Merge tag 'nfs-for-6.13-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client updates from Trond Myklebust:
 "Bugfixes:
   - nfs/localio: fix for a memory corruption in nfs_local_read_done
   - Revert "nfs: don't reuse partially completed requests in
     nfs_lock_and_join_requests"
   - nfsv4:
       - ignore SB_RDONLY when mounting nfs
       - Fix a use-after-free problem in open()
   - sunrpc:
       - clear XPRT_SOCK_UPD_TIMEOUT when reseting the transport
       - timeout and cancel TLS handshake with -ETIMEDOUT
       - fix one UAF issue caused by sunrpc kernel tcp socket
       - Fix a hang in TLS sock_close if sk_write_pending
   - pNFS/blocklayout: Fix device registration issues

  Features and cleanups:
   - localio cleanups from Mike Snitzer
   - Clean up refcounting on the nfs version modules
   - __counted_by() annotations
   - nfs: make processes that are waiting for an I/O lock killable"

* tag 'nfs-for-6.13-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (24 commits)
  fs/nfs/io: make nfs_start_io_*() killable
  nfs/blocklayout: Limit repeat device registration on failure
  nfs/blocklayout: Don't attempt unregister for invalid block device
  sunrpc: fix one UAF issue caused by sunrpc kernel tcp socket
  SUNRPC: timeout and cancel TLS handshake with -ETIMEDOUT
  sunrpc: clear XPRT_SOCK_UPD_TIMEOUT when reset transport
  nfs: ignore SB_RDONLY when mounting nfs
  Revert "nfs: don't reuse partially completed requests in nfs_lock_and_join_requests"
  Revert "fs: nfs: fix missing refcnt by replacing folio_set_private by folio_attach_private"
  nfs/localio: must clear res.replen in nfs_local_read_done
  NFSv4.0: Fix a use-after-free problem in the asynchronous open()
  NFSv4.0: Fix the wake up of the next waiter in nfs_release_seqid()
  SUNRPC: Fix a hang in TLS sock_close if sk_write_pending
  sunrpc: remove newlines from tracepoints
  nfs: Annotate struct pnfs_commit_array with __counted_by()
  nfs/localio: eliminate need for nfs_local_fsync_work forward declaration
  nfs/localio: remove extra indirect nfs_to call to check {read,write}_iter
  nfs/localio: eliminate unnecessary kref in nfs_local_fsync_ctx
  nfs/localio: remove redundant suid/sgid handling
  NFS: Implement get_nfs_version()
  ...

14 files changed, 209 insertions, 145 deletions

diff --git a/fs/nfs/blocklayout/blocklayout.c b/fs/nfs/blocklayout/blocklayout.c
index 0becdec12970..47189476b553 100644
--- a/fs/nfs/blocklayout/blocklayout.c
+++ b/fs/nfs/blocklayout/blocklayout.c
@@ -571,19 +571,32 @@ retry:
 	if (!node)
 		return ERR_PTR(-ENODEV);
 
+	/*
+	 * Devices that are marked unavailable are left in the cache with a
+	 * timeout to avoid sending GETDEVINFO after every LAYOUTGET, or
+	 * constantly attempting to register the device.  Once marked as
+	 * unavailable they must be deleted and never reused.
+	 */
 	if (test_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags)) {
 		unsigned long end = jiffies;
 		unsigned long start = end - PNFS_DEVICE_RETRY_TIMEOUT;
 
 		if (!time_in_range(node->timestamp_unavailable, start, end)) {
+			/* Uncork subsequent GETDEVINFO operations for this device */
 			nfs4_delete_deviceid(node->ld, node->nfs_client, id);
 			goto retry;
 		}
 		goto out_put;
 	}
 
-	if (!bl_register_dev(container_of(node, struct pnfs_block_dev, node)))
+	if (!bl_register_dev(container_of(node, struct pnfs_block_dev, node))) {
+		/*
+		 * If we cannot register, treat this device as transient:
+		 * Make a negative cache entry for the device
+		 */
+		nfs4_mark_deviceid_unavailable(node);
 		goto out_put;
+	}
 
 	return node;
 
diff --git a/fs/nfs/blocklayout/dev.c b/fs/nfs/blocklayout/dev.c
index 6252f4447945..cab8809f0e0f 100644
--- a/fs/nfs/blocklayout/dev.c
+++ b/fs/nfs/blocklayout/dev.c
@@ -20,9 +20,6 @@ static void bl_unregister_scsi(struct pnfs_block_dev *dev)
 	const struct pr_ops *ops = bdev->bd_disk->fops->pr_ops;
 	int status;
 
