Merge branch 'nfs-for-2.6.38' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6

* 'nfs-for-2.6.38' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6: (89 commits)
  NFS fix the setting of exchange id flag
  NFS: Don't use vm_map_ram() in readdir
  NFSv4: Ensure continued open and lockowner name uniqueness
  NFS: Move cl_delegations to the nfs_server struct
  NFS: Introduce nfs_detach_delegations()
  NFS: Move cl_state_owners and related fields to the nfs_server struct
  NFS: Allow walking nfs_client.cl_superblocks list outside client.c
  pnfs: layout roc code
  pnfs: update nfs4_callback_recallany to handle layouts
  pnfs: add CB_LAYOUTRECALL handling
  pnfs: CB_LAYOUTRECALL xdr code
  pnfs: change lo refcounting to atomic_t
  pnfs: check that partial LAYOUTGET return is ignored
  pnfs: add layout to client list before sending rpc
  pnfs: serialize LAYOUTGET(openstateid)
  pnfs: layoutget rpc code cleanup
  pnfs: change how lsegs are removed from layout list
  pnfs: change layout state seqlock to a spinlock
  pnfs: add prefix to struct pnfs_layout_hdr fields
  pnfs: add prefix to struct pnfs_layout_segment fields
  ...

fs/lockd/Makefile

@@ -4,7 +4,7 @@
 
 obj-$(CONFIG_LOCKD) += lockd.o
 
-lockd-objs-y := clntlock.o clntproc.o host.o svc.o svclock.o svcshare.o \
-	        svcproc.o svcsubs.o mon.o xdr.o grace.o
-lockd-objs-$(CONFIG_LOCKD_V4) += xdr4.o svc4proc.o
+lockd-objs-y := clntlock.o clntproc.o clntxdr.o host.o svc.o svclock.o \
+	        svcshare.o svcproc.o svcsubs.o mon.o xdr.o grace.o
+lockd-objs-$(CONFIG_LOCKD_V4) += clnt4xdr.o xdr4.o svc4proc.o
 lockd-objs		      := $(lockd-objs-y)

fs/lockd/clnt4xdr.c

+/*
+ * linux/fs/lockd/clnt4xdr.c
+ *
+ * XDR functions to encode/decode NLM version 4 RPC arguments and results.
+ *
+ * NLM client-side only.
+ *
+ * Copyright (C) 2010, Oracle.  All rights reserved.
+ */
+
+#include <linux/types.h>
+#include <linux/sunrpc/xdr.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/stats.h>
+#include <linux/lockd/lockd.h>
+
+#define NLMDBG_FACILITY		NLMDBG_XDR
+
+#if (NLMCLNT_OHSIZE > XDR_MAX_NETOBJ)
+#  error "NLM host name cannot be larger than XDR_MAX_NETOBJ!"
+#endif
+
+#if (NLMCLNT_OHSIZE > NLM_MAXSTRLEN)
+#  error "NLM host name cannot be larger than NLM's maximum string length!"
+#endif
+
+/*
+ * Declare the space requirements for NLM arguments and replies as
+ * number of 32bit-words
+ */
+#define NLM4_void_sz		(0)
+#define NLM4_cookie_sz		(1+(NLM_MAXCOOKIELEN>>2))
+#define NLM4_caller_sz		(1+(NLMCLNT_OHSIZE>>2))
+#define NLM4_owner_sz		(1+(NLMCLNT_OHSIZE>>2))
+#define NLM4_fhandle_sz		(1+(NFS3_FHSIZE>>2))
+#define NLM4_lock_sz		(5+NLM4_caller_sz+NLM4_owner_sz+NLM4_fhandle_sz)
+#define NLM4_holder_sz		(6+NLM4_owner_sz)
+
+#define NLM4_testargs_sz	(NLM4_cookie_sz+1+NLM4_lock_sz)
+#define NLM4_lockargs_sz	(NLM4_cookie_sz+4+NLM4_lock_sz)
+#define NLM4_cancargs_sz	(NLM4_cookie_sz+2+NLM4_lock_sz)
+#define NLM4_unlockargs_sz	(NLM4_cookie_sz+NLM4_lock_sz)
+
+#define NLM4_testres_sz		(NLM4_cookie_sz+1+NLM4_holder_sz)
+#define NLM4_res_sz		(NLM4_cookie_sz+1)
+#define NLM4_norep_sz		(0)
+
+
+static s64 loff_t_to_s64(loff_t offset)
+{
+	s64 res;
+
+	if (offset >= NLM4_OFFSET_MAX)
+		res = NLM4_OFFSET_MAX;
+	else if (offset <= -NLM4_OFFSET_MAX)
+		res = -NLM4_OFFSET_MAX;
+	else
+		res = offset;
+	return res;
+}
+
+static void nlm4_compute_offsets(const struct nlm_lock *lock,
+				 u64 *l_offset, u64 *l_len)
+{
+	const struct file_lock *fl = &lock->fl;
+
+	BUG_ON(fl->fl_start > NLM4_OFFSET_MAX);
+	BUG_ON(fl->fl_end > NLM4_OFFSET_MAX &&
+				fl->fl_end != OFFSET_MAX);
+
+	*l_offset = loff_t_to_s64(fl->fl_start);
+	if (fl->fl_end == OFFSET_MAX)
+		*l_len = 0;
+	else
+		*l_len = loff_t_to_s64(fl->fl_end - fl->fl_start + 1);
+}
+
+/*
+ * Handle decode buffer overflows out-of-line.
+ */
+static void print_overflow_msg(const char *func, const struct xdr_stream *xdr)
+{
+	dprintk("lockd: %s prematurely hit the end of our receive buffer. "
+		"Remaining buffer length is %tu words.\n",
+		func, xdr->end - xdr->p);
+}
+
+
+/*
+ * Encode/decode NLMv4 basic data types
+ *
+ * Basic NLMv4 data types are defined in Appendix II, section 6.1.4
+ * of RFC 1813: "NFS Version 3 Protocol Specification" and in Chapter
+ * 10 of X/Open's "Protocols for Interworking: XNFS, Version 3W".
+ *
+ * Not all basic data types have their own encoding and decoding
+ * functions.  For run-time efficiency, some data types are encoded
+ * or decoded inline.
+ */
+
+static void encode_bool(struct xdr_stream *xdr, const int value)
+{
+	__be32 *p;
+
+	p = xdr_reserve_space(xdr, 4);
+	*p = value ? xdr_one : xdr_zero;
+}
+
+static void encode_int32(struct xdr_stream *xdr, const s32 value)
+{
+	__be32 *p;
+
+	p = xdr_reserve_space(xdr, 4);
+	*p = cpu_to_be32(value);
+}
+
+/*
+ *	typedef opaque netobj<MAXNETOBJ_SZ>
+ */
+static void encode_netobj(struct xdr_stream *xdr,
+			  const u8 *data, const unsigned int length)
+{
+	__be32 *p;
+
+	BUG_ON(length > XDR_MAX_NETOBJ);
+	p = xdr_reserve_space(xdr, 4 + length);
+	xdr_encode_opaque(p, data, length);
+}
+
+static int decode_netobj(struct xdr_stream *xdr,
+			 struct xdr_netobj *obj)
+{
+	u32 length;
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	length = be32_to_cpup(p++);
+	if (unlikely(length > XDR_MAX_NETOBJ))
+		goto out_size;
+	obj->len = length;
+	obj->data = (u8 *)p;
+	return 0;
+out_size:
+	dprintk("NFS: returned netobj was too long: %u\n", length);
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	netobj cookie;
+ */
+static void encode_cookie(struct xdr_stream *xdr,
+			  const struct nlm_cookie *cookie)
+{
+	BUG_ON(cookie->len > NLM_MAXCOOKIELEN);
+	encode_netobj(xdr, (u8 *)&cookie->data, cookie->len);
+}
+
+static int decode_cookie(struct xdr_stream *xdr,
+			     struct nlm_cookie *cookie)
+{
+	u32 length;
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	length = be32_to_cpup(p++);
+	/* apparently HPUX can return empty cookies */
+	if (length == 0)
+		goto out_hpux;
+	if (length > NLM_MAXCOOKIELEN)
+		goto out_size;
+	p = xdr_inline_decode(xdr, length);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	cookie->len = length;
+	memcpy(cookie->data, p, length);
+	return 0;
+out_hpux:
+	cookie->len = 4;
+	memset(cookie->data, 0, 4);
+	return 0;
+out_size:
+	dprintk("NFS: returned cookie was too long: %u\n", length);
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	netobj fh;
+ */
+static void encode_fh(struct xdr_stream *xdr, const struct nfs_fh *fh)
+{
+	BUG_ON(fh->size > NFS3_FHSIZE);
+	encode_netobj(xdr, (u8 *)&fh->data, fh->size);
+}
+
+/*
+ *	enum nlm4_stats {
+ *		NLM4_GRANTED = 0,
+ *		NLM4_DENIED = 1,
+ *		NLM4_DENIED_NOLOCKS = 2,
+ *		NLM4_BLOCKED = 3,
+ *		NLM4_DENIED_GRACE_PERIOD = 4,
+ *		NLM4_DEADLCK = 5,
+ *		NLM4_ROFS = 6,
+ *		NLM4_STALE_FH = 7,
+ *		NLM4_FBIG = 8,
+ *		NLM4_FAILED = 9
+ *	};
+ *
+ *	struct nlm4_stat {
+ *		nlm4_stats stat;
+ *	};
+ *
+ * NB: we don't swap bytes for the NLM status values.  The upper
+ * layers deal directly with the status value in network byte
+ * order.
+ */
+static void encode_nlm4_stat(struct xdr_stream *xdr,
+			     const __be32 stat)
+{
+	__be32 *p;
+
+	BUG_ON(be32_to_cpu(stat) > NLM_FAILED);
+	p = xdr_reserve_space(xdr, 4);
+	*p = stat;
+}
+
+static int decode_nlm4_stat(struct xdr_stream *xdr, __be32 *stat)
+{
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	if (unlikely(*p > nlm4_failed))
+		goto out_bad_xdr;
+	*stat = *p;
+	return 0;
+out_bad_xdr:
+	dprintk("%s: server returned invalid nlm4_stats value: %u\n",
+			__func__, be32_to_cpup(p));
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	struct nlm4_holder {
+ *		bool	exclusive;
+ *		int32	svid;
+ *		netobj	oh;
+ *		uint64	l_offset;
+ *		uint64	l_len;
+ *	};
+ */
+static void encode_nlm4_holder(struct xdr_stream *xdr,
+			       const struct nlm_res *result)
+{
+	const struct nlm_lock *lock = &result->lock;
+	u64 l_offset, l_len;
+	__be32 *p;
+
+	encode_bool(xdr, lock->fl.fl_type == F_RDLCK);
+	encode_int32(xdr, lock->svid);
+	encode_netobj(xdr, lock->oh.data, lock->oh.len);
+
+	p = xdr_reserve_space(xdr, 4 + 4);
+	nlm4_compute_offsets(lock, &l_offset, &l_len);
+	p = xdr_encode_hyper(p, l_offset);
+	xdr_encode_hyper(p, l_len);
+}
+
+static int decode_nlm4_holder(struct xdr_stream *xdr, struct nlm_res *result)
+{
+	struct nlm_lock *lock = &result->lock;
+	struct file_lock *fl = &lock->fl;
+	u64 l_offset, l_len;
+	u32 exclusive;
+	int error;
+	__be32 *p;
+	s32 end;
+
+	memset(lock, 0, sizeof(*lock));
+	locks_init_lock(fl);
+
+	p = xdr_inline_decode(xdr, 4 + 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	exclusive = be32_to_cpup(p++);
+	lock->svid = be32_to_cpup(p);
+	fl->fl_pid = (pid_t)lock->svid;
+
+	error = decode_netobj(xdr, &lock->oh);
+	if (unlikely(error))
+		goto out;
+
+	p = xdr_inline_decode(xdr, 8 + 8);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+
+	fl->fl_flags = FL_POSIX;
+	fl->fl_type  = exclusive != 0 ? F_WRLCK : F_RDLCK;
+	p = xdr_decode_hyper(p, &l_offset);
+	xdr_decode_hyper(p, &l_len);
+	end = l_offset + l_len - 1;
+
+	fl->fl_start = (loff_t)l_offset;
+	if (l_len == 0 || end < 0)
+		fl->fl_end = OFFSET_MAX;
+	else
+		fl->fl_end = (loff_t)end;
+	error = 0;
+out:
+	return error;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	string caller_name<LM_MAXSTRLEN>;
+ */
+static void encode_caller_name(struct xdr_stream *xdr, const char *name)
+{
+	/* NB: client-side does not set lock->len */
+	u32 length = strlen(name);
+	__be32 *p;
+
+	BUG_ON(length > NLM_MAXSTRLEN);
+	p = xdr_reserve_space(xdr, 4 + length);
+	xdr_encode_opaque(p, name, length);
+}
+
+/*
+ *	struct nlm4_lock {
+ *		string	caller_name<LM_MAXSTRLEN>;
+ *		netobj	fh;
+ *		netobj	oh;
+ *		int32	svid;
+ *		uint64	l_offset;
+ *		uint64	l_len;
+ *	};
+ */
+static void encode_nlm4_lock(struct xdr_stream *xdr,
+			     const struct nlm_lock *lock)
+{
+	u64 l_offset, l_len;
+	__be32 *p;
+
+	encode_caller_name(xdr, lock->caller);
+	encode_fh(xdr, &lock->fh);
+	encode_netobj(xdr, lock->oh.data, lock->oh.len);
+
+	p = xdr_reserve_space(xdr, 4 + 8 + 8);
+	*p++ = cpu_to_be32(lock->svid);
+
+	nlm4_compute_offsets(lock, &l_offset, &l_len);
+	p = xdr_encode_hyper(p, l_offset);
+	xdr_encode_hyper(p, l_len);
+}
+
+
+/*
+ * NLMv4 XDR encode functions
+ *
+ * NLMv4 argument types are defined in Appendix II of RFC 1813:
+ * "NFS Version 3 Protocol Specification" and Chapter 10 of X/Open's
+ * "Protocols for Interworking: XNFS, Version 3W".
+ */
+
+/*
+ *	struct nlm4_testargs {
+ *		netobj cookie;
+ *		bool exclusive;
+ *		struct nlm4_lock alock;
+ *	};
+ */
+static void nlm4_xdr_enc_testargs(struct rpc_rqst *req,
+				  struct xdr_stream *xdr,
+				  const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm4_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm4_lockargs {
+ *		netobj cookie;
+ *		bool block;
+ *		bool exclusive;
+ *		struct nlm4_lock alock;
+ *		bool reclaim;
+ *		int state;
+ *	};
+ */
+static void nlm4_xdr_enc_lockargs(struct rpc_rqst *req,
+				  struct xdr_stream *xdr,
+				  const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, args->block);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm4_lock(xdr, lock);
+	encode_bool(xdr, args->reclaim);
+	encode_int32(xdr, args->state);
+}
+
+/*
+ *	struct nlm4_cancargs {
+ *		netobj cookie;
+ *		bool block;
+ *		bool exclusive;
+ *		struct nlm4_lock alock;
+ *	};
+ */
+static void nlm4_xdr_enc_cancargs(struct rpc_rqst *req,
+				  struct xdr_stream *xdr,
+				  const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, args->block);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm4_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm4_unlockargs {
+ *		netobj cookie;
+ *		struct nlm4_lock alock;
+ *	};
+ */
+static void nlm4_xdr_enc_unlockargs(struct rpc_rqst *req,
+				    struct xdr_stream *xdr,
+				    const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_nlm4_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm4_res {
+ *		netobj cookie;
+ *		nlm4_stat stat;
+ *	};
+ */
+static void nlm4_xdr_enc_res(struct rpc_rqst *req,
+			     struct xdr_stream *xdr,
+			     const struct nlm_res *result)
+{
+	encode_cookie(xdr, &result->cookie);
+	encode_nlm4_stat(xdr, result->status);
+}
+
+/*
+ *	union nlm4_testrply switch (nlm4_stats stat) {
+ *	case NLM4_DENIED:
+ *		struct nlm4_holder holder;
+ *	default:
+ *		void;
+ *	};
+ *
+ *	struct nlm4_testres {
+ *		netobj cookie;
+ *		nlm4_testrply test_stat;
+ *	};
+ */
+static void nlm4_xdr_enc_testres(struct rpc_rqst *req,
+				 struct xdr_stream *xdr,
+				 const struct nlm_res *result)
+{
+	encode_cookie(xdr, &result->cookie);
+	encode_nlm4_stat(xdr, result->status);
+	if (result->status == nlm_lck_denied)
+		encode_nlm4_holder(xdr, result);
+}
+
+
+/*
+ * NLMv4 XDR decode functions
+ *
+ * NLMv4 argument types are defined in Appendix II of RFC 1813:
+ * "NFS Version 3 Protocol Specification" and Chapter 10 of X/Open's
+ * "Protocols for Interworking: XNFS, Version 3W".
+ */
+
+/*
+ *	union nlm4_testrply switch (nlm4_stats stat) {
+ *	case NLM4_DENIED:
+ *		struct nlm4_holder holder;
+ *	default:
+ *		void;
+ *	};
+ *
+ *	struct nlm4_testres {
+ *		netobj cookie;
+ *		nlm4_testrply test_stat;
+ *	};
+ */
+static int decode_nlm4_testrply(struct xdr_stream *xdr,
+				struct nlm_res *result)
+{
+	int error;
+
+	error = decode_nlm4_stat(xdr, &result->status);
+	if (unlikely(error))
+		goto out;
+	if (result->status == nlm_lck_denied)
+		error = decode_nlm4_holder(xdr, result);
+out:
+	return error;
+}
+
+static int nlm4_xdr_dec_testres(struct rpc_rqst *req,
+				struct xdr_stream *xdr,
+				struct nlm_res *result)
+{
+	int error;
+
+	error = decode_cookie(xdr, &result->cookie);
+	if (unlikely(error))
+		goto out;
+	error = decode_nlm4_testrply(xdr, result);
+out:
+	return error;
+}
+
+/*
+ *	struct nlm4_res {
+ *		netobj cookie;
+ *		nlm4_stat stat;
+ *	};
+ */
+static int nlm4_xdr_dec_res(struct rpc_rqst *req,
+			    struct xdr_stream *xdr,
+			    struct nlm_res *result)
+{
+	int error;
+
+	error = decode_cookie(xdr, &result->cookie);
+	if (unlikely(error))
+		goto out;
+	error = decode_nlm4_stat(xdr, &result->status);
+out:
+	return error;
+}
+
+
+/*
+ * For NLM, a void procedure really returns nothing
+ */
+#define nlm4_xdr_dec_norep	NULL
+
+#define PROC(proc, argtype, restype)					\
+[NLMPROC_##proc] = {							\
+	.p_proc      = NLMPROC_##proc,					\
+	.p_encode    = (kxdreproc_t)nlm4_xdr_enc_##argtype,		\
+	.p_decode    = (kxdrdproc_t)nlm4_xdr_dec_##restype,		\
+	.p_arglen    = NLM4_##argtype##_sz,				\
+	.p_replen    = NLM4_##restype##_sz,				\
+	.p_statidx   = NLMPROC_##proc,					\
+	.p_name      = #proc,						\
+	}
+
+static struct rpc_procinfo	nlm4_procedures[] = {
+	PROC(TEST,		testargs,	testres),
+	PROC(LOCK,		lockargs,	res),
+	PROC(CANCEL,		cancargs,	res),
+	PROC(UNLOCK,		unlockargs,	res),
+	PROC(GRANTED,		testargs,	res),
+	PROC(TEST_MSG,		testargs,	norep),
+	PROC(LOCK_MSG,		lockargs,	norep),
+	PROC(CANCEL_MSG,	cancargs,	norep),
+	PROC(UNLOCK_MSG,	unlockargs,	norep),
+	PROC(GRANTED_MSG,	testargs,	norep),
+	PROC(TEST_RES,		testres,	norep),
+	PROC(LOCK_RES,		res,		norep),
+	PROC(CANCEL_RES,	res,		norep),
+	PROC(UNLOCK_RES,	res,		norep),
+	PROC(GRANTED_RES,	res,		norep),
+};
+
+struct rpc_version	nlm_version4 = {
+	.number		= 4,
+	.nrprocs	= ARRAY_SIZE(nlm4_procedures),
+	.procs		= nlm4_procedures,
+};

fs/lockd/clntlock.c

@@ -79,7 +79,7 @@ EXPORT_SYMBOL_GPL(nlmclnt_init);
  */
 void nlmclnt_done(struct nlm_host *host)
 {
-	nlm_release_host(host);
+	nlmclnt_release_host(host);
 	lockd_down();
 }
 EXPORT_SYMBOL_GPL(nlmclnt_done);
@@ -273,7 +273,7 @@ reclaimer(void *ptr)
 	spin_unlock(&nlm_blocked_lock);
 
 	/* Release host handle after use */
-	nlm_release_host(host);
+	nlmclnt_release_host(host);
 	lockd_down();
 	return 0;
 }

fs/lockd/clntproc.c

@@ -58,7 +58,7 @@ static void nlm_put_lockowner(struct nlm_lockowner *lockowner)
 		return;
 	list_del(&lockowner->list);
 	spin_unlock(&lockowner->host->h_lock);
-	nlm_release_host(lockowner->host);
+	nlmclnt_release_host(lockowner->host);
 	kfree(lockowner);
 }
 
@@ -207,22 +207,22 @@ struct nlm_rqst *nlm_alloc_call(struct nlm_host *host)
 		printk("nlm_alloc_call: failed, waiting for memory\n");
 		schedule_timeout_interruptible(5*HZ);
 	}
-	nlm_release_host(host);
+	nlmclnt_release_host(host);
 	return NULL;
 }
 
