Merge tag 'nfs-for-5.1-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client updates from Trond Myklebust:
 "Highlights include:

  Stable fixes:
   - Fixes for NFS I/O request leakages
   - Fix error handling paths in the NFS I/O recoalescing code
   - Reinitialise NFSv4.1 sequence results before retransmitting a
     request
   - Fix a soft lockup in the delegation recovery code
   - Bulk destroy of layouts needs to be safe w.r.t. umount
   - Prevent thundering herd issues when the SUNRPC socket is not
     connected
   - Respect RPC call timeouts when retrying transmission

  Features:
   - Convert rpc auth layer to use xdr_streams
   - Config option to disable insecure RPCSEC_GSS crypto types
   - Reduce size of RPC receive buffers
   - Readdirplus optimization by cache mechanism
   - Convert SUNRPC socket send code to use iov_iter()
   - SUNRPC micro-optimisations to avoid indirect calls
   - Add support for the pNFS LAYOUTERROR operation and use it with the
     pNFS/flexfiles driver
   - Add trace events to report non-zero NFS status codes
   - Various removals of unnecessary dprintks

  Bugfixes and cleanups:
   - Fix a number of sparse warnings and documentation format warnings
   - Fix nfs_parse_devname to not modify it's argument
   - Fix potential corruption of page being written through pNFS/blocks
   - fix xfstest generic/099 failures on nfsv3
   - Avoid NFSv4.1 "false retries" when RPC calls are interrupted
   - Abort I/O early if the pNFS/flexfiles layout segment was
     invalidated
   - Avoid unnecessary pNFS/flexfiles layout invalidations"

* tag 'nfs-for-5.1-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (90 commits)
  SUNRPC: Take the transport send lock before binding+connecting
  SUNRPC: Micro-optimise when the task is known not to be sleeping
  SUNRPC: Check whether the task was transmitted before rebind/reconnect
  SUNRPC: Remove redundant calls to RPC_IS_QUEUED()
  SUNRPC: Clean up
  SUNRPC: Respect RPC call timeouts when retrying transmission
  SUNRPC: Fix up RPC back channel transmission
  SUNRPC: Prevent thundering herd when the socket is not connected
  SUNRPC: Allow dynamic allocation of back channel slots
  NFSv4.1: Bump the default callback session slot count to 16
  SUNRPC: Convert remaining GFP_NOIO, and GFP_NOWAIT sites in sunrpc
  NFS/flexfiles: Clean up mirror DS initialisation
  NFS/flexfiles: Remove dead code in ff_layout_mirror_valid()
  NFS/flexfile: Simplify nfs4_ff_layout_select_ds_stateid()
  NFS/flexfile: Simplify nfs4_ff_layout_ds_version()
  NFS/flexfiles: Simplify ff_layout_get_ds_cred()
  NFS/flexfiles: Simplify nfs4_ff_find_or_create_ds_client()
  NFS/flexfiles: Simplify nfs4_ff_layout_select_ds_fh()
  NFS/flexfiles: Speed up read failover when DSes are down
  NFS/flexfiles: Don't invalidate DS deviceids for being unresponsive
  ...

40 files changed, 1291 insertions, 1095 deletions

diff --git a/fs/lockd/clnt4xdr.c b/fs/lockd/clnt4xdr.c
index 214a2fa1f1e3..7df6324ccb8a 100644
--- a/fs/lockd/clnt4xdr.c
+++ b/fs/lockd/clnt4xdr.c
@@ -75,17 +75,6 @@ static void nlm4_compute_offsets(const struct nlm_lock *lock,
 }
 
 /*
- * 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
@@ -176,7 +165,6 @@ out_size:
 	dprintk("NFS: returned cookie was too long: %u\n", length);
 	return -EIO;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
@@ -236,7 +224,6 @@ out_bad_xdr:
 			__func__, be32_to_cpup(p));
 	return -EIO;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
@@ -309,7 +296,6 @@ static int decode_nlm4_holder(struct xdr_stream *xdr, struct nlm_res *result)
 out:
 	return error;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
diff --git a/fs/lockd/clntxdr.c b/fs/lockd/clntxdr.c
index 747b9c8c940a..4df62f635529 100644
--- a/fs/lockd/clntxdr.c
+++ b/fs/lockd/clntxdr.c
@@ -71,17 +71,6 @@ static void nlm_compute_offsets(const struct nlm_lock *lock,
 }
 