-	if (!test_and_clear_bit(PNFS_BDEV_REGISTERED, &dev->flags))
-		return;
-
 	status = ops->pr_register(bdev, dev->pr_key, 0, false);
 	if (status)
 		trace_bl_pr_key_unreg_err(bdev, dev->pr_key, status);
@@ -58,7 +55,8 @@ static void bl_unregister_dev(struct pnfs_block_dev *dev)
 		return;
 	}
 
-	if (dev->type == PNFS_BLOCK_VOLUME_SCSI)
+	if (dev->type == PNFS_BLOCK_VOLUME_SCSI &&
+		test_and_clear_bit(PNFS_BDEV_REGISTERED, &dev->flags))
 		bl_unregister_scsi(dev);
 }
 
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 03ecc7765615..550ca934c9cf 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -55,9 +55,13 @@
 #define NFSDBG_FACILITY		NFSDBG_CLIENT
 
 static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
-static DEFINE_SPINLOCK(nfs_version_lock);
-static DEFINE_MUTEX(nfs_version_mutex);
-static LIST_HEAD(nfs_versions);
+static DEFINE_RWLOCK(nfs_version_lock);
+
+static struct nfs_subversion *nfs_version_mods[5] = {
+	[2] = NULL,
+	[3] = NULL,
+	[4] = NULL,
+};
 
 /*
  * RPC cruft for NFS
@@ -76,38 +80,38 @@ const struct rpc_program nfs_program = {
 	.pipe_dir_name		= NFS_PIPE_DIRNAME,
 };
 
-static struct nfs_subversion *find_nfs_version(unsigned int version)
+static struct nfs_subversion *__find_nfs_version(unsigned int version)
 {
 	struct nfs_subversion *nfs;
-	spin_lock(&nfs_version_lock);
-
-	list_for_each_entry(nfs, &nfs_versions, list) {
-		if (nfs->rpc_ops->version == version) {
-			spin_unlock(&nfs_version_lock);
-			return nfs;
-		}
-	}
 
-	spin_unlock(&nfs_version_lock);
-	return ERR_PTR(-EPROTONOSUPPORT);
+	read_lock(&nfs_version_lock);
+	nfs = nfs_version_mods[version];
+	read_unlock(&nfs_version_lock);
+	return nfs;
 }
 
-struct nfs_subversion *get_nfs_version(unsigned int version)
+struct nfs_subversion *find_nfs_version(unsigned int version)
 {
-	struct nfs_subversion *nfs = find_nfs_version(version);
+	struct nfs_subversion *nfs = __find_nfs_version(version);
 
-	if (IS_ERR(nfs)) {
-		mutex_lock(&nfs_version_mutex);
-		request_module("nfsv%d", version);
-		nfs = find_nfs_version(version);
-		mutex_unlock(&nfs_version_mutex);
-	}
+	if (!nfs && request_module("nfsv%d", version) == 0)
+		nfs = __find_nfs_version(version);
 
-	if (!IS_ERR(nfs) && !try_module_get(nfs->owner))
+	if (!nfs)
+		return ERR_PTR(-EPROTONOSUPPORT);
+
+	if (!get_nfs_version(nfs))
 		return ERR_PTR(-EAGAIN);
+
 	return nfs;
 }
 
+int get_nfs_version(struct nfs_subversion *nfs)
+{
+	return try_module_get(nfs->owner);
+}
+EXPORT_SYMBOL_GPL(get_nfs_version);
+
 void put_nfs_version(struct nfs_subversion *nfs)
 {
 	module_put(nfs->owner);
@@ -115,23 +119,23 @@ void put_nfs_version(struct nfs_subversion *nfs)
 
 void register_nfs_version(struct nfs_subversion *nfs)
 {
-	spin_lock(&nfs_version_lock);
+	write_lock(&nfs_version_lock);
 
-	list_add(&nfs->list, &nfs_versions);
+	nfs_version_mods[nfs->rpc_ops->version] = nfs;
 	nfs_version[nfs->rpc_ops->version] = nfs->rpc_vers;
 
