unlock();
 	return 0;
 }
 
diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c
index b1f08bd..4f8af63 100644
--- a/net/sunrpc/rpcb_clnt.c
+++ b/net/sunrpc/rpcb_clnt.c
@@ -620,9 +620,10 @@ static struct rpc_task *rpcb_call_async(struct rpc_clnt *rpcb_clnt, struct rpcbi
 static struct rpc_clnt *rpcb_find_transport_owner(struct rpc_clnt *clnt)
 {
 	struct rpc_clnt *parent = clnt->cl_parent;
+	struct rpc_xprt *xprt = rcu_dereference(clnt->cl_xprt);
 
 	while (parent != clnt) {
-		if (parent->cl_xprt != clnt->cl_xprt)
+		if (rcu_dereference(parent->cl_xprt) != xprt)
 			break;
 		if (clnt->cl_autobind)
 			break;
@@ -653,8 +654,12 @@ void rpcb_getport_async(struct rpc_task *task)
 	size_t salen;
 	int status;
 
-	clnt = rpcb_find_transport_owner(task->tk_client);
-	xprt = clnt->cl_xprt;
+	rcu_read_lock();
+	do {
+		clnt = rpcb_find_transport_owner(task->tk_client);
+		xprt = xprt_get(rcu_dereference(clnt->cl_xprt));
+	} while (xprt == NULL);
+	rcu_read_unlock();
 
 	dprintk("RPC: %5u %s(%s, %u, %u, %d)\n",
 		task->tk_pid, __func__,
@@ -667,6 +672,7 @@ void rpcb_getport_async(struct rpc_task *task)
 	if (xprt_test_and_set_binding(xprt)) {
 		dprintk("RPC: %5u %s: waiting for another binder\n",
 			task->tk_pid, __func__);
+		xprt_put(xprt);
 		return;
 	}
 
@@ -734,7 +740,7 @@ void rpcb_getport_async(struct rpc_task *task)
 	switch (bind_version) {
 	case RPCBVERS_4:
 	case RPCBVERS_3:
-		map->r_netid = rpc_peeraddr2str(clnt, RPC_DISPLAY_NETID);
+		map->r_netid = xprt->address_strings[RPC_DISPLAY_NETID];
 		map->r_addr = rpc_sockaddr2uaddr(sap, GFP_ATOMIC);
 		map->r_owner = "";
 		break;
@@ -763,6 +769,7 @@ bailout_release_client:
 bailout_nofree:
 	rpcb_wake_rpcbind_waiters(xprt, status);
 	task->tk_status = status;
+	xprt_put(xprt);
 }
 EXPORT_SYMBOL_GPL(rpcb_getport_async);
 
diff --git a/net/sunrpc/stats.c b/net/sunrpc/stats.c
index 1eb3304..bc2068e 100644
--- a/net/sunrpc/stats.c
+++ b/net/sunrpc/stats.c
@@ -22,6 +22,7 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/svcsock.h>
 #include <linux/sunrpc/metrics.h>
+#include <linux/rcupdate.h>
 
 #include "netns.h"
 
@@ -179,7 +180,7 @@ static void _print_name(struct seq_file *seq, unsigned int op,
 void rpc_print_iostats(struct seq_file *seq, struct rpc_clnt *clnt)
 {
 	struct rpc_iostats *stats = clnt->cl_metrics;
-	struct rpc_xprt *xprt = clnt->cl_xprt;
+	struct rpc_xprt *xprt;
 	unsigned int op, maxproc = clnt->cl_maxproc;
 
 	if (!stats)
@@ -189,8 +190,11 @@ void rpc_print_iostats(struct seq_file *seq, struct rpc_clnt *clnt)
 	seq_printf(seq, "p/v: %u/%u (%s)\n",
 			clnt->cl_prog, clnt->cl_vers, clnt->cl_protname);
 
+	rcu_read_lock();
+	xprt = rcu_dereference(clnt->cl_xprt);
 	if (xprt)
 		xprt->ops->print_stats(xprt, seq);
+	rcu_read_unlock();
 
 	seq_printf(seq, "\tper-op statistics\n");
 	for (op = 0; op < maxproc; op++) {

[PATCH 06/12] SUNRPC: Move clnt->cl_server into struct rpc_xprt

[Chuck Lever]

From: Trond Myklebust <Trond.Myklebust@netapp.com>

When the cl_xprt field is updated, the cl_server field will also have
to change.  Since the contents of cl_server follow the remote endpoint
of cl_xprt, just move that field to the rpc_xprt.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
[ cel: simplify check_gss_callback_principal(), whitespace changes ]
[ cel: forward ported to 3.4 ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---

 fs/nfs/callback.c           |    3 -
 fs/nfs/nfs4proc.c           |    3 +
 include/linux/sunrpc/clnt.h |    1 
 include/linux/sunrpc/xprt.h |    2 +
 net/sunrpc/clnt.c           |   88 +++++++++++++++++++++----------------------
 net/sunrpc/rpc_pipe.c       |    3 +
 net/sunrpc/rpcb_clnt.c      |    4 +-
 net/sunrpc/xprt.c           |   15 +++++++
 8 files changed, 66 insertions(+), 53 deletions(-)

diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 4a122ae..2afe233 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -332,7 +332,6 @@ void nfs_callback_down(int minorversion)
 int
 check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp)
 {
-	struct rpc_clnt *r = clp->cl_rpcclient;
 	char *p = svc_gss_principal(rqstp);
 
