NFS Client Updates for Linux 6.11

New Features:
   * Add support for large folios
   * Implement rpcrdma generic device removal notification
   * Add client support for attribute delegations
   * Use a LAYOUTRETURN during reboot recovery to report layoutstats and errors
   * Improve throughput for random buffered writes
   * Add NVMe support to pnfs/blocklayout
 
 Bugfixes:
   * Fix rpcrdma_reqs_reset()
   * Avoid soft lockups when using UDP
   * Fix an nfs/blocklayout premature PR key unregestration
   * Another fix for EXCHGID4_FLAG_USE_PNFS_DS for DS server
   * Do not extend writes to the entire folio
   * Pass explicit offset and count values to tracepoints
   * Fix a race to wake up sleeping SUNRPC sync tasks
   * Fix gss_status tracepoint output
 
 Cleanups:
   * Add missing MODULE_DESCRIPTION() macros
   * Add blocklayout / SCSI layout tracepoints
   * Remove asm-generic headers from xprtrdma verbs.c
   * Remove unused 'struct mnt_fhstatus'
   * Other delegation related cleanups
   * Other folio related cleanups
   * Other pNFS related cleanups
   * Other xprtrdma cleanups
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCAAdFiEEnZ5MQTpR7cLU7KEp18tUv7ClQOsFAmaZgr0ACgkQ18tUv7Cl
 QOv8FxAAnUyYG7Kdbv+5Ko/SFv0imxCb5DQh2XC/hSHNrlKBlDnqe2PANXR9XocL
 mS0Wry5tZf/T+o+QoKv0HQUdWFlnqKzwclggrekf/lkioU1feWsLe2RzDl1iUh0V
 6fwcCyWXW1mYX2CtCaDe+/ZFcoZOMD+bItNHt/RdDScSnS9Jd8GSyocsVKsqaBx6
 3wub0FJ4UBgYNoX2T3YyK2JwvO9GLaKIQRJV74rjgPJKjcjhptbcb5MKBmOZrF95
 UCcpl4CwvD9RTsSEp0B98UbAFFpk8Nw1tmHF3GmyG/nsrJomDuLKFvbsiq23eHUf
 XeULZIbjMEzU56vjoTglZA4s7JYx17D0vzdPGUqU4mLN3LPm5LtGLBg2uQoPw/xW
 50euLU+ol36mfnQlBsuM/tAXgtoAcT63aNeNRNp8aOL47xA+PC6kWTBK9OaR5+x6
 w+d22Dpy+riMk1TRaAVt0ANcENKELsWRFvxkuWCpQhVoQ1h8LigQJzeggEEK7Sa6
 5u9H6wCTee2wz746uwA43koj1utuyrLq/5S+qEtCY1pbP3U0A+Gh0Xh00OXiYuzL
 TgRdksmiAL8cA51WjSrq6HhGLOUJAYLfbdKaVhW+fULxUVwzWhFFaFbbdiq/e4OR
 0pfqls8UZWICE51GeTfalEidpKZgV/LxU3QOuVoalWBULyj/TeI=
 =avTW
 -----END PGP SIGNATURE-----

Merge tag 'nfs-for-6.11-1' of git://git.linux-nfs.org/projects/anna/linux-nfs

Pull NFS client updates from Anna Schumaker:
 "New Features:
   - Add support for large folios
   - Implement rpcrdma generic device removal notification
   - Add client support for attribute delegations
   - Use a LAYOUTRETURN during reboot recovery to report layoutstats
     and errors
   - Improve throughput for random buffered writes
   - Add NVMe support to pnfs/blocklayout

  Bugfixes:
   - Fix rpcrdma_reqs_reset()
   - Avoid soft lockups when using UDP
   - Fix an nfs/blocklayout premature PR key unregestration
   - Another fix for EXCHGID4_FLAG_USE_PNFS_DS for DS server
   - Do not extend writes to the entire folio
   - Pass explicit offset and count values to tracepoints
   - Fix a race to wake up sleeping SUNRPC sync tasks
   - Fix gss_status tracepoint output

  Cleanups:
   - Add missing MODULE_DESCRIPTION() macros
   - Add blocklayout / SCSI layout tracepoints
   - Remove asm-generic headers from xprtrdma verbs.c
   - Remove unused 'struct mnt_fhstatus'
   - Other delegation related cleanups
   - Other folio related cleanups
   - Other pNFS related cleanups
   - Other xprtrdma cleanups"