-	spin_unlock(&nfs_version_lock);
+	write_unlock(&nfs_version_lock);
 }
 EXPORT_SYMBOL_GPL(register_nfs_version);
 
 void unregister_nfs_version(struct nfs_subversion *nfs)
 {
-	spin_lock(&nfs_version_lock);
+	write_lock(&nfs_version_lock);
 
 	nfs_version[nfs->rpc_ops->version] = NULL;
-	list_del(&nfs->list);
+	nfs_version_mods[nfs->rpc_ops->version] = NULL;
 
-	spin_unlock(&nfs_version_lock);
+	write_unlock(&nfs_version_lock);
 }
 EXPORT_SYMBOL_GPL(unregister_nfs_version);
 
@@ -151,7 +155,7 @@ struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)
 
 	clp->cl_minorversion = cl_init->minorversion;
 	clp->cl_nfs_mod = cl_init->nfs_mod;
-	if (!try_module_get(clp->cl_nfs_mod->owner))
+	if (!get_nfs_version(clp->cl_nfs_mod))
 		goto error_dealloc;
 
 	clp->rpc_ops = clp->cl_nfs_mod->rpc_ops;
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 90079ca134dd..b08dbe96bc57 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -454,8 +454,16 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter,
 	if (user_backed_iter(iter))
 		dreq->flags = NFS_ODIRECT_SHOULD_DIRTY;
 
-	if (!swap)
-		nfs_start_io_direct(inode);
+	if (!swap) {
+		result = nfs_start_io_direct(inode);
+		if (result) {
+			/* release the reference that would usually be
+			 * consumed by nfs_direct_read_schedule_iovec()
+			 */
+			nfs_direct_req_release(dreq);
+			goto out_release;
+		}
+	}
 
 	NFS_I(inode)->read_io += count;
 	requested = nfs_direct_read_schedule_iovec(dreq, iter, iocb->ki_pos);
@@ -1007,7 +1015,14 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, struct iov_iter *iter,
 		requested = nfs_direct_write_schedule_iovec(dreq, iter, pos,
 							    FLUSH_STABLE);
 	} else {
-		nfs_start_io_direct(inode);
+		result = nfs_start_io_direct(inode);
+		if (result) {
+			/* release the reference that would usually be
+			 * consumed by nfs_direct_write_schedule_iovec()
+			 */
+			nfs_direct_req_release(dreq);
+			goto out_release;
+		}
 
 		requested = nfs_direct_write_schedule_iovec(dreq, iter, pos,
 							    FLUSH_COND_STABLE);
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 6800ee92d742..1bb646752e46 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -166,7 +166,10 @@ nfs_file_read(struct kiocb *iocb, struct iov_iter *to)
 		iocb->ki_filp,
 		iov_iter_count(to), (unsigned long) iocb->ki_pos);
 
-	nfs_start_io_read(inode);
+	result = nfs_start_io_read(inode);
+	if (result)
+		return result;
+
 	result = nfs_revalidate_mapping(inode, iocb->ki_filp->f_mapping);
 	if (!result) {
 		result = generic_file_read_iter(iocb, to);
@@ -187,7 +190,10 @@ nfs_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe
 
 	dprintk("NFS: splice_read(%pD2, %zu@%llu)\n", in, len, *ppos);
 
-	nfs_start_io_read(inode);
+	result = nfs_start_io_read(inode);
+	if (result)
+		return result;
+
 	result = nfs_revalidate_mapping(inode, in->f_mapping);
 	if (!result) {
 		result = filemap_splice_read(in, ppos, pipe, len, flags);
@@ -668,7 +674,9 @@ ssize_t nfs_file_write(struct kiocb *iocb, struct iov_iter *from)
 	nfs_clear_invalid_mapping(file->f_mapping);
 
 	since = filemap_sample_wb_err(file->f_mapping);
-	nfs_start_io_write(inode);
+	error = nfs_start_io_write(inode);
+	if (error)
+		return error;
 	result = generic_write_checks(iocb, from);
 	if (result > 0)
 		result = generic_perform_write(iocb, from);
diff --git a/fs/nfs/fs_context.c b/fs/nfs/fs_context.c
index 7e000d782e28..b069385eea17 100644
--- a/fs/nfs/fs_context.c
+++ b/fs/nfs/fs_context.c
@@ -1467,7 +1467,7 @@ static int nfs_fs_context_validate(struct fs_context *fc)
 