 	if (rqstp->rq_authop->flavour != RPC_AUTH_GSS)
@@ -353,7 +352,7 @@ check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp)
 	if (memcmp(p, "nfs@", 4) != 0)
 		return 0;
 	p += 4;
-	if (strcmp(p, r->cl_server) != 0)
+	if (strcmp(p, clp->cl_hostname) != 0)
 		return 0;
 	return 1;
 }
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 3ec09fb..d43bcda 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1100,6 +1100,7 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data
 	if (state == NULL)
 		goto err_put_inode;
 	if (data->o_res.delegation_type != 0) {
+		struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 		int delegation_flags = 0;
 
 		rcu_read_lock();
@@ -1111,7 +1112,7 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data
 			pr_err_ratelimited("NFS: Broken NFSv4 server %s is "
 					"returning a delegation for "
 					"OPEN(CLAIM_DELEGATE_CUR)\n",
-					NFS_CLIENT(inode)->cl_server);
+					clp->cl_hostname);
 		} else if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0)
 			nfs_inode_set_delegation(state->inode,
 					data->owner->so_cred,
diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index e3d12b4..acd5502 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -41,7 +41,6 @@ struct rpc_clnt {
 				cl_vers,	/* RPC version number */
 				cl_maxproc;	/* max procedure number */
 
-	const char *		cl_server;	/* server machine name */
 	const char *		cl_protname;	/* protocol name */
 	struct rpc_auth *	cl_auth;	/* authenticator */
 	struct rpc_stat *	cl_stats;	/* per-program statistics */
diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h
index ea712f9..77d278d 100644
--- a/include/linux/sunrpc/xprt.h
+++ b/include/linux/sunrpc/xprt.h
@@ -229,6 +229,7 @@ struct rpc_xprt {
 	} stat;
 
 	struct net		*xprt_net;
+	const char		*servername;
 	const char		*address_strings[RPC_DISPLAY_MAX];
 };
 
@@ -258,6 +259,7 @@ struct xprt_create {
 	struct sockaddr *	srcaddr;	/* optional local address */
 	struct sockaddr *	dstaddr;	/* remote peer address */
 	size_t			addrlen;
+	const char		*servername;
 	struct svc_xprt		*bc_xprt;	/* NFSv4.1 backchannel */
 };
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index f043c7b..84bebb1 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -249,15 +249,8 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru
 	struct rpc_clnt		*clnt = NULL;
 	struct rpc_auth		*auth;
 	int err;
-	size_t len;
 
 	/* sanity check the name before trying to print it */
-	err = -EINVAL;
-	len = strlen(args->servername);
-	if (len > RPC_MAXNETNAMELEN)
-		goto out_no_rpciod;
-	len++;
-
 	dprintk("RPC:       creating %s client for %s (xprt %p)\n",
 			program->name, args->servername, xprt);
 
@@ -280,10 +273,6 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru
 		goto out_err;
 	clnt->cl_parent = clnt;
 
-	clnt->cl_server = kstrdup(args->servername, GFP_KERNEL);
-	if (clnt->cl_server == NULL)
-		goto out_no_server;
-
 	rcu_assign_pointer(clnt->cl_xprt, xprt);
 	clnt->cl_procinfo = version->procs;
 	clnt->cl_maxproc  = version->nrprocs;
@@ -347,8 +336,6 @@ out_no_path:
 out_no_principal:
 	rpc_free_iostats(clnt->cl_metrics);
 out_no_stats:
-	kfree(clnt->cl_server);
-out_no_server:
 	kfree(clnt);
 out_err:
 	xprt_put(xprt);
@@ -378,6 +365,7 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 		.srcaddr = args->saddress,
 		.dstaddr = args->address,
 		.addrlen = args->addrsize,
+		.servername = args->servername,
 		.bc_xprt = args->bc_xprt,
 	};
 	char servername[48];
@@ -386,7 +374,7 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 	 * If the caller chooses not to specify a hostname, whip
 	 * up a string representation of the passed-in address.
 	 */
-	if (args->servername == NULL) {
+	if (xprtargs.servername == NULL) {
 		struct sockaddr_un *sun =
 				(struct sockaddr_un *)args->address;
 		struct sockaddr_in *sin =
@@ -413,7 +401,7 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 			 * address family isn't recognized. */
 			return ERR_PTR(-EINVAL);
 		}
-		args->servername = servername;
+		xprtargs.servername = servername;
 	}
 
 	xprt = xprt_create_transport(&xprtargs);
@@ -472,9 +460,6 @@ rpc_clone_client(struct rpc_clnt *clnt)
 	new = kmemdup(clnt, sizeof(*new), GFP_KERNEL);
 	if (!new)
 		goto out_no_clnt;
-	new->cl_server = kstrdup(clnt->cl_server, GFP_KERNEL);
-	if (new->cl_server == NULL)
-		goto out_no_server;
 	new->cl_parent = clnt;
 	/* Turn off autobind on clones */
 	new->cl_autobind = 0;
@@ -512,8 +497,6 @@ out_no_transport:
 out_no_principal:
 	rpc_free_iostats(new->cl_metrics);
 out_no_stats:
-	kfree(new->cl_server);
-out_no_server:
 	kfree(new);
 out_no_clnt:
 	dprintk("RPC:       %s: returned error %d\n", __func__, err);
@@ -558,8 +541,9 @@ EXPORT_SYMBOL_GPL(rpc_killall_tasks);
  */
 void rpc_shutdown_client(struct rpc_clnt *clnt)
 {
-	dprintk("RPC:       shutting down %s client for %s\n",
-			clnt->cl_protname, clnt->cl_server);
+	dprintk_rcu("RPC:       shutting down %s client for %s\n",
+			clnt->cl_protname,
+			rcu_dereference(clnt->cl_xprt)->servername);
 