* tag 'nfs-for-6.11-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (63 commits)
  SUNRPC: Fixup gss_status tracepoint error output
  SUNRPC: Fix a race to wake a sync task
  nfs: split nfs_read_folio
  nfs: pass explicit offset/count to trace events
  nfs: do not extend writes to the entire folio
  nfs/blocklayout: add support for NVMe
  nfs: remove nfs_page_length
  nfs: remove the unused max_deviceinfo_size field from struct pnfs_layoutdriver_type
  nfs: don't reuse partially completed requests in nfs_lock_and_join_requests
  nfs: move nfs_wait_on_request to write.c
  nfs: fold nfs_page_group_lock_subrequests into nfs_lock_and_join_requests
  nfs: fold nfs_folio_find_and_lock_request into nfs_lock_and_join_requests
  nfs: simplify nfs_folio_find_and_lock_request
  nfs: remove nfs_folio_private_request
  nfs: remove dead code for the old swap over NFS implementation
  NFSv4.1 another fix for EXCHGID4_FLAG_USE_PNFS_DS for DS server
  nfs: Block on write congestion
  nfs: Properly initialize server->writeback
  nfs: Drop pointless check from nfs_commit_release_pages()
  nfs/blocklayout: SCSI layout trace points for reservation key reg/unreg
  ...

fs/nfs/blocklayout/blocklayout.c

@@ -564,25 +564,32 @@ bl_find_get_deviceid(struct nfs_server *server,
 		gfp_t gfp_mask)
 {
 	struct nfs4_deviceid_node *node;
-	unsigned long start, end;
+	int err = -ENODEV;
 
 retry:
 	node = nfs4_find_get_deviceid(server, id, cred, gfp_mask);
 	if (!node)
 		return ERR_PTR(-ENODEV);
 
-	if (test_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags) == 0)
+	if (test_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags)) {
-		return node;
+		unsigned long end = jiffies;
+		unsigned long start = end - PNFS_DEVICE_RETRY_TIMEOUT;
 
-	end = jiffies;
+		if (!time_in_range(node->timestamp_unavailable, start, end)) {
-	start = end - PNFS_DEVICE_RETRY_TIMEOUT;
+			nfs4_delete_deviceid(node->ld, node->nfs_client, id);
-	if (!time_in_range(node->timestamp_unavailable, start, end)) {
+			goto retry;
-		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)))
+		goto out_put;
+
+	return node;
+
+out_put:
 	nfs4_put_deviceid_node(node);
-	return ERR_PTR(-ENODEV);
+	return ERR_PTR(err);
 }
 
 static int

fs/nfs/blocklayout/blocklayout.h

@@ -104,20 +104,26 @@ struct pnfs_block_dev {
 	u64				start;
 	u64				len;
 
+	enum pnfs_block_volume_type	type;
 	u32				nr_children;
 	struct pnfs_block_dev		*children;
 	u64				chunk_size;
 
 	struct file			*bdev_file;
 	u64				disk_offset;
+	unsigned long			flags;
 
 	u64				pr_key;
-	bool				pr_registered;
 
 	bool (*map)(struct pnfs_block_dev *dev, u64 offset,
 			struct pnfs_block_dev_map *map);
 };
 
+/* pnfs_block_dev flag bits */
+enum {
+	PNFS_BDEV_REGISTERED = 0,
+};
+
 /* sector_t fields are all in 512-byte sectors */
 struct pnfs_block_extent {
 	union {
@@ -172,6 +178,7 @@ struct bl_msg_hdr {
 #define BL_DEVICE_REQUEST_ERR          0x2 /* User level process fails */
 
 /* dev.c */
+bool bl_register_dev(struct pnfs_block_dev *d);
 struct nfs4_deviceid_node *bl_alloc_deviceid_node(struct nfs_server *server,
 		struct pnfs_device *pdev, gfp_t gfp_mask);
 void bl_free_deviceid_node(struct nfs4_deviceid_node *d);

fs/nfs/blocklayout/dev.c

@@ -10,12 +10,83 @@
 #include <linux/pr.h>
 
 #include "blocklayout.h"
+#include "../nfs4trace.h"
 
 #define NFSDBG_FACILITY		NFSDBG_PNFS_LD
 
+static void bl_unregister_scsi(struct pnfs_block_dev *dev)
+{
+	struct block_device *bdev = file_bdev(dev->bdev_file);
+	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);
+	else
+		trace_bl_pr_key_unreg(bdev, dev->pr_key);
+}
+
+static bool bl_register_scsi(struct pnfs_block_dev *dev)
+{
+	struct block_device *bdev = file_bdev(dev->bdev_file);
+	const struct pr_ops *ops = bdev->bd_disk->fops->pr_ops;
+	int status;
+
+	if (test_and_set_bit(PNFS_BDEV_REGISTERED, &dev->flags))
+		return true;
+
+	status = ops->pr_register(bdev, 0, dev->pr_key, true);
+	if (status) {
+		trace_bl_pr_key_reg_err(bdev, dev->pr_key, status);
+		return false;
+	}
+	trace_bl_pr_key_reg(bdev, dev->pr_key);
+	return true;
+}
+
+static void bl_unregister_dev(struct pnfs_block_dev *dev)
+{
+	u32 i;
+
+	if (dev->nr_children) {
+		for (i = 0; i < dev->nr_children; i++)
+			bl_unregister_dev(&dev->children[i]);
+		return;
+	}
+
+	if (dev->type == PNFS_BLOCK_VOLUME_SCSI)
+		bl_unregister_scsi(dev);
+}
+
+bool bl_register_dev(struct pnfs_block_dev *dev)
+{
+	u32 i;
+
+	if (dev->nr_children) {
+		for (i = 0; i < dev->nr_children; i++) {
+			if (!bl_register_dev(&dev->children[i])) {
+				while (i > 0)
+					bl_unregister_dev(&dev->children[--i]);
+				return false;
+			}
+		}
+		return true;
+	}
+
+	if (dev->type == PNFS_BLOCK_VOLUME_SCSI)
+		return bl_register_scsi(dev);
+	return true;
+}
+
 static void
 bl_free_device(struct pnfs_block_dev *dev)
 {
+	bl_unregister_dev(dev);
+
 	if (dev->nr_children) {
 		int i;
 