-void nlm_release_call(struct nlm_rqst *call)
+void nlmclnt_release_call(struct nlm_rqst *call)
 {
 	if (!atomic_dec_and_test(&call->a_count))
 		return;
-	nlm_release_host(call->a_host);
+	nlmclnt_release_host(call->a_host);
 	nlmclnt_release_lockargs(call);
 	kfree(call);
 }
 
 static void nlmclnt_rpc_release(void *data)
 {
-	nlm_release_call(data);
+	nlmclnt_release_call(data);
 }
 
 static int nlm_wait_on_grace(wait_queue_head_t *queue)
@@ -436,7 +436,7 @@ nlmclnt_test(struct nlm_rqst *req, struct file_lock *fl)
 			status = nlm_stat_to_errno(req->a_res.status);
 	}
 out:
-	nlm_release_call(req);
+	nlmclnt_release_call(req);
 	return status;
 }
 
@@ -593,7 +593,7 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl)
 out_unblock:
 	nlmclnt_finish_block(block);
 out:
-	nlm_release_call(req);
+	nlmclnt_release_call(req);
 	return status;
 out_unlock:
 	/* Fatal error: ensure that we remove the lock altogether */
@@ -694,7 +694,7 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl)
 	/* What to do now? I'm out of my depth... */
 	status = -ENOLCK;
 out:
-	nlm_release_call(req);
+	nlmclnt_release_call(req);
 	return status;
 }
 
@@ -755,7 +755,7 @@ static int nlmclnt_cancel(struct nlm_host *host, int block, struct file_lock *fl
 			NLMPROC_CANCEL, &nlmclnt_cancel_ops);
 	if (status == 0 && req->a_res.status == nlm_lck_denied)
 		status = -ENOLCK;
-	nlm_release_call(req);
+	nlmclnt_release_call(req);
 	return status;
 }
 

fs/lockd/clntxdr.c

+/*
+ * linux/fs/lockd/clntxdr.c
+ *
+ * XDR functions to encode/decode NLM version 3 RPC arguments and results.
+ * NLM version 3 is backwards compatible with NLM versions 1 and 2.
+ *
+ * NLM client-side only.
+ *
+ * Copyright (C) 2010, Oracle.  All rights reserved.
+ */
+
+#include <linux/types.h>
+#include <linux/sunrpc/xdr.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/stats.h>
+#include <linux/lockd/lockd.h>
+
+#define NLMDBG_FACILITY		NLMDBG_XDR
+
+#if (NLMCLNT_OHSIZE > XDR_MAX_NETOBJ)
+#  error "NLM host name cannot be larger than XDR_MAX_NETOBJ!"
+#endif
+
+/*
+ * Declare the space requirements for NLM arguments and replies as
+ * number of 32bit-words
+ */
+#define NLM_cookie_sz		(1+(NLM_MAXCOOKIELEN>>2))
+#define NLM_caller_sz		(1+(NLMCLNT_OHSIZE>>2))
+#define NLM_owner_sz		(1+(NLMCLNT_OHSIZE>>2))
+#define NLM_fhandle_sz		(1+(NFS2_FHSIZE>>2))
+#define NLM_lock_sz		(3+NLM_caller_sz+NLM_owner_sz+NLM_fhandle_sz)
+#define NLM_holder_sz		(4+NLM_owner_sz)
+
+#define NLM_testargs_sz		(NLM_cookie_sz+1+NLM_lock_sz)
+#define NLM_lockargs_sz		(NLM_cookie_sz+4+NLM_lock_sz)
+#define NLM_cancargs_sz		(NLM_cookie_sz+2+NLM_lock_sz)
+#define NLM_unlockargs_sz	(NLM_cookie_sz+NLM_lock_sz)
+
+#define NLM_testres_sz		(NLM_cookie_sz+1+NLM_holder_sz)
+#define NLM_res_sz		(NLM_cookie_sz+1)
+#define NLM_norep_sz		(0)
+
+
+static s32 loff_t_to_s32(loff_t offset)
+{
+	s32 res;
+
+	if (offset >= NLM_OFFSET_MAX)
+		res = NLM_OFFSET_MAX;
+	else if (offset <= -NLM_OFFSET_MAX)
+		res = -NLM_OFFSET_MAX;
+	else
+		res = offset;
+	return res;
+}
+
+static void nlm_compute_offsets(const struct nlm_lock *lock,
+				u32 *l_offset, u32 *l_len)
+{
+	const struct file_lock *fl = &lock->fl;
+
+	BUG_ON(fl->fl_start > NLM_OFFSET_MAX);
+	BUG_ON(fl->fl_end > NLM_OFFSET_MAX &&
+				fl->fl_end != OFFSET_MAX);
+
+	*l_offset = loff_t_to_s32(fl->fl_start);
+	if (fl->fl_end == OFFSET_MAX)
+		*l_len = 0;
+	else
+		*l_len = loff_t_to_s32(fl->fl_end - fl->fl_start + 1);
+}
+
+/*
+ * Handle decode buffer overflows out-of-line.
+ */
+static void print_overflow_msg(const char *func, const struct xdr_stream *xdr)
+{
+	dprintk("lockd: %s prematurely hit the end of our receive buffer. "
+		"Remaining buffer length is %tu words.\n",
+		func, xdr->end - xdr->p);
+}
+
+
+/*
+ * Encode/decode NLMv3 basic data types
+ *
+ * Basic NLMv3 data types are not defined in an IETF standards
+ * document.  X/Open has a description of these data types that
+ * is useful.  See Chapter 10 of "Protocols for Interworking:
+ * XNFS, Version 3W".
+ *
+ * Not all basic data types have their own encoding and decoding
+ * functions.  For run-time efficiency, some data types are encoded
+ * or decoded inline.
+ */
+
+static void encode_bool(struct xdr_stream *xdr, const int value)
+{
+	__be32 *p;
+
+	p = xdr_reserve_space(xdr, 4);
+	*p = value ? xdr_one : xdr_zero;
+}
+
+static void encode_int32(struct xdr_stream *xdr, const s32 value)
+{
+	__be32 *p;
+
+	p = xdr_reserve_space(xdr, 4);
+	*p = cpu_to_be32(value);
+}
+
+/*
+ *	typedef opaque netobj<MAXNETOBJ_SZ>
+ */
+static void encode_netobj(struct xdr_stream *xdr,
+			  const u8 *data, const unsigned int length)
+{
+	__be32 *p;
+
+	BUG_ON(length > XDR_MAX_NETOBJ);
+	p = xdr_reserve_space(xdr, 4 + length);
+	xdr_encode_opaque(p, data, length);
+}
+
+static int decode_netobj(struct xdr_stream *xdr,
+			 struct xdr_netobj *obj)
+{
+	u32 length;
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	length = be32_to_cpup(p++);
+	if (unlikely(length > XDR_MAX_NETOBJ))
+		goto out_size;
+	obj->len = length;
+	obj->data = (u8 *)p;
+	return 0;
+out_size:
+	dprintk("NFS: returned netobj was too long: %u\n", length);
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	netobj cookie;
+ */
+static void encode_cookie(struct xdr_stream *xdr,
+			  const struct nlm_cookie *cookie)
+{
+	BUG_ON(cookie->len > NLM_MAXCOOKIELEN);
+	encode_netobj(xdr, (u8 *)&cookie->data, cookie->len);
+}
+
+static int decode_cookie(struct xdr_stream *xdr,
+			 struct nlm_cookie *cookie)
+{
+	u32 length;
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	length = be32_to_cpup(p++);
+	/* apparently HPUX can return empty cookies */
+	if (length == 0)
+		goto out_hpux;
+	if (length > NLM_MAXCOOKIELEN)
+		goto out_size;
+	p = xdr_inline_decode(xdr, length);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	cookie->len = length;
+	memcpy(cookie->data, p, length);
+	return 0;
+out_hpux:
+	cookie->len = 4;
+	memset(cookie->data, 0, 4);
+	return 0;
+out_size:
+	dprintk("NFS: returned cookie was too long: %u\n", length);
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	netobj fh;
+ */
+static void encode_fh(struct xdr_stream *xdr, const struct nfs_fh *fh)
+{
+	BUG_ON(fh->size != NFS2_FHSIZE);
+	encode_netobj(xdr, (u8 *)&fh->data, NFS2_FHSIZE);
+}
+
+/*
+ *	enum nlm_stats {
+ *		LCK_GRANTED = 0,
+ *		LCK_DENIED = 1,
+ *		LCK_DENIED_NOLOCKS = 2,
+ *		LCK_BLOCKED = 3,
+ *		LCK_DENIED_GRACE_PERIOD = 4
+ *	};
+ *
+ *
+ *	struct nlm_stat {
+ *		nlm_stats stat;
+ *	};
+ *
+ * NB: we don't swap bytes for the NLM status values.  The upper
+ * layers deal directly with the status value in network byte
+ * order.
+ */
+
+static void encode_nlm_stat(struct xdr_stream *xdr,
+			    const __be32 stat)
+{
+	__be32 *p;
+
+	BUG_ON(be32_to_cpu(stat) > NLM_LCK_DENIED_GRACE_PERIOD);
+	p = xdr_reserve_space(xdr, 4);
+	*p = stat;
+}
+
+static int decode_nlm_stat(struct xdr_stream *xdr,
+			   __be32 *stat)
+{
+	__be32 *p;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	if (unlikely(*p > nlm_lck_denied_grace_period))
+		goto out_enum;
+	*stat = *p;
+	return 0;
+out_enum:
+	dprintk("%s: server returned invalid nlm_stats value: %u\n",
+		__func__, be32_to_cpup(p));
+	return -EIO;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	struct nlm_holder {
+ *		bool exclusive;
+ *		int uppid;
+ *		netobj oh;
+ *		unsigned l_offset;
+ *		unsigned l_len;
+ *	};
+ */
+static void encode_nlm_holder(struct xdr_stream *xdr,
+			      const struct nlm_res *result)
+{
+	const struct nlm_lock *lock = &result->lock;
+	u32 l_offset, l_len;
+	__be32 *p;
+
+	encode_bool(xdr, lock->fl.fl_type == F_RDLCK);
+	encode_int32(xdr, lock->svid);
+	encode_netobj(xdr, lock->oh.data, lock->oh.len);
+
+	p = xdr_reserve_space(xdr, 4 + 4);
+	nlm_compute_offsets(lock, &l_offset, &l_len);
+	*p++ = cpu_to_be32(l_offset);
+	*p   = cpu_to_be32(l_len);
+}
+
+static int decode_nlm_holder(struct xdr_stream *xdr, struct nlm_res *result)
+{
+	struct nlm_lock *lock = &result->lock;
+	struct file_lock *fl = &lock->fl;
+	u32 exclusive, l_offset, l_len;
+	int error;
+	__be32 *p;
+	s32 end;
+
+	memset(lock, 0, sizeof(*lock));
+	locks_init_lock(fl);
+
+	p = xdr_inline_decode(xdr, 4 + 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+	exclusive = be32_to_cpup(p++);
+	lock->svid = be32_to_cpup(p);
+	fl->fl_pid = (pid_t)lock->svid;
+
+	error = decode_netobj(xdr, &lock->oh);
+	if (unlikely(error))
+		goto out;
+
+	p = xdr_inline_decode(xdr, 4 + 4);
+	if (unlikely(p == NULL))
+		goto out_overflow;
+
+	fl->fl_flags = FL_POSIX;
+	fl->fl_type  = exclusive != 0 ? F_WRLCK : F_RDLCK;
+	l_offset = be32_to_cpup(p++);
+	l_len = be32_to_cpup(p);
+	end = l_offset + l_len - 1;
+
+	fl->fl_start = (loff_t)l_offset;
+	if (l_len == 0 || end < 0)
+		fl->fl_end = OFFSET_MAX;
+	else
+		fl->fl_end = (loff_t)end;
+	error = 0;
+out:
+	return error;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
+
+/*
+ *	string caller_name<LM_MAXSTRLEN>;
+ */
+static void encode_caller_name(struct xdr_stream *xdr, const char *name)
+{
+	/* NB: client-side does not set lock->len */
+	u32 length = strlen(name);
+	__be32 *p;
+
+	BUG_ON(length > NLM_MAXSTRLEN);
+	p = xdr_reserve_space(xdr, 4 + length);
+	xdr_encode_opaque(p, name, length);
+}
+
+/*
+ *	struct nlm_lock {
+ *		string caller_name<LM_MAXSTRLEN>;
+ *		netobj fh;
+ *		netobj oh;
+ *		int uppid;
+ *		unsigned l_offset;
+ *		unsigned l_len;
+ *	};
+ */
+static void encode_nlm_lock(struct xdr_stream *xdr,
+			    const struct nlm_lock *lock)
+{
+	u32 l_offset, l_len;
+	__be32 *p;
+
+	encode_caller_name(xdr, lock->caller);
+	encode_fh(xdr, &lock->fh);
+	encode_netobj(xdr, lock->oh.data, lock->oh.len);
+
+	p = xdr_reserve_space(xdr, 4 + 4 + 4);
+	*p++ = cpu_to_be32(lock->svid);
+
+	nlm_compute_offsets(lock, &l_offset, &l_len);
+	*p++ = cpu_to_be32(l_offset);
+	*p   = cpu_to_be32(l_len);
+}
+
+
+/*
+ * NLMv3 XDR encode functions
+ *
+ * NLMv3 argument types are defined in Chapter 10 of The Open Group's
+ * "Protocols for Interworking: XNFS, Version 3W".
+ */
+
+/*
+ *	struct nlm_testargs {
+ *		netobj cookie;
+ *		bool exclusive;
+ *		struct nlm_lock alock;
+ *	};
+ */
+static void nlm_xdr_enc_testargs(struct rpc_rqst *req,
+				 struct xdr_stream *xdr,
+				 const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm_lockargs {
+ *		netobj cookie;
+ *		bool block;
+ *		bool exclusive;
+ *		struct nlm_lock alock;
+ *		bool reclaim;
+ *		int state;
+ *	};
+ */
+static void nlm_xdr_enc_lockargs(struct rpc_rqst *req,
+				 struct xdr_stream *xdr,
+				 const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, args->block);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm_lock(xdr, lock);
+	encode_bool(xdr, args->reclaim);
+	encode_int32(xdr, args->state);
+}
+
+/*
+ *	struct nlm_cancargs {
+ *		netobj cookie;
+ *		bool block;
+ *		bool exclusive;
+ *		struct nlm_lock alock;
+ *	};
+ */
+static void nlm_xdr_enc_cancargs(struct rpc_rqst *req,
+				 struct xdr_stream *xdr,
+				 const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_bool(xdr, args->block);
+	encode_bool(xdr, lock->fl.fl_type == F_WRLCK);
+	encode_nlm_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm_unlockargs {
+ *		netobj cookie;
+ *		struct nlm_lock alock;
+ *	};
+ */
+static void nlm_xdr_enc_unlockargs(struct rpc_rqst *req,
+				   struct xdr_stream *xdr,
+				   const struct nlm_args *args)
+{
+	const struct nlm_lock *lock = &args->lock;
+
+	encode_cookie(xdr, &args->cookie);
+	encode_nlm_lock(xdr, lock);
+}
+
+/*
+ *	struct nlm_res {
+ *		netobj cookie;
+ *		nlm_stat stat;
+ *	};
+ */
+static void nlm_xdr_enc_res(struct rpc_rqst *req,
+			    struct xdr_stream *xdr,
+			    const struct nlm_res *result)
+{
+	encode_cookie(xdr, &result->cookie);
+	encode_nlm_stat(xdr, result->status);
+}
+
+/*
+ *	union nlm_testrply switch (nlm_stats stat) {
+ *	case LCK_DENIED:
+ *		struct nlm_holder holder;
+ *	default:
+ *		void;
+ *	};
+ *
+ *	struct nlm_testres {
+ *		netobj cookie;
+ *		nlm_testrply test_stat;
+ *	};
+ */
+static void encode_nlm_testrply(struct xdr_stream *xdr,
+				const struct nlm_res *result)
+{
+	if (result->status == nlm_lck_denied)
+		encode_nlm_holder(xdr, result);
+}
+
+static void nlm_xdr_enc_testres(struct rpc_rqst *req,
+				struct xdr_stream *xdr,
+				const struct nlm_res *result)
+{
+	encode_cookie(xdr, &result->cookie);
+	encode_nlm_stat(xdr, result->status);
+	encode_nlm_testrply(xdr, result);
+}
+
+
+/*
+ * NLMv3 XDR decode functions
+ *
+ * NLMv3 result types are defined in Chapter 10 of The Open Group's
+ * "Protocols for Interworking: XNFS, Version 3W".
+ */
+
+/*
+ *	union nlm_testrply switch (nlm_stats stat) {
+ *	case LCK_DENIED:
+ *		struct nlm_holder holder;
+ *	default:
+ *		void;
+ *	};
+ *
+ *	struct nlm_testres {
+ *		netobj cookie;
+ *		nlm_testrply test_stat;
+ *	};
+ */
+static int decode_nlm_testrply(struct xdr_stream *xdr,
+			       struct nlm_res *result)
+{
+	int error;
+
+	error = decode_nlm_stat(xdr, &result->status);
+	if (unlikely(error))
+		goto out;
+	if (result->status == nlm_lck_denied)
+		error = decode_nlm_holder(xdr, result);
+out:
+	return error;
+}
+
+static int nlm_xdr_dec_testres(struct rpc_rqst *req,
+			       struct xdr_stream *xdr,
+			       struct nlm_res *result)
+{
+	int error;
+
+	error = decode_cookie(xdr, &result->cookie);
+	if (unlikely(error))
+		goto out;
+	error = decode_nlm_testrply(xdr, result);
+out:
+	return error;
+}
+
+/*
+ *	struct nlm_res {
+ *		netobj cookie;
+ *		nlm_stat stat;
+ *	};
+ */
+static int nlm_xdr_dec_res(struct rpc_rqst *req,
+			   struct xdr_stream *xdr,
+			   struct nlm_res *result)
+{
+	int error;
+
+	error = decode_cookie(xdr, &result->cookie);
+	if (unlikely(error))
+		goto out;
+	error = decode_nlm_stat(xdr, &result->status);
+out:
+	return error;
+}
+
+
+/*
+ * For NLM, a void procedure really returns nothing
+ */
+#define nlm_xdr_dec_norep	NULL
+
+#define PROC(proc, argtype, restype)	\
+[NLMPROC_##proc] = {							\
+	.p_proc      = NLMPROC_##proc,					\
+	.p_encode    = (kxdreproc_t)nlm_xdr_enc_##argtype,		\
+	.p_decode    = (kxdrdproc_t)nlm_xdr_dec_##restype,		\
+	.p_arglen    = NLM_##argtype##_sz,				\
+	.p_replen    = NLM_##restype##_sz,				\
+	.p_statidx   = NLMPROC_##proc,					\
+	.p_name      = #proc,						\
+	}
+
+static struct rpc_procinfo	nlm_procedures[] = {
+	PROC(TEST,		testargs,	testres),
+	PROC(LOCK,		lockargs,	res),
+	PROC(CANCEL,		cancargs,	res),
+	PROC(UNLOCK,		unlockargs,	res),
+	PROC(GRANTED,		testargs,	res),
+	PROC(TEST_MSG,		testargs,	norep),
+	PROC(LOCK_MSG,		lockargs,	norep),
+	PROC(CANCEL_MSG,	cancargs,	norep),
+	PROC(UNLOCK_MSG,	unlockargs,	norep),
+	PROC(GRANTED_MSG,	testargs,	norep),
+	PROC(TEST_RES,		testres,	norep),
+	PROC(LOCK_RES,		res,		norep),
+	PROC(CANCEL_RES,	res,		norep),
+	PROC(UNLOCK_RES,	res,		norep),
+	PROC(GRANTED_RES,	res,		norep),
+};
+
+static struct rpc_version	nlm_version1 = {
+		.number		= 1,
+		.nrprocs	= ARRAY_SIZE(nlm_procedures),
+		.procs		= nlm_procedures,
+};
+
+static struct rpc_version	nlm_version3 = {
+		.number		= 3,
+		.nrprocs	= ARRAY_SIZE(nlm_procedures),
+		.procs		= nlm_procedures,
+};
+
+static struct rpc_version	*nlm_versions[] = {
+	[1] = &nlm_version1,
+	[3] = &nlm_version3,
+#ifdef CONFIG_LOCKD_V4
+	[4] = &nlm_version4,
+#endif
+};
+
+static struct rpc_stat		nlm_rpc_stats;
+
+struct rpc_program		nlm_program = {
+		.name		= "lockd",
+		.number		= NLM_PROGRAM,
+		.nrvers		= ARRAY_SIZE(nlm_versions),
+		.version	= nlm_versions,
+		.stats		= &nlm_rpc_stats,
+};

fs/lockd/host.c

@@ -25,9 +25,22 @@
 #define NLM_HOST_EXPIRE		(300 * HZ)
 #define NLM_HOST_COLLECT	(120 * HZ)
 
-static struct hlist_head	nlm_hosts[NLM_HOST_NRHASH];
+static struct hlist_head	nlm_server_hosts[NLM_HOST_NRHASH];
+static struct hlist_head	nlm_client_hosts[NLM_HOST_NRHASH];
+
+#define for_each_host(host, pos, chain, table) \
+	for ((chain) = (table); \
+	     (chain) < (table) + NLM_HOST_NRHASH; ++(chain)) \
+		hlist_for_each_entry((host), (pos), (chain), h_hash)
+
+#define for_each_host_safe(host, pos, next, chain, table) \
+	for ((chain) = (table); \
+	     (chain) < (table) + NLM_HOST_NRHASH; ++(chain)) \
+		hlist_for_each_entry_safe((host), (pos), (next), \
+						(chain), h_hash)
+
 static unsigned long		next_gc;
-static int			nrhosts;
+static unsigned long		nrhosts;
 static DEFINE_MUTEX(nlm_host_mutex);
 
 static void			nlm_gc_hosts(void);
@@ -40,8 +53,6 @@ struct nlm_lookup_host_info {
 	const u32		version;	/* NLM version to search for */
 	const char		*hostname;	/* remote's hostname */
 	const size_t		hostname_len;	/* it's length */
-	const struct sockaddr	*src_sap;	/* our address (optional) */
-	const size_t		src_len;	/* it's length */
 	const int		noresvport;	/* use non-priv port */
 };
 