 /*
- * 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
@@ -173,7 +162,6 @@ out_size:
 	dprintk("NFS: returned cookie was too long: %u\n", length);
 	return -EIO;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
@@ -231,7 +219,6 @@ out_enum:
 		__func__, be32_to_cpup(p));
 	return -EIO;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
@@ -303,7 +290,6 @@ static int decode_nlm_holder(struct xdr_stream *xdr, struct nlm_res *result)
 out:
 	return error;
 out_overflow:
-	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
 
diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c
index a87a56273407..06233bfa6d73 100644
--- a/fs/nfs/callback_xdr.c
+++ b/fs/nfs/callback_xdr.c
@@ -72,16 +72,6 @@ static int nfs4_encode_void(struct svc_rqst *rqstp, __be32 *p)
 	return xdr_ressize_check(rqstp, p);
 }
 
-static __be32 *read_buf(struct xdr_stream *xdr, size_t nbytes)
-{
-	__be32 *p;
-
-	p = xdr_inline_decode(xdr, nbytes);
-	if (unlikely(p == NULL))
-		printk(KERN_WARNING "NFS: NFSv4 callback reply buffer overflowed!\n");
-	return p;
-}
-
 static __be32 decode_string(struct xdr_stream *xdr, unsigned int *len,
 		const char **str, size_t maxlen)
 {
@@ -98,13 +88,13 @@ static __be32 decode_fh(struct xdr_stream *xdr, struct nfs_fh *fh)
 {
 	__be32 *p;
 
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	fh->size = ntohl(*p);
 	if (fh->size > NFS4_FHSIZE)
 		return htonl(NFS4ERR_BADHANDLE);
-	p = read_buf(xdr, fh->size);
+	p = xdr_inline_decode(xdr, fh->size);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	memcpy(&fh->data[0], p, fh->size);
@@ -117,11 +107,11 @@ static __be32 decode_bitmap(struct xdr_stream *xdr, uint32_t *bitmap)
 	__be32 *p;
 	unsigned int attrlen;
 
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	attrlen = ntohl(*p);
-	p = read_buf(xdr, attrlen << 2);
+	p = xdr_inline_decode(xdr, attrlen << 2);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	if (likely(attrlen > 0))
@@ -135,7 +125,7 @@ static __be32 decode_stateid(struct xdr_stream *xdr, nfs4_stateid *stateid)
 {
 	__be32 *p;
 
-	p = read_buf(xdr, NFS4_STATEID_SIZE);
+	p = xdr_inline_decode(xdr, NFS4_STATEID_SIZE);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	memcpy(stateid->data, p, NFS4_STATEID_SIZE);
@@ -156,7 +146,7 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound
 	status = decode_string(xdr, &hdr->taglen, &hdr->tag, CB_OP_TAGLEN_MAXSZ);
 	if (unlikely(status != 0))
 		return status;
-	p = read_buf(xdr, 12);
+	p = xdr_inline_decode(xdr, 12);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	hdr->minorversion = ntohl(*p++);
@@ -176,7 +166,7 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound
 static __be32 decode_op_hdr(struct xdr_stream *xdr, unsigned int *op)
 {
 	__be32 *p;
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE_HDR);
 	*op = ntohl(*p);
@@ -205,7 +195,7 @@ static __be32 decode_recall_args(struct svc_rqst *rqstp,
 	status = decode_delegation_stateid(xdr, &args->stateid);
 	if (unlikely(status != 0))
 		return status;
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 	args->truncate = ntohl(*p);
@@ -227,7 +217,7 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp,
 	__be32 status = 0;
 	uint32_t iomode;
 
-	p = read_buf(xdr, 4 * sizeof(uint32_t));
+	p = xdr_inline_decode(xdr, 4 * sizeof(uint32_t));
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_BADXDR);
 