@@ -23,17 +94,6 @@ bl_free_device(struct pnfs_block_dev *dev)
 			bl_free_device(&dev->children[i]);
 		kfree(dev->children);
 	} else {
-		if (dev->pr_registered) {
-			const struct pr_ops *ops =
-				file_bdev(dev->bdev_file)->bd_disk->fops->pr_ops;
-			int error;
-
-			error = ops->pr_register(file_bdev(dev->bdev_file),
-				dev->pr_key, 0, false);
-			if (error)
-				pr_err("failed to unregister PR key.\n");
-		}
-
 		if (dev->bdev_file)
 			fput(dev->bdev_file);
 	}
@@ -314,7 +374,7 @@ bl_open_path(struct pnfs_block_volume *v, const char *prefix)
 	bdev_file = bdev_file_open_by_path(devname, BLK_OPEN_READ | BLK_OPEN_WRITE,
 					NULL, NULL);
 	if (IS_ERR(bdev_file)) {
-		pr_warn("pNFS: failed to open device %s (%ld)\n",
+		dprintk("failed to open device %s (%ld)\n",
 			devname, PTR_ERR(bdev_file));
 	}
 
@@ -327,8 +387,9 @@ bl_parse_scsi(struct nfs_server *server, struct pnfs_block_dev *d,
 		struct pnfs_block_volume *volumes, int idx, gfp_t gfp_mask)
 {
 	struct pnfs_block_volume *v = &volumes[idx];
-	struct file *bdev_file;
+	struct block_device *bdev;
 	const struct pr_ops *ops;
+	struct file *bdev_file;
 	int error;
 
 	if (!bl_validate_designator(v))
@@ -344,35 +405,30 @@ bl_parse_scsi(struct nfs_server *server, struct pnfs_block_dev *d,
 	if (IS_ERR(bdev_file))
 		bdev_file = bl_open_path(v, "wwn-0x");
 	if (IS_ERR(bdev_file))
+		bdev_file = bl_open_path(v, "nvme-eui.");
+	if (IS_ERR(bdev_file)) {
+		pr_warn("pNFS: no device found for volume %*phN\n",
+			v->scsi.designator_len, v->scsi.designator);
 		return PTR_ERR(bdev_file);
+	}
 	d->bdev_file = bdev_file;
+	bdev = file_bdev(bdev_file);
 
-	d->len = bdev_nr_bytes(file_bdev(d->bdev_file));
+	d->len = bdev_nr_bytes(bdev);
 	d->map = bl_map_simple;
 	d->pr_key = v->scsi.pr_key;
 
 	if (d->len == 0)
 		return -ENODEV;
 
-	pr_info("pNFS: using block device %s (reservation key 0x%llx)\n",
+	ops = bdev->bd_disk->fops->pr_ops;
-		file_bdev(d->bdev_file)->bd_disk->disk_name, d->pr_key);
-
-	ops = file_bdev(d->bdev_file)->bd_disk->fops->pr_ops;
 	if (!ops) {
 		pr_err("pNFS: block device %s does not support reservations.",
-				file_bdev(d->bdev_file)->bd_disk->disk_name);
+				bdev->bd_disk->disk_name);
 		error = -EINVAL;
 		goto out_blkdev_put;
 	}
 
-	error = ops->pr_register(file_bdev(d->bdev_file), 0, d->pr_key, true);
-	if (error) {
-		pr_err("pNFS: failed to register key for block device %s.",
-				file_bdev(d->bdev_file)->bd_disk->disk_name);
-		goto out_blkdev_put;
-	}
-
-	d->pr_registered = true;
 	return 0;
 
 out_blkdev_put:
@@ -458,7 +514,9 @@ static int
 bl_parse_deviceid(struct nfs_server *server, struct pnfs_block_dev *d,
 		struct pnfs_block_volume *volumes, int idx, gfp_t gfp_mask)
 {
-	switch (volumes[idx].type) {
+	d->type = volumes[idx].type;
+
+	switch (d->type) {
 	case PNFS_BLOCK_VOLUME_SIMPLE:
 		return bl_parse_simple(server, d, volumes, idx, gfp_mask);
 	case PNFS_BLOCK_VOLUME_SLICE:
@@ -470,7 +528,7 @@ bl_parse_deviceid(struct nfs_server *server, struct pnfs_block_dev *d,
 	case PNFS_BLOCK_VOLUME_SCSI:
 		return bl_parse_scsi(server, d, volumes, idx, gfp_mask);
 	default:
-		dprintk("unsupported volume type: %d\n", volumes[idx].type);
+		dprintk("unsupported volume type: %d\n", d->type);
 		return -EIO;
 	}
 }