 	while (!list_empty(&clnt->cl_tasks)) {
 		rpc_killall_tasks(clnt);
@@ -577,11 +561,11 @@ EXPORT_SYMBOL_GPL(rpc_shutdown_client);
 static void
 rpc_free_client(struct rpc_clnt *clnt)
 {
-	dprintk("RPC:       destroying %s client for %s\n",
-			clnt->cl_protname, clnt->cl_server);
+	dprintk_rcu("RPC:       destroying %s client for %s\n",
+			clnt->cl_protname,
+			rcu_dereference(clnt->cl_xprt)->servername);
 	if (clnt->cl_parent != clnt)
 		rpc_release_client(clnt->cl_parent);
-	kfree(clnt->cl_server);
 	rpc_unregister_client(clnt);
 	rpc_clnt_remove_pipedir(clnt);
 	rpc_free_iostats(clnt->cl_metrics);
@@ -1669,8 +1653,11 @@ call_timeout(struct rpc_task *task)
 	}
 	if (RPC_IS_SOFT(task)) {
 		if (clnt->cl_chatty)
+			rcu_read_lock();
 			printk(KERN_NOTICE "%s: server %s not responding, timed out\n",
-				clnt->cl_protname, clnt->cl_server);
+				clnt->cl_protname,
+				rcu_dereference(clnt->cl_xprt)->servername);
+			rcu_read_unlock();
 		if (task->tk_flags & RPC_TASK_TIMEOUT)
 			rpc_exit(task, -ETIMEDOUT);
 		else
@@ -1680,9 +1667,13 @@ call_timeout(struct rpc_task *task)
 
 	if (!(task->tk_flags & RPC_CALL_MAJORSEEN)) {
 		task->tk_flags |= RPC_CALL_MAJORSEEN;
-		if (clnt->cl_chatty)
+		if (clnt->cl_chatty) {
+			rcu_read_lock();
 			printk(KERN_NOTICE "%s: server %s not responding, still trying\n",
-			clnt->cl_protname, clnt->cl_server);
+			clnt->cl_protname,
+			rcu_dereference(clnt->cl_xprt)->servername);
+			rcu_read_unlock();
+		}
 	}
 	rpc_force_rebind(clnt);
 	/*
@@ -1711,9 +1702,13 @@ call_decode(struct rpc_task *task)
 	dprint_status(task);
 
 	if (task->tk_flags & RPC_CALL_MAJORSEEN) {
-		if (clnt->cl_chatty)
+		if (clnt->cl_chatty) {
+			rcu_read_lock();
 			printk(KERN_NOTICE "%s: server %s OK\n",
-				clnt->cl_protname, clnt->cl_server);
+				clnt->cl_protname,
+				rcu_dereference(clnt->cl_xprt)->servername);
+			rcu_read_unlock();
+		}
 		task->tk_flags &= ~RPC_CALL_MAJORSEEN;
 	}
 
@@ -1791,6 +1786,7 @@ rpc_encode_header(struct rpc_task *task)
 static __be32 *
 rpc_verify_header(struct rpc_task *task)
 {
+	struct rpc_clnt *clnt = task->tk_client;
 	struct kvec *iov = &task->tk_rqstp->rq_rcv_buf.head[0];
 	int len = task->tk_rqstp->rq_rcv_buf.len >> 2;
 	__be32	*p = iov->iov_base;
@@ -1863,8 +1859,11 @@ rpc_verify_header(struct rpc_task *task)
 			task->tk_action = call_bind;
 			goto out_retry;
 		case RPC_AUTH_TOOWEAK:
+			rcu_read_lock();
 			printk(KERN_NOTICE "RPC: server %s requires stronger "
-			       "authentication.\n", task->tk_client->cl_server);
+			       "authentication.\n",
+			       rcu_dereference(clnt->cl_xprt)->servername);
+			rcu_read_unlock();
 			break;
 		default:
 			dprintk("RPC: %5u %s: unknown auth error: %x\n",
@@ -1887,28 +1886,27 @@ rpc_verify_header(struct rpc_task *task)
 	case RPC_SUCCESS:
 		return p;
 	case RPC_PROG_UNAVAIL:
-		dprintk("RPC: %5u %s: program %u is unsupported by server %s\n",
-				task->tk_pid, __func__,
-				(unsigned int)task->tk_client->cl_prog,
-				task->tk_client->cl_server);
+		dprintk_rcu("RPC: %5u %s: program %u is unsupported "
+				"by server %s\n", task->tk_pid, __func__,
+				(unsigned int)clnt->cl_prog,
+				rcu_dereference(clnt->cl_xprt)->servername);
 		error = -EPFNOSUPPORT;
 		goto out_err;
 	case RPC_PROG_MISMATCH:
-		dprintk("RPC: %5u %s: program %u, version %u unsupported by "
-				"server %s\n", task->tk_pid, __func__,
-				(unsigned int)task->tk_client->cl_prog,
-				(unsigned int)task->tk_client->cl_vers,
-				task->tk_client->cl_server);
+		dprintk_rcu("RPC: %5u %s: program %u, version %u unsupported "
+				"by server %s\n", task->tk_pid, __func__,
+				(unsigned int)clnt->cl_prog,
+				(unsigned int)clnt->cl_vers,
+				rcu_dereference(clnt->cl_xprt)->servername);
 		error = -EPROTONOSUPPORT;
 		goto out_err;
 	case RPC_PROC_UNAVAIL:
-		dprintk("RPC: %5u %s: proc %s unsupported by program %u, "
+		dprintk_rcu("RPC: %5u %s: proc %s unsupported by program %u, "
 				"version %u on server %s\n",
 				task->tk_pid, __func__,
 				rpc_proc_name(task),
-				task->tk_client->cl_prog,
-				task->tk_client->cl_vers,
-				task->tk_client->cl_server);
+				clnt->cl_prog, clnt->cl_vers,
+				rcu_dereference(clnt->cl_xprt)->servername);
 		error = -EOPNOTSUPP;
 		goto out_err;
 	case RPC_GARBAGE_ARGS:
@@ -1922,7 +1920,7 @@ rpc_verify_header(struct rpc_task *task)
 	}
 