@@ -88,127 +99,83 @@ static unsigned int nlm_hash_address(const struct sockaddr *sap)
 }
 
 /*
- * Common host lookup routine for server & client
+ * Allocate and initialize an nlm_host.  Common to both client and server.
  */
-static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni)
+static struct nlm_host *nlm_alloc_host(struct nlm_lookup_host_info *ni,
+				       struct nsm_handle *nsm)
 {
-	struct hlist_head *chain;
-	struct hlist_node *pos;
-	struct nlm_host	*host;
-	struct nsm_handle *nsm = NULL;
-
-	mutex_lock(&nlm_host_mutex);
+	struct nlm_host *host = NULL;
+	unsigned long now = jiffies;
 
-	if (time_after_eq(jiffies, next_gc))
-		nlm_gc_hosts();
-
-	/* We may keep several nlm_host objects for a peer, because each
-	 * nlm_host is identified by
-	 * (address, protocol, version, server/client)
-	 * We could probably simplify this a little by putting all those
-	 * different NLM rpc_clients into one single nlm_host object.
-	 * This would allow us to have one nlm_host per address.
-	 */
-	chain = &nlm_hosts[nlm_hash_address(ni->sap)];
-	hlist_for_each_entry(host, pos, chain, h_hash) {
-		if (!rpc_cmp_addr(nlm_addr(host), ni->sap))
-			continue;
-
-		/* See if we have an NSM handle for this client */
-		if (!nsm)
-			nsm = host->h_nsmhandle;
-
-		if (host->h_proto != ni->protocol)
-			continue;
-		if (host->h_version != ni->version)
-			continue;
-		if (host->h_server != ni->server)
-			continue;
-		if (ni->server && ni->src_len != 0 &&
-		    !rpc_cmp_addr(nlm_srcaddr(host), ni->src_sap))
-			continue;
-
-		/* Move to head of hash chain. */
-		hlist_del(&host->h_hash);
-		hlist_add_head(&host->h_hash, chain);
-
-		nlm_get_host(host);
-		dprintk("lockd: nlm_lookup_host found host %s (%s)\n",
-				host->h_name, host->h_addrbuf);
-		goto out;
-	}
-
-	/*
-	 * The host wasn't in our hash table.  If we don't
-	 * have an NSM handle for it yet, create one.
-	 */
-	if (nsm)
+	if (nsm != NULL)
 		atomic_inc(&nsm->sm_count);
 	else {
 		host = NULL;
 		nsm = nsm_get_handle(ni->sap, ni->salen,
 					ni->hostname, ni->hostname_len);
-		if (!nsm) {
-			dprintk("lockd: nlm_lookup_host failed; "
-				"no nsm handle\n");
+		if (unlikely(nsm == NULL)) {
+			dprintk("lockd: %s failed; no nsm handle\n",
+				__func__);
 			goto out;
 		}
 	}
 
-	host = kzalloc(sizeof(*host), GFP_KERNEL);
-	if (!host) {
+	host = kmalloc(sizeof(*host), GFP_KERNEL);
+	if (unlikely(host == NULL)) {
+		dprintk("lockd: %s failed; no memory\n", __func__);
 		nsm_release(nsm);
-		dprintk("lockd: nlm_lookup_host failed; no memory\n");
 		goto out;
 	}
-	host->h_name	   = nsm->sm_name;
-	host->h_addrbuf    = nsm->sm_addrbuf;
+
 	memcpy(nlm_addr(host), ni->sap, ni->salen);
-	host->h_addrlen = ni->salen;
+	host->h_addrlen    = ni->salen;
 	rpc_set_port(nlm_addr(host), 0);
-	memcpy(nlm_srcaddr(host), ni->src_sap, ni->src_len);
-	host->h_srcaddrlen = ni->src_len;
+	host->h_srcaddrlen = 0;
+
+	host->h_rpcclnt    = NULL;
+	host->h_name	   = nsm->sm_name;
 	host->h_version    = ni->version;
 	host->h_proto      = ni->protocol;
-	host->h_rpcclnt    = NULL;
-	mutex_init(&host->h_mutex);
-	host->h_nextrebind = jiffies + NLM_HOST_REBIND;
-	host->h_expires    = jiffies + NLM_HOST_EXPIRE;
-	atomic_set(&host->h_count, 1);
+	host->h_reclaiming = 0;
+	host->h_server     = ni->server;
+	host->h_noresvport = ni->noresvport;
+	host->h_inuse      = 0;
 	init_waitqueue_head(&host->h_gracewait);
 	init_rwsem(&host->h_rwsem);
-	host->h_state      = 0;			/* pseudo NSM state */
-	host->h_nsmstate   = 0;			/* real NSM state */
-	host->h_nsmhandle  = nsm;
-	host->h_server	   = ni->server;
-	host->h_noresvport = ni->noresvport;
-	hlist_add_head(&host->h_hash, chain);
+	host->h_state      = 0;
+	host->h_nsmstate   = 0;
+	host->h_pidcount   = 0;
+	atomic_set(&host->h_count, 1);
+	mutex_init(&host->h_mutex);
+	host->h_nextrebind = now + NLM_HOST_REBIND;
+	host->h_expires    = now + NLM_HOST_EXPIRE;
 	INIT_LIST_HEAD(&host->h_lockowners);
 	spin_lock_init(&host->h_lock);
 	INIT_LIST_HEAD(&host->h_granted);
 	INIT_LIST_HEAD(&host->h_reclaim);
-
-	nrhosts++;
-
-	dprintk("lockd: nlm_lookup_host created host %s\n",
-			host->h_name);
+	host->h_nsmhandle  = nsm;
+	host->h_addrbuf    = nsm->sm_addrbuf;
 
 out:
-	mutex_unlock(&nlm_host_mutex);
 	return host;
 }
 
 /*
- * Destroy a host
+ * Destroy an nlm_host and free associated resources
+ *
+ * Caller must hold nlm_host_mutex.
  */
-static void
-nlm_destroy_host(struct nlm_host *host)
+static void nlm_destroy_host_locked(struct nlm_host *host)
 {
 	struct rpc_clnt	*clnt;
 
+	dprintk("lockd: destroy host %s\n", host->h_name);
+
 	BUG_ON(!list_empty(&host->h_lockowners));
 	BUG_ON(atomic_read(&host->h_count));
 
+	hlist_del_init(&host->h_hash);
+
 	nsm_unmonitor(host);
 	nsm_release(host->h_nsmhandle);
 
@@ -216,6 +183,8 @@ nlm_destroy_host(struct nlm_host *host)
 	if (clnt != NULL)
 		rpc_shutdown_client(clnt);
 	kfree(host);
+
+	nrhosts--;
 }
 
 /**
@@ -249,12 +218,76 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap,
 		.hostname_len	= strlen(hostname),
 		.noresvport	= noresvport,
 	};
+	struct hlist_head *chain;
+	struct hlist_node *pos;
+	struct nlm_host	*host;
+	struct nsm_handle *nsm = NULL;
 
 	dprintk("lockd: %s(host='%s', vers=%u, proto=%s)\n", __func__,
 			(hostname ? hostname : "<none>"), version,
 			(protocol == IPPROTO_UDP ? "udp" : "tcp"));
 
-	return nlm_lookup_host(&ni);
+	mutex_lock(&nlm_host_mutex);
+
+	chain = &nlm_client_hosts[nlm_hash_address(sap)];
+	hlist_for_each_entry(host, pos, chain, h_hash) {
+		if (!rpc_cmp_addr(nlm_addr(host), sap))
+			continue;
+
+		/* Same address. Share an NSM handle if we already have one */
+		if (nsm == NULL)
+			nsm = host->h_nsmhandle;
+
+		if (host->h_proto != protocol)
+			continue;
+		if (host->h_version != version)
+			continue;
+
+		nlm_get_host(host);
+		dprintk("lockd: %s found host %s (%s)\n", __func__,
+			host->h_name, host->h_addrbuf);
+		goto out;
+	}
+
+	host = nlm_alloc_host(&ni, nsm);
+	if (unlikely(host == NULL))
+		goto out;
+
+	hlist_add_head(&host->h_hash, chain);
+	nrhosts++;
+
+	dprintk("lockd: %s created host %s (%s)\n", __func__,
+		host->h_name, host->h_addrbuf);
+
+out:
+	mutex_unlock(&nlm_host_mutex);
+	return host;
+}
+
+/**
+ * nlmclnt_release_host - release client nlm_host
+ * @host: nlm_host to release
+ *
+ */
+void nlmclnt_release_host(struct nlm_host *host)
+{
+	if (host == NULL)
+		return;
+
+	dprintk("lockd: release client host %s\n", host->h_name);
+
+	BUG_ON(atomic_read(&host->h_count) < 0);
+	BUG_ON(host->h_server);
+
+	if (atomic_dec_and_test(&host->h_count)) {
+		BUG_ON(!list_empty(&host->h_lockowners));
+		BUG_ON(!list_empty(&host->h_granted));
+		BUG_ON(!list_empty(&host->h_reclaim));
+
+		mutex_lock(&nlm_host_mutex);
+		nlm_destroy_host_locked(host);
+		mutex_unlock(&nlm_host_mutex);
+	}
 }
 
 /**
@@ -279,12 +312,18 @@ struct nlm_host *nlmsvc_lookup_host(const struct svc_rqst *rqstp,
 				    const char *hostname,
 				    const size_t hostname_len)
 {
+	struct hlist_head *chain;
+	struct hlist_node *pos;
+	struct nlm_host	*host = NULL;
+	struct nsm_handle *nsm = NULL;
 	struct sockaddr_in sin = {
 		.sin_family	= AF_INET,
 	};
 	struct sockaddr_in6 sin6 = {
 		.sin6_family	= AF_INET6,
 	};
+	struct sockaddr *src_sap;
+	size_t src_len = rqstp->rq_addrlen;
 	struct nlm_lookup_host_info ni = {
 		.server		= 1,
 		.sap		= svc_addr(rqstp),
@@ -293,27 +332,91 @@ struct nlm_host *nlmsvc_lookup_host(const struct svc_rqst *rqstp,
 		.version	= rqstp->rq_vers,
 		.hostname	= hostname,
 		.hostname_len	= hostname_len,
-		.src_len	= rqstp->rq_addrlen,
 	};
 
 	dprintk("lockd: %s(host='%*s', vers=%u, proto=%s)\n", __func__,
 			(int)hostname_len, hostname, rqstp->rq_vers,
 			(rqstp->rq_prot == IPPROTO_UDP ? "udp" : "tcp"));
 
+	mutex_lock(&nlm_host_mutex);
+
 	switch (ni.sap->sa_family) {
 	case AF_INET:
 		sin.sin_addr.s_addr = rqstp->rq_daddr.addr.s_addr;
-		ni.src_sap = (struct sockaddr *)&sin;
+		src_sap = (struct sockaddr *)&sin;
 		break;
 	case AF_INET6:
 		ipv6_addr_copy(&sin6.sin6_addr, &rqstp->rq_daddr.addr6);
-		ni.src_sap = (struct sockaddr *)&sin6;
+		src_sap = (struct sockaddr *)&sin6;
 		break;
 	default:
-		return NULL;
+		dprintk("lockd: %s failed; unrecognized address family\n",
+			__func__);
+		goto out;
+	}
+
+	if (time_after_eq(jiffies, next_gc))
+		nlm_gc_hosts();
+
+	chain = &nlm_server_hosts[nlm_hash_address(ni.sap)];
+	hlist_for_each_entry(host, pos, chain, h_hash) {
+		if (!rpc_cmp_addr(nlm_addr(host), ni.sap))
+			continue;
+
+		/* Same address. Share an NSM handle if we already have one */
+		if (nsm == NULL)
+			nsm = host->h_nsmhandle;
+
+		if (host->h_proto != ni.protocol)
+			continue;
+		if (host->h_version != ni.version)
+			continue;
+		if (!rpc_cmp_addr(nlm_srcaddr(host), src_sap))
+			continue;
+
+		/* Move to head of hash chain. */
+		hlist_del(&host->h_hash);
+		hlist_add_head(&host->h_hash, chain);
+
+		nlm_get_host(host);
+		dprintk("lockd: %s found host %s (%s)\n",
+			__func__, host->h_name, host->h_addrbuf);
+		goto out;
 	}
 
-	return nlm_lookup_host(&ni);
+	host = nlm_alloc_host(&ni, nsm);
+	if (unlikely(host == NULL))
+		goto out;
+
+	memcpy(nlm_srcaddr(host), src_sap, src_len);
+	host->h_srcaddrlen = src_len;
+	hlist_add_head(&host->h_hash, chain);
+	nrhosts++;
+
+	dprintk("lockd: %s created host %s (%s)\n",
+		__func__, host->h_name, host->h_addrbuf);
+
+out:
+	mutex_unlock(&nlm_host_mutex);
+	return host;
+}
+
+/**
+ * nlmsvc_release_host - release server nlm_host
+ * @host: nlm_host to release
+ *
+ * Host is destroyed later in nlm_gc_host().
+ */
+void nlmsvc_release_host(struct nlm_host *host)
+{
+	if (host == NULL)
+		return;
+
+	dprintk("lockd: release server host %s\n", host->h_name);
+
+	BUG_ON(atomic_read(&host->h_count) < 0);
+	BUG_ON(!host->h_server);
+	atomic_dec(&host->h_count);
 }
 
 /*
@@ -413,20 +516,28 @@ struct nlm_host * nlm_get_host(struct nlm_host *host)
 	return host;
 }
 
-/*
- * Release NLM host after use
- */
-void nlm_release_host(struct nlm_host *host)
+static struct nlm_host *next_host_state(struct hlist_head *cache,
+					struct nsm_handle *nsm,
+					const struct nlm_reboot *info)
 {
-	if (host != NULL) {
-		dprintk("lockd: release host %s\n", host->h_name);
-		BUG_ON(atomic_read(&host->h_count) < 0);
-		if (atomic_dec_and_test(&host->h_count)) {
-			BUG_ON(!list_empty(&host->h_lockowners));
-			BUG_ON(!list_empty(&host->h_granted));
-			BUG_ON(!list_empty(&host->h_reclaim));
+	struct nlm_host *host = NULL;
+	struct hlist_head *chain;
+	struct hlist_node *pos;
+
+	mutex_lock(&nlm_host_mutex);
+	for_each_host(host, pos, chain, cache) {
+		if (host->h_nsmhandle == nsm
+		    && host->h_nsmstate != info->state) {
+			host->h_nsmstate = info->state;
+			host->h_state++;
+
+			nlm_get_host(host);
+			goto out;
 		}
 	}
+out:
+	mutex_unlock(&nlm_host_mutex);
+	return host;
 }
 
 /**
@@ -438,8 +549,6 @@ void nlm_release_host(struct nlm_host *host)
  */
 void nlm_host_rebooted(const struct nlm_reboot *info)
 {
-	struct hlist_head *chain;
-	struct hlist_node *pos;
 	struct nsm_handle *nsm;
 	struct nlm_host	*host;
 
@@ -452,32 +561,15 @@ void nlm_host_rebooted(const struct nlm_reboot *info)
 	 * lock for this.
 	 * To avoid processing a host several times, we match the nsmstate.
 	 */
-again:	mutex_lock(&nlm_host_mutex);
-	for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) {
-		hlist_for_each_entry(host, pos, chain, h_hash) {
-			if (host->h_nsmhandle == nsm
-			 && host->h_nsmstate != info->state) {
-				host->h_nsmstate = info->state;
-				host->h_state++;
-
-				nlm_get_host(host);
-				mutex_unlock(&nlm_host_mutex);
-
-				if (host->h_server) {
-					/* We're server for this guy, just ditch
-					 * all the locks he held. */
-					nlmsvc_free_host_resources(host);
-				} else {
-					/* He's the server, initiate lock recovery. */
-					nlmclnt_recovery(host);
-				}
-
-				nlm_release_host(host);
-				goto again;
-			}
-		}
+	while ((host = next_host_state(nlm_server_hosts, nsm, info)) != NULL) {
+		nlmsvc_free_host_resources(host);
+		nlmsvc_release_host(host);
 	}
-	mutex_unlock(&nlm_host_mutex);
+	while ((host = next_host_state(nlm_client_hosts, nsm, info)) != NULL) {
+		nlmclnt_recovery(host);
+		nlmclnt_release_host(host);
+	}
+
 	nsm_release(nsm);
 }
 
@@ -497,13 +589,11 @@ nlm_shutdown_hosts(void)
 
 	/* First, make all hosts eligible for gc */
 	dprintk("lockd: nuking all hosts...\n");
-	for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) {
-		hlist_for_each_entry(host, pos, chain, h_hash) {
-			host->h_expires = jiffies - 1;
-			if (host->h_rpcclnt) {
-				rpc_shutdown_client(host->h_rpcclnt);
-				host->h_rpcclnt = NULL;
-			}
+	for_each_host(host, pos, chain, nlm_server_hosts) {
+		host->h_expires = jiffies - 1;
+		if (host->h_rpcclnt) {
+			rpc_shutdown_client(host->h_rpcclnt);
+			host->h_rpcclnt = NULL;
 		}
 	}
 
@@ -512,15 +602,13 @@ nlm_shutdown_hosts(void)
 	mutex_unlock(&nlm_host_mutex);
 
 	/* complain if any hosts are left */
-	if (nrhosts) {
+	if (nrhosts != 0) {
 		printk(KERN_WARNING "lockd: couldn't shutdown host module!\n");
-		dprintk("lockd: %d hosts left:\n", nrhosts);
-		for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) {
-			hlist_for_each_entry(host, pos, chain, h_hash) {
-				dprintk("       %s (cnt %d use %d exp %ld)\n",
-					host->h_name, atomic_read(&host->h_count),
-					host->h_inuse, host->h_expires);
-			}
+		dprintk("lockd: %lu hosts left:\n", nrhosts);
+		for_each_host(host, pos, chain, nlm_server_hosts) {
+			dprintk("       %s (cnt %d use %d exp %ld)\n",
+				host->h_name, atomic_read(&host->h_count),
+				host->h_inuse, host->h_expires);
 		}
 	}
 }
@@ -538,29 +626,22 @@ nlm_gc_hosts(void)
 	struct nlm_host	*host;
 
 	dprintk("lockd: host garbage collection\n");
-	for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) {
-		hlist_for_each_entry(host, pos, chain, h_hash)
-			host->h_inuse = 0;
-	}
+	for_each_host(host, pos, chain, nlm_server_hosts)
+		host->h_inuse = 0;
 
 	/* Mark all hosts that hold locks, blocks or shares */
 	nlmsvc_mark_resources();
 
-	for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) {
-		hlist_for_each_entry_safe(host, pos, next, chain, h_hash) {
-			if (atomic_read(&host->h_count) || host->h_inuse
-			 || time_before(jiffies, host->h_expires)) {
-				dprintk("nlm_gc_hosts skipping %s (cnt %d use %d exp %ld)\n",
-					host->h_name, atomic_read(&host->h_count),
-					host->h_inuse, host->h_expires);
-				continue;
-			}
-			dprintk("lockd: delete host %s\n", host->h_name);
-			hlist_del_init(&host->h_hash);
-
-			nlm_destroy_host(host);
-			nrhosts--;
+	for_each_host_safe(host, pos, next, chain, nlm_server_hosts) {
+		if (atomic_read(&host->h_count) || host->h_inuse
+		 || time_before(jiffies, host->h_expires)) {
+			dprintk("nlm_gc_hosts skipping %s "
+				"(cnt %d use %d exp %ld)\n",
+				host->h_name, atomic_read(&host->h_count),
+				host->h_inuse, host->h_expires);
+			continue;
 		}
+		nlm_destroy_host_locked(host);
 	}
 
 	next_gc = jiffies + NLM_HOST_COLLECT;

fs/lockd/mon.c

@@ -401,26 +401,22 @@ void nsm_release(struct nsm_handle *nsm)
  * Status Monitor wire protocol.
  */
 
-static int encode_nsm_string(struct xdr_stream *xdr, const char *string)
+static void encode_nsm_string(struct xdr_stream *xdr, const char *string)
 {
 	const u32 len = strlen(string);
 	__be32 *p;
 
-	if (unlikely(len > SM_MAXSTRLEN))
-		return -EIO;
-	p = xdr_reserve_space(xdr, sizeof(u32) + len);
-	if (unlikely(p == NULL))
-		return -EIO;
+	BUG_ON(len > SM_MAXSTRLEN);
+	p = xdr_reserve_space(xdr, 4 + len);
 	xdr_encode_opaque(p, string, len);
-	return 0;
 }
 
 /*
  * "mon_name" specifies the host to be monitored.
  */
-static int encode_mon_name(struct xdr_stream *xdr, const struct nsm_args *argp)
+static void encode_mon_name(struct xdr_stream *xdr, const struct nsm_args *argp)
 {
-	return encode_nsm_string(xdr, argp->mon_name);
+	encode_nsm_string(xdr, argp->mon_name);
 }
 
 /*
@@ -429,35 +425,25 @@ static int encode_mon_name(struct xdr_stream *xdr, const struct nsm_args *argp)
  * (via the NLMPROC_SM_NOTIFY call) that the state of host "mon_name"
  * has changed.
  */
-static int encode_my_id(struct xdr_stream *xdr, const struct nsm_args *argp)
+static void encode_my_id(struct xdr_stream *xdr, const struct nsm_args *argp)
 {
-	int status;
 	__be32 *p;
 