fs/nfs/callback.h

@@ -46,14 +46,15 @@ struct cb_compound_hdr_res {
 
 struct cb_getattrargs {
 	struct nfs_fh fh;
-	uint32_t bitmap[2];
+	uint32_t bitmap[3];
 };
 
 struct cb_getattrres {
 	__be32 status;
-	uint32_t bitmap[2];
+	uint32_t bitmap[3];
 	uint64_t size;
 	uint64_t change_attr;
+	struct timespec64 atime;
 	struct timespec64 ctime;
 	struct timespec64 mtime;
 };

fs/nfs/callback_proc.c

@@ -37,7 +37,7 @@ __be32 nfs4_callback_getattr(void *argp, void *resp,
 	if (!cps->clp) /* Always set for v4.0. Set in cb_sequence for v4.1 */
 		goto out;
 
-	res->bitmap[0] = res->bitmap[1] = 0;
+	memset(res->bitmap, 0, sizeof(res->bitmap));
 	res->status = htonl(NFS4ERR_BADHANDLE);
 
 	dprintk_rcu("NFS: GETATTR callback request from %s\n",
@@ -59,12 +59,16 @@ __be32 nfs4_callback_getattr(void *argp, void *resp,
 	res->change_attr = delegation->change_attr;
 	if (nfs_have_writebacks(inode))
 		res->change_attr++;
+	res->atime = inode_get_atime(inode);
 	res->ctime = inode_get_ctime(inode);
 	res->mtime = inode_get_mtime(inode);
-	res->bitmap[0] = (FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE) &
+	res->bitmap[0] = (FATTR4_WORD0_CHANGE | FATTR4_WORD0_SIZE) &
-		args->bitmap[0];
+			 args->bitmap[0];
-	res->bitmap[1] = (FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY) &
+	res->bitmap[1] = (FATTR4_WORD1_TIME_ACCESS |
-		args->bitmap[1];
+			  FATTR4_WORD1_TIME_METADATA |
+			  FATTR4_WORD1_TIME_MODIFY) & args->bitmap[1];
+	res->bitmap[2] = (FATTR4_WORD2_TIME_DELEG_ACCESS |
+			  FATTR4_WORD2_TIME_DELEG_MODIFY) & args->bitmap[2];
 	res->status = 0;
 out_iput:
 	rcu_read_unlock();
@@ -319,9 +323,10 @@ static u32 initiate_bulk_draining(struct nfs_client *clp,
 	int stat;
 
 	if (args->cbl_recall_type == RETURN_FSID)
-		stat = pnfs_destroy_layouts_byfsid(clp, &args->cbl_fsid, true);
+		stat = pnfs_layout_destroy_byfsid(clp, &args->cbl_fsid,
+						  PNFS_LAYOUT_BULK_RETURN);
 	else
-		stat = pnfs_destroy_layouts_byclid(clp, true);
+		stat = pnfs_layout_destroy_byclid(clp, PNFS_LAYOUT_BULK_RETURN);
 	if (stat != 0)
 		return NFS4ERR_DELAY;
 	return NFS4ERR_NOMATCHING_LAYOUT;

fs/nfs/callback_xdr.c

@@ -25,8 +25,9 @@
 #define CB_OP_GETATTR_BITMAP_MAXSZ	(4 * 4) // bitmap length, 3 bitmaps
 #define CB_OP_GETATTR_RES_MAXSZ		(CB_OP_HDR_RES_MAXSZ + \
 					 CB_OP_GETATTR_BITMAP_MAXSZ + \
-					 /* change, size, ctime, mtime */\
+					 /* change, size, atime, ctime,
-					 (2 + 2 + 3 + 3) * 4)
+					  * mtime, deleg_atime, deleg_mtime */\
+					 (2 + 2 + 3 + 3 + 3 + 3 + 3) * 4)
 #define CB_OP_RECALL_RES_MAXSZ		(CB_OP_HDR_RES_MAXSZ)
 
 #if defined(CONFIG_NFS_V4_1)
@@ -635,6 +636,13 @@ static __be32 encode_attr_time(struct xdr_stream *xdr, const struct timespec64 *
 	return 0;
 }
 
+static __be32 encode_attr_atime(struct xdr_stream *xdr, const uint32_t *bitmap, const struct timespec64 *time)
+{
+	if (!(bitmap[1] & FATTR4_WORD1_TIME_ACCESS))
+		return 0;
+	return encode_attr_time(xdr,time);
+}
+
 static __be32 encode_attr_ctime(struct xdr_stream *xdr, const uint32_t *bitmap, const struct timespec64 *time)
 {
 	if (!(bitmap[1] & FATTR4_WORD1_TIME_METADATA))
@@ -649,6 +657,24 @@ static __be32 encode_attr_mtime(struct xdr_stream *xdr, const uint32_t *bitmap,
 	return encode_attr_time(xdr,time);
 }
 