 out_garbage:
-	task->tk_client->cl_stats->rpcgarbage++;
+	clnt->cl_stats->rpcgarbage++;
 	if (task->tk_garb_retry) {
 		task->tk_garb_retry--;
 		dprintk("RPC: %5u %s: retrying\n",
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index bbd636b..3c9b16f 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -385,7 +385,8 @@ rpc_show_info(struct seq_file *m, void *v)
 	struct rpc_clnt *clnt = m->private;
 
 	rcu_read_lock();
-	seq_printf(m, "RPC server: %s\n", clnt->cl_server);
+	seq_printf(m, "RPC server: %s\n",
+			rcu_dereference(clnt->cl_xprt)->servername);
 	seq_printf(m, "service: %s (%d) version %d\n", clnt->cl_protname,
 			clnt->cl_prog, clnt->cl_vers);
 	seq_printf(m, "address: %s\n", rpc_peeraddr2str(clnt, RPC_DISPLAY_ADDR));
diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c
index 4f8af63..e699ff0 100644
--- a/net/sunrpc/rpcb_clnt.c
+++ b/net/sunrpc/rpcb_clnt.c
@@ -663,7 +663,7 @@ void rpcb_getport_async(struct rpc_task *task)
 
 	dprintk("RPC: %5u %s(%s, %u, %u, %d)\n",
 		task->tk_pid, __func__,
-		clnt->cl_server, clnt->cl_prog, clnt->cl_vers, xprt->prot);
+		xprt->servername, clnt->cl_prog, clnt->cl_vers, xprt->prot);
 
 	/* Put self on the wait queue to ensure we get notified if
 	 * some other task is already attempting to bind the port */
@@ -714,7 +714,7 @@ void rpcb_getport_async(struct rpc_task *task)
 	dprintk("RPC: %5u %s: trying rpcbind version %u\n",
 		task->tk_pid, __func__, bind_version);
 
-	rpcb_clnt = rpcb_create(xprt->xprt_net, clnt->cl_server, sap, salen,
+	rpcb_clnt = rpcb_create(xprt->xprt_net, xprt->servername, sap, salen,
 				xprt->prot, bind_version);
 	if (IS_ERR(rpcb_clnt)) {
 		status = PTR_ERR(rpcb_clnt);
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 32e3794..0cbcd1a 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -66,6 +66,7 @@ static void	 xprt_init(struct rpc_xprt *xprt, struct net *net);
 static void	xprt_request_init(struct rpc_task *, struct rpc_xprt *);
 static void	xprt_connect_status(struct rpc_task *task);
 static int      __xprt_get_cong(struct rpc_xprt *, struct rpc_task *);
+static void	 xprt_destroy(struct rpc_xprt *xprt);
 
 static DEFINE_SPINLOCK(xprt_list_lock);
 static LIST_HEAD(xprt_list);
@@ -751,7 +752,7 @@ static void xprt_connect_status(struct rpc_task *task)
 	default:
 		dprintk("RPC: %5u xprt_connect_status: error %d connecting to "
 				"server %s\n", task->tk_pid, -task->tk_status,
-				task->tk_client->cl_server);
+				xprt->servername);
 		xprt_release_write(xprt, task);
 		task->tk_status = -EIO;
 	}
@@ -1229,6 +1230,17 @@ found:
 			    (unsigned long)xprt);
 	else
 		init_timer(&xprt->timer);
+
+	if (strlen(args->servername) > RPC_MAXNETNAMELEN) {
+		xprt_destroy(xprt);
+		return ERR_PTR(-EINVAL);
+	}
+	xprt->servername = kstrdup(args->servername, GFP_KERNEL);
+	if (xprt->servername == NULL) {
+		xprt_destroy(xprt);
+		return ERR_PTR(-ENOMEM);
+	}
+
 	dprintk("RPC:       created transport %p with %u slots\n", xprt,
 			xprt->max_reqs);
 out:
@@ -1251,6 +1263,7 @@ static void xprt_destroy(struct rpc_xprt *xprt)
 	rpc_destroy_wait_queue(&xprt->sending);
 	rpc_destroy_wait_queue(&xprt->backlog);
 	cancel_work_sync(&xprt->task_cleanup);
+	kfree(xprt->servername);
 	/*
 	 * Tear down transport state and free the rpc_xprt
 	 */

[PATCH 07/12] SUNRPC: Add API to acquire source address

[Chuck Lever]

NFSv4.0 clients must send endpoint information for their callback
service to NFSv4.0 servers during their first contact with a server.
Traditionally on Linux, user space provides the callback endpoint IP
address via the "clientaddr=" mount option.

During an NFSv4 migration event, it is possible that an FSID may be
migrated to a destination server that is accessible via a different
source IP address than the source server was.  The client must update
callback endpoint information on the destination server so that it can
maintain leases and allow delegation.

Without a new "clientaddr=" option from user space, however, the
kernel itself must construct an appropriate IP address for the
callback update.  Provide an API in the RPC client for upper layer
RPC consumers to acquire a source address for a remote.