 	/* Load the NFS protocol module if we haven't done so yet */
 	if (!ctx->nfs_mod) {
-		nfs_mod = get_nfs_version(ctx->version);
+		nfs_mod = find_nfs_version(ctx->version);
 		if (IS_ERR(nfs_mod)) {
 			ret = PTR_ERR(nfs_mod);
 			goto out_version_unavailable;
@@ -1541,7 +1541,7 @@ static int nfs_fs_context_dup(struct fs_context *fc, struct fs_context *src_fc)
 	}
 	nfs_copy_fh(ctx->mntfh, src->mntfh);
 
-	__module_get(ctx->nfs_mod->owner);
+	get_nfs_version(ctx->nfs_mod);
 	ctx->client_address		= NULL;
 	ctx->mount_server.hostname	= NULL;
 	ctx->nfs_server.export_path	= NULL;
@@ -1633,7 +1633,7 @@ static int nfs_init_fs_context(struct fs_context *fc)
 		}
 
 		ctx->nfs_mod = nfss->nfs_client->cl_nfs_mod;
-		__module_get(ctx->nfs_mod->owner);
+		get_nfs_version(ctx->nfs_mod);
 	} else {
 		/* defaults */
 		ctx->timeo		= NFS_UNSPEC_TIMEO;
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 430733e3eff2..e564bd11ba60 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -6,13 +6,14 @@
 #include "nfs4_fs.h"
 #include <linux/fs_context.h>
 #include <linux/security.h>
+#include <linux/compiler_attributes.h>
 #include <linux/crc32.h>
 #include <linux/sunrpc/addr.h>
 #include <linux/nfs_page.h>
 #include <linux/nfslocalio.h>
 #include <linux/wait_bit.h>
 
-#define NFS_SB_MASK (SB_RDONLY|SB_NOSUID|SB_NODEV|SB_NOEXEC|SB_SYNCHRONOUS)
+#define NFS_SB_MASK (SB_NOSUID|SB_NODEV|SB_NOEXEC|SB_SYNCHRONOUS)
 
 extern const struct export_operations nfs_export_ops;
 
@@ -516,11 +517,11 @@ extern const struct netfs_request_ops nfs_netfs_ops;
 #endif
 
 /* io.c */
-extern void nfs_start_io_read(struct inode *inode);
+extern __must_check int nfs_start_io_read(struct inode *inode);
 extern void nfs_end_io_read(struct inode *inode);
-extern void nfs_start_io_write(struct inode *inode);
+extern  __must_check int nfs_start_io_write(struct inode *inode);
 extern void nfs_end_io_write(struct inode *inode);
-extern void nfs_start_io_direct(struct inode *inode);
+extern __must_check int nfs_start_io_direct(struct inode *inode);
 extern void nfs_end_io_direct(struct inode *inode);
 
 static inline bool nfs_file_io_is_buffered(struct nfs_inode *nfsi)
diff --git a/fs/nfs/io.c b/fs/nfs/io.c
index b5551ed8f648..3388faf2acb9 100644
--- a/fs/nfs/io.c
+++ b/fs/nfs/io.c
@@ -39,19 +39,28 @@ static void nfs_block_o_direct(struct nfs_inode *nfsi, struct inode *inode)
  * Note that buffered writes and truncates both take a write lock on
  * inode->i_rwsem, meaning that those are serialised w.r.t. the reads.
  */
-void
+int
 nfs_start_io_read(struct inode *inode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
+	int err;
+
 	/* Be an optimist! */
-	down_read(&inode->i_rwsem);
+	err = down_read_killable(&inode->i_rwsem);
+	if (err)
+		return err;
 	if (test_bit(NFS_INO_ODIRECT, &nfsi->flags) == 0)
-		return;
+		return 0;
 	up_read(&inode->i_rwsem);
+
 	/* Slow path.... */
-	down_write(&inode->i_rwsem);
+	err = down_write_killable(&inode->i_rwsem);
+	if (err)
+		return err;
 	nfs_block_o_direct(nfsi, inode);
 	downgrade_write(&inode->i_rwsem);
+
+	return 0;
 }
 