-	status = encode_nsm_string(xdr, utsname()->nodename);
-	if (unlikely(status != 0))
-		return status;
-	p = xdr_reserve_space(xdr, 3 * sizeof(u32));
-	if (unlikely(p == NULL))
-		return -EIO;
-	*p++ = htonl(argp->prog);
-	*p++ = htonl(argp->vers);
-	*p++ = htonl(argp->proc);
-	return 0;
+	encode_nsm_string(xdr, utsname()->nodename);
+	p = xdr_reserve_space(xdr, 4 + 4 + 4);
+	*p++ = cpu_to_be32(argp->prog);
+	*p++ = cpu_to_be32(argp->vers);
+	*p = cpu_to_be32(argp->proc);
 }
 
 /*
  * The "mon_id" argument specifies the non-private arguments
  * of an NSMPROC_MON or NSMPROC_UNMON call.
  */
-static int encode_mon_id(struct xdr_stream *xdr, const struct nsm_args *argp)
+static void encode_mon_id(struct xdr_stream *xdr, const struct nsm_args *argp)
 {
-	int status;
-
-	status = encode_mon_name(xdr, argp);
-	if (unlikely(status != 0))
-		return status;
-	return encode_my_id(xdr, argp);
+	encode_mon_name(xdr, argp);
+	encode_my_id(xdr, argp);
 }
 
 /*
@@ -465,68 +451,56 @@ static int encode_mon_id(struct xdr_stream *xdr, const struct nsm_args *argp)
  * by the NSMPROC_MON call. This information will be supplied in the
  * NLMPROC_SM_NOTIFY call.
  */
-static int encode_priv(struct xdr_stream *xdr, const struct nsm_args *argp)
+static void encode_priv(struct xdr_stream *xdr, const struct nsm_args *argp)
 {
 	__be32 *p;
 
 	p = xdr_reserve_space(xdr, SM_PRIV_SIZE);
-	if (unlikely(p == NULL))
-		return -EIO;
 	xdr_encode_opaque_fixed(p, argp->priv->data, SM_PRIV_SIZE);
-	return 0;
 }
 
-static int xdr_enc_mon(struct rpc_rqst *req, __be32 *p,
-		       const struct nsm_args *argp)
+static void nsm_xdr_enc_mon(struct rpc_rqst *req, struct xdr_stream *xdr,
+			    const struct nsm_args *argp)
 {
-	struct xdr_stream xdr;
-	int status;
-
-	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
-	status = encode_mon_id(&xdr, argp);
-	if (unlikely(status))
-		return status;
-	return encode_priv(&xdr, argp);
+	encode_mon_id(xdr, argp);
+	encode_priv(xdr, argp);
 }
 
-static int xdr_enc_unmon(struct rpc_rqst *req, __be32 *p,
-			 const struct nsm_args *argp)
+static void nsm_xdr_enc_unmon(struct rpc_rqst *req, struct xdr_stream *xdr,
+			      const struct nsm_args *argp)
 {
-	struct xdr_stream xdr;
-
-	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
-	return encode_mon_id(&xdr, argp);
+	encode_mon_id(xdr, argp);
 }
 
-static int xdr_dec_stat_res(struct rpc_rqst *rqstp, __be32 *p,
-			    struct nsm_res *resp)
+static int nsm_xdr_dec_stat_res(struct rpc_rqst *rqstp,
+				struct xdr_stream *xdr,
+				struct nsm_res *resp)
 {
-	struct xdr_stream xdr;
+	__be32 *p;
 
-	xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p);
-	p = xdr_inline_decode(&xdr, 2 * sizeof(u32));
+	p = xdr_inline_decode(xdr, 4 + 4);
 	if (unlikely(p == NULL))
 		return -EIO;
-	resp->status = ntohl(*p++);
-	resp->state = ntohl(*p);
+	resp->status = be32_to_cpup(p++);
+	resp->state = be32_to_cpup(p);
 
-	dprintk("lockd: xdr_dec_stat_res status %d state %d\n",
-			resp->status, resp->state);
+	dprintk("lockd: %s status %d state %d\n",
+		__func__, resp->status, resp->state);
 	return 0;
 }
 
-static int xdr_dec_stat(struct rpc_rqst *rqstp, __be32 *p,
-			struct nsm_res *resp)
+static int nsm_xdr_dec_stat(struct rpc_rqst *rqstp,
+			    struct xdr_stream *xdr,
+			    struct nsm_res *resp)
 {
-	struct xdr_stream xdr;
+	__be32 *p;
 
-	xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p);
-	p = xdr_inline_decode(&xdr, sizeof(u32));
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return -EIO;
-	resp->state = ntohl(*p);
+	resp->state = be32_to_cpup(p);
 
-	dprintk("lockd: xdr_dec_stat state %d\n", resp->state);
+	dprintk("lockd: %s state %d\n", __func__, resp->state);
 	return 0;
 }
 
@@ -542,8 +516,8 @@ static int xdr_dec_stat(struct rpc_rqst *rqstp, __be32 *p,
 static struct rpc_procinfo	nsm_procedures[] = {
 [NSMPROC_MON] = {
 		.p_proc		= NSMPROC_MON,
-		.p_encode	= (kxdrproc_t)xdr_enc_mon,
-		.p_decode	= (kxdrproc_t)xdr_dec_stat_res,
+		.p_encode	= (kxdreproc_t)nsm_xdr_enc_mon,
+		.p_decode	= (kxdrdproc_t)nsm_xdr_dec_stat_res,
 		.p_arglen	= SM_mon_sz,
 		.p_replen	= SM_monres_sz,
 		.p_statidx	= NSMPROC_MON,
@@ -551,8 +525,8 @@ static struct rpc_procinfo	nsm_procedures[] = {
 	},
 [NSMPROC_UNMON] = {
 		.p_proc		= NSMPROC_UNMON,
-		.p_encode	= (kxdrproc_t)xdr_enc_unmon,
-		.p_decode	= (kxdrproc_t)xdr_dec_stat,
+		.p_encode	= (kxdreproc_t)nsm_xdr_enc_unmon,
+		.p_decode	= (kxdrdproc_t)nsm_xdr_dec_stat,
 		.p_arglen	= SM_mon_id_sz,
 		.p_replen	= SM_unmonres_sz,
 		.p_statidx	= NSMPROC_UNMON,

fs/lockd/svc4proc.c

@@ -51,7 +51,7 @@ nlm4svc_retrieve_args(struct svc_rqst *rqstp, struct nlm_args *argp,
 	return 0;
 
 no_locks:
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
  	if (error)
 		return error;	
 	return nlm_lck_denied_nolocks;
@@ -92,7 +92,7 @@ nlm4svc_proc_test(struct svc_rqst *rqstp, struct nlm_args *argp,
 	else
 		dprintk("lockd: TEST4        status %d\n", ntohl(resp->status));
 
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
 }
@@ -134,7 +134,7 @@ nlm4svc_proc_lock(struct svc_rqst *rqstp, struct nlm_args *argp,
 	else
 		dprintk("lockd: LOCK         status %d\n", ntohl(resp->status));
 
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
 }
@@ -164,7 +164,7 @@ nlm4svc_proc_cancel(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = nlmsvc_cancel_blocked(file, &argp->lock);
 
 	dprintk("lockd: CANCEL        status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -197,7 +197,7 @@ nlm4svc_proc_unlock(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = nlmsvc_unlock(file, &argp->lock);
 
 	dprintk("lockd: UNLOCK        status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -229,7 +229,7 @@ static void nlm4svc_callback_exit(struct rpc_task *task, void *data)
 
 static void nlm4svc_callback_release(void *data)
 {
-	nlm_release_call(data);
+	nlmsvc_release_call(data);
 }
 
 static const struct rpc_call_ops nlm4svc_callback_ops = {
@@ -261,7 +261,7 @@ static __be32 nlm4svc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_args
 
 	stat = func(rqstp, argp, &call->a_res);
 	if (stat != 0) {
-		nlm_release_call(call);
+		nlmsvc_release_call(call);
 		return stat;
 	}
 
@@ -334,7 +334,7 @@ nlm4svc_proc_share(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = nlmsvc_share_file(host, file, argp);
 
 	dprintk("lockd: SHARE         status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -367,7 +367,7 @@ nlm4svc_proc_unshare(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = nlmsvc_unshare_file(host, file, argp);
 
 	dprintk("lockd: UNSHARE       status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -399,7 +399,7 @@ nlm4svc_proc_free_all(struct svc_rqst *rqstp, struct nlm_args *argp,
 		return rpc_success;
 
 	nlmsvc_free_host_resources(host);
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	return rpc_success;
 }
 

fs/lockd/svclock.c

@@ -46,6 +46,7 @@ static void	nlmsvc_remove_block(struct nlm_block *block);
 static int nlmsvc_setgrantargs(struct nlm_rqst *call, struct nlm_lock *lock);
 static void nlmsvc_freegrantargs(struct nlm_rqst *call);
 static const struct rpc_call_ops nlmsvc_grant_ops;
+static const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie);
 
 /*
  * The list of blocked locks to retry
@@ -233,7 +234,7 @@ nlmsvc_create_block(struct svc_rqst *rqstp, struct nlm_host *host,
 failed_free:
 	kfree(block);
 failed:
-	nlm_release_call(call);
+	nlmsvc_release_call(call);
 	return NULL;
 }
 
@@ -266,7 +267,7 @@ static void nlmsvc_free_block(struct kref *kref)
 	mutex_unlock(&file->f_mutex);
 
 	nlmsvc_freegrantargs(block->b_call);
-	nlm_release_call(block->b_call);
+	nlmsvc_release_call(block->b_call);
 	nlm_release_file(block->b_file);
 	kfree(block->b_fl);
 	kfree(block);
@@ -934,3 +935,32 @@ nlmsvc_retry_blocked(void)
 
 	return timeout;
 }
+
+#ifdef RPC_DEBUG
+static const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie)
+{
+	/*
+	 * We can get away with a static buffer because we're only
+	 * called with BKL held.
+	 */
+	static char buf[2*NLM_MAXCOOKIELEN+1];
+	unsigned int i, len = sizeof(buf);
+	char *p = buf;
+
+	len--;	/* allow for trailing \0 */
+	if (len < 3)
+		return "???";
+	for (i = 0 ; i < cookie->len ; i++) {
+		if (len < 2) {
+			strcpy(p-3, "...");
+			break;
+		}
+		sprintf(p, "%02x", cookie->data[i]);
+		p += 2;
+		len -= 2;
+	}
+	*p = '\0';
+
+	return buf;
+}
+#endif

fs/lockd/svcproc.c

@@ -80,7 +80,7 @@ nlmsvc_retrieve_args(struct svc_rqst *rqstp, struct nlm_args *argp,
 	return 0;
 
 no_locks:
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	if (error)
 		return error;
 	return nlm_lck_denied_nolocks;
@@ -122,7 +122,7 @@ nlmsvc_proc_test(struct svc_rqst *rqstp, struct nlm_args *argp,
 		dprintk("lockd: TEST          status %d vers %d\n",
 			ntohl(resp->status), rqstp->rq_vers);
 
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
 }
@@ -164,7 +164,7 @@ nlmsvc_proc_lock(struct svc_rqst *rqstp, struct nlm_args *argp,
 	else
 		dprintk("lockd: LOCK         status %d\n", ntohl(resp->status));
 
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rc;
 }
@@ -194,7 +194,7 @@ nlmsvc_proc_cancel(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = cast_status(nlmsvc_cancel_blocked(file, &argp->lock));
 
 	dprintk("lockd: CANCEL        status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -227,7 +227,7 @@ nlmsvc_proc_unlock(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = cast_status(nlmsvc_unlock(file, &argp->lock));
 
 	dprintk("lockd: UNLOCK        status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -257,9 +257,17 @@ static void nlmsvc_callback_exit(struct rpc_task *task, void *data)
 			-task->tk_status);
 }
 
+void nlmsvc_release_call(struct nlm_rqst *call)
+{
+	if (!atomic_dec_and_test(&call->a_count))
+		return;
+	nlmsvc_release_host(call->a_host);
+	kfree(call);
+}
+
 static void nlmsvc_callback_release(void *data)
 {
-	nlm_release_call(data);
+	nlmsvc_release_call(data);
 }
 
 static const struct rpc_call_ops nlmsvc_callback_ops = {
@@ -291,7 +299,7 @@ static __be32 nlmsvc_callback(struct svc_rqst *rqstp, u32 proc, struct nlm_args
 
 	stat = func(rqstp, argp, &call->a_res);
 	if (stat != 0) {
-		nlm_release_call(call);
+		nlmsvc_release_call(call);
 		return stat;
 	}
 
@@ -366,7 +374,7 @@ nlmsvc_proc_share(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = cast_status(nlmsvc_share_file(host, file, argp));
 
 	dprintk("lockd: SHARE         status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -399,7 +407,7 @@ nlmsvc_proc_unshare(struct svc_rqst *rqstp, struct nlm_args *argp,
 	resp->status = cast_status(nlmsvc_unshare_file(host, file, argp));
 
 	dprintk("lockd: UNSHARE       status %d\n", ntohl(resp->status));
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	nlm_release_file(file);
 	return rpc_success;
 }
@@ -431,7 +439,7 @@ nlmsvc_proc_free_all(struct svc_rqst *rqstp, struct nlm_args *argp,
 		return rpc_success;
 
 	nlmsvc_free_host_resources(host);
-	nlm_release_host(host);
+	nlmsvc_release_host(host);
 	return rpc_success;
 }
 

fs/lockd/xdr.c

@@ -148,37 +148,6 @@ nlm_decode_lock(__be32 *p, struct nlm_lock *lock)
 	return p;
 }
 
-/*
- * Encode a lock as part of an NLM call
- */
-static __be32 *
-nlm_encode_lock(__be32 *p, struct nlm_lock *lock)
-{
-	struct file_lock	*fl = &lock->fl;
-	__s32			start, len;
-
-	if (!(p = xdr_encode_string(p, lock->caller))
-	 || !(p = nlm_encode_fh(p, &lock->fh))
-	 || !(p = nlm_encode_oh(p, &lock->oh)))
-		return NULL;
-
-	if (fl->fl_start > NLM_OFFSET_MAX
-	 || (fl->fl_end > NLM_OFFSET_MAX && fl->fl_end != OFFSET_MAX))
-		return NULL;
-
-	start = loff_t_to_s32(fl->fl_start);
-	if (fl->fl_end == OFFSET_MAX)
-		len = 0;
-	else
-		len = loff_t_to_s32(fl->fl_end - fl->fl_start + 1);
-
-	*p++ = htonl(lock->svid);
-	*p++ = htonl(start);
-	*p++ = htonl(len);
-
-	return p;
-}
-
 /*
  * Encode result of a TEST/TEST_MSG call
  */
@@ -372,259 +341,3 @@ nlmsvc_encode_void(struct svc_rqst *rqstp, __be32 *p, void *dummy)
 {
 	return xdr_ressize_check(rqstp, p);
 }
-
-/*
- * Now, the client side XDR functions
- */
-#ifdef NLMCLNT_SUPPORT_SHARES
-static int
-nlmclt_decode_void(struct rpc_rqst *req, u32 *p, void *ptr)
-{
-	return 0;
-}
-#endif
-
-static int
-nlmclt_encode_testargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_decode_testres(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm_decode_cookie(p, &resp->cookie)))
-		return -EIO;
-	resp->status = *p++;
-	if (resp->status == nlm_lck_denied) {
-		struct file_lock	*fl = &resp->lock.fl;
-		u32			excl;
-		s32			start, len, end;
-
-		memset(&resp->lock, 0, sizeof(resp->lock));
-		locks_init_lock(fl);
-		excl = ntohl(*p++);
-		resp->lock.svid = ntohl(*p++);
-		fl->fl_pid = (pid_t)resp->lock.svid;
-		if (!(p = nlm_decode_oh(p, &resp->lock.oh)))
-			return -EIO;
-
-		fl->fl_flags = FL_POSIX;
-		fl->fl_type  = excl? F_WRLCK : F_RDLCK;
-		start = ntohl(*p++);
-		len = ntohl(*p++);
-		end = start + len - 1;
-
-		fl->fl_start = s32_to_loff_t(start);
-		if (len == 0 || end < 0)
-			fl->fl_end = OFFSET_MAX;
-		else
-			fl->fl_end = s32_to_loff_t(end);
-	}
-	return 0;
-}
-
-
-static int
-nlmclt_encode_lockargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = argp->block? xdr_one : xdr_zero;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm_encode_lock(p, lock)))
-		return -EIO;
-	*p++ = argp->reclaim? xdr_one : xdr_zero;
-	*p++ = htonl(argp->state);
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_encode_cancargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = argp->block? xdr_one : xdr_zero;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_encode_unlockargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	if (!(p = nlm_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_encode_res(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm_encode_cookie(p, &resp->cookie)))
-		return -EIO;
-	*p++ = resp->status;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_encode_testres(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm_encode_testres(p, resp)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlmclt_decode_res(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm_decode_cookie(p, &resp->cookie)))
-		return -EIO;
-	resp->status = *p++;
-	return 0;
-}
-
-#if (NLMCLNT_OHSIZE > XDR_MAX_NETOBJ)
-#  error "NLM host name cannot be larger than XDR_MAX_NETOBJ!"
-#endif
-
-/*
- * Buffer requirements for NLM
- */
-#define NLM_void_sz		0
-#define NLM_cookie_sz		1+XDR_QUADLEN(NLM_MAXCOOKIELEN)
-#define NLM_caller_sz		1+XDR_QUADLEN(NLMCLNT_OHSIZE)
-#define NLM_owner_sz		1+XDR_QUADLEN(NLMCLNT_OHSIZE)
-#define NLM_fhandle_sz		1+XDR_QUADLEN(NFS2_FHSIZE)
-#define NLM_lock_sz		3+NLM_caller_sz+NLM_owner_sz+NLM_fhandle_sz
-#define NLM_holder_sz		4+NLM_owner_sz
-
-#define NLM_testargs_sz		NLM_cookie_sz+1+NLM_lock_sz
-#define NLM_lockargs_sz		NLM_cookie_sz+4+NLM_lock_sz
-#define NLM_cancargs_sz		NLM_cookie_sz+2+NLM_lock_sz
-#define NLM_unlockargs_sz	NLM_cookie_sz+NLM_lock_sz
-
-#define NLM_testres_sz		NLM_cookie_sz+1+NLM_holder_sz
-#define NLM_res_sz		NLM_cookie_sz+1
-#define NLM_norep_sz		0
-
-/*
- * For NLM, a void procedure really returns nothing
- */
-#define nlmclt_decode_norep	NULL
-
-#define PROC(proc, argtype, restype)	\
-[NLMPROC_##proc] = {							\
-	.p_proc      = NLMPROC_##proc,					\
-	.p_encode    = (kxdrproc_t) nlmclt_encode_##argtype,		\
-	.p_decode    = (kxdrproc_t) nlmclt_decode_##restype,		\
-	.p_arglen    = NLM_##argtype##_sz,				\
-	.p_replen    = NLM_##restype##_sz,				\
-	.p_statidx   = NLMPROC_##proc,					\
-	.p_name      = #proc,						\
-	}
-
-static struct rpc_procinfo	nlm_procedures[] = {
-    PROC(TEST,		testargs,	testres),
-    PROC(LOCK,		lockargs,	res),
-    PROC(CANCEL,	cancargs,	res),
-    PROC(UNLOCK,	unlockargs,	res),
-    PROC(GRANTED,	testargs,	res),
-    PROC(TEST_MSG,	testargs,	norep),
-    PROC(LOCK_MSG,	lockargs,	norep),
-    PROC(CANCEL_MSG,	cancargs,	norep),
-    PROC(UNLOCK_MSG,	unlockargs,	norep),
-    PROC(GRANTED_MSG,	testargs,	norep),
-    PROC(TEST_RES,	testres,	norep),
-    PROC(LOCK_RES,	res,		norep),
-    PROC(CANCEL_RES,	res,		norep),
-    PROC(UNLOCK_RES,	res,		norep),
-    PROC(GRANTED_RES,	res,		norep),
-#ifdef NLMCLNT_SUPPORT_SHARES
-    PROC(SHARE,		shareargs,	shareres),
-    PROC(UNSHARE,	shareargs,	shareres),
-    PROC(NM_LOCK,	lockargs,	res),
-    PROC(FREE_ALL,	notify,		void),
-#endif
-};
-
-static struct rpc_version	nlm_version1 = {
-		.number		= 1,
-		.nrprocs	= 16,
-		.procs		= nlm_procedures,
-};
-
-static struct rpc_version	nlm_version3 = {
-		.number		= 3,
-		.nrprocs	= 24,
-		.procs		= nlm_procedures,
-};
-
-static struct rpc_version *	nlm_versions[] = {
-	[1] = &nlm_version1,
-	[3] = &nlm_version3,
-#ifdef 	CONFIG_LOCKD_V4
-	[4] = &nlm_version4,
-#endif
-};
-
-static struct rpc_stat		nlm_stats;
-
-struct rpc_program		nlm_program = {
-		.name		= "lockd",
-		.number		= NLM_PROGRAM,
-		.nrvers		= ARRAY_SIZE(nlm_versions),
-		.version	= nlm_versions,
-		.stats		= &nlm_stats,
-};
-
-#ifdef RPC_DEBUG
-const char *nlmdbg_cookie2a(const struct nlm_cookie *cookie)
-{
-	/*
-	 * We can get away with a static buffer because we're only
-	 * called with BKL held.
-	 */
-	static char buf[2*NLM_MAXCOOKIELEN+1];
-	unsigned int i, len = sizeof(buf);
-	char *p = buf;
-
-	len--;	/* allow for trailing \0 */
-	if (len < 3)
-		return "???";
-	for (i = 0 ; i < cookie->len ; i++) {
-		if (len < 2) {
-			strcpy(p-3, "...");
-			break;
-		}
-		sprintf(p, "%02x", cookie->data[i]);
-		p += 2;
-		len -= 2;
-	}
-	*p = '\0';
-
-	return buf;
-}
-#endif

fs/lockd/xdr4.c

@@ -93,15 +93,6 @@ nlm4_decode_fh(__be32 *p, struct nfs_fh *f)
 	return p + XDR_QUADLEN(f->size);
 }
 