The mechanism used by the mount.nfs command is copied: set up a
connected UDP socket to the designated remote, then scrape the source
address off the socket.  We are careful to select the correct network
namespace when setting up the temporary UDP socket.

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

 include/linux/sunrpc/clnt.h |    1 
 net/sunrpc/clnt.c           |  149 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 150 insertions(+), 0 deletions(-)

diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index acd5502..523547e 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -161,6 +161,7 @@ size_t		rpc_max_payload(struct rpc_clnt *);
 void		rpc_force_rebind(struct rpc_clnt *);
 size_t		rpc_peeraddr(struct rpc_clnt *, struct sockaddr *, size_t);
 const char	*rpc_peeraddr2str(struct rpc_clnt *, enum rpc_display_format_t);
+int		rpc_localaddr(struct rpc_clnt *, struct sockaddr *, size_t);
 
 size_t		rpc_ntop(const struct sockaddr *, char *, const size_t);
 size_t		rpc_pton(struct net *, const char *, const size_t,
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 84bebb1..064e55bb 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -898,6 +898,155 @@ const char *rpc_peeraddr2str(struct rpc_clnt *clnt,
 }
 EXPORT_SYMBOL_GPL(rpc_peeraddr2str);
 
+static const struct sockaddr_in rpc_inaddr_loopback = {
+	.sin_family		= AF_INET,
+	.sin_addr.s_addr	= htonl(INADDR_ANY),
+};
+
+static const struct sockaddr_in6 rpc_in6addr_loopback = {
+	.sin6_family		= AF_INET6,
+	.sin6_addr		= IN6ADDR_ANY_INIT,
+};
+
+/*
+ * Try a getsockname() on a connected datagram socket.  Using a
+ * connected datagram socket prevents leaving a socket in TIME_WAIT.
+ * This conserves the ephemeral port number space.
+ *
+ * Returns zero and fills in "buf" if successful; otherwise, a
+ * negative errno is returned.
+ */
+static int rpc_sockname(struct net *net, struct sockaddr *sap, size_t salen,
+			struct sockaddr *buf, int buflen)
+{
+	struct socket *sock;
+	int err;
+
+	err = __sock_create(net, sap->sa_family,
+				SOCK_DGRAM, IPPROTO_UDP, &sock, 1);
+	if (err < 0) {
+		dprintk("RPC:       can't create UDP socket (%d)\n", err);
+		goto out;
+	}
+
+	switch (sap->sa_family) {
+	case AF_INET:
+		err = kernel_bind(sock,
+				(struct sockaddr *)&rpc_inaddr_loopback,
+				sizeof(rpc_inaddr_loopback));
+		break;
+	case AF_INET6:
+		err = kernel_bind(sock,
+				(struct sockaddr *)&rpc_in6addr_loopback,
+				sizeof(rpc_in6addr_loopback));
+		break;
+	default:
+		err = -EAFNOSUPPORT;
+		goto out;
+	}
+	if (err < 0) {
+		dprintk("RPC:       can't bind UDP socket (%d)\n", err);
+		goto out_release;
+	}
+
+	err = kernel_connect(sock, sap, salen, 0);
+	if (err < 0) {
+		dprintk("RPC:       can't connect UDP socket (%d)\n", err);
+		goto out_release;
+	}
+
+	err = kernel_getsockname(sock, buf, &buflen);
+	if (err < 0) {
+		dprintk("RPC:       getsockname failed (%d)\n", err);
+		goto out_release;
+	}
+
+	err = 0;
+	if (buf->sa_family == AF_INET6) {
+		struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)buf;
+		sin6->sin6_scope_id = 0;
+	}
+	dprintk("RPC:       %s succeeded\n", __func__);
+
+out_release:
+	sock_release(sock);
+out:
+	return err;
+}
+
+/*
+ * Scraping a connected socket failed, so we don't have a useable
+ * local address.  Fallback: generate an address that will prevent
+ * the server from calling us back.
+ *
+ * Returns zero and fills in "buf" if successful; otherwise, a
+ * negative errno is returned.
+ */
+static int rpc_anyaddr(int family, struct sockaddr *buf, size_t buflen)
+{
+	switch (family) {
+	case AF_INET:
+		if (buflen < sizeof(rpc_inaddr_loopback))
+			return -EINVAL;
+		memcpy(buf, &rpc_inaddr_loopback,
+				sizeof(rpc_inaddr_loopback));
+		break;
+	case AF_INET6:
+		if (buflen < sizeof(rpc_in6addr_loopback))
+			return -EINVAL;
+		memcpy(buf, &rpc_in6addr_loopback,
+				sizeof(rpc_in6addr_loopback));
+	default:
+		dprintk("RPC:       %s: address family not supported\n",
+			__func__);
+		return -EAFNOSUPPORT;
+	}
+	dprintk("RPC:       %s: succeeded\n", __func__);
+	return 0;
+}
+
+/**
+ * rpc_localaddr - discover local endpoint address for an RPC client
+ * @clnt: RPC client structure
+ * @buf: target buffer
+ * @buflen: size of target buffer, in bytes
+ *
+ * Returns zero and fills in "buf" and "buflen" if successful;
+ * otherwise, a negative errno is returned.
+ *
+ * This works even if the underlying transport is not currently connected,
+ * or if the upper layer never previously provided a source address.
+ *
+ * The result of this function call is transient: multiple calls in
+ * succession may give different results, depending on how local
+ * networking configuration changes over time.
+ */
+int rpc_localaddr(struct rpc_clnt *clnt, struct sockaddr *buf, size_t buflen)
+{
+	struct sockaddr_storage address;
+	struct sockaddr *sap = (struct sockaddr *)&address;
+	struct rpc_xprt *xprt;
+	struct net *net;
+	size_t salen;
+	int err;
+
+	rcu_read_lock();
+	xprt = rcu_dereference(clnt->cl_xprt);
+	salen = xprt->addrlen;
+	memcpy(sap, &xprt->addr, salen);
+	net = get_net(xprt->xprt_net);
+	rcu_read_unlock();
+
+	rpc_set_port(sap, 0);
+	err = rpc_sockname(net, sap, salen, buf, buflen);
+	put_net(net);
+	if (err != 0)
+		/* Couldn't discover local address, return ANYADDR */
+		return rpc_anyaddr(sap->sa_family, buf, buflen);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rpc_localaddr);
+
 void
 rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize)
 {