 /**
@@ -74,11 +83,15 @@ nfs_end_io_read(struct inode *inode)
  * Declare that a buffered read operation is about to start, and ensure
  * that we block all direct I/O.
  */
-void
+int
 nfs_start_io_write(struct inode *inode)
 {
-	down_write(&inode->i_rwsem);
-	nfs_block_o_direct(NFS_I(inode), inode);
+	int err;
+
+	err = down_write_killable(&inode->i_rwsem);
+	if (!err)
+		nfs_block_o_direct(NFS_I(inode), inode);
+	return err;
 }
 
 /**
@@ -119,19 +132,28 @@ static void nfs_block_buffered(struct nfs_inode *nfsi, struct inode *inode)
  * Note that buffered writes and truncates both take a write lock on
  * inode->i_rwsem, meaning that those are serialised w.r.t. O_DIRECT.
  */
-void
+int
 nfs_start_io_direct(struct inode *inode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
+	int err;
+
 	/* Be an optimist! */
-	down_read(&inode->i_rwsem);
+	err = down_read_killable(&inode->i_rwsem);
+	if (err)
+		return err;
 	if (test_bit(NFS_INO_ODIRECT, &nfsi->flags) != 0)
-		return;
+		return 0;
 	up_read(&inode->i_rwsem);
+
 	/* Slow path.... */
-	down_write(&inode->i_rwsem);
+	err = down_write_killable(&inode->i_rwsem);
+	if (err)
+		return err;
 	nfs_block_buffered(nfsi, inode);
 	downgrade_write(&inode->i_rwsem);
+
+	return 0;
 }
 
 /**
diff --git a/fs/nfs/localio.c b/fs/nfs/localio.c
index 8f0ce82a677e..4b8618cf114c 100644
--- a/fs/nfs/localio.c
+++ b/fs/nfs/localio.c
@@ -42,10 +42,8 @@ struct nfs_local_fsync_ctx {
 	struct nfsd_file	*localio;
 	struct nfs_commit_data	*data;
 	struct work_struct	work;
-	struct kref		kref;
 	struct completion	*done;
 };
-static void nfs_local_fsync_work(struct work_struct *work);
 
 static bool localio_enabled __read_mostly = true;
 module_param(localio_enabled, bool, 0644);
@@ -274,7 +272,7 @@ nfs_local_iocb_free(struct nfs_local_kiocb *iocb)
 
 static struct nfs_local_kiocb *
 nfs_local_iocb_alloc(struct nfs_pgio_header *hdr,
-		     struct nfsd_file *localio, gfp_t flags)
+		     struct file *file, gfp_t flags)
 {
 	struct nfs_local_kiocb *iocb;
 
@@ -287,9 +285,8 @@ nfs_local_iocb_alloc(struct nfs_pgio_header *hdr,
 		kfree(iocb);
 		return NULL;
 	}
-	init_sync_kiocb(&iocb->kiocb, nfs_to->nfsd_file_file(localio));
+	init_sync_kiocb(&iocb->kiocb, file);
 	iocb->kiocb.ki_pos = hdr->args.offset;
-	iocb->localio = localio;
 	iocb->hdr = hdr;
 	iocb->kiocb.ki_flags &= ~IOCB_APPEND;
 	return iocb;
@@ -354,6 +351,12 @@ nfs_local_read_done(struct nfs_local_kiocb *iocb, long status)
 
 	nfs_local_pgio_done(hdr, status);
 
+	/*
+	 * Must clear replen otherwise NFSv3 data corruption will occur
+	 * if/when switching from LOCALIO back to using normal RPC.
+	 */
+	hdr->res.replen = 0;
+
 	if (hdr->res.count != hdr->args.count ||
 	    hdr->args.offset + hdr->res.count >= i_size_read(file_inode(filp)))
 		hdr->res.eof = true;
@@ -390,13 +393,19 @@ nfs_do_local_read(struct nfs_pgio_header *hdr,
 		  const struct rpc_call_ops *call_ops)
 {
 	struct nfs_local_kiocb *iocb;
+	struct file *file = nfs_to->nfsd_file_file(localio);
+
+	/* Don't support filesystems without read_iter */
+	if (!file->f_op->read_iter)
+		return -EAGAIN;
 
 	dprintk("%s: vfs_read count=%u pos=%llu\n",
 		__func__, hdr->args.count, hdr->args.offset);
 