-static __be32 *
-nlm4_encode_fh(__be32 *p, struct nfs_fh *f)
-{
-	*p++ = htonl(f->size);
-	if (f->size) p[XDR_QUADLEN(f->size)-1] = 0; /* don't leak anything */
-	memcpy(p, f->data, f->size);
-	return p + XDR_QUADLEN(f->size);
-}
-
 /*
  * Encode and decode owner handle
  */
@@ -111,12 +102,6 @@ nlm4_decode_oh(__be32 *p, struct xdr_netobj *oh)
 	return xdr_decode_netobj(p, oh);
 }
 
-static __be32 *
-nlm4_encode_oh(__be32 *p, struct xdr_netobj *oh)
-{
-	return xdr_encode_netobj(p, oh);
-}
-
 static __be32 *
 nlm4_decode_lock(__be32 *p, struct nlm_lock *lock)
 {
@@ -149,38 +134,6 @@ nlm4_decode_lock(__be32 *p, struct nlm_lock *lock)
 	return p;
 }
 
-/*
- * Encode a lock as part of an NLM call
- */
-static __be32 *
-nlm4_encode_lock(__be32 *p, struct nlm_lock *lock)
-{
-	struct file_lock	*fl = &lock->fl;
-	__s64			start, len;
-
-	if (!(p = xdr_encode_string(p, lock->caller))
-	 || !(p = nlm4_encode_fh(p, &lock->fh))
-	 || !(p = nlm4_encode_oh(p, &lock->oh)))
-		return NULL;
-
-	if (fl->fl_start > NLM4_OFFSET_MAX
-	 || (fl->fl_end > NLM4_OFFSET_MAX && fl->fl_end != OFFSET_MAX))
-		return NULL;
-
-	*p++ = htonl(lock->svid);
-
-	start = loff_t_to_s64(fl->fl_start);
-	if (fl->fl_end == OFFSET_MAX)
-		len = 0;
-	else
-		len = loff_t_to_s64(fl->fl_end - fl->fl_start + 1);
-
-	p = xdr_encode_hyper(p, start);
-	p = xdr_encode_hyper(p, len);
-
-	return p;
-}
-
 /*
  * Encode result of a TEST/TEST_MSG call
  */
@@ -379,211 +332,3 @@ nlm4svc_encode_void(struct svc_rqst *rqstp, __be32 *p, void *dummy)
 {
 	return xdr_ressize_check(rqstp, p);
 }
-
-/*
- * Now, the client side XDR functions
- */
-#ifdef NLMCLNT_SUPPORT_SHARES
-static int
-nlm4clt_decode_void(struct rpc_rqst *req, __be32 *p, void *ptr)
-{
-	return 0;
-}
-#endif
-
-static int
-nlm4clt_encode_testargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm4_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm4_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_decode_testres(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm4_decode_cookie(p, &resp->cookie)))
-		return -EIO;
-	resp->status = *p++;
-	if (resp->status == nlm_lck_denied) {
-		struct file_lock	*fl = &resp->lock.fl;
-		u32			excl;
-		__u64			start, len;
-		__s64			end;
-
-		memset(&resp->lock, 0, sizeof(resp->lock));
-		locks_init_lock(fl);
-		excl = ntohl(*p++);
-		resp->lock.svid = ntohl(*p++);
-		fl->fl_pid = (pid_t)resp->lock.svid;
-		if (!(p = nlm4_decode_oh(p, &resp->lock.oh)))
-			return -EIO;
-
-		fl->fl_flags = FL_POSIX;
-		fl->fl_type  = excl? F_WRLCK : F_RDLCK;
-		p = xdr_decode_hyper(p, &start);
-		p = xdr_decode_hyper(p, &len);
-		end = start + len - 1;
-
-		fl->fl_start = s64_to_loff_t(start);
-		if (len == 0 || end < 0)
-			fl->fl_end = OFFSET_MAX;
-		else
-			fl->fl_end = s64_to_loff_t(end);
-	}
-	return 0;
-}
-
-
-static int
-nlm4clt_encode_lockargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm4_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = argp->block? xdr_one : xdr_zero;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm4_encode_lock(p, lock)))
-		return -EIO;
-	*p++ = argp->reclaim? xdr_one : xdr_zero;
-	*p++ = htonl(argp->state);
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_encode_cancargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm4_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	*p++ = argp->block? xdr_one : xdr_zero;
-	*p++ = (lock->fl.fl_type == F_WRLCK)? xdr_one : xdr_zero;
-	if (!(p = nlm4_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_encode_unlockargs(struct rpc_rqst *req, __be32 *p, nlm_args *argp)
-{
-	struct nlm_lock	*lock = &argp->lock;
-
-	if (!(p = nlm4_encode_cookie(p, &argp->cookie)))
-		return -EIO;
-	if (!(p = nlm4_encode_lock(p, lock)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_encode_res(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm4_encode_cookie(p, &resp->cookie)))
-		return -EIO;
-	*p++ = resp->status;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_encode_testres(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm4_encode_testres(p, resp)))
-		return -EIO;
-	req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
-	return 0;
-}
-
-static int
-nlm4clt_decode_res(struct rpc_rqst *req, __be32 *p, struct nlm_res *resp)
-{
-	if (!(p = nlm4_decode_cookie(p, &resp->cookie)))
-		return -EIO;
-	resp->status = *p++;
-	return 0;
-}
-
-#if (NLMCLNT_OHSIZE > XDR_MAX_NETOBJ)
-#  error "NLM host name cannot be larger than XDR_MAX_NETOBJ!"
-#endif
-
-#if (NLMCLNT_OHSIZE > NLM_MAXSTRLEN)
-#  error "NLM host name cannot be larger than NLM's maximum string length!"
-#endif
-
-/*
- * Buffer requirements for NLM
- */
-#define NLM4_void_sz		0
-#define NLM4_cookie_sz		1+XDR_QUADLEN(NLM_MAXCOOKIELEN)
-#define NLM4_caller_sz		1+XDR_QUADLEN(NLMCLNT_OHSIZE)
-#define NLM4_owner_sz		1+XDR_QUADLEN(NLMCLNT_OHSIZE)
-#define NLM4_fhandle_sz		1+XDR_QUADLEN(NFS3_FHSIZE)
-#define NLM4_lock_sz		5+NLM4_caller_sz+NLM4_owner_sz+NLM4_fhandle_sz
-#define NLM4_holder_sz		6+NLM4_owner_sz
-
-#define NLM4_testargs_sz	NLM4_cookie_sz+1+NLM4_lock_sz
-#define NLM4_lockargs_sz	NLM4_cookie_sz+4+NLM4_lock_sz
-#define NLM4_cancargs_sz	NLM4_cookie_sz+2+NLM4_lock_sz
-#define NLM4_unlockargs_sz	NLM4_cookie_sz+NLM4_lock_sz
-
-#define NLM4_testres_sz		NLM4_cookie_sz+1+NLM4_holder_sz
-#define NLM4_res_sz		NLM4_cookie_sz+1
-#define NLM4_norep_sz		0
-
-/*
- * For NLM, a void procedure really returns nothing
- */
-#define nlm4clt_decode_norep	NULL
-
-#define PROC(proc, argtype, restype)					\
-[NLMPROC_##proc] = {							\
-	.p_proc      = NLMPROC_##proc,					\
-	.p_encode    = (kxdrproc_t) nlm4clt_encode_##argtype,		\
-	.p_decode    = (kxdrproc_t) nlm4clt_decode_##restype,		\
-	.p_arglen    = NLM4_##argtype##_sz,				\
-	.p_replen    = NLM4_##restype##_sz,				\
-	.p_statidx   = NLMPROC_##proc,					\
-	.p_name      = #proc,						\
-	}
-
-static struct rpc_procinfo	nlm4_procedures[] = {
-    PROC(TEST,		testargs,	testres),
-    PROC(LOCK,		lockargs,	res),
-    PROC(CANCEL,	cancargs,	res),
-    PROC(UNLOCK,	unlockargs,	res),
-    PROC(GRANTED,	testargs,	res),
-    PROC(TEST_MSG,	testargs,	norep),
-    PROC(LOCK_MSG,	lockargs,	norep),
-    PROC(CANCEL_MSG,	cancargs,	norep),
-    PROC(UNLOCK_MSG,	unlockargs,	norep),
-    PROC(GRANTED_MSG,	testargs,	norep),
-    PROC(TEST_RES,	testres,	norep),
-    PROC(LOCK_RES,	res,		norep),
-    PROC(CANCEL_RES,	res,		norep),
-    PROC(UNLOCK_RES,	res,		norep),
-    PROC(GRANTED_RES,	res,		norep),
-#ifdef NLMCLNT_SUPPORT_SHARES
-    PROC(SHARE,		shareargs,	shareres),
-    PROC(UNSHARE,	shareargs,	shareres),
-    PROC(NM_LOCK,	lockargs,	res),
-    PROC(FREE_ALL,	notify,		void),
-#endif
-};
-
-struct rpc_version	nlm_version4 = {
-	.number		= 4,
-	.nrprocs	= 24,
-	.procs		= nlm4_procedures,
-};

fs/nfs/callback.c

@@ -16,9 +16,7 @@
 #include <linux/freezer.h>
 #include <linux/kthread.h>
 #include <linux/sunrpc/svcauth_gss.h>
-#if defined(CONFIG_NFS_V4_1)
 #include <linux/sunrpc/bc_xprt.h>
-#endif
 
 #include <net/inet_sock.h>
 
@@ -136,6 +134,33 @@ nfs4_callback_up(struct svc_serv *serv)
 }
 
 #if defined(CONFIG_NFS_V4_1)
+/*
+ *  * CB_SEQUENCE operations will fail until the callback sessionid is set.
+ *   */
+int nfs4_set_callback_sessionid(struct nfs_client *clp)
+{
+	struct svc_serv *serv = clp->cl_rpcclient->cl_xprt->bc_serv;
+	struct nfs4_sessionid *bc_sid;
+
+	if (!serv->sv_bc_xprt)
+		return -EINVAL;
+
+	/* on success freed in xprt_free */
+	bc_sid = kmalloc(sizeof(struct nfs4_sessionid), GFP_KERNEL);
+	if (!bc_sid)
+		return -ENOMEM;
+	memcpy(bc_sid->data, &clp->cl_session->sess_id.data,
+		NFS4_MAX_SESSIONID_LEN);
+	spin_lock_bh(&serv->sv_cb_lock);
+	serv->sv_bc_xprt->xpt_bc_sid = bc_sid;
+	spin_unlock_bh(&serv->sv_cb_lock);
+	dprintk("%s set xpt_bc_sid=%u:%u:%u:%u for sv_bc_xprt %p\n", __func__,
+		((u32 *)bc_sid->data)[0], ((u32 *)bc_sid->data)[1],
+		((u32 *)bc_sid->data)[2], ((u32 *)bc_sid->data)[3],
+		serv->sv_bc_xprt);
+	return 0;
+}
+
 /*
  * The callback service for NFSv4.1 callbacks
  */
@@ -177,30 +202,38 @@ nfs41_callback_svc(void *vrqstp)
 struct svc_rqst *
 nfs41_callback_up(struct svc_serv *serv, struct rpc_xprt *xprt)
 {
-	struct svc_xprt *bc_xprt;
-	struct svc_rqst *rqstp = ERR_PTR(-ENOMEM);
+	struct svc_rqst *rqstp;
+	int ret;
 
-	dprintk("--> %s\n", __func__);
-	/* Create a svc_sock for the service */
-	bc_xprt = svc_sock_create(serv, xprt->prot);
-	if (!bc_xprt)
+	/*
+	 * Create an svc_sock for the back channel service that shares the
+	 * fore channel connection.
+	 * Returns the input port (0) and sets the svc_serv bc_xprt on success
+	 */
+	ret = svc_create_xprt(serv, "tcp-bc", &init_net, PF_INET, 0,
+			      SVC_SOCK_ANONYMOUS);
+	if (ret < 0) {
+		rqstp = ERR_PTR(ret);
 		goto out;
+	}
 
 	/*
 	 * Save the svc_serv in the transport so that it can
 	 * be referenced when the session backchannel is initialized
 	 */
-	serv->bc_xprt = bc_xprt;
 	xprt->bc_serv = serv;
 
 	INIT_LIST_HEAD(&serv->sv_cb_list);
 	spin_lock_init(&serv->sv_cb_lock);
 	init_waitqueue_head(&serv->sv_cb_waitq);
 	rqstp = svc_prepare_thread(serv, &serv->sv_pools[0]);
-	if (IS_ERR(rqstp))
-		svc_sock_destroy(bc_xprt);
+	if (IS_ERR(rqstp)) {
+		svc_xprt_put(serv->sv_bc_xprt);
+		serv->sv_bc_xprt = NULL;
+	}
 out:
-	dprintk("--> %s return %p\n", __func__, rqstp);
+	dprintk("--> %s return %ld\n", __func__,
+		IS_ERR(rqstp) ? PTR_ERR(rqstp) : 0);
 	return rqstp;
 }
 
@@ -233,6 +266,10 @@ static inline void nfs_callback_bc_serv(u32 minorversion, struct rpc_xprt *xprt,
 		struct nfs_callback_data *cb_info)
 {
 }
+int nfs4_set_callback_sessionid(struct nfs_client *clp)
+{
+	return 0;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 /*
@@ -328,6 +365,9 @@ static int check_gss_callback_principal(struct nfs_client *clp,
 	struct rpc_clnt *r = clp->cl_rpcclient;
 	char *p = svc_gss_principal(rqstp);
 
+	/* No RPC_AUTH_GSS on NFSv4.1 back channel yet */
+	if (clp->cl_minorversion != 0)
+		return SVC_DROP;
 	/*
 	 * It might just be a normal user principal, in which case
 	 * userspace won't bother to tell us the name at all.
@@ -345,6 +385,23 @@ static int check_gss_callback_principal(struct nfs_client *clp,
 	return SVC_OK;
 }
 
+/* pg_authenticate method helper */
+static struct nfs_client *nfs_cb_find_client(struct svc_rqst *rqstp)
+{
+	struct nfs4_sessionid *sessionid = bc_xprt_sid(rqstp);
+	int is_cb_compound = rqstp->rq_proc == CB_COMPOUND ? 1 : 0;
+
+	dprintk("--> %s rq_proc %d\n", __func__, rqstp->rq_proc);
+	if (svc_is_backchannel(rqstp))
+		/* Sessionid (usually) set after CB_NULL ping */
+		return nfs4_find_client_sessionid(svc_addr(rqstp), sessionid,
+						  is_cb_compound);
+	else
+		/* No callback identifier in pg_authenticate */
+		return nfs4_find_client_no_ident(svc_addr(rqstp));
+}
+
+/* pg_authenticate method for nfsv4 callback threads. */
 static int nfs_callback_authenticate(struct svc_rqst *rqstp)
 {
 	struct nfs_client *clp;
@@ -352,7 +409,7 @@ static int nfs_callback_authenticate(struct svc_rqst *rqstp)
 	int ret = SVC_OK;
 
 	/* Don't talk to strangers */
-	clp = nfs_find_client(svc_addr(rqstp), 4);
+	clp = nfs_cb_find_client(rqstp);
 	if (clp == NULL)
 		return SVC_DROP;
 

fs/nfs/callback.h

@@ -34,10 +34,17 @@ enum nfs4_callback_opnum {
 	OP_CB_ILLEGAL = 10044,
 };
 
+struct cb_process_state {
+	__be32			drc_status;
+	struct nfs_client	*clp;
+	struct nfs4_sessionid	*svc_sid; /* v4.1 callback service sessionid */
+};
+
 struct cb_compound_hdr_arg {
 	unsigned int taglen;
 	const char *tag;
 	unsigned int minorversion;
+	unsigned int cb_ident; /* v4.0 callback identifier */
 	unsigned nops;
 };
 
@@ -103,14 +110,23 @@ struct cb_sequenceres {
 	uint32_t			csr_target_highestslotid;
 };
 
-extern unsigned nfs4_callback_sequence(struct cb_sequenceargs *args,
-				       struct cb_sequenceres *res);
+extern __be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
+				       struct cb_sequenceres *res,
+				       struct cb_process_state *cps);
 
 extern int nfs41_validate_delegation_stateid(struct nfs_delegation *delegation,
 					     const nfs4_stateid *stateid);
 
 #define RCA4_TYPE_MASK_RDATA_DLG	0
 #define RCA4_TYPE_MASK_WDATA_DLG	1
+#define RCA4_TYPE_MASK_DIR_DLG         2
+#define RCA4_TYPE_MASK_FILE_LAYOUT     3
+#define RCA4_TYPE_MASK_BLK_LAYOUT      4
+#define RCA4_TYPE_MASK_OBJ_LAYOUT_MIN  8
+#define RCA4_TYPE_MASK_OBJ_LAYOUT_MAX  9
+#define RCA4_TYPE_MASK_OTHER_LAYOUT_MIN 12
+#define RCA4_TYPE_MASK_OTHER_LAYOUT_MAX 15
+#define RCA4_TYPE_MASK_ALL 0xf31f
 
 struct cb_recallanyargs {
 	struct sockaddr	*craa_addr;
@@ -118,25 +134,52 @@ struct cb_recallanyargs {
 	uint32_t	craa_type_mask;
 };
 
-extern unsigned nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy);
+extern __be32 nfs4_callback_recallany(struct cb_recallanyargs *args,
+					void *dummy,
+					struct cb_process_state *cps);
 
 struct cb_recallslotargs {
 	struct sockaddr	*crsa_addr;
 	uint32_t	crsa_target_max_slots;
 };
-extern unsigned nfs4_callback_recallslot(struct cb_recallslotargs *args,
-					  void *dummy);
+extern __be32 nfs4_callback_recallslot(struct cb_recallslotargs *args,
+					 void *dummy,
+					 struct cb_process_state *cps);
+
+struct cb_layoutrecallargs {
+	struct sockaddr		*cbl_addr;
+	uint32_t		cbl_recall_type;
+	uint32_t		cbl_layout_type;
+	uint32_t		cbl_layoutchanged;
+	union {
+		struct {
+			struct nfs_fh		cbl_fh;
+			struct pnfs_layout_range cbl_range;
+			nfs4_stateid		cbl_stateid;
+		};
+		struct nfs_fsid		cbl_fsid;
+	};
+};
 
-#endif /* CONFIG_NFS_V4_1 */
+extern unsigned nfs4_callback_layoutrecall(
+	struct cb_layoutrecallargs *args,
+	void *dummy, struct cb_process_state *cps);
 
-extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res);
-extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy);
+extern void nfs4_check_drain_bc_complete(struct nfs4_session *ses);
+extern void nfs4_cb_take_slot(struct nfs_client *clp);
+#endif /* CONFIG_NFS_V4_1 */
 
+extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args,
+				    struct cb_getattrres *res,
+				    struct cb_process_state *cps);
+extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
+				   struct cb_process_state *cps);
 #ifdef CONFIG_NFS_V4
 extern int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt);
 extern void nfs_callback_down(int minorversion);
 extern int nfs4_validate_delegation_stateid(struct nfs_delegation *delegation,
 					    const nfs4_stateid *stateid);
+extern int nfs4_set_callback_sessionid(struct nfs_client *clp);
 #endif /* CONFIG_NFS_V4 */
 /*
  * nfs41: Callbacks are expected to not cause substantial latency,

fs/nfs/callback_proc.c

@@ -12,30 +12,33 @@
 #include "callback.h"
 #include "delegation.h"
 #include "internal.h"
+#include "pnfs.h"
 
 #ifdef NFS_DEBUG
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
 #endif
- 
-__be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res)
+
+__be32 nfs4_callback_getattr(struct cb_getattrargs *args,
+			     struct cb_getattrres *res,
+			     struct cb_process_state *cps)
 {
-	struct nfs_client *clp;
 	struct nfs_delegation *delegation;
 	struct nfs_inode *nfsi;
 	struct inode *inode;
 
+	res->status = htonl(NFS4ERR_OP_NOT_IN_SESSION);
+	if (!cps->clp) /* Always set for v4.0. Set in cb_sequence for v4.1 */
+		goto out;
+
 	res->bitmap[0] = res->bitmap[1] = 0;
 	res->status = htonl(NFS4ERR_BADHANDLE);
-	clp = nfs_find_client(args->addr, 4);
-	if (clp == NULL)
-		goto out;
 
 	dprintk("NFS: GETATTR callback request from %s\n",
-		rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
+		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
 
-	inode = nfs_delegation_find_inode(clp, &args->fh);
+	inode = nfs_delegation_find_inode(cps->clp, &args->fh);
 	if (inode == NULL)
-		goto out_putclient;
+		goto out;
 	nfsi = NFS_I(inode);
 	rcu_read_lock();
 	delegation = rcu_dereference(nfsi->delegation);
@@ -55,49 +58,41 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *
 out_iput:
 	rcu_read_unlock();
 	iput(inode);
-out_putclient:
-	nfs_put_client(clp);
 out:
 	dprintk("%s: exit with status = %d\n", __func__, ntohl(res->status));
 	return res->status;
 }
 
-__be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
+__be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
+			    struct cb_process_state *cps)
 {
-	struct nfs_client *clp;
 	struct inode *inode;
 	__be32 res;
 	