+static __be32 encode_attr_delegatime(struct xdr_stream *xdr,
+				     const uint32_t *bitmap,
+				     const struct timespec64 *time)
+{
+	if (!(bitmap[2] & FATTR4_WORD2_TIME_DELEG_ACCESS))
+		return 0;
+	return encode_attr_time(xdr,time);
+}
+
+static __be32 encode_attr_delegmtime(struct xdr_stream *xdr,
+				     const uint32_t *bitmap,
+				     const struct timespec64 *time)
+{
+	if (!(bitmap[2] & FATTR4_WORD2_TIME_DELEG_MODIFY))
+		return 0;
+	return encode_attr_time(xdr,time);
+}
+
 static __be32 encode_compound_hdr_res(struct xdr_stream *xdr, struct cb_compound_hdr_res *hdr)
 {
 	__be32 status;
@@ -697,12 +723,21 @@ static __be32 encode_getattr_res(struct svc_rqst *rqstp, struct xdr_stream *xdr,
 	if (unlikely(status != 0))
 		goto out;
 	status = encode_attr_size(xdr, res->bitmap, res->size);
+	if (unlikely(status != 0))
+		goto out;
+	status = encode_attr_atime(xdr, res->bitmap, &res->atime);
 	if (unlikely(status != 0))
 		goto out;
 	status = encode_attr_ctime(xdr, res->bitmap, &res->ctime);
 	if (unlikely(status != 0))
 		goto out;
 	status = encode_attr_mtime(xdr, res->bitmap, &res->mtime);
+	if (unlikely(status != 0))
+		goto out;
+	status = encode_attr_delegatime(xdr, res->bitmap, &res->atime);
+	if (unlikely(status != 0))
+		goto out;
+	status = encode_attr_delegmtime(xdr, res->bitmap, &res->mtime);
 	*savep = htonl((unsigned int)((char *)xdr->p - (char *)(savep+1)));
 out:
 	return status;

fs/nfs/client.c

@@ -994,6 +994,9 @@ struct nfs_server *nfs_alloc_server(void)
 
 	server->change_attr_type = NFS4_CHANGE_TYPE_IS_UNDEFINED;
 
+	init_waitqueue_head(&server->write_congestion_wait);
+	atomic_long_set(&server->writeback, 0);
+
 	ida_init(&server->openowner_id);
 	ida_init(&server->lockowner_id);
 	pnfs_init_server(server);

fs/nfs/delegation.c

@@ -82,11 +82,10 @@ static void nfs_mark_return_delegation(struct nfs_server *server,
 	set_bit(NFS4CLNT_DELEGRETURN, &server->nfs_client->cl_state);
 }
 
-static bool
+static bool nfs4_is_valid_delegation(const struct nfs_delegation *delegation,
-nfs4_is_valid_delegation(const struct nfs_delegation *delegation,
+				     fmode_t type)
-		fmode_t flags)
 {
-	if (delegation != NULL && (delegation->type & flags) == flags &&
+	if (delegation != NULL && (delegation->type & type) == type &&
 	    !test_bit(NFS_DELEGATION_REVOKED, &delegation->flags) &&
 	    !test_bit(NFS_DELEGATION_RETURNING, &delegation->flags))
 		return true;
@@ -103,19 +102,22 @@ struct nfs_delegation *nfs4_get_valid_delegation(const struct inode *inode)
 	return NULL;
 }
 
-static int
+static int nfs4_do_check_delegation(struct inode *inode, fmode_t type,
-nfs4_do_check_delegation(struct inode *inode, fmode_t flags, bool mark)
+				    int flags, bool mark)
 {
 	struct nfs_delegation *delegation;
 	int ret = 0;
 
-	flags &= FMODE_READ|FMODE_WRITE;
+	type &= FMODE_READ|FMODE_WRITE;
 	rcu_read_lock();
 	delegation = rcu_dereference(NFS_I(inode)->delegation);
-	if (nfs4_is_valid_delegation(delegation, flags)) {
+	if (nfs4_is_valid_delegation(delegation, type)) {
 		if (mark)
 			nfs_mark_delegation_referenced(delegation);
 		ret = 1;
+		if ((flags & NFS_DELEGATION_FLAG_TIME) &&
+		    !test_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags))
+			ret = 0;
 	}
 	rcu_read_unlock();
 	return ret;
@@ -124,22 +126,23 @@ nfs4_do_check_delegation(struct inode *inode, fmode_t flags, bool mark)
  * nfs4_have_delegation - check if inode has a delegation, mark it
  * NFS_DELEGATION_REFERENCED if there is one.
  * @inode: inode to check
- * @flags: delegation types to check for
+ * @type: delegation types to check for
+ * @flags: various modifiers
  *
  * Returns one if inode has the indicated delegation, otherwise zero.
  */
-int nfs4_have_delegation(struct inode *inode, fmode_t flags)
+int nfs4_have_delegation(struct inode *inode, fmode_t type, int flags)
 {
-	return nfs4_do_check_delegation(inode, flags, true);
+	return nfs4_do_check_delegation(inode, type, flags, true);
 }
 