-	iocb = nfs_local_iocb_alloc(hdr, localio, GFP_KERNEL);
+	iocb = nfs_local_iocb_alloc(hdr, file, GFP_KERNEL);
 	if (iocb == NULL)
 		return -ENOMEM;
+	iocb->localio = localio;
 
 	nfs_local_pgio_init(hdr, call_ops);
 	hdr->res.eof = false;
@@ -521,12 +530,7 @@ nfs_local_write_done(struct nfs_local_kiocb *iocb, long status)
 	}
 	if (status < 0)
 		nfs_reset_boot_verifier(inode);
-	else if (nfs_should_remove_suid(inode)) {
-		/* Deal with the suid/sgid bit corner case */
-		spin_lock(&inode->i_lock);
-		nfs_set_cache_invalid(inode, NFS_INO_INVALID_MODE);
-		spin_unlock(&inode->i_lock);
-	}
+
 	nfs_local_pgio_done(hdr, status);
 }
 
@@ -564,14 +568,20 @@ nfs_do_local_write(struct nfs_pgio_header *hdr,
 		   const struct rpc_call_ops *call_ops)
 {
 	struct nfs_local_kiocb *iocb;
+	struct file *file = nfs_to->nfsd_file_file(localio);
+
+	/* Don't support filesystems without write_iter */
+	if (!file->f_op->write_iter)
+		return -EAGAIN;
 
 	dprintk("%s: vfs_write count=%u pos=%llu %s\n",
 		__func__, hdr->args.count, hdr->args.offset,
 		(hdr->args.stable == NFS_UNSTABLE) ?  "unstable" : "stable");
 
-	iocb = nfs_local_iocb_alloc(hdr, localio, GFP_NOIO);
+	iocb = nfs_local_iocb_alloc(hdr, file, GFP_NOIO);
 	if (iocb == NULL)
 		return -ENOMEM;
+	iocb->localio = localio;
 
 	switch (hdr->args.stable) {
 	default:
@@ -597,16 +607,9 @@ int nfs_local_doio(struct nfs_client *clp, struct nfsd_file *localio,
 		   const struct rpc_call_ops *call_ops)
 {
 	int status = 0;
-	struct file *filp = nfs_to->nfsd_file_file(localio);
 
 	if (!hdr->args.count)
 		return 0;
-	/* Don't support filesystems without read_iter/write_iter */
-	if (!filp->f_op->read_iter || !filp->f_op->write_iter) {
-		nfs_local_disable(clp);
-		status = -EAGAIN;
-		goto out;
-	}
 
 	switch (hdr->rw_mode) {
 	case FMODE_READ:
@@ -620,8 +623,10 @@ int nfs_local_doio(struct nfs_client *clp, struct nfsd_file *localio,
 			hdr->rw_mode);
 		status = -EINVAL;
 	}
-out:
+
 	if (status != 0) {
+		if (status == -EAGAIN)
+			nfs_local_disable(clp);
 		nfs_to_nfsd_file_put_local(localio);
 		hdr->task.tk_status = status;
 		nfs_local_hdr_release(hdr, call_ops);
@@ -678,40 +683,12 @@ nfs_local_release_commit_data(struct nfsd_file *localio,
 	call_ops->rpc_release(data);
 }
 
-static struct nfs_local_fsync_ctx *
-nfs_local_fsync_ctx_alloc(struct nfs_commit_data *data,
-			  struct nfsd_file *localio, gfp_t flags)
-{
-	struct nfs_local_fsync_ctx *ctx = kmalloc(sizeof(*ctx), flags);
-
-	if (ctx != NULL) {
-		ctx->localio = localio;
-		ctx->data = data;
-		INIT_WORK(&ctx->work, nfs_local_fsync_work);
-		kref_init(&ctx->kref);
-		ctx->done = NULL;
-	}
-	return ctx;
-}
-
-static void
-nfs_local_fsync_ctx_kref_free(struct kref *kref)
-{
-	kfree(container_of(kref, struct nfs_local_fsync_ctx, kref));
-}
-
-static void
-nfs_local_fsync_ctx_put(struct nfs_local_fsync_ctx *ctx)
-{
-	kref_put(&ctx->kref, nfs_local_fsync_ctx_kref_free);
-}
-
 static void
 nfs_local_fsync_ctx_free(struct nfs_local_fsync_ctx *ctx)
 {
 	nfs_local_release_commit_data(ctx->localio, ctx->data,
 				      ctx->data->task.tk_ops);
-	nfs_local_fsync_ctx_put(ctx);
+	kfree(ctx);
 }
 
 static void
@@ -730,6 +707,21 @@ nfs_local_fsync_work(struct work_struct *work)
 	nfs_local_fsync_ctx_free(ctx);
 }
 
