Merge tag 'nfs-for-5.3-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

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

  Stable fixes:

   - Fix a page lock leak in nfs_pageio_resend()

   - Ensure O_DIRECT reports an error if the bytes read/written is 0

   - Don't handle errors if the bind/connect succeeded

   - Revert "NFSv4/flexfiles: Abort I/O early if the layout segment was
     invalidat ed"

  Bugfixes:

   - Don't refresh attributes with mounted-on-file information

   - Fix return values for nfs4_file_open() and nfs_finish_open()

   - Fix pnfs layoutstats reporting of I/O errors

   - Don't use soft RPC calls for pNFS/flexfiles I/O, and don't abort
     for soft I/O errors when the user specifies a hard mount.

   - Various fixes to the error handling in sunrpc

   - Don't report writepage()/writepages() errors twice"

* tag 'nfs-for-5.3-3' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  NFS: remove set but not used variable 'mapping'
  NFSv2: Fix write regression
  NFSv2: Fix eof handling
  NFS: Fix writepage(s) error handling to not report errors twice
  NFS: Fix spurious EIO read errors
  pNFS/flexfiles: Don't time out requests on hard mounts
  SUNRPC: Handle connection breakages correctly in call_status()
  Revert "NFSv4/flexfiles: Abort I/O early if the layout segment was invalidated"
  SUNRPC: Handle EADDRINUSE and ENOBUFS correctly
  pNFS/flexfiles: Turn off soft RPC calls
  SUNRPC: Don't handle errors if the bind/connect succeeded
  NFS: On fatal writeback errors, we need to call nfs_inode_remove_request()
  NFS: Fix initialisation of I/O result struct in nfs_pgio_rpcsetup
  NFS: Ensure O_DIRECT reports an error if the bytes read/written is 0
  NFSv4/pnfs: Fix a page lock leak in nfs_pageio_resend()
  NFSv4: Fix return value in nfs_finish_open()
  NFSv4: Fix return values for nfs4_file_open()
  NFS: Don't refresh attributes with mounted-on-file information

fs/nfs/dir.c

@@ -1487,7 +1487,7 @@ static int nfs_finish_open(struct nfs_open_context *ctx,
 	if (S_ISREG(file->f_path.dentry->d_inode->i_mode))
 		nfs_file_set_open_context(file, ctx);
 	else
-		err = -ESTALE;
+		err = -EOPENSTALE;
 out:
 	return err;
 }

fs/nfs/direct.c

@@ -401,15 +401,21 @@ static void nfs_direct_read_completion(struct nfs_pgio_header *hdr)
 	unsigned long bytes = 0;
 	struct nfs_direct_req *dreq = hdr->dreq;
 
-	if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
-		goto out_put;
-
 	spin_lock(&dreq->lock);
-	if (test_bit(NFS_IOHDR_ERROR, &hdr->flags) && (hdr->good_bytes == 0))
+	if (test_bit(NFS_IOHDR_ERROR, &hdr->flags))
 		dreq->error = hdr->error;
-	else
+
+	if (test_bit(NFS_IOHDR_REDO, &hdr->flags)) {
+		spin_unlock(&dreq->lock);
+		goto out_put;
+	}
+
+	if (hdr->good_bytes != 0)
 		nfs_direct_good_bytes(dreq, hdr);
 
+	if (test_bit(NFS_IOHDR_EOF, &hdr->flags))
+		dreq->error = 0;
+
 	spin_unlock(&dreq->lock);
 
 	while (!list_empty(&hdr->pages)) {
@@ -782,16 +788,19 @@ static void nfs_direct_write_completion(struct nfs_pgio_header *hdr)
 	bool request_commit = false;
 	struct nfs_page *req = nfs_list_entry(hdr->pages.next);
 