 /*
  * nfs4_check_delegation - check if inode has a delegation, do not mark
  * NFS_DELEGATION_REFERENCED if it has one.
  */
-int nfs4_check_delegation(struct inode *inode, fmode_t flags)
+int nfs4_check_delegation(struct inode *inode, fmode_t type)
 {
-	return nfs4_do_check_delegation(inode, flags, false);
+	return nfs4_do_check_delegation(inode, type, 0, false);
 }
 
 static int nfs_delegation_claim_locks(struct nfs4_state *state, const nfs4_stateid *stateid)
@@ -221,11 +224,12 @@ again:
  * @type: delegation type
  * @stateid: delegation stateid
  * @pagemod_limit: write delegation "space_limit"
+ * @deleg_type: raw delegation type
  *
  */
 void nfs_inode_reclaim_delegation(struct inode *inode, const struct cred *cred,
 				  fmode_t type, const nfs4_stateid *stateid,
-				  unsigned long pagemod_limit)
+				  unsigned long pagemod_limit, u32 deleg_type)
 {
 	struct nfs_delegation *delegation;
 	const struct cred *oldcred = NULL;
@@ -239,6 +243,14 @@ void nfs_inode_reclaim_delegation(struct inode *inode, const struct cred *cred,
 		delegation->pagemod_limit = pagemod_limit;
 		oldcred = delegation->cred;
 		delegation->cred = get_cred(cred);
+		switch (deleg_type) {
+		case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
+		case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
+			set_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags);
+			break;
+		default:
+			clear_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags);
+		}
 		clear_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags);
 		if (test_and_clear_bit(NFS_DELEGATION_REVOKED,
 				       &delegation->flags))
@@ -250,11 +262,13 @@ void nfs_inode_reclaim_delegation(struct inode *inode, const struct cred *cred,
 	} else {
 		rcu_read_unlock();
 		nfs_inode_set_delegation(inode, cred, type, stateid,
-					 pagemod_limit);
+					 pagemod_limit, deleg_type);
 	}
 }
 
-static int nfs_do_return_delegation(struct inode *inode, struct nfs_delegation *delegation, int issync)
+static int nfs_do_return_delegation(struct inode *inode,
+				    struct nfs_delegation *delegation,
+				    int issync)
 {
 	const struct cred *cred;
 	int res = 0;
@@ -263,9 +277,8 @@ static int nfs_do_return_delegation(struct inode *inode, struct nfs_delegation *
 		spin_lock(&delegation->lock);
 		cred = get_cred(delegation->cred);
 		spin_unlock(&delegation->lock);
-		res = nfs4_proc_delegreturn(inode, cred,
+		res = nfs4_proc_delegreturn(inode, cred, &delegation->stateid,
-				&delegation->stateid,
+					    delegation, issync);
-				issync);
 		put_cred(cred);
 	}
 	return res;
@@ -418,13 +431,13 @@ nfs_update_inplace_delegation(struct nfs_delegation *delegation,
  * @type: delegation type
  * @stateid: delegation stateid
  * @pagemod_limit: write delegation "space_limit"
+ * @deleg_type: raw delegation type
  *
  * Returns zero on success, or a negative errno value.
  */
 int nfs_inode_set_delegation(struct inode *inode, const struct cred *cred,
-				  fmode_t type,
+			     fmode_t type, const nfs4_stateid *stateid,
-				  const nfs4_stateid *stateid,
+			     unsigned long pagemod_limit, u32 deleg_type)
-				  unsigned long pagemod_limit)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs_client *clp = server->nfs_client;
@@ -444,6 +457,11 @@ int nfs_inode_set_delegation(struct inode *inode, const struct cred *cred,
 	delegation->cred = get_cred(cred);
 	delegation->inode = inode;
 	delegation->flags = 1<<NFS_DELEGATION_REFERENCED;
+	switch (deleg_type) {
+	case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
+	case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
+		delegation->flags |= BIT(NFS_DELEGATION_DELEGTIME);
+	}
 	delegation->test_gen = 0;
 	spin_lock_init(&delegation->lock);
 
@@ -508,6 +526,11 @@ add_new:
 	atomic_long_inc(&nfs_active_delegations);
 
 	trace_nfs4_set_delegation(inode, type);
+
+	/* If we hold writebacks and have delegated mtime then update */
+	if (deleg_type == NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG &&
+	    nfs_have_writebacks(inode))
+		nfs_update_delegated_mtime(inode);
 out:
 	spin_unlock(&clp->cl_lock);
 	if (delegation != NULL)

fs/nfs/delegation.h

@@ -38,12 +38,15 @@ enum {
 	NFS_DELEGATION_TEST_EXPIRED,
 	NFS_DELEGATION_INODE_FREEING,
 	NFS_DELEGATION_RETURN_DELAYED,
+	NFS_DELEGATION_DELEGTIME,
 };
 