+static struct nfs_local_fsync_ctx *
+nfs_local_fsync_ctx_alloc(struct nfs_commit_data *data,
+			  struct nfsd_file *localio, gfp_t flags)
+{
+	struct nfs_local_fsync_ctx *ctx = kmalloc(sizeof(*ctx), flags);
+
+	if (ctx != NULL) {
+		ctx->localio = localio;
+		ctx->data = data;
+		INIT_WORK(&ctx->work, nfs_local_fsync_work);
+		ctx->done = NULL;
+	}
+	return ctx;
+}
+
 int nfs_local_commit(struct nfsd_file *localio,
 		     struct nfs_commit_data *data,
 		     const struct rpc_call_ops *call_ops, int how)
@@ -744,7 +736,7 @@ int nfs_local_commit(struct nfsd_file *localio,
 	}
 
 	nfs_local_init_commit(data, call_ops);
-	kref_get(&ctx->kref);
+
 	if (how & FLUSH_SYNC) {
 		DECLARE_COMPLETION_ONSTACK(done);
 		ctx->done = &done;
@@ -752,6 +744,6 @@ int nfs_local_commit(struct nfsd_file *localio,
 		wait_for_completion(&done);
 	} else
 		queue_work(nfsiod_workqueue, &ctx->work);
-	nfs_local_fsync_ctx_put(ctx);
+
 	return 0;
 }
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index e7494cdd957e..2d53574da605 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -182,7 +182,7 @@ struct vfsmount *nfs_d_automount(struct path *path)
 	ctx->version		= client->rpc_ops->version;
 	ctx->minorversion	= client->cl_minorversion;
 	ctx->nfs_mod		= client->cl_nfs_mod;
-	__module_get(ctx->nfs_mod->owner);
+	get_nfs_version(ctx->nfs_mod);
 
 	ret = client->rpc_ops->submount(fc, server);
 	if (ret < 0) {
diff --git a/fs/nfs/nfs.h b/fs/nfs/nfs.h
index 0d3ce0460e35..8a5f51be013a 100644
--- a/fs/nfs/nfs.h
+++ b/fs/nfs/nfs.h
@@ -19,10 +19,10 @@ struct nfs_subversion {
 	const struct nfs_rpc_ops *rpc_ops;	/* NFS operations */
 	const struct super_operations *sops;	/* NFS Super operations */
 	const struct xattr_handler * const *xattr;	/* NFS xattr handlers */
-	struct list_head list;		/* List of NFS versions */
 };
 
-struct nfs_subversion *get_nfs_version(unsigned int);
+struct nfs_subversion *find_nfs_version(unsigned int);
+int get_nfs_version(struct nfs_subversion *);
 void put_nfs_version(struct nfs_subversion *);
 void register_nfs_version(struct nfs_subversion *);
 void unregister_nfs_version(struct nfs_subversion *);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 9d40319e063d..405f17e6e0b4 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2603,12 +2603,14 @@ static void nfs4_open_release(void *calldata)
 	struct nfs4_opendata *data = calldata;
 	struct nfs4_state *state = NULL;
 
+	/* In case of error, no cleanup! */
+	if (data->rpc_status != 0 || !data->rpc_done) {
+		nfs_release_seqid(data->o_arg.seqid);
+		goto out_free;
+	}
 	/* If this request hasn't been cancelled, do nothing */
 	if (!data->cancelled)
 		goto out_free;
-	/* In case of error, no cleanup! */
-	if (data->rpc_status != 0 || !data->rpc_done)
-		goto out_free;
 	/* In case we need an open_confirm, no cleanup! */
 	if (data->o_res.rflags & NFS4_OPEN_RESULT_CONFIRM)
 		goto out_free;
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index dafd61186557..9a9f60a2291b 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1083,14 +1083,12 @@ void nfs_release_seqid(struct nfs_seqid *seqid)
 		return;
 	sequence = seqid->sequence;
 	spin_lock(&sequence->lock);
-	list_del_init(&seqid->list);
-	if (!list_empty(&sequence->list)) {
-		struct nfs_seqid *next;
-
-		next = list_first_entry(&sequence->list,
-				struct nfs_seqid, list);
+	if (list_is_first(&seqid->list, &sequence->list) &&
+	    !list_is_singular(&sequence->list)) {
+		struct nfs_seqid *next = list_next_entry(seqid, list);
 		rpc_wake_up_queued_task(&sequence->wait, next->task);
 	}
+	list_del_init(&seqid->list);
 	spin_unlock(&sequence->lock);
 }
 