-	if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
-		goto out_put;
-
 	nfs_init_cinfo_from_dreq(&cinfo, dreq);
 
 	spin_lock(&dreq->lock);
 
 	if (test_bit(NFS_IOHDR_ERROR, &hdr->flags))
 		dreq->error = hdr->error;
-	if (dreq->error == 0) {
+
+	if (test_bit(NFS_IOHDR_REDO, &hdr->flags)) {
+		spin_unlock(&dreq->lock);
+		goto out_put;
+	}
+
+	if (hdr->good_bytes != 0) {
 		nfs_direct_good_bytes(dreq, hdr);
 		if (nfs_write_need_commit(hdr)) {
 			if (dreq->flags == NFS_ODIRECT_RESCHED_WRITES)

fs/nfs/flexfilelayout/flexfilelayout.c

@@ -8,6 +8,7 @@
  */
 
 #include <linux/nfs_fs.h>
+#include <linux/nfs_mount.h>
 #include <linux/nfs_page.h>
 #include <linux/module.h>
 #include <linux/sched/mm.h>
@@ -928,7 +929,9 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
 	pgm = &pgio->pg_mirrors[0];
 	pgm->pg_bsize = mirror->mirror_ds->ds_versions[0].rsize;
 
-	pgio->pg_maxretrans = io_maxretrans;
+	if (NFS_SERVER(pgio->pg_inode)->flags &
+			(NFS_MOUNT_SOFT|NFS_MOUNT_SOFTERR))
+		pgio->pg_maxretrans = io_maxretrans;
 	return;
 out_nolseg:
 	if (pgio->pg_error < 0)
@@ -940,6 +943,7 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
 			pgio->pg_lseg);
 	pnfs_put_lseg(pgio->pg_lseg);
 	pgio->pg_lseg = NULL;
+	pgio->pg_maxretrans = 0;
 	nfs_pageio_reset_read_mds(pgio);
 }
 
@@ -1000,7 +1004,9 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
 		pgm->pg_bsize = mirror->mirror_ds->ds_versions[0].wsize;
 	}
 
-	pgio->pg_maxretrans = io_maxretrans;
+	if (NFS_SERVER(pgio->pg_inode)->flags &
+			(NFS_MOUNT_SOFT|NFS_MOUNT_SOFTERR))
+		pgio->pg_maxretrans = io_maxretrans;
 	return;
 
 out_mds:
@@ -1010,6 +1016,7 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
 			pgio->pg_lseg);
 	pnfs_put_lseg(pgio->pg_lseg);
 	pgio->pg_lseg = NULL;
+	pgio->pg_maxretrans = 0;
 	nfs_pageio_reset_write_mds(pgio);
 }
 
@@ -1148,8 +1155,6 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task,
 		break;
 	case -NFS4ERR_RETRY_UNCACHED_REP:
 		break;
-	case -EAGAIN:
-		return -NFS4ERR_RESET_TO_PNFS;
 	/* Invalidate Layout errors */
 	case -NFS4ERR_PNFS_NO_LAYOUT:
 	case -ESTALE:           /* mapped NFS4ERR_STALE */
@@ -1210,7 +1215,6 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task,
 	case -EBADHANDLE:
 	case -ELOOP:
 	case -ENOSPC:
-	case -EAGAIN:
 		break;
 	case -EJUKEBOX:
 		nfs_inc_stats(lseg->pls_layout->plh_inode, NFSIOS_DELAY);
@@ -1445,16 +1449,6 @@ static void ff_layout_read_prepare_v4(struct rpc_task *task, void *data)
 	ff_layout_read_prepare_common(task, hdr);
 }
 