[PATCH 08/12] commit 6f38b4ba433ac6494f83cb73dd07dcbde797e1e0

[Chuck Lever]

NFS: Make clientaddr= optional

    For NFSv4 mounts, the clientaddr= mount option has always been
    required.  Now we have rpc_localaddr() in the kernel, which was
    modeled after the same logic in the mount.nfs command that constructs
    the clientaddr= mount option.  If user space doesn't provide a
    clientaddr= mount option, the kernel can now construct its own.

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

 fs/nfs/client.c |   15 +++++++++++++++
 fs/nfs/super.c  |    6 ------
 2 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 994899d..add92d9 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1344,6 +1344,7 @@ int nfs4_init_client(struct nfs_client *clp,
 		     rpc_authflavor_t authflavour,
 		     int noresvport)
 {
+	char buf[INET6_ADDRSTRLEN + 1];
 	int error;
 
 	if (clp->cl_cons_state == NFS_CS_READY) {
@@ -1359,6 +1360,20 @@ int nfs4_init_client(struct nfs_client *clp,
 				      1, noresvport);
 	if (error < 0)
 		goto error;
+
+	/* If no clientaddr= option was specified, find a usable cb address */
+	if (ip_addr == NULL) {
+		struct sockaddr_storage cb_addr;
+		struct sockaddr *sap = (struct sockaddr *)&cb_addr;
+
+		error = rpc_localaddr(clp->cl_rpcclient, sap, sizeof(cb_addr));
+		if (error < 0)
+			goto error;
+		error = rpc_ntop(sap, buf, sizeof(buf));
+		if (error < 0)
+			goto error;
+		ip_addr = (const char *)buf;
+	}
 	strlcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr));
 
 	error = nfs_idmap_new(clp);
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 72c7f63..8ae8fb2 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2526,12 +2526,6 @@ static int nfs4_validate_text_mount_data(void *options,
 		return -EINVAL;
 	}
 
-	if (args->client_address == NULL) {
-		dfprintk(MOUNT,
-			 "NFS4: mount program didn't pass callback address\n");
-		return -EINVAL;
-	}
-
 	return nfs_parse_devname(dev_name,
 				   &args->nfs_server.hostname,
 				   NFS4_MAXNAMLEN,

[PATCH 09/12] NFS: Add a client-side function to display NFS file handles

[Chuck Lever]

For debugging, introduce a simplistic function to print NFS file
handles on the system console.  The main function is hooked into the
dprintk debugging facility, but you can directly call the helper,
_nfs_display_fhandle(), if you want to print a handle unconditionally.

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

 fs/nfs/inode.c         |   45 +++++++++++++++++++++++++++++++++++++++++++++
 include/linux/nfs_fs.h |   14 ++++++++++++++
 2 files changed, 59 insertions(+), 0 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 6c66259..99a4f52 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1046,6 +1046,51 @@ struct nfs_fh *nfs_alloc_fhandle(void)
 }
 
 /**
+ * _nfs_display_fhandle - display an NFS file handle on the console
+ *
+ * @fh: file handle to display
+ * @caption: display caption
+ *
+ * For debugging only.
+ */
+#ifdef RPC_DEBUG
+void _nfs_display_fhandle(const struct nfs_fh *fh, const char *caption)
+{
+	unsigned short i;
+
+	if (fh->size == 0 || fh == NULL) {
+		printk(KERN_DEFAULT "%s at %p is empty\n", caption, fh);
+		return;
+	}
+
+	printk(KERN_DEFAULT "%s at %p is %u bytes:\n", caption, fh, fh->size);
+	for (i = 0; i < fh->size; i += 16) {
+		__be32 *pos = (__be32 *)&fh->data[i];
+
+		switch ((fh->size - i - 1) >> 2) {
+		case 0:
+			printk(KERN_DEFAULT " %08x\n",
+				be32_to_cpup(pos));
+			break;
+		case 1:
+			printk(KERN_DEFAULT " %08x %08x\n",
+				be32_to_cpup(pos), be32_to_cpup(pos + 1));
+			break;
+		case 2:
+			printk(KERN_DEFAULT " %08x %08x %08x\n",
+				be32_to_cpup(pos), be32_to_cpup(pos + 1),
+				be32_to_cpup(pos + 2));
+			break;
+		default:
+			printk(KERN_DEFAULT " %08x %08x %08x %08x\n",
+				be32_to_cpup(pos), be32_to_cpup(pos + 1),
+				be32_to_cpup(pos + 2), be32_to_cpup(pos + 3));
+		}
+	}
+}
+#endif
+
+/**
  * nfs_inode_attrs_need_update - check if the inode attributes need updating
  * @inode - pointer to inode
  * @fattr - attributes
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index 8c29950..c07a757 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -395,6 +395,20 @@ static inline void nfs_free_fhandle(const struct nfs_fh *fh)
 	kfree(fh);
 }
 
+#ifdef RPC_DEBUG
+extern void _nfs_display_fhandle(const struct nfs_fh *fh, const char *caption);
+#define nfs_display_fhandle(fh, caption)			\
+	do {							\
+		if (unlikely(nfs_debug & NFSDBG_FACILITY))	\
+			_nfs_display_fhandle(fh, caption);	\
+	} while (0)
+#else
+static inline void nfs_display_fhandle(const struct nfs_fh *fh,
+				       const char *caption)
+{
+}
+#endif
+
 /*
  * linux/fs/nfs/nfsroot.c
  */

[PATCH 10/12] NFS: Simplify arguments of encode_renew()

[Chuck Lever]

Clean up: pass just the clientid4 to encode_renew().  This enables it
to be used by callers who might not have an full nfs_client.