 int nfs_inode_set_delegation(struct inode *inode, const struct cred *cred,
-		fmode_t type, const nfs4_stateid *stateid, unsigned long pagemod_limit);
+			     fmode_t type, const nfs4_stateid *stateid,
+			     unsigned long pagemod_limit, u32 deleg_type);
 void nfs_inode_reclaim_delegation(struct inode *inode, const struct cred *cred,
-		fmode_t type, const nfs4_stateid *stateid, unsigned long pagemod_limit);
+				  fmode_t type, const nfs4_stateid *stateid,
+				  unsigned long pagemod_limit, u32 deleg_type);
 int nfs4_inode_return_delegation(struct inode *inode);
 void nfs4_inode_return_delegation_on_close(struct inode *inode);
 int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid);
@@ -67,7 +70,9 @@ void nfs_test_expired_all_delegations(struct nfs_client *clp);
 void nfs_reap_expired_delegations(struct nfs_client *clp);
 
 /* NFSv4 delegation-related procedures */
-int nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred, const nfs4_stateid *stateid, int issync);
+int nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
+			  const nfs4_stateid *stateid,
+			  struct nfs_delegation *delegation, int issync);
 int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid);
 int nfs4_lock_delegation_recall(struct file_lock *fl, struct nfs4_state *state, const nfs4_stateid *stateid);
 bool nfs4_copy_delegation_stateid(struct inode *inode, fmode_t flags, nfs4_stateid *dst, const struct cred **cred);
@@ -75,8 +80,8 @@ bool nfs4_refresh_delegation_stateid(nfs4_stateid *dst, struct inode *inode);
 
 struct nfs_delegation *nfs4_get_valid_delegation(const struct inode *inode);
 void nfs_mark_delegation_referenced(struct nfs_delegation *delegation);
-int nfs4_have_delegation(struct inode *inode, fmode_t flags);
+int nfs4_have_delegation(struct inode *inode, fmode_t type, int flags);
-int nfs4_check_delegation(struct inode *inode, fmode_t flags);
+int nfs4_check_delegation(struct inode *inode, fmode_t type);
 bool nfs4_delegation_flush_on_close(const struct inode *inode);
 void nfs_inode_find_delegation_state_and_recover(struct inode *inode,
 		const nfs4_stateid *stateid);
@@ -84,9 +89,37 @@ int nfs4_inode_make_writeable(struct inode *inode);
 
 #endif
 
+#define NFS_DELEGATION_FLAG_TIME	BIT(1)
+
+void nfs_update_delegated_atime(struct inode *inode);
+void nfs_update_delegated_mtime(struct inode *inode);
+void nfs_update_delegated_mtime_locked(struct inode *inode);
+
+static inline int nfs_have_read_or_write_delegation(struct inode *inode)
+{
+	return NFS_PROTO(inode)->have_delegation(inode, FMODE_READ, 0);
+}
+
+static inline int nfs_have_write_delegation(struct inode *inode)
+{
+	return NFS_PROTO(inode)->have_delegation(inode, FMODE_WRITE, 0);
+}
+
 static inline int nfs_have_delegated_attributes(struct inode *inode)
 {
-	return NFS_PROTO(inode)->have_delegation(inode, FMODE_READ);
+	return NFS_PROTO(inode)->have_delegation(inode, FMODE_READ, 0);
+}
+
+static inline int nfs_have_delegated_atime(struct inode *inode)
+{
+	return NFS_PROTO(inode)->have_delegation(inode, FMODE_READ,
+						 NFS_DELEGATION_FLAG_TIME);
+}
+
+static inline int nfs_have_delegated_mtime(struct inode *inode)
+{
+	return NFS_PROTO(inode)->have_delegation(inode, FMODE_WRITE,
+						 NFS_DELEGATION_FLAG_TIME);
 }
 
 #endif

fs/nfs/dir.c

@@ -1437,7 +1437,7 @@ static void nfs_set_verifier_locked(struct dentry *dentry, unsigned long verf)
 
 	if (!dir || !nfs_verify_change_attribute(dir, verf))
 		return;
-	if (inode && NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
+	if (inode && NFS_PROTO(inode)->have_delegation(inode, FMODE_READ, 0))
 		nfs_set_verifier_delegated(&verf);
 	dentry->d_time = verf;
 }

fs/nfs/file.c

@@ -339,6 +339,7 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,
 			   loff_t pos, unsigned len, struct page **pagep,
 			   void **fsdata)
 {
+	fgf_t fgp = FGP_WRITEBEGIN;
 	struct folio *folio;
 	int once_thru = 0;
 	int ret;
@@ -346,8 +347,9 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,
 	dfprintk(PAGECACHE, "NFS: write_begin(%pD2(%lu), %u@%lld)\n",
 		file, mapping->host->i_ino, len, (long long) pos);
 