-	res = htonl(NFS4ERR_BADHANDLE);
-	clp = nfs_find_client(args->addr, 4);
-	if (clp == NULL)
+	res = htonl(NFS4ERR_OP_NOT_IN_SESSION);
+	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",
-		rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
-
-	do {
-		struct nfs_client *prev = clp;
-
-		inode = nfs_delegation_find_inode(clp, &args->fh);
-		if (inode != NULL) {
-			/* Set up a helper thread to actually return the delegation */
-			switch (nfs_async_inode_return_delegation(inode, &args->stateid)) {
-				case 0:
-					res = 0;
-					break;
-				case -ENOENT:
-					if (res != 0)
-						res = htonl(NFS4ERR_BAD_STATEID);
-					break;
-				default:
-					res = htonl(NFS4ERR_RESOURCE);
-			}
-			iput(inode);
-		}
-		clp = nfs_find_client_next(prev);
-		nfs_put_client(prev);
-	} while (clp != NULL);
+		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
+
+	res = htonl(NFS4ERR_BADHANDLE);
+	inode = nfs_delegation_find_inode(cps->clp, &args->fh);
+	if (inode == NULL)
+		goto out;
+	/* Set up a helper thread to actually return the delegation */
+	switch (nfs_async_inode_return_delegation(inode, &args->stateid)) {
+	case 0:
+		res = 0;
+		break;
+	case -ENOENT:
+		if (res != 0)
+			res = htonl(NFS4ERR_BAD_STATEID);
+		break;
+	default:
+		res = htonl(NFS4ERR_RESOURCE);
+	}
+	iput(inode);
 out:
 	dprintk("%s: exit with status = %d\n", __func__, ntohl(res));
 	return res;
@@ -113,6 +108,139 @@ int nfs4_validate_delegation_stateid(struct nfs_delegation *delegation, const nf
 
 #if defined(CONFIG_NFS_V4_1)
 
+static u32 initiate_file_draining(struct nfs_client *clp,
+				  struct cb_layoutrecallargs *args)
+{
+	struct pnfs_layout_hdr *lo;
+	struct inode *ino;
+	bool found = false;
+	u32 rv = NFS4ERR_NOMATCHING_LAYOUT;
+	LIST_HEAD(free_me_list);
+
+	spin_lock(&clp->cl_lock);
+	list_for_each_entry(lo, &clp->cl_layouts, plh_layouts) {
+		if (nfs_compare_fh(&args->cbl_fh,
+				   &NFS_I(lo->plh_inode)->fh))
+			continue;
+		ino = igrab(lo->plh_inode);
+		if (!ino)
+			continue;
+		found = true;
+		/* Without this, layout can be freed as soon
+		 * as we release cl_lock.
+		 */
+		get_layout_hdr(lo);
+		break;
+	}
+	spin_unlock(&clp->cl_lock);
+	if (!found)
+		return NFS4ERR_NOMATCHING_LAYOUT;
+
+	spin_lock(&ino->i_lock);
+	if (test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags) ||
+	    mark_matching_lsegs_invalid(lo, &free_me_list,
+					args->cbl_range.iomode))
+		rv = NFS4ERR_DELAY;
+	else
+		rv = NFS4ERR_NOMATCHING_LAYOUT;
+	pnfs_set_layout_stateid(lo, &args->cbl_stateid, true);
+	spin_unlock(&ino->i_lock);
+	pnfs_free_lseg_list(&free_me_list);
+	put_layout_hdr(lo);
+	iput(ino);
+	return rv;
+}
+
+static u32 initiate_bulk_draining(struct nfs_client *clp,
+				  struct cb_layoutrecallargs *args)
+{
+	struct pnfs_layout_hdr *lo;
+	struct inode *ino;
+	u32 rv = NFS4ERR_NOMATCHING_LAYOUT;
+	struct pnfs_layout_hdr *tmp;
+	LIST_HEAD(recall_list);
+	LIST_HEAD(free_me_list);
+	struct pnfs_layout_range range = {
+		.iomode = IOMODE_ANY,
+		.offset = 0,
+		.length = NFS4_MAX_UINT64,
+	};
+
+	spin_lock(&clp->cl_lock);
+	list_for_each_entry(lo, &clp->cl_layouts, plh_layouts) {
+		if ((args->cbl_recall_type == RETURN_FSID) &&
+		    memcmp(&NFS_SERVER(lo->plh_inode)->fsid,
+			   &args->cbl_fsid, sizeof(struct nfs_fsid)))
+			continue;
+		if (!igrab(lo->plh_inode))
+			continue;
+		get_layout_hdr(lo);
+		BUG_ON(!list_empty(&lo->plh_bulk_recall));
+		list_add(&lo->plh_bulk_recall, &recall_list);
+	}
+	spin_unlock(&clp->cl_lock);
+	list_for_each_entry_safe(lo, tmp,
+				 &recall_list, plh_bulk_recall) {
+		ino = lo->plh_inode;
+		spin_lock(&ino->i_lock);
+		set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
+		if (mark_matching_lsegs_invalid(lo, &free_me_list, range.iomode))
+			rv = NFS4ERR_DELAY;
+		list_del_init(&lo->plh_bulk_recall);
+		spin_unlock(&ino->i_lock);
+		put_layout_hdr(lo);
+		iput(ino);
+	}
+	pnfs_free_lseg_list(&free_me_list);
+	return rv;
+}
+
+static u32 do_callback_layoutrecall(struct nfs_client *clp,
+				    struct cb_layoutrecallargs *args)
+{
+	u32 res = NFS4ERR_DELAY;
+
+	dprintk("%s enter, type=%i\n", __func__, args->cbl_recall_type);
+	if (test_and_set_bit(NFS4CLNT_LAYOUTRECALL, &clp->cl_state))
+		goto out;
+	if (args->cbl_recall_type == RETURN_FILE)
+		res = initiate_file_draining(clp, args);
+	else
+		res = initiate_bulk_draining(clp, args);
+	clear_bit(NFS4CLNT_LAYOUTRECALL, &clp->cl_state);
+out:
+	dprintk("%s returning %i\n", __func__, res);
+	return res;
+
+}
+
+__be32 nfs4_callback_layoutrecall(struct cb_layoutrecallargs *args,
+				  void *dummy, struct cb_process_state *cps)
+{
+	u32 res;
+
+	dprintk("%s: -->\n", __func__);
+
+	if (cps->clp)
+		res = do_callback_layoutrecall(cps->clp, args);
+	else
+		res = NFS4ERR_OP_NOT_IN_SESSION;
+
+	dprintk("%s: exit with status = %d\n", __func__, res);
+	return cpu_to_be32(res);
+}
+
+static void pnfs_recall_all_layouts(struct nfs_client *clp)
+{
+	struct cb_layoutrecallargs args;
+
+	/* Pretend we got a CB_LAYOUTRECALL(ALL) */
+	memset(&args, 0, sizeof(args));
+	args.cbl_recall_type = RETURN_ALL;
+	/* FIXME we ignore errors, what should we do? */
+	do_callback_layoutrecall(clp, &args);
+}
+
 int nfs41_validate_delegation_stateid(struct nfs_delegation *delegation, const nfs4_stateid *stateid)
 {
 	if (delegation == NULL)
@@ -184,42 +312,6 @@ validate_seqid(struct nfs4_slot_table *tbl, struct cb_sequenceargs * args)
 	return htonl(NFS4ERR_SEQ_MISORDERED);
 }
 
-/*
- * Returns a pointer to a held 'struct nfs_client' that matches the server's
- * address, major version number, and session ID.  It is the caller's
- * responsibility to release the returned reference.
- *
- * Returns NULL if there are no connections with sessions, or if no session
- * matches the one of interest.
- */
- static struct nfs_client *find_client_with_session(
-	const struct sockaddr *addr, u32 nfsversion,
-	struct nfs4_sessionid *sessionid)
-{
-	struct nfs_client *clp;
-
-	clp = nfs_find_client(addr, 4);
-	if (clp == NULL)
-		return NULL;
-
-	do {
-		struct nfs_client *prev = clp;
-
-		if (clp->cl_session != NULL) {
-			if (memcmp(clp->cl_session->sess_id.data,
-					sessionid->data,
-					NFS4_MAX_SESSIONID_LEN) == 0) {
-				/* Returns a held reference to clp */
-				return clp;
-			}
-		}
-		clp = nfs_find_client_next(prev);
-		nfs_put_client(prev);
-	} while (clp != NULL);
-
-	return NULL;
-}
-
 /*
  * For each referring call triple, check the session's slot table for
  * a match.  If the slot is in use and the sequence numbers match, the
@@ -276,20 +368,34 @@ static bool referring_call_exists(struct nfs_client *clp,
 }
 
 __be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
-				struct cb_sequenceres *res)
+			      struct cb_sequenceres *res,
+			      struct cb_process_state *cps)
 {
 	struct nfs_client *clp;
 	int i;
 	__be32 status;
 
+	cps->clp = NULL;
+
 	status = htonl(NFS4ERR_BADSESSION);
-	clp = find_client_with_session(args->csa_addr, 4, &args->csa_sessionid);
+	/* Incoming session must match the callback session */
+	if (memcmp(&args->csa_sessionid, cps->svc_sid, NFS4_MAX_SESSIONID_LEN))
+		goto out;
+
+	clp = nfs4_find_client_sessionid(args->csa_addr,
+					 &args->csa_sessionid, 1);
 	if (clp == NULL)
 		goto out;
 
+	/* state manager is resetting the session */
+	if (test_bit(NFS4_SESSION_DRAINING, &clp->cl_session->session_state)) {
+		status = NFS4ERR_DELAY;
+		goto out;
+	}
+
 	status = validate_seqid(&clp->cl_session->bc_slot_table, args);
 	if (status)
-		goto out_putclient;
+		goto out;
 
 	/*
 	 * Check for pending referring calls.  If a match is found, a
@@ -298,7 +404,7 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
 	 */
 	if (referring_call_exists(clp, args->csa_nrclists, args->csa_rclists)) {
 		status = htonl(NFS4ERR_DELAY);
-		goto out_putclient;
+		goto out;
 	}
 
 	memcpy(&res->csr_sessionid, &args->csa_sessionid,
@@ -307,83 +413,93 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args,
 	res->csr_slotid = args->csa_slotid;
 	res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1;
 	res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1;
+	nfs4_cb_take_slot(clp);
+	cps->clp = clp; /* put in nfs4_callback_compound */
 
-out_putclient:
-	nfs_put_client(clp);
 out:
 	for (i = 0; i < args->csa_nrclists; i++)
 		kfree(args->csa_rclists[i].rcl_refcalls);
 	kfree(args->csa_rclists);
 
-	if (status == htonl(NFS4ERR_RETRY_UNCACHED_REP))
-		res->csr_status = 0;
-	else
+	if (status == htonl(NFS4ERR_RETRY_UNCACHED_REP)) {
+		cps->drc_status = status;
+		status = 0;
+	} else
 		res->csr_status = status;
+
 	dprintk("%s: exit with status = %d res->csr_status %d\n", __func__,
 		ntohl(status), ntohl(res->csr_status));
 	return status;
 }
 
-__be32 nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy)
+static bool
+validate_bitmap_values(unsigned long mask)
+{
+	return (mask & ~RCA4_TYPE_MASK_ALL) == 0;
+}
+
+__be32 nfs4_callback_recallany(struct cb_recallanyargs *args, void *dummy,
+			       struct cb_process_state *cps)
 {
-	struct nfs_client *clp;
 	__be32 status;
 	fmode_t flags = 0;
 
-	status = htonl(NFS4ERR_OP_NOT_IN_SESSION);
-	clp = nfs_find_client(args->craa_addr, 4);
-	if (clp == NULL)
+	status = cpu_to_be32(NFS4ERR_OP_NOT_IN_SESSION);
+	if (!cps->clp) /* set in cb_sequence */
 		goto out;
 
 	dprintk("NFS: RECALL_ANY callback request from %s\n",
-		rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
+		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
+
+	status = cpu_to_be32(NFS4ERR_INVAL);
+	if (!validate_bitmap_values(args->craa_type_mask))
+		goto out;
 
+	status = cpu_to_be32(NFS4_OK);
 	if (test_bit(RCA4_TYPE_MASK_RDATA_DLG, (const unsigned long *)
 		     &args->craa_type_mask))
 		flags = FMODE_READ;
 	if (test_bit(RCA4_TYPE_MASK_WDATA_DLG, (const unsigned long *)
 		     &args->craa_type_mask))
 		flags |= FMODE_WRITE;
-
+	if (test_bit(RCA4_TYPE_MASK_FILE_LAYOUT, (const unsigned long *)
+		     &args->craa_type_mask))
+		pnfs_recall_all_layouts(cps->clp);
 	if (flags)
-		nfs_expire_all_delegation_types(clp, flags);
-	status = htonl(NFS4_OK);
+		nfs_expire_all_delegation_types(cps->clp, flags);
 out:
 	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
 	return status;
 }
 
 /* Reduce the fore channel's max_slots to the target value */
-__be32 nfs4_callback_recallslot(struct cb_recallslotargs *args, void *dummy)
+__be32 nfs4_callback_recallslot(struct cb_recallslotargs *args, void *dummy,
+				struct cb_process_state *cps)
 {
-	struct nfs_client *clp;
 	struct nfs4_slot_table *fc_tbl;
 	__be32 status;
 
 	status = htonl(NFS4ERR_OP_NOT_IN_SESSION);
-	clp = nfs_find_client(args->crsa_addr, 4);
-	if (clp == NULL)
+	if (!cps->clp) /* set in cb_sequence */
 		goto out;
 
 	dprintk("NFS: CB_RECALL_SLOT request from %s target max slots %d\n",
-		rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR),
+		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR),
 		args->crsa_target_max_slots);
 
-	fc_tbl = &clp->cl_session->fc_slot_table;
+	fc_tbl = &cps->clp->cl_session->fc_slot_table;
 
 	status = htonl(NFS4ERR_BAD_HIGH_SLOT);
 	if (args->crsa_target_max_slots > fc_tbl->max_slots ||
 	    args->crsa_target_max_slots < 1)
-		goto out_putclient;
+		goto out;
 
 	status = htonl(NFS4_OK);
 	if (args->crsa_target_max_slots == fc_tbl->max_slots)
-		goto out_putclient;
+		goto out;
 
 	fc_tbl->target_max_slots = args->crsa_target_max_slots;
-	nfs41_handle_recall_slot(clp);
-out_putclient:
-	nfs_put_client(clp);	/* balance nfs_find_client */
+	nfs41_handle_recall_slot(cps->clp);
 out:
 	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
 	return status;

fs/nfs/callback_xdr.c

@@ -10,8 +10,10 @@
 #include <linux/nfs4.h>
 #include <linux/nfs_fs.h>
 #include <linux/slab.h>
+#include <linux/sunrpc/bc_xprt.h>
 #include "nfs4_fs.h"
 #include "callback.h"
+#include "internal.h"
 
 #define CB_OP_TAGLEN_MAXSZ	(512)
 #define CB_OP_HDR_RES_MAXSZ	(2 + CB_OP_TAGLEN_MAXSZ)
@@ -22,6 +24,7 @@
 #define CB_OP_RECALL_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ)
 
 #if defined(CONFIG_NFS_V4_1)
+#define CB_OP_LAYOUTRECALL_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ)
 #define CB_OP_SEQUENCE_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ + \
 					4 + 1 + 3)
 #define CB_OP_RECALLANY_RES_MAXSZ	(CB_OP_HDR_RES_MAXSZ)
@@ -33,7 +36,8 @@
 /* Internal error code */
 #define NFS4ERR_RESOURCE_HDR	11050
 
-typedef __be32 (*callback_process_op_t)(void *, void *);
+typedef __be32 (*callback_process_op_t)(void *, void *,
+					struct cb_process_state *);
 typedef __be32 (*callback_decode_arg_t)(struct svc_rqst *, struct xdr_stream *, void *);
 typedef __be32 (*callback_encode_res_t)(struct svc_rqst *, struct xdr_stream *, void *);
 
@@ -160,7 +164,7 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound
 	hdr->minorversion = ntohl(*p++);
 	/* Check minor version is zero or one. */
 	if (hdr->minorversion <= 1) {
-		p++;	/* skip callback_ident */
+		hdr->cb_ident = ntohl(*p++); /* ignored by v4.1 */
 	} else {
 		printk(KERN_WARNING "%s: NFSv4 server callback with "
 			"illegal minor version %u!\n",
@@ -220,6 +224,66 @@ static __be32 decode_recall_args(struct svc_rqst *rqstp, struct xdr_stream *xdr,
 
 #if defined(CONFIG_NFS_V4_1)
 
+static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp,
+				       struct xdr_stream *xdr,
+				       struct cb_layoutrecallargs *args)
+{
+	__be32 *p;
+	__be32 status = 0;
+	uint32_t iomode;
+
+	args->cbl_addr = svc_addr(rqstp);
+	p = read_buf(xdr, 4 * sizeof(uint32_t));
+	if (unlikely(p == NULL)) {
+		status = htonl(NFS4ERR_BADXDR);
+		goto out;
+	}
+
+	args->cbl_layout_type = ntohl(*p++);
+	/* Depite the spec's xdr, iomode really belongs in the FILE switch,
+	 * as it is unuseable and ignored with the other types.
+	 */
+	iomode = ntohl(*p++);
+	args->cbl_layoutchanged = ntohl(*p++);
+	args->cbl_recall_type = ntohl(*p++);
+
+	if (args->cbl_recall_type == RETURN_FILE) {
+		args->cbl_range.iomode = iomode;
+		status = decode_fh(xdr, &args->cbl_fh);
+		if (unlikely(status != 0))
+			goto out;
+
+		p = read_buf(xdr, 2 * sizeof(uint64_t));
+		if (unlikely(p == NULL)) {
+			status = htonl(NFS4ERR_BADXDR);
+			goto out;
+		}
+		p = xdr_decode_hyper(p, &args->cbl_range.offset);
+		p = xdr_decode_hyper(p, &args->cbl_range.length);
+		status = decode_stateid(xdr, &args->cbl_stateid);
+		if (unlikely(status != 0))
+			goto out;
+	} else if (args->cbl_recall_type == RETURN_FSID) {
+		p = read_buf(xdr, 2 * sizeof(uint64_t));
+		if (unlikely(p == NULL)) {
+			status = htonl(NFS4ERR_BADXDR);
+			goto out;
+		}
+		p = xdr_decode_hyper(p, &args->cbl_fsid.major);
+		p = xdr_decode_hyper(p, &args->cbl_fsid.minor);
+	} else if (args->cbl_recall_type != RETURN_ALL) {
+		status = htonl(NFS4ERR_BADXDR);
+		goto out;
+	}
+	dprintk("%s: ltype 0x%x iomode %d changed %d recall_type %d\n",
+		__func__,
+		args->cbl_layout_type, iomode,
+		args->cbl_layoutchanged, args->cbl_recall_type);
+out:
+	dprintk("%s: exit with status = %d\n", __func__, ntohl(status));
+	return status;
+}
+
 static __be32 decode_sessionid(struct xdr_stream *xdr,
 				 struct nfs4_sessionid *sid)
 {
@@ -574,10 +638,10 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op)
 	case OP_CB_SEQUENCE:
 	case OP_CB_RECALL_ANY:
 	case OP_CB_RECALL_SLOT:
+	case OP_CB_LAYOUTRECALL:
 		*op = &callback_ops[op_nr];
 		break;
 
-	case OP_CB_LAYOUTRECALL:
 	case OP_CB_NOTIFY_DEVICEID:
 	case OP_CB_NOTIFY:
 	case OP_CB_PUSH_DELEG:
@@ -593,6 +657,37 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op)
 	return htonl(NFS_OK);
 }
 
+static void nfs4_callback_free_slot(struct nfs4_session *session)
+{
+	struct nfs4_slot_table *tbl = &session->bc_slot_table;
+
+	spin_lock(&tbl->slot_tbl_lock);
+	/*
+	 * Let the state manager know callback processing done.
+	 * A single slot, so highest used slotid is either 0 or -1
+	 */
+	tbl->highest_used_slotid--;
+	nfs4_check_drain_bc_complete(session);
+	spin_unlock(&tbl->slot_tbl_lock);
+}
+
+static void nfs4_cb_free_slot(struct nfs_client *clp)
+{
+	if (clp && clp->cl_session)
+		nfs4_callback_free_slot(clp->cl_session);
+}
+
+/* A single slot, so highest used slotid is either 0 or -1 */
+void nfs4_cb_take_slot(struct nfs_client *clp)
+{
+	struct nfs4_slot_table *tbl = &clp->cl_session->bc_slot_table;
+
+	spin_lock(&tbl->slot_tbl_lock);
+	tbl->highest_used_slotid++;
+	BUG_ON(tbl->highest_used_slotid != 0);
+	spin_unlock(&tbl->slot_tbl_lock);
+}
+
 #else /* CONFIG_NFS_V4_1 */
 
 static __be32
@@ -601,6 +696,9 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op)
 	return htonl(NFS4ERR_MINOR_VERS_MISMATCH);
 }
 
+static void nfs4_cb_free_slot(struct nfs_client *clp)
+{
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 static __be32
@@ -621,7 +719,8 @@ preprocess_nfs4_op(unsigned int op_nr, struct callback_op **op)
 static __be32 process_op(uint32_t minorversion, int nop,
 		struct svc_rqst *rqstp,
 		struct xdr_stream *xdr_in, void *argp,
-		struct xdr_stream *xdr_out, void *resp, int* drc_status)
+		struct xdr_stream *xdr_out, void *resp,
+		struct cb_process_state *cps)
 {
 	struct callback_op *op = &callback_ops[0];
 	unsigned int op_nr;
@@ -644,8 +743,8 @@ static __be32 process_op(uint32_t minorversion, int nop,
 	if (status)
 		goto encode_hdr;
 
-	if (*drc_status) {
-		status = *drc_status;
+	if (cps->drc_status) {
+		status = cps->drc_status;
 		goto encode_hdr;
 	}
 
@@ -653,16 +752,10 @@ static __be32 process_op(uint32_t minorversion, int nop,
 	if (maxlen > 0 && maxlen < PAGE_SIZE) {
 		status = op->decode_args(rqstp, xdr_in, argp);
 		if (likely(status == 0))
-			status = op->process_op(argp, resp);
+			status = op->process_op(argp, resp, cps);
 	} else
 		status = htonl(NFS4ERR_RESOURCE);
 