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

 fs/nfs/nfs4xdr.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index ea01e31..d53c21e 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1619,13 +1619,14 @@ static void encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, co
 	hdr->replen += decode_rename_maxsz;
 }
 
-static void encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid, struct compound_hdr *hdr)
+static void encode_renew(struct xdr_stream *xdr, clientid4 clid,
+			 struct compound_hdr *hdr)
 {
 	__be32 *p;
 
 	p = reserve_space(xdr, 12);
 	*p++ = cpu_to_be32(OP_RENEW);
-	xdr_encode_hyper(p, client_stateid->cl_clientid);
+	xdr_encode_hyper(p, clid);
 	hdr->nops++;
 	hdr->replen += decode_renew_maxsz;
 }
@@ -2650,7 +2651,7 @@ static void nfs4_xdr_enc_renew(struct rpc_rqst *req, struct xdr_stream *xdr,
 	};
 
 	encode_compound_hdr(xdr, req, &hdr);
-	encode_renew(xdr, clp, &hdr);
+	encode_renew(xdr, clp->cl_clientid, &hdr);
 	encode_nops(&hdr);
 }
 

[PATCH 11/12] NFS: Introduce NFS_ATTR_FATTR_V4_LOCATIONS

[Chuck Lever]

The Linux NFS client must distinguish between referral events (which
it currently supports) and migration events (which it does not yet
support).

In both types of events, an fs_locations array is returned.  But upper
layers, not the XDR layer, should make the distinction between a
referral and a migration.  There really isn't a way for an XDR decoder
function to distinguish the two, in general.

Slightly adjust the FATTR flags returned by decode_fs_locations()
to set NFS_ATTR_FATTR_V4_LOCATIONS only if a non-empty locations
array was returned from the server.  Then have logic in nfs4proc.c
distinguish whether the locations array is for a referral or
something else.

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

 fs/nfs/nfs4proc.c       |    6 +++---
 fs/nfs/nfs4xdr.c        |    2 +-
 include/linux/nfs_xdr.h |   11 ++++++-----
 3 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index d43bcda..ced56d0 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -79,6 +79,7 @@ static int _nfs4_proc_open(struct nfs4_opendata *data);
 static int _nfs4_recover_proc_open(struct nfs4_opendata *data);
 static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *);
 static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *, struct nfs4_state *);
+static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr);
 static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr);
 static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 			    struct nfs_fattr *fattr, struct iattr *sattr,
@@ -2340,7 +2341,6 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 	return nfs4_map_errors(status);
 }
 
-static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr);
 /*
  * Get locations and (maybe) other attributes of a referral.
  * Note that we'll actually follow the referral later when
@@ -4797,11 +4797,11 @@ static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr)
 	if (!(((fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) ||
 	       (fattr->valid & NFS_ATTR_FATTR_FILEID)) &&
 	      (fattr->valid & NFS_ATTR_FATTR_FSID) &&
-	      (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL)))
+	      (fattr->valid & NFS_ATTR_FATTR_V4_LOCATIONS)))
 		return;
 
 	fattr->valid |= NFS_ATTR_FATTR_TYPE | NFS_ATTR_FATTR_MODE |
-		NFS_ATTR_FATTR_NLINK;
+		NFS_ATTR_FATTR_NLINK | NFS_ATTR_FATTR_V4_REFERRAL;
 	fattr->mode = S_IFDIR | S_IRUGO | S_IXUGO;
 	fattr->nlink = 2;
 }
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index d53c21e..b78cd08 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -3618,7 +3618,7 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st
 			res->nlocations++;
 	}
 	if (res->nlocations != 0)
-		status = NFS_ATTR_FATTR_V4_REFERRAL;
+		status = NFS_ATTR_FATTR_V4_LOCATIONS;
 out:
 	dprintk("%s: fs_locations done, error = %d\n", __func__, status);
 	return status;
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index adbc84a..07eb9c9 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -88,11 +88,12 @@ struct nfs_fattr {
 #define NFS_ATTR_FATTR_PRECTIME		(1U << 16)
 #define NFS_ATTR_FATTR_CHANGE		(1U << 17)
 #define NFS_ATTR_FATTR_PRECHANGE	(1U << 18)
-#define NFS_ATTR_FATTR_V4_REFERRAL	(1U << 19)	/* NFSv4 referral */
-#define NFS_ATTR_FATTR_MOUNTPOINT	(1U << 20)	/* Treat as mountpoint */
-#define NFS_ATTR_FATTR_MOUNTED_ON_FILEID		(1U << 21)
-#define NFS_ATTR_FATTR_OWNER_NAME	(1U << 22)
-#define NFS_ATTR_FATTR_GROUP_NAME	(1U << 23)
+#define NFS_ATTR_FATTR_V4_LOCATIONS	(1U << 19)
+#define NFS_ATTR_FATTR_V4_REFERRAL	(1U << 20)
+#define NFS_ATTR_FATTR_MOUNTPOINT	(1U << 21)
+#define NFS_ATTR_FATTR_MOUNTED_ON_FILEID (1U << 22)
+#define NFS_ATTR_FATTR_OWNER_NAME	(1U << 23)
+#define NFS_ATTR_FATTR_GROUP_NAME	(1U << 24)
 
 #define NFS_ATTR_FATTR (NFS_ATTR_FATTR_TYPE \
 		| NFS_ATTR_FATTR_MODE \

[PATCH 12/12] NFS: Request fh_expire_type attribute in "server caps" operation

[Chuck Lever]

The fh_expire_type file attribute is a filesystem wide attribute that
consists of flags that indicate what characteristics file handles
on this FSID have.