-static void
-ff_layout_io_prepare_transmit(struct rpc_task *task,
-		void *data)
-{
-	struct nfs_pgio_header *hdr = data;
-
-	if (!pnfs_is_valid_lseg(hdr->lseg))
-		rpc_exit(task, -EAGAIN);
-}
-
 static void ff_layout_read_call_done(struct rpc_task *task, void *data)
 {
 	struct nfs_pgio_header *hdr = data;
@@ -1740,7 +1734,6 @@ static void ff_layout_commit_release(void *data)
 
 static const struct rpc_call_ops ff_layout_read_call_ops_v3 = {
 	.rpc_call_prepare = ff_layout_read_prepare_v3,
-	.rpc_call_prepare_transmit = ff_layout_io_prepare_transmit,
 	.rpc_call_done = ff_layout_read_call_done,
 	.rpc_count_stats = ff_layout_read_count_stats,
 	.rpc_release = ff_layout_read_release,
@@ -1748,7 +1741,6 @@ static const struct rpc_call_ops ff_layout_read_call_ops_v3 = {
 
 static const struct rpc_call_ops ff_layout_read_call_ops_v4 = {
 	.rpc_call_prepare = ff_layout_read_prepare_v4,
-	.rpc_call_prepare_transmit = ff_layout_io_prepare_transmit,
 	.rpc_call_done = ff_layout_read_call_done,
 	.rpc_count_stats = ff_layout_read_count_stats,
 	.rpc_release = ff_layout_read_release,
@@ -1756,7 +1748,6 @@ static const struct rpc_call_ops ff_layout_read_call_ops_v4 = {
 
 static const struct rpc_call_ops ff_layout_write_call_ops_v3 = {
 	.rpc_call_prepare = ff_layout_write_prepare_v3,
-	.rpc_call_prepare_transmit = ff_layout_io_prepare_transmit,
 	.rpc_call_done = ff_layout_write_call_done,
 	.rpc_count_stats = ff_layout_write_count_stats,
 	.rpc_release = ff_layout_write_release,
@@ -1764,7 +1755,6 @@ static const struct rpc_call_ops ff_layout_write_call_ops_v3 = {
 
 static const struct rpc_call_ops ff_layout_write_call_ops_v4 = {
 	.rpc_call_prepare = ff_layout_write_prepare_v4,
-	.rpc_call_prepare_transmit = ff_layout_io_prepare_transmit,
 	.rpc_call_done = ff_layout_write_call_done,
 	.rpc_count_stats = ff_layout_write_count_stats,
 	.rpc_release = ff_layout_write_release,

fs/nfs/inode.c

@@ -1403,12 +1403,21 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat
 	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
 		return 0;
 
+	/* No fileid? Just exit */
+	if (!(fattr->valid & NFS_ATTR_FATTR_FILEID))
+		return 0;
 	/* Has the inode gone and changed behind our back? */
-	if ((fattr->valid & NFS_ATTR_FATTR_FILEID) && nfsi->fileid != fattr->fileid)
+	if (nfsi->fileid != fattr->fileid) {
+		/* Is this perhaps the mounted-on fileid? */
+		if ((fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) &&
+		    nfsi->fileid == fattr->mounted_on_fileid)
+			return 0;
 		return -ESTALE;
+	}
 	if ((fattr->valid & NFS_ATTR_FATTR_TYPE) && (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT))
 		return -ESTALE;
 
+
 	if (!nfs_file_has_buffered_writers(nfsi)) {
 		/* Verify a few of the more important attributes */
 		if ((fattr->valid & NFS_ATTR_FATTR_CHANGE) != 0 && !inode_eq_iversion_raw(inode, fattr->change_attr))
@@ -1768,18 +1777,6 @@ int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fa
 EXPORT_SYMBOL_GPL(nfs_post_op_update_inode_force_wcc);
 
 
-static inline bool nfs_fileid_valid(struct nfs_inode *nfsi,
-				    struct nfs_fattr *fattr)
-{
-	bool ret1 = true, ret2 = true;
-
-	if (fattr->valid & NFS_ATTR_FATTR_FILEID)
-		ret1 = (nfsi->fileid == fattr->fileid);
-	if (fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID)
-		ret2 = (nfsi->fileid == fattr->mounted_on_fileid);
-	return ret1 || ret2;
-}
-
 /*
  * Many nfs protocol calls return the new file attributes after
  * an operation.  Here we update the inode to reflect the state
@@ -1810,7 +1807,15 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 			nfs_display_fhandle_hash(NFS_FH(inode)),
 			atomic_read(&inode->i_count), fattr->valid);
 