@@ -245,14 +235,14 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp,
 		if (unlikely(status != 0))
 			return status;
 
-		p = read_buf(xdr, 2 * sizeof(uint64_t));
+		p = xdr_inline_decode(xdr, 2 * sizeof(uint64_t));
 		if (unlikely(p == NULL))
 			return htonl(NFS4ERR_BADXDR);
 		p = xdr_decode_hyper(p, &args->cbl_range.offset);
 		p = xdr_decode_hyper(p, &args->cbl_range.length);
 		return decode_layout_stateid(xdr, &args->cbl_stateid);
 	} else if (args->cbl_recall_type == RETURN_FSID) {
-		p = read_buf(xdr, 2 * sizeof(uint64_t));
+		p = xdr_inline_decode(xdr, 2 * sizeof(uint64_t));
 		if (unlikely(p == NULL))
 			return htonl(NFS4ERR_BADXDR);
 		p = xdr_decode_hyper(p, &args->cbl_fsid.major);
@@ -275,7 +265,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp,
 	args->ndevs = 0;
 
 	/* Num of device notifications */
-	p = read_buf(xdr, sizeof(uint32_t));
+	p = xdr_inline_decode(xdr, sizeof(uint32_t));
 	if (unlikely(p == NULL)) {
 		status = htonl(NFS4ERR_BADXDR);
 		goto out;
@@ -298,7 +288,8 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp,
 	for (i = 0; i < n; i++) {
 		struct cb_devicenotifyitem *dev = &args->devs[i];
 
-		p = read_buf(xdr, (4 * sizeof(uint32_t)) + NFS4_DEVICEID4_SIZE);
+		p = xdr_inline_decode(xdr, (4 * sizeof(uint32_t)) +
+				      NFS4_DEVICEID4_SIZE);
 		if (unlikely(p == NULL)) {
 			status = htonl(NFS4ERR_BADXDR);
 			goto err;
@@ -329,7 +320,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp,
 		p += XDR_QUADLEN(NFS4_DEVICEID4_SIZE);
 
 		if (dev->cbd_layout_type == NOTIFY_DEVICEID4_CHANGE) {
-			p = read_buf(xdr, sizeof(uint32_t));
+			p = xdr_inline_decode(xdr, sizeof(uint32_t));
 			if (unlikely(p == NULL)) {
 				status = htonl(NFS4ERR_BADXDR);
 				goto err;
@@ -359,7 +350,7 @@ static __be32 decode_sessionid(struct xdr_stream *xdr,
 {
 	__be32 *p;
 
-	p = read_buf(xdr, NFS4_MAX_SESSIONID_LEN);
+	p = xdr_inline_decode(xdr, NFS4_MAX_SESSIONID_LEN);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 
@@ -379,13 +370,13 @@ static __be32 decode_rc_list(struct xdr_stream *xdr,
 		goto out;
 
 	status = htonl(NFS4ERR_RESOURCE);
-	p = read_buf(xdr, sizeof(uint32_t));
+	p = xdr_inline_decode(xdr, sizeof(uint32_t));
 	if (unlikely(p == NULL))
 		goto out;
 
 	rc_list->rcl_nrefcalls = ntohl(*p++);
 	if (rc_list->rcl_nrefcalls) {
-		p = read_buf(xdr,
+		p = xdr_inline_decode(xdr,
 			     rc_list->rcl_nrefcalls * 2 * sizeof(uint32_t));
 		if (unlikely(p == NULL))
 			goto out;
@@ -418,7 +409,7 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp,
 	if (status)
 		return status;
 
-	p = read_buf(xdr, 5 * sizeof(uint32_t));
+	p = xdr_inline_decode(xdr, 5 * sizeof(uint32_t));
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_RESOURCE);
 
@@ -461,7 +452,7 @@ static __be32 decode_recallany_args(struct svc_rqst *rqstp,
 	uint32_t bitmap[2];
 	__be32 *p, status;
 
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_BADXDR);
 	args->craa_objs_to_keep = ntohl(*p++);
@@ -480,7 +471,7 @@ static __be32 decode_recallslot_args(struct svc_rqst *rqstp,
 	struct cb_recallslotargs *args = argp;
 	__be32 *p;
 