-	/* Only set by OP_CB_SEQUENCE processing */
-	if (status == htonl(NFS4ERR_RETRY_UNCACHED_REP)) {
-		*drc_status = status;
-		status = 0;
-	}
-
 encode_hdr:
 	res = encode_op_hdr(xdr_out, op_nr, status);
 	if (unlikely(res))
@@ -681,8 +774,11 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 	struct cb_compound_hdr_arg hdr_arg = { 0 };
 	struct cb_compound_hdr_res hdr_res = { NULL };
 	struct xdr_stream xdr_in, xdr_out;
-	__be32 *p;
-	__be32 status, drc_status = 0;
+	__be32 *p, status;
+	struct cb_process_state cps = {
+		.drc_status = 0,
+		.clp = NULL,
+	};
 	unsigned int nops = 0;
 
 	dprintk("%s: start\n", __func__);
@@ -696,6 +792,13 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 	if (status == __constant_htonl(NFS4ERR_RESOURCE))
 		return rpc_garbage_args;
 
+	if (hdr_arg.minorversion == 0) {
+		cps.clp = nfs4_find_client_ident(hdr_arg.cb_ident);
+		if (!cps.clp)
+			return rpc_drop_reply;
+	} else
+		cps.svc_sid = bc_xprt_sid(rqstp);
+
 	hdr_res.taglen = hdr_arg.taglen;
 	hdr_res.tag = hdr_arg.tag;
 	if (encode_compound_hdr_res(&xdr_out, &hdr_res) != 0)
@@ -703,7 +806,7 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 
 	while (status == 0 && nops != hdr_arg.nops) {
 		status = process_op(hdr_arg.minorversion, nops, rqstp,
-				    &xdr_in, argp, &xdr_out, resp, &drc_status);
+				    &xdr_in, argp, &xdr_out, resp, &cps);
 		nops++;
 	}
 
@@ -716,6 +819,8 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r
 
 	*hdr_res.status = status;
 	*hdr_res.nops = htonl(nops);
+	nfs4_cb_free_slot(cps.clp);
+	nfs_put_client(cps.clp);
 	dprintk("%s: done, status = %u\n", __func__, ntohl(status));
 	return rpc_success;
 }
@@ -739,6 +844,12 @@ static struct callback_op callback_ops[] = {
 		.res_maxsize = CB_OP_RECALL_RES_MAXSZ,
 	},
 #if defined(CONFIG_NFS_V4_1)
+	[OP_CB_LAYOUTRECALL] = {
+		.process_op = (callback_process_op_t)nfs4_callback_layoutrecall,
+		.decode_args =
+			(callback_decode_arg_t)decode_layoutrecall_args,
+		.res_maxsize = CB_OP_LAYOUTRECALL_RES_MAXSZ,
+	},
 	[OP_CB_SEQUENCE] = {
 		.process_op = (callback_process_op_t)nfs4_callback_sequence,
 		.decode_args = (callback_decode_arg_t)decode_cb_sequence_args,

fs/nfs/client.c

@@ -56,6 +56,30 @@ static DEFINE_SPINLOCK(nfs_client_lock);
 static LIST_HEAD(nfs_client_list);
 static LIST_HEAD(nfs_volume_list);
 static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
+#ifdef CONFIG_NFS_V4
+static DEFINE_IDR(cb_ident_idr); /* Protected by nfs_client_lock */
+
+/*
+ * Get a unique NFSv4.0 callback identifier which will be used
+ * by the V4.0 callback service to lookup the nfs_client struct
+ */
+static int nfs_get_cb_ident_idr(struct nfs_client *clp, int minorversion)
+{
+	int ret = 0;
+
+	if (clp->rpc_ops->version != 4 || minorversion != 0)
+		return ret;
+retry:
+	if (!idr_pre_get(&cb_ident_idr, GFP_KERNEL))
+		return -ENOMEM;
+	spin_lock(&nfs_client_lock);
+	ret = idr_get_new(&cb_ident_idr, clp, &clp->cl_cb_ident);
+	spin_unlock(&nfs_client_lock);
+	if (ret == -EAGAIN)
+		goto retry;
+	return ret;
+}
+#endif /* CONFIG_NFS_V4 */
 
 /*
  * RPC cruft for NFS
@@ -144,7 +168,10 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_
 	clp->cl_proto = cl_init->proto;
 
 #ifdef CONFIG_NFS_V4
-	INIT_LIST_HEAD(&clp->cl_delegations);
+	err = nfs_get_cb_ident_idr(clp, cl_init->minorversion);
+	if (err)
+		goto error_cleanup;
+
 	spin_lock_init(&clp->cl_lock);
 	INIT_DELAYED_WORK(&clp->cl_renewd, nfs4_renew_state);
 	rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client");
@@ -170,21 +197,17 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_
 }
 
 #ifdef CONFIG_NFS_V4
-/*
- * Clears/puts all minor version specific parts from an nfs_client struct
- * reverting it to minorversion 0.
- */
-static void nfs4_clear_client_minor_version(struct nfs_client *clp)
-{
 #ifdef CONFIG_NFS_V4_1
-	if (nfs4_has_session(clp)) {
+static void nfs4_shutdown_session(struct nfs_client *clp)
+{
+	if (nfs4_has_session(clp))
 		nfs4_destroy_session(clp->cl_session);
-		clp->cl_session = NULL;
-	}
-
-	clp->cl_mvops = nfs_v4_minor_ops[0];
-#endif /* CONFIG_NFS_V4_1 */
 }
+#else /* CONFIG_NFS_V4_1 */
+static void nfs4_shutdown_session(struct nfs_client *clp)
+{
+}
+#endif /* CONFIG_NFS_V4_1 */
 
 /*
  * Destroy the NFS4 callback service
@@ -199,17 +222,49 @@ static void nfs4_shutdown_client(struct nfs_client *clp)
 {
 	if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))
 		nfs4_kill_renewd(clp);
-	nfs4_clear_client_minor_version(clp);
+	nfs4_shutdown_session(clp);
 	nfs4_destroy_callback(clp);
 	if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
 		nfs_idmap_delete(clp);
 
 	rpc_destroy_wait_queue(&clp->cl_rpcwaitq);
 }
+
+/* idr_remove_all is not needed as all id's are removed by nfs_put_client */
+void nfs_cleanup_cb_ident_idr(void)
+{
+	idr_destroy(&cb_ident_idr);
+}
+
+/* nfs_client_lock held */
+static void nfs_cb_idr_remove_locked(struct nfs_client *clp)
+{
+	if (clp->cl_cb_ident)
+		idr_remove(&cb_ident_idr, clp->cl_cb_ident);
+}
+
+static void pnfs_init_server(struct nfs_server *server)
+{
+	rpc_init_wait_queue(&server->roc_rpcwaitq, "pNFS ROC");
+}
+
 #else
 static void nfs4_shutdown_client(struct nfs_client *clp)
 {
 }
+
+void nfs_cleanup_cb_ident_idr(void)
+{
+}
+
+static void nfs_cb_idr_remove_locked(struct nfs_client *clp)
+{
+}
+
+static void pnfs_init_server(struct nfs_server *server)
+{
+}
+
 #endif /* CONFIG_NFS_V4 */
 
 /*
@@ -248,6 +303,7 @@ void nfs_put_client(struct nfs_client *clp)
 
 	if (atomic_dec_and_lock(&clp->cl_count, &nfs_client_lock)) {
 		list_del(&clp->cl_share_link);
+		nfs_cb_idr_remove_locked(clp);
 		spin_unlock(&nfs_client_lock);
 
 		BUG_ON(!list_empty(&clp->cl_superblocks));
@@ -363,70 +419,28 @@ static int nfs_sockaddr_cmp(const struct sockaddr *sa1,
 	return 0;
 }
 
-/*
- * Find a client by IP address and protocol version
- * - returns NULL if no such client
- */
-struct nfs_client *nfs_find_client(const struct sockaddr *addr, u32 nfsversion)
-{
-	struct nfs_client *clp;
-
-	spin_lock(&nfs_client_lock);
-	list_for_each_entry(clp, &nfs_client_list, cl_share_link) {
-		struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr;
-
-		/* Don't match clients that failed to initialise properly */
-		if (!(clp->cl_cons_state == NFS_CS_READY ||
-		      clp->cl_cons_state == NFS_CS_SESSION_INITING))
-			continue;
-
-		/* Different NFS versions cannot share the same nfs_client */
-		if (clp->rpc_ops->version != nfsversion)
-			continue;
-
-		/* Match only the IP address, not the port number */
-		if (!nfs_sockaddr_match_ipaddr(addr, clap))
-			continue;
-
-		atomic_inc(&clp->cl_count);
-		spin_unlock(&nfs_client_lock);
-		return clp;
-	}
-	spin_unlock(&nfs_client_lock);
-	return NULL;
-}
-
-/*
- * Find a client by IP address and protocol version
- * - returns NULL if no such client
- */
-struct nfs_client *nfs_find_client_next(struct nfs_client *clp)
+/* Common match routine for v4.0 and v4.1 callback services */
+bool
+nfs4_cb_match_client(const struct sockaddr *addr, struct nfs_client *clp,
+		     u32 minorversion)
 {
-	struct sockaddr *sap = (struct sockaddr *)&clp->cl_addr;
-	u32 nfsvers = clp->rpc_ops->version;
+	struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr;
 
-	spin_lock(&nfs_client_lock);
-	list_for_each_entry_continue(clp, &nfs_client_list, cl_share_link) {
-		struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr;
+	/* Don't match clients that failed to initialise */
+	if (!(clp->cl_cons_state == NFS_CS_READY ||
+	    clp->cl_cons_state == NFS_CS_SESSION_INITING))
+		return false;
 
-		/* Don't match clients that failed to initialise properly */
-		if (clp->cl_cons_state != NFS_CS_READY)
-			continue;
+	/* Match the version and minorversion */
+	if (clp->rpc_ops->version != 4 ||
+	    clp->cl_minorversion != minorversion)
+		return false;
 
-		/* Different NFS versions cannot share the same nfs_client */
-		if (clp->rpc_ops->version != nfsvers)
-			continue;
+	/* Match only the IP address, not the port number */
+	if (!nfs_sockaddr_match_ipaddr(addr, clap))
+		return false;
 
-		/* Match only the IP address, not the port number */
-		if (!nfs_sockaddr_match_ipaddr(sap, clap))
-			continue;
-
-		atomic_inc(&clp->cl_count);
-		spin_unlock(&nfs_client_lock);
-		return clp;
-	}
-	spin_unlock(&nfs_client_lock);
-	return NULL;
+	return true;
 }
 
 /*
@@ -988,6 +1002,27 @@ static void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_serve
 	target->options = source->options;
 }
 
+static void nfs_server_insert_lists(struct nfs_server *server)
+{
+	struct nfs_client *clp = server->nfs_client;
+
+	spin_lock(&nfs_client_lock);
+	list_add_tail_rcu(&server->client_link, &clp->cl_superblocks);
+	list_add_tail(&server->master_link, &nfs_volume_list);
+	spin_unlock(&nfs_client_lock);
+
+}
+
+static void nfs_server_remove_lists(struct nfs_server *server)
+{
+	spin_lock(&nfs_client_lock);
+	list_del_rcu(&server->client_link);
+	list_del(&server->master_link);
+	spin_unlock(&nfs_client_lock);
+
+	synchronize_rcu();
+}
+
 /*
  * Allocate and initialise a server record
  */
@@ -1004,6 +1039,7 @@ static struct nfs_server *nfs_alloc_server(void)
 	/* Zero out the NFS state stuff */
 	INIT_LIST_HEAD(&server->client_link);
 	INIT_LIST_HEAD(&server->master_link);
+	INIT_LIST_HEAD(&server->delegations);
 
 	atomic_set(&server->active, 0);
 
@@ -1019,6 +1055,8 @@ static struct nfs_server *nfs_alloc_server(void)
 		return NULL;
 	}
 
+	pnfs_init_server(server);
+
 	return server;
 }
 
@@ -1029,11 +1067,8 @@ void nfs_free_server(struct nfs_server *server)
 {
 	dprintk("--> nfs_free_server()\n");
 
+	nfs_server_remove_lists(server);
 	unset_pnfs_layoutdriver(server);
-	spin_lock(&nfs_client_lock);
-	list_del(&server->client_link);
-	list_del(&server->master_link);
-	spin_unlock(&nfs_client_lock);
 
 	if (server->destroy != NULL)
 		server->destroy(server);
@@ -1108,11 +1143,7 @@ struct nfs_server *nfs_create_server(const struct nfs_parsed_mount_data *data,
 		(unsigned long long) server->fsid.major,
 		(unsigned long long) server->fsid.minor);
 
-	spin_lock(&nfs_client_lock);
-	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
-	list_add_tail(&server->master_link, &nfs_volume_list);
-	spin_unlock(&nfs_client_lock);
-
+	nfs_server_insert_lists(server);
 	server->mount_time = jiffies;
 	nfs_free_fattr(fattr);
 	return server;
@@ -1124,6 +1155,101 @@ struct nfs_server *nfs_create_server(const struct nfs_parsed_mount_data *data,
 }
 
 #ifdef CONFIG_NFS_V4
+/*
+ * NFSv4.0 callback thread helper
+ *
+ * Find a client by IP address, protocol version, and minorversion
+ *
+ * Called from the pg_authenticate method. The callback identifier
+ * is not used as it has not been decoded.
+ *
+ * Returns NULL if no such client
+ */
+struct nfs_client *
+nfs4_find_client_no_ident(const struct sockaddr *addr)
+{
+	struct nfs_client *clp;
+
+	spin_lock(&nfs_client_lock);
+	list_for_each_entry(clp, &nfs_client_list, cl_share_link) {
+		if (nfs4_cb_match_client(addr, clp, 0) == false)
+			continue;
+		atomic_inc(&clp->cl_count);
+		spin_unlock(&nfs_client_lock);
+		return clp;
+	}
+	spin_unlock(&nfs_client_lock);
+	return NULL;
+}
+
+/*
+ * NFSv4.0 callback thread helper
+ *
+ * Find a client by callback identifier
+ */
+struct nfs_client *
+nfs4_find_client_ident(int cb_ident)
+{
+	struct nfs_client *clp;
+
+	spin_lock(&nfs_client_lock);
+	clp = idr_find(&cb_ident_idr, cb_ident);
+	if (clp)
+		atomic_inc(&clp->cl_count);
+	spin_unlock(&nfs_client_lock);
+	return clp;
+}
+
+#if defined(CONFIG_NFS_V4_1)
+/*
+ * NFSv4.1 callback thread helper
+ * For CB_COMPOUND calls, find a client by IP address, protocol version,
+ * minorversion, and sessionID
+ *
+ * CREATE_SESSION triggers a CB_NULL ping from servers. The callback service
+ * sessionid can only be set after the CREATE_SESSION return, so a CB_NULL
+ * can arrive before the callback sessionid is set. For CB_NULL calls,
+ * find a client by IP address protocol version, and minorversion.
+ *
+ * Returns NULL if no such client
+ */
+struct nfs_client *
+nfs4_find_client_sessionid(const struct sockaddr *addr,
+			   struct nfs4_sessionid *sid, int is_cb_compound)
+{
+	struct nfs_client *clp;
+
+	spin_lock(&nfs_client_lock);
+	list_for_each_entry(clp, &nfs_client_list, cl_share_link) {
+		if (nfs4_cb_match_client(addr, clp, 1) == false)
+			continue;
+
+		if (!nfs4_has_session(clp))
+			continue;
+
+		/* Match sessionid unless cb_null call*/
+		if (is_cb_compound && (memcmp(clp->cl_session->sess_id.data,
+		    sid->data, NFS4_MAX_SESSIONID_LEN) != 0))
+			continue;
+
+		atomic_inc(&clp->cl_count);
+		spin_unlock(&nfs_client_lock);
+		return clp;
+	}
+	spin_unlock(&nfs_client_lock);
+	return NULL;
+}
+
+#else /* CONFIG_NFS_V4_1 */
+
+struct nfs_client *
+nfs4_find_client_sessionid(const struct sockaddr *addr,
+			   struct nfs4_sessionid *sid, int is_cb_compound)
+{
+	return NULL;
+}
+#endif /* CONFIG_NFS_V4_1 */
+
 /*
  * Initialize the NFS4 callback service
  */
@@ -1342,11 +1468,7 @@ static int nfs4_server_common_setup(struct nfs_server *server,
 	if (server->namelen == 0 || server->namelen > NFS4_MAXNAMLEN)
 		server->namelen = NFS4_MAXNAMLEN;
 
-	spin_lock(&nfs_client_lock);
-	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
-	list_add_tail(&server->master_link, &nfs_volume_list);
-	spin_unlock(&nfs_client_lock);
-
+	nfs_server_insert_lists(server);
 	server->mount_time = jiffies;
 out:
 	nfs_free_fattr(fattr);
@@ -1551,11 +1673,7 @@ struct nfs_server *nfs_clone_server(struct nfs_server *source,
 	if (error < 0)
 		goto out_free_server;
 
-	spin_lock(&nfs_client_lock);
-	list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
-	list_add_tail(&server->master_link, &nfs_volume_list);
-	spin_unlock(&nfs_client_lock);
-
+	nfs_server_insert_lists(server);
 	server->mount_time = jiffies;
 
 	nfs_free_fattr(fattr_fsinfo);

fs/nfs/delegation.c

@@ -40,11 +40,23 @@ static void nfs_free_delegation(struct nfs_delegation *delegation)
 	call_rcu(&delegation->rcu, nfs_free_delegation_callback);
 }
 
+/**
+ * nfs_mark_delegation_referenced - set delegation's REFERENCED flag
+ * @delegation: delegation to process
+ *
+ */
 void nfs_mark_delegation_referenced(struct nfs_delegation *delegation)
 {
 	set_bit(NFS_DELEGATION_REFERENCED, &delegation->flags);
 }
 
+/**
+ * nfs_have_delegation - check if inode has a delegation
+ * @inode: inode to check
+ * @flags: delegation types to check for
+ *
+ * Returns one if inode has the indicated delegation, otherwise zero.
+ */
 int nfs_have_delegation(struct inode *inode, fmode_t flags)
 {
 	struct nfs_delegation *delegation;
@@ -119,10 +131,15 @@ static int nfs_delegation_claim_opens(struct inode *inode, const nfs4_stateid *s
 	return 0;
 }
 
-/*
- * Set up a delegation on an inode
+/**
+ * nfs_inode_reclaim_delegation - process a delegation reclaim request
+ * @inode: inode to process
+ * @cred: credential to use for request
+ * @res: new delegation state from server
+ *
  */
-void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
+void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred,
+				  struct nfs_openres *res)
 {
 	struct nfs_delegation *delegation;
 	struct rpc_cred *oldcred = NULL;
@@ -175,38 +192,52 @@ static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation
 	return inode;
 }
 
-static struct nfs_delegation *nfs_detach_delegation_locked(struct nfs_inode *nfsi,
-							   const nfs4_stateid *stateid,
-							   struct nfs_client *clp)
+static struct nfs_delegation *
+nfs_detach_delegation_locked(struct nfs_inode *nfsi,
+			     struct nfs_server *server)
 {
 	struct nfs_delegation *delegation =
 		rcu_dereference_protected(nfsi->delegation,
-					  lockdep_is_held(&clp->cl_lock));
+				lockdep_is_held(&server->nfs_client->cl_lock));
 
 	if (delegation == NULL)
 		goto nomatch;
+
 	spin_lock(&delegation->lock);
-	if (stateid != NULL && memcmp(delegation->stateid.data, stateid->data,
-				sizeof(delegation->stateid.data)) != 0)
-		goto nomatch_unlock;
 	list_del_rcu(&delegation->super_list);
 	delegation->inode = NULL;
 	nfsi->delegation_state = 0;
 	rcu_assign_pointer(nfsi->delegation, NULL);
 	spin_unlock(&delegation->lock);
 	return delegation;
-nomatch_unlock:
-	spin_unlock(&delegation->lock);
 nomatch:
 	return NULL;
 }
 
-/*
- * Set up a delegation on an inode
+static struct nfs_delegation *nfs_detach_delegation(struct nfs_inode *nfsi,
+						    struct nfs_server *server)
+{
+	struct nfs_client *clp = server->nfs_client;
+	struct nfs_delegation *delegation;
+
+	spin_lock(&clp->cl_lock);
+	delegation = nfs_detach_delegation_locked(nfsi, server);
+	spin_unlock(&clp->cl_lock);
+	return delegation;
+}
+
+/**
+ * nfs_inode_set_delegation - set up a delegation on an inode
+ * @inode: inode to which delegation applies
+ * @cred: cred to use for subsequent delegation processing
+ * @res: new delegation state from server
+ *
+ * Returns zero on success, or a negative errno value.
  */
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
+	struct nfs_server *server = NFS_SERVER(inode);
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation, *old_delegation;
 	struct nfs_delegation *freeme = NULL;
@@ -227,7 +258,7 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct
 
 	spin_lock(&clp->cl_lock);
 	old_delegation = rcu_dereference_protected(nfsi->delegation,
-						   lockdep_is_held(&clp->cl_lock));
+					lockdep_is_held(&clp->cl_lock));
 	if (old_delegation != NULL) {
 		if (memcmp(&delegation->stateid, &old_delegation->stateid,
 					sizeof(old_delegation->stateid)) == 0 &&
@@ -246,9 +277,9 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct
 			delegation = NULL;
 			goto out;
 		}
-		freeme = nfs_detach_delegation_locked(nfsi, NULL, clp);
+		freeme = nfs_detach_delegation_locked(nfsi, server);
 	}
-	list_add_rcu(&delegation->super_list, &clp->cl_delegations);
+	list_add_rcu(&delegation->super_list, &server->delegations);
 	nfsi->delegation_state = delegation->type;
 	rcu_assign_pointer(nfsi->delegation, delegation);
 	delegation = NULL;
@@ -290,73 +321,85 @@ static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegat
 	return err;
 }
 