-	if (!nfs_fileid_valid(nfsi, fattr)) {
+	/* No fileid? Just exit */
+	if (!(fattr->valid & NFS_ATTR_FATTR_FILEID))
+		return 0;
+	/* Has the inode gone and changed behind our back? */
+	if (nfsi->fileid != fattr->fileid) {
+		/* Is this perhaps the mounted-on fileid? */
+		if ((fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) &&
+		    nfsi->fileid == fattr->mounted_on_fileid)
+			return 0;
 		printk(KERN_ERR "NFS: server %s error: fileid changed\n"
 			"fsid %s: expected fileid 0x%Lx, got 0x%Lx\n",
 			NFS_SERVER(inode)->nfs_client->cl_hostname,

fs/nfs/internal.h

@@ -775,3 +775,13 @@ static inline bool nfs_error_is_fatal(int err)
 	}
 }
 
+static inline bool nfs_error_is_fatal_on_server(int err)
+{
+	switch (err) {
+	case 0:
+	case -ERESTARTSYS:
+	case -EINTR:
+		return false;
+	}
+	return nfs_error_is_fatal(err);
+}

fs/nfs/nfs4file.c

@@ -73,13 +73,13 @@ nfs4_file_open(struct inode *inode, struct file *filp)
 	if (IS_ERR(inode)) {
 		err = PTR_ERR(inode);
 		switch (err) {
-		case -EPERM:
-		case -EACCES:
-		case -EDQUOT:
-		case -ENOSPC:
-		case -EROFS:
-			goto out_put_ctx;
 		default:
+			goto out_put_ctx;
+		case -ENOENT:
+		case -ESTALE:
+		case -EISDIR:
+		case -ENOTDIR:
+		case -ELOOP:
 			goto out_drop;
 		}
 	}

fs/nfs/pagelist.c

@@ -590,7 +590,7 @@ static void nfs_pgio_rpcsetup(struct nfs_pgio_header *hdr,
 	}
 
 	hdr->res.fattr   = &hdr->fattr;
-	hdr->res.count   = count;
+	hdr->res.count   = 0;
 	hdr->res.eof     = 0;
 	hdr->res.verf    = &hdr->verf;
 	nfs_fattr_init(&hdr->fattr);
@@ -1251,20 +1251,23 @@ static void nfs_pageio_complete_mirror(struct nfs_pageio_descriptor *desc,
 int nfs_pageio_resend(struct nfs_pageio_descriptor *desc,
 		      struct nfs_pgio_header *hdr)
 {
-	LIST_HEAD(failed);
+	LIST_HEAD(pages);
 
 	desc->pg_io_completion = hdr->io_completion;
 	desc->pg_dreq = hdr->dreq;
-	while (!list_empty(&hdr->pages)) {
-		struct nfs_page *req = nfs_list_entry(hdr->pages.next);
+	list_splice_init(&hdr->pages, &pages);
+	while (!list_empty(&pages)) {
+		struct nfs_page *req = nfs_list_entry(pages.next);
 
 		if (!nfs_pageio_add_request(desc, req))
-			nfs_list_move_request(req, &failed);
+			break;
 	}
 	nfs_pageio_complete(desc);
-	if (!list_empty(&failed)) {
-		list_move(&failed, &hdr->pages);
-		return desc->pg_error < 0 ? desc->pg_error : -EIO;
+	if (!list_empty(&pages)) {
+		int err = desc->pg_error < 0 ? desc->pg_error : -EIO;
+		hdr->completion_ops->error_cleanup(&pages, err);
+		nfs_set_pgio_error(hdr, err, hdr->io_start);
+		return err;
 	}
 	return 0;
 }