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_BADXDR);
 	args->crsa_target_highest_slotid = ntohl(*p++);
@@ -492,14 +483,14 @@ static __be32 decode_lockowner(struct xdr_stream *xdr, struct cb_notify_lock_arg
 	__be32		*p;
 	unsigned int	len;
 
-	p = read_buf(xdr, 12);
+	p = xdr_inline_decode(xdr, 12);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_BADXDR);
 
 	p = xdr_decode_hyper(p, &args->cbnl_owner.clientid);
 	len = be32_to_cpu(*p);
 
-	p = read_buf(xdr, len);
+	p = xdr_inline_decode(xdr, len);
 	if (unlikely(p == NULL))
 		return htonl(NFS4ERR_BADXDR);
 
@@ -537,7 +528,7 @@ static __be32 decode_write_response(struct xdr_stream *xdr,
 	__be32 *p;
 
 	/* skip the always zero field */
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		goto out;
 	p++;
@@ -577,7 +568,7 @@ static __be32 decode_offload_args(struct svc_rqst *rqstp,
 		return status;
 
 	/* decode status */
-	p = read_buf(xdr, 4);
+	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		goto out;
 	args->error = ntohl(*p++);
@@ -943,10 +934,11 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp)
 	};
 	unsigned int nops = 0;
 
-	xdr_init_decode(&xdr_in, &rqstp->rq_arg, rqstp->rq_arg.head[0].iov_base);
+	xdr_init_decode(&xdr_in, &rqstp->rq_arg,
+			rqstp->rq_arg.head[0].iov_base, NULL);
 
 	p = (__be32*)((char *)rqstp->rq_res.head[0].iov_base + rqstp->rq_res.head[0].iov_len);
-	xdr_init_encode(&xdr_out, &rqstp->rq_res, p);
+	xdr_init_encode(&xdr_out, &rqstp->rq_res, p, NULL);
 
 	status = decode_compound_hdr_arg(&xdr_in, &hdr_arg);
 	if (status == htonl(NFS4ERR_RESOURCE))
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 885363ca8569..2f6b447cdd82 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -229,6 +229,8 @@ static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation
 	spin_lock(&delegation->lock);
 	if (delegation->inode != NULL)
 		inode = igrab(delegation->inode);
+	if (!inode)
+		set_bit(NFS_DELEGATION_INODE_FREEING, &delegation->flags);
 	spin_unlock(&delegation->lock);
 	return inode;
 }
@@ -681,7 +683,7 @@ void nfs_expire_all_delegations(struct nfs_client *clp)
 
 /**
  * nfs_super_return_all_delegations - return delegations for one superblock
- * @sb: sb to process
+ * @server: pointer to nfs_server to process
  *
  */
 void nfs_server_return_all_delegations(struct nfs_server *server)
@@ -944,10 +946,11 @@ 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_RETURNING,
-						&delegation->flags))
-				continue;
-			if (test_bit(NFS_DELEGATION_NEED_RECLAIM,
+			if (test_bit(NFS_DELEGATION_INODE_FREEING,
+						&delegation->flags) ||
+			    test_bit(NFS_DELEGATION_RETURNING,
+						&delegation->flags) ||
+			    test_bit(NFS_DELEGATION_NEED_RECLAIM,
 						&delegation->flags) == 0)
 				continue;
 			if (!nfs_sb_active(server->super))
@@ -1053,10 +1056,11 @@ 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_RETURNING,
-						&delegation->flags))
-				continue;
-			if (test_bit(NFS_DELEGATION_TEST_EXPIRED,
+			if (test_bit(NFS_DELEGATION_INODE_FREEING,
+						&delegation->flags) ||
+			    test_bit(NFS_DELEGATION_RETURNING,
+						&delegation->flags) ||
+			    test_bit(NFS_DELEGATION_TEST_EXPIRED,
 						&delegation->flags) == 0)
 				continue;
 			if (!nfs_sb_active(server->super))
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index dcbf3394ba0e..35b4b02c1ae0 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -34,6 +34,7 @@ enum {
 	NFS_DELEGATION_RETURNING,
 	NFS_DELEGATION_REVOKED,
 	NFS_DELEGATION_TEST_EXPIRED,
+	NFS_DELEGATION_INODE_FREEING,
 };
 
 int nfs_inode_set_delegation(struct inode *inode, const struct cred *cred,
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 6bf4471850c8..a71d0b42d160 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -139,12 +139,19 @@ struct nfs_cache_array {
 	struct nfs_cache_array_entry array[0];
 };
 