-/*
- * Return all delegations that have been marked for return
+/**
+ * nfs_client_return_marked_delegations - return previously marked delegations
+ * @clp: nfs_client to process
+ *
+ * Returns zero on success, or a negative errno value.
  */
 int nfs_client_return_marked_delegations(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation;
+	struct nfs_server *server;
 	struct inode *inode;
 	int err = 0;
 
 restart:
 	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
-		if (!test_and_clear_bit(NFS_DELEGATION_RETURN, &delegation->flags))
-			continue;
-		inode = nfs_delegation_grab_inode(delegation);
-		if (inode == NULL)
-			continue;
-		spin_lock(&clp->cl_lock);
-		delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL, clp);
-		spin_unlock(&clp->cl_lock);
-		rcu_read_unlock();
-		if (delegation != NULL) {
-			filemap_flush(inode->i_mapping);
-			err = __nfs_inode_return_delegation(inode, delegation, 0);
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		list_for_each_entry_rcu(delegation, &server->delegations,
+								super_list) {
+			if (!test_and_clear_bit(NFS_DELEGATION_RETURN,
+							&delegation->flags))
+				continue;
+			inode = nfs_delegation_grab_inode(delegation);
+			if (inode == NULL)
+				continue;
+			delegation = nfs_detach_delegation(NFS_I(inode),
+								server);
+			rcu_read_unlock();
+
+			if (delegation != NULL) {
+				filemap_flush(inode->i_mapping);
+				err = __nfs_inode_return_delegation(inode,
+								delegation, 0);
+			}
+			iput(inode);
+			if (!err)
+				goto restart;
+			set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
+			return err;
 		}
-		iput(inode);
-		if (!err)
-			goto restart;
-		set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
-		return err;
 	}
 	rcu_read_unlock();
 	return 0;
 }
 
-/*
- * This function returns the delegation without reclaiming opens
- * or protecting against delegation reclaims.
- * It is therefore really only safe to be called from
- * nfs4_clear_inode()
+/**
+ * nfs_inode_return_delegation_noreclaim - return delegation, don't reclaim opens
+ * @inode: inode to process
+ *
+ * Does not protect against delegation reclaims, therefore really only safe
+ * to be called from nfs4_clear_inode().
  */
 void nfs_inode_return_delegation_noreclaim(struct inode *inode)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
+	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 
 	if (rcu_access_pointer(nfsi->delegation) != NULL) {
-		spin_lock(&clp->cl_lock);
-		delegation = nfs_detach_delegation_locked(nfsi, NULL, clp);
-		spin_unlock(&clp->cl_lock);
+		delegation = nfs_detach_delegation(nfsi, server);
 		if (delegation != NULL)
 			nfs_do_return_delegation(inode, delegation, 0);
 	}
 }
 
+/**
+ * nfs_inode_return_delegation - synchronously return a delegation
+ * @inode: inode to process
+ *
+ * Returns zero on success, or a negative errno value.
+ */
 int nfs_inode_return_delegation(struct inode *inode)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
+	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int err = 0;
 
 	if (rcu_access_pointer(nfsi->delegation) != NULL) {
-		spin_lock(&clp->cl_lock);
-		delegation = nfs_detach_delegation_locked(nfsi, NULL, clp);
-		spin_unlock(&clp->cl_lock);
+		delegation = nfs_detach_delegation(nfsi, server);
 		if (delegation != NULL) {
 			nfs_wb_all(inode);
 			err = __nfs_inode_return_delegation(inode, delegation, 1);
@@ -365,46 +408,61 @@ int nfs_inode_return_delegation(struct inode *inode)
 	return err;
 }
 
-static void nfs_mark_return_delegation(struct nfs_client *clp, struct nfs_delegation *delegation)
+static void nfs_mark_return_delegation(struct nfs_delegation *delegation)
 {
+	struct nfs_client *clp = NFS_SERVER(delegation->inode)->nfs_client;
+
 	set_bit(NFS_DELEGATION_RETURN, &delegation->flags);
 	set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
 }
 
-/*
- * Return all delegations associated to a super block
+/**
+ * nfs_super_return_all_delegations - return delegations for one superblock
+ * @sb: sb to process
+ *
  */
 void nfs_super_return_all_delegations(struct super_block *sb)
 {
-	struct nfs_client *clp = NFS_SB(sb)->nfs_client;
+	struct nfs_server *server = NFS_SB(sb);
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs_delegation *delegation;
 
 	if (clp == NULL)
 		return;
+
 	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
+	list_for_each_entry_rcu(delegation, &server->delegations, super_list) {
 		spin_lock(&delegation->lock);
-		if (delegation->inode != NULL && delegation->inode->i_sb == sb)
-			set_bit(NFS_DELEGATION_RETURN, &delegation->flags);
+		set_bit(NFS_DELEGATION_RETURN, &delegation->flags);
 		spin_unlock(&delegation->lock);
 	}
 	rcu_read_unlock();
+
 	if (nfs_client_return_marked_delegations(clp) != 0)
 		nfs4_schedule_state_manager(clp);
 }
 
-static
-void nfs_client_mark_return_all_delegation_types(struct nfs_client *clp, fmode_t flags)
+static void nfs_mark_return_all_delegation_types(struct nfs_server *server,
+						 fmode_t flags)
 {
 	struct nfs_delegation *delegation;
 
-	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
+	list_for_each_entry_rcu(delegation, &server->delegations, super_list) {
 		if ((delegation->type == (FMODE_READ|FMODE_WRITE)) && !(flags & FMODE_WRITE))
 			continue;
 		if (delegation->type & flags)
-			nfs_mark_return_delegation(clp, delegation);
+			nfs_mark_return_delegation(delegation);
 	}
+}
+
+static void nfs_client_mark_return_all_delegation_types(struct nfs_client *clp,
+							fmode_t flags)
+{
+	struct nfs_server *server;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
+		nfs_mark_return_all_delegation_types(server, flags);
 	rcu_read_unlock();
 }
 
@@ -419,19 +477,32 @@ static void nfs_delegation_run_state_manager(struct nfs_client *clp)
 		nfs4_schedule_state_manager(clp);
 }
 
+/**
+ * nfs_expire_all_delegation_types
+ * @clp: client to process
+ * @flags: delegation types to expire
+ *
+ */
 void nfs_expire_all_delegation_types(struct nfs_client *clp, fmode_t flags)
 {
 	nfs_client_mark_return_all_delegation_types(clp, flags);
 	nfs_delegation_run_state_manager(clp);
 }
 
+/**
+ * nfs_expire_all_delegations
+ * @clp: client to process
+ *
+ */
 void nfs_expire_all_delegations(struct nfs_client *clp)
 {
 	nfs_expire_all_delegation_types(clp, FMODE_READ|FMODE_WRITE);
 }
 
-/*
- * Return all delegations following an NFS4ERR_CB_PATH_DOWN error.
+/**
+ * nfs_handle_cb_pathdown - return all delegations after NFS4ERR_CB_PATH_DOWN
+ * @clp: client to process
+ *
  */
 void nfs_handle_cb_pathdown(struct nfs_client *clp)
 {
@@ -440,29 +511,43 @@ void nfs_handle_cb_pathdown(struct nfs_client *clp)
 	nfs_client_mark_return_all_delegations(clp);
 }
 
-static void nfs_client_mark_return_unreferenced_delegations(struct nfs_client *clp)
+static void nfs_mark_return_unreferenced_delegations(struct nfs_server *server)
 {
 	struct nfs_delegation *delegation;
 
-	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
+	list_for_each_entry_rcu(delegation, &server->delegations, super_list) {
 		if (test_and_clear_bit(NFS_DELEGATION_REFERENCED, &delegation->flags))
 			continue;
-		nfs_mark_return_delegation(clp, delegation);
+		nfs_mark_return_delegation(delegation);
 	}
-	rcu_read_unlock();
 }
 
+/**
+ * nfs_expire_unreferenced_delegations - Eliminate unused delegations
+ * @clp: nfs_client to process
+ *
+ */
 void nfs_expire_unreferenced_delegations(struct nfs_client *clp)
 {
-	nfs_client_mark_return_unreferenced_delegations(clp);
+	struct nfs_server *server;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
+		nfs_mark_return_unreferenced_delegations(server);
+	rcu_read_unlock();
+
 	nfs_delegation_run_state_manager(clp);
 }
 
-/*
- * Asynchronous delegation recall!
+/**
+ * nfs_async_inode_return_delegation - asynchronously return a delegation
+ * @inode: inode to process
+ * @stateid: state ID information from CB_RECALL arguments
+ *
+ * Returns zero on success, or a negative errno value.
  */
-int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid)
+int nfs_async_inode_return_delegation(struct inode *inode,
+				      const nfs4_stateid *stateid)
 {
 	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_delegation *delegation;
@@ -474,22 +559,21 @@ int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *s
 		rcu_read_unlock();
 		return -ENOENT;
 	}
-
-	nfs_mark_return_delegation(clp, delegation);
+	nfs_mark_return_delegation(delegation);
 	rcu_read_unlock();
+
 	nfs_delegation_run_state_manager(clp);
 	return 0;
 }
 
-/*
- * Retrieve the inode associated with a delegation
- */
-struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle)
+static struct inode *
+nfs_delegation_find_inode_server(struct nfs_server *server,
+				 const struct nfs_fh *fhandle)
 {
 	struct nfs_delegation *delegation;
 	struct inode *res = NULL;
-	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
+
+	list_for_each_entry_rcu(delegation, &server->delegations, super_list) {
 		spin_lock(&delegation->lock);
 		if (delegation->inode != NULL &&
 		    nfs_compare_fh(fhandle, &NFS_I(delegation->inode)->fh) == 0) {
@@ -499,49 +583,121 @@ struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs
 		if (res != NULL)
 			break;
 	}
+	return res;
+}
+
+/**
+ * nfs_delegation_find_inode - retrieve the inode associated with a delegation
+ * @clp: client state handle
+ * @fhandle: filehandle from a delegation recall
+ *
+ * Returns pointer to inode matching "fhandle," or NULL if a matching inode
+ * cannot be found.
+ */
+struct inode *nfs_delegation_find_inode(struct nfs_client *clp,
+					const struct nfs_fh *fhandle)
+{
+	struct nfs_server *server;
+	struct inode *res = NULL;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		res = nfs_delegation_find_inode_server(server, fhandle);
+		if (res != NULL)
+			break;
+	}
 	rcu_read_unlock();
 	return res;
 }
 
-/*
- * Mark all delegations as needing to be reclaimed
+static void nfs_delegation_mark_reclaim_server(struct nfs_server *server)
+{
+	struct nfs_delegation *delegation;
+
+	list_for_each_entry_rcu(delegation, &server->delegations, super_list)
+		set_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags);
+}
+
+/**
+ * nfs_delegation_mark_reclaim - mark all delegations as needing to be reclaimed
+ * @clp: nfs_client to process
+ *
  */
 void nfs_delegation_mark_reclaim(struct nfs_client *clp)
 {
-	struct nfs_delegation *delegation;
+	struct nfs_server *server;
+
 	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list)
-		set_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags);
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
+		nfs_delegation_mark_reclaim_server(server);
 	rcu_read_unlock();
 }
 
-/*
- * Reap all unclaimed delegations after reboot recovery is done
+/**
+ * nfs_delegation_reap_unclaimed - reap unclaimed delegations after reboot recovery is done
+ * @clp: nfs_client to process
+ *
  */
 void nfs_delegation_reap_unclaimed(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation;
+	struct nfs_server *server;
 	struct inode *inode;
+
 restart:
 	rcu_read_lock();
-	list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) {
-		if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0)
-			continue;
-		inode = nfs_delegation_grab_inode(delegation);
-		if (inode == NULL)
-			continue;
-		spin_lock(&clp->cl_lock);
-		delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL, clp);
-		spin_unlock(&clp->cl_lock);
-		rcu_read_unlock();
-		if (delegation != NULL)
-			nfs_free_delegation(delegation);
-		iput(inode);
-		goto restart;
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		list_for_each_entry_rcu(delegation, &server->delegations,
+								super_list) {
+			if (test_bit(NFS_DELEGATION_NEED_RECLAIM,
+						&delegation->flags) == 0)
+				continue;
+			inode = nfs_delegation_grab_inode(delegation);
+			if (inode == NULL)
+				continue;
+			delegation = nfs_detach_delegation(NFS_I(inode),
+								server);
+			rcu_read_unlock();
+
+			if (delegation != NULL)
+				nfs_free_delegation(delegation);
+			iput(inode);
+			goto restart;
+		}
 	}
 	rcu_read_unlock();
 }
 
+/**
+ * nfs_delegations_present - check for existence of delegations
+ * @clp: client state handle
+ *
+ * Returns one if there are any nfs_delegation structures attached
+ * to this nfs_client.
+ */
+int nfs_delegations_present(struct nfs_client *clp)
+{
+	struct nfs_server *server;
+	int ret = 0;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
+		if (!list_empty(&server->delegations)) {
+			ret = 1;
+			break;
+		}
+	rcu_read_unlock();
+	return ret;
+}
+
+/**
+ * nfs4_copy_delegation_stateid - Copy inode's state ID information
+ * @dst: stateid data structure to fill in
+ * @inode: inode to check
+ *
+ * Returns one and fills in "dst->data" * if inode had a delegation,
+ * otherwise zero is returned.
+ */
 int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);

fs/nfs/delegation.h

@@ -44,6 +44,7 @@ void nfs_expire_all_delegation_types(struct nfs_client *clp, fmode_t flags);
 void nfs_expire_unreferenced_delegations(struct nfs_client *clp);
 void nfs_handle_cb_pathdown(struct nfs_client *clp);
 int nfs_client_return_marked_delegations(struct nfs_client *clp);
+int nfs_delegations_present(struct nfs_client *clp);
 
 void nfs_delegation_mark_reclaim(struct nfs_client *clp);
 void nfs_delegation_reap_unclaimed(struct nfs_client *clp);

fs/nfs/dir.c

@@ -33,8 +33,8 @@
 #include <linux/namei.h>
 #include <linux/mount.h>
 #include <linux/sched.h>
-#include <linux/vmalloc.h>
 #include <linux/kmemleak.h>
+#include <linux/xattr.h>
 
 #include "delegation.h"
 #include "iostat.h"
@@ -125,9 +125,10 @@ const struct inode_operations nfs4_dir_inode_operations = {
 	.permission	= nfs_permission,
 	.getattr	= nfs_getattr,
 	.setattr	= nfs_setattr,
-	.getxattr       = nfs4_getxattr,
-	.setxattr       = nfs4_setxattr,
-	.listxattr      = nfs4_listxattr,
+	.getxattr	= generic_getxattr,
+	.setxattr	= generic_setxattr,
+	.listxattr	= generic_listxattr,
+	.removexattr	= generic_removexattr,
 };
 
 #endif /* CONFIG_NFS_V4 */
@@ -172,7 +173,7 @@ struct nfs_cache_array {
 	struct nfs_cache_array_entry array[0];
 };
 
-typedef __be32 * (*decode_dirent_t)(struct xdr_stream *, struct nfs_entry *, struct nfs_server *, int);
+typedef int (*decode_dirent_t)(struct xdr_stream *, struct nfs_entry *, int);
 typedef struct {
 	struct file	*file;
 	struct page	*page;
@@ -378,14 +379,14 @@ int nfs_readdir_xdr_filler(struct page **pages, nfs_readdir_descriptor_t *desc,
 	return error;
 }
 
-/* Fill in an entry based on the xdr code stored in desc->page */
-static
-int xdr_decode(nfs_readdir_descriptor_t *desc, struct nfs_entry *entry, struct xdr_stream *stream)
+static int xdr_decode(nfs_readdir_descriptor_t *desc,
+		      struct nfs_entry *entry, struct xdr_stream *xdr)
 {
-	__be32 *p = desc->decode(stream, entry, NFS_SERVER(desc->file->f_path.dentry->d_inode), desc->plus);
-	if (IS_ERR(p))
-		return PTR_ERR(p);
+	int error;
 
+	error = desc->decode(xdr, entry, desc->plus);
+	if (error)
+		return error;
 	entry->fattr->time_start = desc->timestamp;
 	entry->fattr->gencount = desc->gencount;
 	return 0;
@@ -459,25 +460,26 @@ void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry)
 /* Perform conversion from xdr to cache array */
 static
 int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *entry,
-				void *xdr_page, struct page *page, unsigned int buflen)
+				struct page **xdr_pages, struct page *page, unsigned int buflen)
 {
 	struct xdr_stream stream;
-	struct xdr_buf buf;
-	__be32 *ptr = xdr_page;
+	struct xdr_buf buf = {
+		.pages = xdr_pages,
+		.page_len = buflen,
+		.buflen = buflen,
+		.len = buflen,
+	};
+	struct page *scratch;
 	struct nfs_cache_array *array;
 	unsigned int count = 0;
 	int status;
 
-	buf.head->iov_base = xdr_page;
-	buf.head->iov_len = buflen;
-	buf.tail->iov_len = 0;
-	buf.page_base = 0;
-	buf.page_len = 0;
-	buf.buflen = buf.head->iov_len;
-	buf.len = buf.head->iov_len;
-
-	xdr_init_decode(&stream, &buf, ptr);
+	scratch = alloc_page(GFP_KERNEL);
+	if (scratch == NULL)
+		return -ENOMEM;
 
+	xdr_init_decode(&stream, &buf, NULL);
+	xdr_set_scratch_buffer(&stream, page_address(scratch), PAGE_SIZE);
 
 	do {
 		status = xdr_decode(desc, entry, &stream);
@@ -506,6 +508,8 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en
 		} else
 			status = PTR_ERR(array);
 	}
+
+	put_page(scratch);
 	return status;
 }
 
@@ -521,7 +525,6 @@ static
 void nfs_readdir_free_large_page(void *ptr, struct page **pages,
 		unsigned int npages)
 {
-	vm_unmap_ram(ptr, npages);
 	nfs_readdir_free_pagearray(pages, npages);
 }
 
@@ -530,9 +533,8 @@ void nfs_readdir_free_large_page(void *ptr, struct page **pages,
  * to nfs_readdir_free_large_page
  */
 static
-void *nfs_readdir_large_page(struct page **pages, unsigned int npages)
+int nfs_readdir_large_page(struct page **pages, unsigned int npages)
 {
-	void *ptr;
 	unsigned int i;
 
 	for (i = 0; i < npages; i++) {
@@ -541,13 +543,11 @@ void *nfs_readdir_large_page(struct page **pages, unsigned int npages)
 			goto out_freepages;
 		pages[i] = page;
 	}
+	return 0;
 
-	ptr = vm_map_ram(pages, npages, 0, PAGE_KERNEL);
-	if (!IS_ERR_OR_NULL(ptr))
-		return ptr;
 out_freepages:
 	nfs_readdir_free_pagearray(pages, i);
-	return NULL;
+	return -ENOMEM;
 }
 
 static
@@ -566,6 +566,7 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
 	entry.eof = 0;
 	entry.fh = nfs_alloc_fhandle();
 	entry.fattr = nfs_alloc_fattr();
+	entry.server = NFS_SERVER(inode);
 	if (entry.fh == NULL || entry.fattr == NULL)
 		goto out;
 
@@ -577,8 +578,8 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
 	memset(array, 0, sizeof(struct nfs_cache_array));
 	array->eof_index = -1;
 
-	pages_ptr = nfs_readdir_large_page(pages, array_size);
-	if (!pages_ptr)
+	status = nfs_readdir_large_page(pages, array_size);
+	if (status < 0)
 		goto out_release_array;
 	do {
 		unsigned int pglen;
@@ -587,7 +588,7 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
 		if (status < 0)
 			break;
 		pglen = status;
-		status = nfs_readdir_page_filler(desc, &entry, pages_ptr, page, pglen);
+		status = nfs_readdir_page_filler(desc, &entry, pages, page, pglen);
 		if (status < 0) {
 			if (status == -ENOSPC)
 				status = 0;
@@ -1221,7 +1222,7 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
 		goto out_unblock_sillyrename;
 	}
 	inode = nfs_fhget(dentry->d_sb, fhandle, fattr);
-	res = (struct dentry *)inode;
+	res = ERR_CAST(inode);
 	if (IS_ERR(res))
 		goto out_unblock_sillyrename;
 
@@ -1355,8 +1356,7 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 	if (nd->flags & LOOKUP_CREATE) {
 		attr.ia_mode = nd->intent.open.create_mode;
 		attr.ia_valid = ATTR_MODE;
-		if (!IS_POSIXACL(dir))
-			attr.ia_mode &= ~current_umask();
+		attr.ia_mode &= ~current_umask();
 	} else {
 		open_flags &= ~(O_EXCL | O_CREAT);
 		attr.ia_valid = 0;

fs/nfs/idmap.c

@@ -238,7 +238,7 @@ int nfs_map_gid_to_group(struct nfs_client *clp, __u32 gid, char *buf, size_t bu
 	return nfs_idmap_lookup_name(gid, "group", buf, buflen);
 }
 
-#else  /* CONFIG_NFS_USE_IDMAPPER not defined */
+#else  /* CONFIG_NFS_USE_NEW_IDMAPPER not defined */
 
 #include <linux/module.h>
 #include <linux/mutex.h>

fs/nfs/inode.c

@@ -1410,9 +1410,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
  */
 void nfs4_evict_inode(struct inode *inode)
 {
+	pnfs_destroy_layout(NFS_I(inode));
 	truncate_inode_pages(&inode->i_data, 0);
 	end_writeback(inode);
-	pnfs_destroy_layout(NFS_I(inode));
 	/* If we are holding a delegation, return it! */
 	nfs_inode_return_delegation_noreclaim(inode);
 	/* First call standard NFS clear_inode() code */
@@ -1619,6 +1619,7 @@ static void __exit exit_nfs_fs(void)
 #ifdef CONFIG_PROC_FS
 	rpc_proc_unregister("nfs");
 #endif
+	nfs_cleanup_cb_ident_idr();
 	unregister_nfs_fs();
 	nfs_fs_proc_exit();
 	nfsiod_stop();