fs/nfs/pnfs_nfs.c

@@ -627,11 +627,16 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv,
 			/* Add this address as an alias */
 			rpc_clnt_add_xprt(clp->cl_rpcclient, &xprt_args,
 					rpc_clnt_test_and_add_xprt, NULL);
-		} else
-			clp = get_v3_ds_connect(mds_srv,
-					(struct sockaddr *)&da->da_addr,
-					da->da_addrlen, IPPROTO_TCP,
-					timeo, retrans);
+			continue;
+		}
+		clp = get_v3_ds_connect(mds_srv,
+				(struct sockaddr *)&da->da_addr,
+				da->da_addrlen, IPPROTO_TCP,
+				timeo, retrans);
+		if (IS_ERR(clp))
+			continue;
+		clp->cl_rpcclient->cl_softerr = 0;
+		clp->cl_rpcclient->cl_softrtry = 0;
 	}
 
 	if (IS_ERR(clp)) {

fs/nfs/proc.c

@@ -594,7 +594,8 @@ static int nfs_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 		/* Emulate the eof flag, which isn't normally needed in NFSv2
 		 * as it is guaranteed to always return the file attributes
 		 */
-		if (hdr->args.offset + hdr->res.count >= hdr->res.fattr->size)
+		if ((hdr->res.count == 0 && hdr->args.count > 0) ||
+		    hdr->args.offset + hdr->res.count >= hdr->res.fattr->size)
 			hdr->res.eof = 1;
 	}
 	return 0;
@@ -615,8 +616,10 @@ static int nfs_proc_pgio_rpc_prepare(struct rpc_task *task,
 
 static int nfs_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	if (task->tk_status >= 0)
+	if (task->tk_status >= 0) {
+		hdr->res.count = hdr->args.count;
 		nfs_writeback_update_inode(hdr);
+	}
 	return 0;
 }
 

fs/nfs/read.c

@@ -91,19 +91,25 @@ void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio)
 }
 EXPORT_SYMBOL_GPL(nfs_pageio_reset_read_mds);
 
-static void nfs_readpage_release(struct nfs_page *req)
+static void nfs_readpage_release(struct nfs_page *req, int error)
 {
 	struct inode *inode = d_inode(nfs_req_openctx(req)->dentry);
+	struct page *page = req->wb_page;
 
 	dprintk("NFS: read done (%s/%llu %d@%lld)\n", inode->i_sb->s_id,
 		(unsigned long long)NFS_FILEID(inode), req->wb_bytes,
 		(long long)req_offset(req));
 
+	if (nfs_error_is_fatal_on_server(error) && error != -ETIMEDOUT)
+		SetPageError(page);
 	if (nfs_page_group_sync_on_bit(req, PG_UNLOCKPAGE)) {
-		if (PageUptodate(req->wb_page))
-			nfs_readpage_to_fscache(inode, req->wb_page, 0);
+		struct address_space *mapping = page_file_mapping(page);
 
-		unlock_page(req->wb_page);
+		if (PageUptodate(page))
+			nfs_readpage_to_fscache(inode, page, 0);
+		else if (!PageError(page) && !PagePrivate(page))
+			generic_error_remove_page(mapping, page);
+		unlock_page(page);
 	}
 	nfs_release_request(req);
 }
@@ -131,7 +137,7 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
 			     &nfs_async_read_completion_ops);
 	if (!nfs_pageio_add_request(&pgio, new)) {
 		nfs_list_remove_request(new);
-		nfs_readpage_release(new);
+		nfs_readpage_release(new, pgio.pg_error);
 	}
 	nfs_pageio_complete(&pgio);
 
@@ -153,6 +159,7 @@ static void nfs_page_group_set_uptodate(struct nfs_page *req)
 static void nfs_read_completion(struct nfs_pgio_header *hdr)
 {
 	unsigned long bytes = 0;
+	int error;
 