+struct readdirvec {
+	unsigned long nr;
+	unsigned long index;
+	struct page *pages[NFS_MAX_READDIR_RAPAGES];
+};
+
 typedef int (*decode_dirent_t)(struct xdr_stream *, struct nfs_entry *, bool);
 typedef struct {
 	struct file	*file;
 	struct page	*page;
 	struct dir_context *ctx;
 	unsigned long	page_index;
+	struct readdirvec pvec;
 	u64		*dir_cookie;
 	u64		last_cookie;
 	loff_t		current_index;
@@ -524,6 +531,10 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en
 	struct nfs_cache_array *array;
 	unsigned int count = 0;
 	int status;
+	int max_rapages = NFS_MAX_READDIR_RAPAGES;
+
+	desc->pvec.index = desc->page_index;
+	desc->pvec.nr = 0;
 
 	scratch = alloc_page(GFP_KERNEL);
 	if (scratch == NULL)
@@ -548,20 +559,40 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en
 		if (desc->plus)
 			nfs_prime_dcache(file_dentry(desc->file), entry);
 
-		status = nfs_readdir_add_to_array(entry, page);
+		status = nfs_readdir_add_to_array(entry, desc->pvec.pages[desc->pvec.nr]);
+		if (status == -ENOSPC) {
+			desc->pvec.nr++;
+			if (desc->pvec.nr == max_rapages)
+				break;
+			status = nfs_readdir_add_to_array(entry, desc->pvec.pages[desc->pvec.nr]);
+		}
 		if (status != 0)
 			break;
 	} while (!entry->eof);
 
+	/*
+	 * page and desc->pvec.pages[0] are valid, don't need to check
+	 * whether or not to be NULL.
+	 */
+	copy_highpage(page, desc->pvec.pages[0]);
+
 out_nopages:
 	if (count == 0 || (status == -EBADCOOKIE && entry->eof != 0)) {
-		array = kmap(page);
+		array = kmap_atomic(desc->pvec.pages[desc->pvec.nr]);
 		array->eof_index = array->size;
 		status = 0;
-		kunmap(page);
+		kunmap_atomic(array);
 	}
 
 	put_page(scratch);
+
+	/*
+	 * desc->pvec.nr > 0 means at least one page was completely filled,
+	 * we should return -ENOSPC. Otherwise function
+	 * nfs_readdir_xdr_to_array will enter infinite loop.
+	 */
+	if (desc->pvec.nr > 0)
+		return -ENOSPC;
 	return status;
 }
 
@@ -574,8 +605,8 @@ void nfs_readdir_free_pages(struct page **pages, unsigned int npages)
 }
 
 /*
- * nfs_readdir_large_page will allocate pages that must be freed with a call
- * to nfs_readdir_free_pagearray
+ * nfs_readdir_alloc_pages() will allocate pages that must be freed with a call
+ * to nfs_readdir_free_pages()
  */
 static
 int nfs_readdir_alloc_pages(struct page **pages, unsigned int npages)
@@ -595,6 +626,24 @@ out_freepages:
 	return -ENOMEM;
 }
 
+/*
+ * nfs_readdir_rapages_init initialize rapages by nfs_cache_array structure.
+ */
+static
+void nfs_readdir_rapages_init(nfs_readdir_descriptor_t *desc)
+{
+	struct nfs_cache_array *array;
+	int max_rapages = NFS_MAX_READDIR_RAPAGES;
+	int index;
+
+	for (index = 0; index < max_rapages; index++) {
+		array = kmap_atomic(desc->pvec.pages[index]);
+		memset(array, 0, sizeof(struct nfs_cache_array));
+		array->eof_index = -1;
+		kunmap_atomic(array);
+	}
+}
+
 static
 int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page, struct inode *inode)
 {
@@ -605,6 +654,12 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
 	int status = -ENOMEM;
 	unsigned int array_size = ARRAY_SIZE(pages);
 