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index ead2dc55952d..50fa539611f5 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -144,6 +144,31 @@ static void nfs_io_completion_put(struct nfs_io_completion *ioc)
 		kref_put(&ioc->refcount, nfs_io_completion_release);
 }
 
+static void
+nfs_page_set_inode_ref(struct nfs_page *req, struct inode *inode)
+{
+	if (!test_and_set_bit(PG_INODE_REF, &req->wb_flags)) {
+		kref_get(&req->wb_kref);
+		atomic_long_inc(&NFS_I(inode)->nrequests);
+	}
+}
+
+static int
+nfs_cancel_remove_inode(struct nfs_page *req, struct inode *inode)
+{
+	int ret;
+
+	if (!test_bit(PG_REMOVE, &req->wb_flags))
+		return 0;
+	ret = nfs_page_group_lock(req);
+	if (ret)
+		return ret;
+	if (test_and_clear_bit(PG_REMOVE, &req->wb_flags))
+		nfs_page_set_inode_ref(req, inode);
+	nfs_page_group_unlock(req);
+	return 0;
+}
+
 /**
  * nfs_folio_find_head_request - find head request associated with a folio
  * @folio: pointer to folio
@@ -540,7 +565,6 @@ static struct nfs_page *nfs_lock_and_join_requests(struct folio *folio)
 	struct inode *inode = folio->mapping->host;
 	struct nfs_page *head, *subreq;
 	struct nfs_commit_info cinfo;
-	bool removed;
 	int ret;
 
 	/*
@@ -565,18 +589,18 @@ retry:
 		goto retry;
 	}
 
-	ret = nfs_page_group_lock(head);
+	ret = nfs_cancel_remove_inode(head, inode);
 	if (ret < 0)
 		goto out_unlock;
 
-	removed = test_bit(PG_REMOVE, &head->wb_flags);
+	ret = nfs_page_group_lock(head);
+	if (ret < 0)
+		goto out_unlock;
 
 	/* lock each request in the page group */
 	for (subreq = head->wb_this_page;
 	     subreq != head;
 	     subreq = subreq->wb_this_page) {
-		if (test_bit(PG_REMOVE, &subreq->wb_flags))
-			removed = true;
 		ret = nfs_page_group_lock_subreq(head, subreq);
 		if (ret < 0)
 			goto out_unlock;
@@ -584,21 +608,6 @@ retry:
 
 	nfs_page_group_unlock(head);
 
-	/*
-	 * If PG_REMOVE is set on any request, I/O on that request has
-	 * completed, but some requests were still under I/O at the time
-	 * we locked the head request.
-	 *
-	 * In that case the above wait for all requests means that all I/O
-	 * has now finished, and we can restart from a clean slate.  Let the
-	 * old requests go away and start from scratch instead.
-	 */
-	if (removed) {
-		nfs_unroll_locks(head, head);
-		nfs_unlock_and_release_request(head);
-		goto retry;
-	}
-
 	nfs_init_cinfo_from_inode(&cinfo, inode);
 	nfs_join_page_group(head, &cinfo, inode);
 	return head;
@@ -772,7 +781,8 @@ static void nfs_inode_add_request(struct nfs_page *req)
 	nfs_lock_request(req);
 	spin_lock(&mapping->i_private_lock);
 	set_bit(PG_MAPPED, &req->wb_flags);
-	folio_attach_private(folio, req);
+	folio_set_private(folio);
+	folio->private = req;
 	spin_unlock(&mapping->i_private_lock);
 	atomic_long_inc(&nfsi->nrequests);
 	/* this a head request for a page group - mark it as having an
@@ -796,7 +806,8 @@ static void nfs_inode_remove_request(struct nfs_page *req)
 
 		spin_lock(&mapping->i_private_lock);
 		if (likely(folio)) {
-			folio_detach_private(folio);
+			folio->private = NULL;
+			folio_clear_private(folio);
 			clear_bit(PG_MAPPED, &req->wb_head->wb_flags);
 		}
 		spin_unlock(&mapping->i_private_lock);