 	if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
 		goto out;
@@ -179,14 +186,19 @@ static void nfs_read_completion(struct nfs_pgio_header *hdr)
 				zero_user_segment(page, start, end);
 			}
 		}
+		error = 0;
 		bytes += req->wb_bytes;
 		if (test_bit(NFS_IOHDR_ERROR, &hdr->flags)) {
 			if (bytes <= hdr->good_bytes)
 				nfs_page_group_set_uptodate(req);
+			else {
+				error = hdr->error;
+				xchg(&nfs_req_openctx(req)->error, error);
+			}
 		} else
 			nfs_page_group_set_uptodate(req);
 		nfs_list_remove_request(req);
-		nfs_readpage_release(req);
+		nfs_readpage_release(req, error);
 	}
 out:
 	hdr->release(hdr);
@@ -213,7 +225,7 @@ nfs_async_read_error(struct list_head *head, int error)
 	while (!list_empty(head)) {
 		req = nfs_list_entry(head->next);
 		nfs_list_remove_request(req);
-		nfs_readpage_release(req);
+		nfs_readpage_release(req, error);
 	}
 }
 
@@ -337,8 +349,13 @@ int nfs_readpage(struct file *file, struct page *page)
 			goto out;
 	}
 
+	xchg(&ctx->error, 0);
 	error = nfs_readpage_async(ctx, inode, page);
-
+	if (!error) {
+		error = wait_on_page_locked_killable(page);
+		if (!PageUptodate(page) && !error)
+			error = xchg(&ctx->error, 0);
+	}
 out:
 	put_nfs_open_context(ctx);
 	return error;
@@ -372,8 +389,8 @@ readpage_async_filler(void *data, struct page *page)
 		zero_user_segment(page, len, PAGE_SIZE);
 	if (!nfs_pageio_add_request(desc->pgio, new)) {
 		nfs_list_remove_request(new);
-		nfs_readpage_release(new);
 		error = desc->pgio->pg_error;
+		nfs_readpage_release(new, error);
 		goto out;
 	}
 	return 0;

fs/nfs/write.c

@@ -57,6 +57,7 @@ static const struct rpc_call_ops nfs_commit_ops;
 static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops;
 static const struct nfs_commit_completion_ops nfs_commit_completion_ops;
 static const struct nfs_rw_ops nfs_rw_write_ops;
+static void nfs_inode_remove_request(struct nfs_page *req);
 static void nfs_clear_request_commit(struct nfs_page *req);
 static void nfs_init_cinfo_from_inode(struct nfs_commit_info *cinfo,
 				      struct inode *inode);
@@ -591,23 +592,13 @@ nfs_lock_and_join_requests(struct page *page)
 
 static void nfs_write_error(struct nfs_page *req, int error)
 {
+	nfs_set_pageerror(page_file_mapping(req->wb_page));
 	nfs_mapping_set_error(req->wb_page, error);
+	nfs_inode_remove_request(req);
 	nfs_end_page_writeback(req);
 	nfs_release_request(req);
 }
 
-static bool
-nfs_error_is_fatal_on_server(int err)
-{
-	switch (err) {
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-		return false;
-	}
-	return nfs_error_is_fatal(err);
-}
-
 /*
  * Find an associated nfs write request, and prepare to flush it out
  * May return an error if the user signalled nfs_wait_on_request().
@@ -615,7 +606,6 @@ nfs_error_is_fatal_on_server(int err)
 static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 				struct page *page)
 {
-	struct address_space *mapping;
 	struct nfs_page *req;
 	int ret = 0;
 
@@ -630,12 +620,11 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 	WARN_ON_ONCE(test_bit(PG_CLEAN, &req->wb_flags));
 
 	/* If there is a fatal error that covers this write, just exit */
-	ret = 0;
-	mapping = page_file_mapping(page);
-	if (test_bit(AS_ENOSPC, &mapping->flags) ||
-	    test_bit(AS_EIO, &mapping->flags))
+	ret = pgio->pg_error;
+	if (nfs_error_is_fatal_on_server(ret))
 		goto out_launder;
 