+	/*
+	 * This means we hit readdir rdpages miss, the preallocated rdpages
+	 * are useless, the preallocate rdpages should be reinitialized.
+	 */
+	nfs_readdir_rapages_init(desc);
+
 	entry.prev_cookie = 0;
 	entry.cookie = desc->last_cookie;
 	entry.eof = 0;
@@ -664,9 +719,24 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page)
 	struct inode	*inode = file_inode(desc->file);
 	int ret;
 
-	ret = nfs_readdir_xdr_to_array(desc, page, inode);
-	if (ret < 0)
-		goto error;
+	/*
+	 * If desc->page_index in range desc->pvec.index and
+	 * desc->pvec.index + desc->pvec.nr, we get readdir cache hit.
+	 */
+	if (desc->page_index >= desc->pvec.index &&
+		desc->page_index < (desc->pvec.index + desc->pvec.nr)) {
+		/*
+		 * page and desc->pvec.pages[x] are valid, don't need to check
+		 * whether or not to be NULL.
+		 */
+		copy_highpage(page, desc->pvec.pages[desc->page_index - desc->pvec.index]);
+		ret = 0;
+	} else {
+		ret = nfs_readdir_xdr_to_array(desc, page, inode);
+		if (ret < 0)
+			goto error;
+	}
+
 	SetPageUptodate(page);
 
 	if (invalidate_inode_pages2_range(inode->i_mapping, page->index + 1, -1) < 0) {
@@ -831,6 +901,7 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx)
 			*desc = &my_desc;
 	struct nfs_open_dir_context *dir_ctx = file->private_data;
 	int res = 0;
+	int max_rapages = NFS_MAX_READDIR_RAPAGES;
 
 	dfprintk(FILE, "NFS: readdir(%pD2) starting at cookie %llu\n",
 			file, (long long)ctx->pos);
@@ -850,6 +921,12 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx)
 	desc->decode = NFS_PROTO(inode)->decode_dirent;
 	desc->plus = nfs_use_readdirplus(inode, ctx);
 
+	res = nfs_readdir_alloc_pages(desc->pvec.pages, max_rapages);
+	if (res < 0)
+		return -ENOMEM;
+
+	nfs_readdir_rapages_init(desc);
+
 	if (ctx->pos == 0 || nfs_attribute_cache_expired(inode))
 		res = nfs_revalidate_mapping(inode, file->f_mapping);
 	if (res < 0)
@@ -885,6 +962,7 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx)
 			break;
 	} while (!desc->eof);
 out:
+	nfs_readdir_free_pages(desc->pvec.pages, max_rapages);
 	if (res > 0)
 		res = 0;
 	dfprintk(FILE, "NFS: readdir(%pD2) returns %d\n", file, res);
@@ -945,7 +1023,7 @@ static int nfs_fsync_dir(struct file *filp, loff_t start, loff_t end,
 
 /**
  * nfs_force_lookup_revalidate - Mark the directory as having changed
- * @dir - pointer to directory inode
+ * @dir: pointer to directory inode
  *
  * This forces the revalidation code in nfs_lookup_revalidate() to do a
  * full lookup on all child dentries of 'dir' whenever a change occurs
@@ -1649,7 +1727,7 @@ nfs4_do_lookup_revalidate(struct inode *dir, struct dentry *dentry,
 reval_dentry:
 	if (flags & LOOKUP_RCU)
 		return -ECHILD;
-	return nfs_lookup_revalidate_dentry(dir, dentry, inode);;
+	return nfs_lookup_revalidate_dentry(dir, dentry, inode);
 
 full_reval:
 	return nfs_do_lookup_revalidate(dir, dentry, flags);
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 33824a0a57bf..0fd811ac08b5 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -428,7 +428,7 @@ out_put:
 	hdr->release(hdr);
 }
 