Our client doesn't support volatile file handles.  It should find
out early (say, at mount time) whether the server is going to play
shenanighans with file handles during a migration.

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

 fs/nfs/nfs4proc.c         |    1 +
 fs/nfs/nfs4xdr.c          |   26 ++++++++++++++++++++++++++
 include/linux/nfs_fs_sb.h |    3 +++
 include/linux/nfs_xdr.h   |    1 +
 4 files changed, 31 insertions(+), 0 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index ced56d0..07bd1f3 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2220,6 +2220,7 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
 		server->cache_consistency_bitmask[0] &= FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE;
 		server->cache_consistency_bitmask[1] &= FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY;
 		server->acl_bitmask = res.acl_bitmask;
+		server->fh_expire_type = res.fh_expire_type;
 	}
 
 	return status;
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index b78cd08..d8d5b61 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -2634,6 +2634,7 @@ static void nfs4_xdr_enc_server_caps(struct rpc_rqst *req,
 	encode_sequence(xdr, &args->seq_args, &hdr);
 	encode_putfh(xdr, args->fhandle, &hdr);
 	encode_getattr_one(xdr, FATTR4_WORD0_SUPPORTED_ATTRS|
+			   FATTR4_WORD0_FH_EXPIRE_TYPE|
 			   FATTR4_WORD0_LINK_SUPPORT|
 			   FATTR4_WORD0_SYMLINK_SUPPORT|
 			   FATTR4_WORD0_ACLSUPPORT, &hdr);
@@ -3181,6 +3182,28 @@ out_overflow:
 	return -EIO;
 }
 
+static int decode_attr_fh_expire_type(struct xdr_stream *xdr,
+				      uint32_t *bitmap, uint32_t *type)
+{
+	__be32 *p;
+
+	*type = 0;
+	if (unlikely(bitmap[0] & (FATTR4_WORD0_FH_EXPIRE_TYPE - 1U)))
+		return -EIO;
+	if (likely(bitmap[0] & FATTR4_WORD0_FH_EXPIRE_TYPE)) {
+		p = xdr_inline_decode(xdr, 4);
+		if (unlikely(!p))
+			goto out_overflow;
+		*type = be32_to_cpup(p);
+		bitmap[0] &= ~FATTR4_WORD0_FH_EXPIRE_TYPE;
+	}
+	dprintk("%s: expire type=0x%x\n", __func__, *type);
+	return 0;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
 static int decode_attr_change(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *change)
 {
 	__be32 *p;
@@ -4229,6 +4252,9 @@ static int decode_server_caps(struct xdr_stream *xdr, struct nfs4_server_caps_re
 		goto xdr_error;
 	if ((status = decode_attr_supported(xdr, bitmap, res->attr_bitmask)) != 0)
 		goto xdr_error;
+	if ((status = decode_attr_fh_expire_type(xdr, bitmap,
+						 &res->fh_expire_type)) != 0)
+		goto xdr_error;
 	if ((status = decode_attr_link_support(xdr, bitmap, &res->has_links)) != 0)
 		goto xdr_error;
 	if ((status = decode_attr_symlink_support(xdr, bitmap, &res->has_symlinks)) != 0)
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 3bf4766..c2595d7 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -146,6 +146,9 @@ struct nfs_server {
 	u32			acl_bitmask;	/* V4 bitmask representing the ACEs
 						   that are supported on this
 						   filesystem */
+	u32			fh_expire_type;	/* V4 bitmask representing file
+						   handle volatility type for
+						   this filesystem */
 	struct pnfs_layoutdriver_type  *pnfs_curr_ld; /* Active layout driver */
 	struct rpc_wait_queue	roc_rpcwaitq;
 	void			*pnfs_ld_data;	/* per mount point data */
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 07eb9c9..211b280 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -977,6 +977,7 @@ struct nfs4_server_caps_res {
 	u32				acl_bitmask;
 	u32				has_links;
 	u32				has_symlinks;
+	u32				fh_expire_type;
 	struct nfs4_sequence_res	seq_res;
 };
 

Re: [PATCH 08/12] commit 6f38b4ba433ac6494f83cb73dd07dcbde797e1e0

[Christoph Hellwig]

On Fri, Mar 02, 2012 at 04:59:31PM -0500, Chuck Lever wrote:
> NFS: Make clientaddr= optional
> 
>     For NFSv4 mounts, the clientaddr= mount option has always been
>     required.  Now we have rpc_localaddr() in the kernel, which was
>     modeled after the same logic in the mount.nfs command that constructs
>     the clientaddr= mount option.  If user space doesn't provide a
>     clientaddr= mount option, the kernel can now construct its own.
> 
>     Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Something messed up formatting here.

Re: [PATCH 08/12] commit 6f38b4ba433ac6494f83cb73dd07dcbde797e1e0

[Myklebust, Trond]
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