+	ret = 0;
 	if (!nfs_pageio_add_request(pgio, req)) {
 		ret = pgio->pg_error;
 		/*
@@ -647,6 +636,7 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 		} else
 			ret = -EAGAIN;
 		nfs_redirty_request(req);
+		pgio->pg_error = 0;
 	} else
 		nfs_add_stats(page_file_mapping(page)->host,
 				NFSIOS_WRITEPAGES, 1);
@@ -666,7 +656,7 @@ static int nfs_do_writepage(struct page *page, struct writeback_control *wbc,
 	ret = nfs_page_async_flush(pgio, page);
 	if (ret == -EAGAIN) {
 		redirty_page_for_writepage(wbc, page);
-		ret = 0;
+		ret = AOP_WRITEPAGE_ACTIVATE;
 	}
 	return ret;
 }
@@ -685,10 +675,11 @@ static int nfs_writepage_locked(struct page *page,
 	nfs_pageio_init_write(&pgio, inode, 0,
 				false, &nfs_async_write_completion_ops);
 	err = nfs_do_writepage(page, wbc, &pgio);
+	pgio.pg_error = 0;
 	nfs_pageio_complete(&pgio);
 	if (err < 0)
 		return err;
-	if (pgio.pg_error < 0)
+	if (nfs_error_is_fatal(pgio.pg_error))
 		return pgio.pg_error;
 	return 0;
 }
@@ -698,7 +689,8 @@ int nfs_writepage(struct page *page, struct writeback_control *wbc)
 	int ret;
 
 	ret = nfs_writepage_locked(page, wbc);
-	unlock_page(page);
+	if (ret != AOP_WRITEPAGE_ACTIVATE)
+		unlock_page(page);
 	return ret;
 }
 
@@ -707,7 +699,8 @@ static int nfs_writepages_callback(struct page *page, struct writeback_control *
 	int ret;
 
 	ret = nfs_do_writepage(page, wbc, data);
-	unlock_page(page);
+	if (ret != AOP_WRITEPAGE_ACTIVATE)
+		unlock_page(page);
 	return ret;
 }
 
@@ -733,13 +726,14 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 				&nfs_async_write_completion_ops);
 	pgio.pg_io_completion = ioc;
 	err = write_cache_pages(mapping, wbc, nfs_writepages_callback, &pgio);
+	pgio.pg_error = 0;
 	nfs_pageio_complete(&pgio);
 	nfs_io_completion_put(ioc);
 
 	if (err < 0)
 		goto out_err;
 	err = pgio.pg_error;
-	if (err < 0)
+	if (nfs_error_is_fatal(err))
 		goto out_err;
 	return 0;
 out_err:

include/linux/sunrpc/sched.h

@@ -98,7 +98,6 @@ typedef void			(*rpc_action)(struct rpc_task *);
 
 struct rpc_call_ops {
 	void (*rpc_call_prepare)(struct rpc_task *, void *);
-	void (*rpc_call_prepare_transmit)(struct rpc_task *, void *);
 	void (*rpc_call_done)(struct rpc_task *, void *);
 	void (*rpc_count_stats)(struct rpc_task *, void *);
 	void (*rpc_release)(void *);

net/sunrpc/clnt.c

@@ -1970,6 +1970,7 @@ call_bind(struct rpc_task *task)
 static void
 call_bind_status(struct rpc_task *task)
 {
+	struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt;
 	int status = -EIO;
 
 	if (rpc_task_transmitted(task)) {
@@ -1977,14 +1978,15 @@ call_bind_status(struct rpc_task *task)
 		return;
 	}
 
-	if (task->tk_status >= 0) {
-		dprint_status(task);
+	dprint_status(task);
+	trace_rpc_bind_status(task);
+	if (task->tk_status >= 0)
+		goto out_next;
+	if (xprt_bound(xprt)) {
 		task->tk_status = 0;
-		task->tk_action = call_connect;
-		return;
+		goto out_next;
 	}
 