-static void nfs_read_sync_pgio_error(struct list_head *head)
+static void nfs_read_sync_pgio_error(struct list_head *head, int error)
 {
 	struct nfs_page *req;
 
@@ -664,8 +664,7 @@ static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq)
 
 	list_for_each_entry_safe(req, tmp, &reqs, wb_list) {
 		if (!nfs_pageio_add_request(&desc, req)) {
-			nfs_list_remove_request(req);
-			nfs_list_add_request(req, &failed);
+			nfs_list_move_request(req, &failed);
 			spin_lock(&cinfo.inode->i_lock);
 			dreq->flags = 0;
 			if (desc.pg_error < 0)
@@ -821,7 +820,7 @@ out_put:
 	hdr->release(hdr);
 }
 
-static void nfs_write_sync_pgio_error(struct list_head *head)
+static void nfs_write_sync_pgio_error(struct list_head *head, int error)
 {
 	struct nfs_page *req;
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 29553fdba8af..4899b85f9b3c 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -89,8 +89,8 @@ EXPORT_SYMBOL_GPL(nfs_file_release);
 
 /**
  * nfs_revalidate_size - Revalidate the file size
- * @inode - pointer to inode struct
- * @file - pointer to struct file
+ * @inode: pointer to inode struct
+ * @filp: pointer to struct file
  *
  * Revalidates the file length. This is basically a wrapper around
  * nfs_revalidate_inode() that takes into account the fact that we may
@@ -276,6 +276,12 @@ EXPORT_SYMBOL_GPL(nfs_file_fsync);
  * then a modify/write/read cycle when writing to a page in the
  * page cache.
  *
+ * Some pNFS layout drivers can only read/write at a certain block
+ * granularity like all block devices and therefore we must perform
+ * read/modify/write whenever a page hasn't read yet and the data
+ * to be written there is not aligned to a block boundary and/or
+ * smaller than the block size.
+ *
  * The modify/write/read cycle may occur if a page is read before
  * being completely filled by the writer.  In this situation, the
  * page must be completely written to stable storage on the server
@@ -291,26 +297,32 @@ EXPORT_SYMBOL_GPL(nfs_file_fsync);
  * and that the new data won't completely replace the old data in
  * that range of the file.
  */
-static int nfs_want_read_modify_write(struct file *file, struct page *page,
-			loff_t pos, unsigned len)
+static bool nfs_full_page_write(struct page *page, loff_t pos, unsigned int len)
 {
 	unsigned int pglen = nfs_page_length(page);
 	unsigned int offset = pos & (PAGE_SIZE - 1);
 	unsigned int end = offset + len;
 
-	if (pnfs_ld_read_whole_page(file->f_mapping->host)) {
-		if (!PageUptodate(page))
-			return 1;
-		return 0;
-	}
+	return !pglen || (end >= pglen && !offset);
+}
 
-	if ((file->f_mode & FMODE_READ) &&	/* open for read? */
-	    !PageUptodate(page) &&		/* Uptodate? */
-	    !PagePrivate(page) &&		/* i/o request already? */
-	    pglen &&				/* valid bytes of file? */
-	    (end < pglen || offset))		/* replace all valid bytes? */
-		return 1;
-	return 0;
+static bool nfs_want_read_modify_write(struct file *file, struct page *page,
+			loff_t pos, unsigned int len)
+{
+	/*
+	 * Up-to-date pages, those with ongoing or full-page write
+	 * don't need read/modify/write
+	 */
+	if (PageUptodate(page) || PagePrivate(page) ||
+	    nfs_full_page_write(page, pos, len))
+		return false;
+
+	if (pnfs_ld_read_whole_page(file->f_mapping->host))
+		return true;
+	/* Open for reading too? */
+	if (file->f_mode & FMODE_READ)
+		return true;
+	return false;
 }
 
 /*
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index 63abe705f4ca..f9264e1922a2 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -410,7 +410,7 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh,
 	for (i = 0; i < fls->mirror_array_cnt; i++) {
 		struct nfs4_ff_layout_mirror *mirror;
 		struct cred *kcred;
-		const struct cred *cred;
+		const struct cred __rcu *cred;
 		kuid_t uid;
 		kgid_t gid;