-	trace_rpc_bind_status(task);
 	switch (task->tk_status) {
 	case -ENOMEM:
 		dprintk("RPC: %5u rpcbind out of memory\n", task->tk_pid);
@@ -2003,6 +2005,9 @@ call_bind_status(struct rpc_task *task)
 		task->tk_rebind_retry--;
 		rpc_delay(task, 3*HZ);
 		goto retry_timeout;
+	case -ENOBUFS:
+		rpc_delay(task, HZ >> 2);
+		goto retry_timeout;
 	case -EAGAIN:
 		goto retry_timeout;
 	case -ETIMEDOUT:
@@ -2026,7 +2031,6 @@ call_bind_status(struct rpc_task *task)
 	case -ENETDOWN:
 	case -EHOSTUNREACH:
 	case -ENETUNREACH:
-	case -ENOBUFS:
 	case -EPIPE:
 		dprintk("RPC: %5u remote rpcbind unreachable: %d\n",
 				task->tk_pid, task->tk_status);
@@ -2043,7 +2047,9 @@ call_bind_status(struct rpc_task *task)
 
 	rpc_call_rpcerror(task, status);
 	return;
-
+out_next:
+	task->tk_action = call_connect;
+	return;
 retry_timeout:
 	task->tk_status = 0;
 	task->tk_action = call_bind;
@@ -2090,6 +2096,7 @@ call_connect(struct rpc_task *task)
 static void
 call_connect_status(struct rpc_task *task)
 {
+	struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt;
 	struct rpc_clnt *clnt = task->tk_client;
 	int status = task->tk_status;
 
@@ -2099,8 +2106,17 @@ call_connect_status(struct rpc_task *task)
 	}
 
 	dprint_status(task);
-
 	trace_rpc_connect_status(task);
+
+	if (task->tk_status == 0) {
+		clnt->cl_stats->netreconn++;
+		goto out_next;
+	}
+	if (xprt_connected(xprt)) {
+		task->tk_status = 0;
+		goto out_next;
+	}
+
 	task->tk_status = 0;
 	switch (status) {
 	case -ECONNREFUSED:
@@ -2117,8 +2133,6 @@ call_connect_status(struct rpc_task *task)
 	case -ENETDOWN:
 	case -ENETUNREACH:
 	case -EHOSTUNREACH:
-	case -EADDRINUSE:
-	case -ENOBUFS:
 	case -EPIPE:
 		xprt_conditional_disconnect(task->tk_rqstp->rq_xprt,
 					    task->tk_rqstp->rq_connect_cookie);
@@ -2127,17 +2141,20 @@ call_connect_status(struct rpc_task *task)
 		/* retry with existing socket, after a delay */
 		rpc_delay(task, 3*HZ);
 		/* fall through */
+	case -EADDRINUSE:
 	case -ENOTCONN:
 	case -EAGAIN:
 	case -ETIMEDOUT:
 		goto out_retry;
-	case 0:
-		clnt->cl_stats->netreconn++;
-		task->tk_action = call_transmit;
-		return;
+	case -ENOBUFS:
+		rpc_delay(task, HZ >> 2);
+		goto out_retry;
 	}
 	rpc_call_rpcerror(task, status);
 	return;
+out_next:
+	task->tk_action = call_transmit;
+	return;
 out_retry:
 	/* Check for timeouts before looping back to call_bind */
 	task->tk_action = call_bind;
@@ -2365,7 +2382,7 @@ call_status(struct rpc_task *task)
 	case -ECONNABORTED:
 	case -ENOTCONN:
 		rpc_force_rebind(clnt);
-		/* fall through */
+		break;
 	case -EADDRINUSE:
 		rpc_delay(task, 3*HZ);
 		/* fall through */

net/sunrpc/xprt.c

@@ -1408,13 +1408,6 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task)
 			status = -EBADMSG;
 			goto out_dequeue;
 		}
-		if (task->tk_ops->rpc_call_prepare_transmit) {
-			task->tk_ops->rpc_call_prepare_transmit(task,
-					task->tk_calldata);
-			status = task->tk_status;
-			if (status < 0)
-				goto out_dequeue;
-		}
 		if (RPC_SIGNALLED(task)) {
 			status = -ERESTARTSYS;
 			goto out_dequeue;