+	fgp |= fgf_set_order(len);
 start:
-	folio = __filemap_get_folio(mapping, pos >> PAGE_SHIFT, FGP_WRITEBEGIN,
+	folio = __filemap_get_folio(mapping, pos >> PAGE_SHIFT, fgp,
 				    mapping_gfp_mask(mapping));
 	if (IS_ERR(folio))
 		return PTR_ERR(folio);
@@ -425,7 +427,7 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,
 static void nfs_invalidate_folio(struct folio *folio, size_t offset,
 				size_t length)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 	dfprintk(PAGECACHE, "NFS: invalidate_folio(%lu, %zu, %zu)\n",
 		 folio->index, offset, length);
 
@@ -434,7 +436,7 @@ static void nfs_invalidate_folio(struct folio *folio, size_t offset,
 	/* Cancel any unstarted writes on this page */
 	nfs_wb_folio_cancel(inode, folio);
 	folio_wait_private_2(folio); /* [DEPRECATED] */
-	trace_nfs_invalidate_folio(inode, folio);
+	trace_nfs_invalidate_folio(inode, folio_pos(folio) + offset, length);
 }
 
 /*
@@ -452,7 +454,7 @@ static bool nfs_release_folio(struct folio *folio, gfp_t gfp)
 		if ((current_gfp_context(gfp) & GFP_KERNEL) != GFP_KERNEL ||
 		    current_is_kswapd())
 			return false;
-		if (nfs_wb_folio(folio_file_mapping(folio)->host, folio) < 0)
+		if (nfs_wb_folio(folio->mapping->host, folio) < 0)
 			return false;
 	}
 	return nfs_fscache_release_folio(folio, gfp);
@@ -502,7 +504,8 @@ static int nfs_launder_folio(struct folio *folio)
 
 	folio_wait_private_2(folio); /* [DEPRECATED] */
 	ret = nfs_wb_folio(inode, folio);
-	trace_nfs_launder_folio_done(inode, folio, ret);
+	trace_nfs_launder_folio_done(inode, folio_pos(folio),
+			folio_size(folio), ret);
 	return ret;
 }
 
@@ -604,7 +607,7 @@ static vm_fault_t nfs_vm_page_mkwrite(struct vm_fault *vmf)
 			   TASK_KILLABLE|TASK_FREEZABLE_UNSAFE);
 
 	folio_lock(folio);
-	mapping = folio_file_mapping(folio);
+	mapping = folio->mapping;
 	if (mapping != inode->i_mapping)
 		goto out_unlock;
 
@@ -730,7 +733,7 @@ do_getlk(struct file *filp, int cmd, struct file_lock *fl, int is_local)
 	}
 	fl->c.flc_type = saved_type;
 
-	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
+	if (nfs_have_read_or_write_delegation(inode))
 		goto out_noconflict;
 
 	if (is_local)
@@ -813,7 +816,7 @@ do_setlk(struct file *filp, int cmd, struct file_lock *fl, int is_local)
 	 * This makes locking act as a cache coherency point.
 	 */
 	nfs_sync_mapping(filp->f_mapping);
-	if (!NFS_PROTO(inode)->have_delegation(inode, FMODE_READ)) {
+	if (!nfs_have_read_or_write_delegation(inode)) {
 		nfs_zap_caches(inode);
 		if (mapping_mapped(filp->f_mapping))
 			nfs_revalidate_mapping(inode, filp->f_mapping);

fs/nfs/filelayout/filelayout.c

@@ -1110,7 +1110,6 @@ static const struct pnfs_commit_ops filelayout_commit_ops = {
 	.clear_request_commit	= pnfs_generic_clear_request_commit,
 	.scan_commit_lists	= pnfs_generic_scan_commit_lists,
 	.recover_commit_reqs	= pnfs_generic_recover_commit_reqs,
-	.search_commit_reqs	= pnfs_generic_search_commit_reqs,
 	.commit_pagelist	= filelayout_commit_pagelist,
 };
 

fs/nfs/flexfilelayout/flexfilelayout.c

@@ -2548,7 +2548,7 @@ ff_layout_set_layoutdriver(struct nfs_server *server,
 		const struct nfs_fh *dummy)
 {
 #if IS_ENABLED(CONFIG_NFS_V4_2)
-	server->caps |= NFS_CAP_LAYOUTSTATS;
+	server->caps |= NFS_CAP_LAYOUTSTATS | NFS_CAP_REBOOT_LAYOUTRETURN;
 #endif
 	return 0;
 }

fs/nfs/fscache.c

@@ -341,7 +341,7 @@ void nfs_netfs_initiate_read(struct nfs_pgio_header *hdr)
 
 int nfs_netfs_folio_unlock(struct folio *folio)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 
 	/*
 	 * If fscache is enabled, netfs will unlock pages.

fs/nfs/inode.c

@@ -190,9 +190,8 @@ static bool nfs_has_xattr_cache(const struct nfs_inode *nfsi)
 void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
-	bool have_delegation = NFS_PROTO(inode)->have_delegation(inode, FMODE_READ);
 
-	if (have_delegation) {
+	if (nfs_have_delegated_attributes(inode)) {
 		if (!(flags & NFS_INO_REVAL_FORCED))
 			flags &= ~(NFS_INO_INVALID_MODE |
 				   NFS_INO_INVALID_OTHER |
@@ -276,6 +275,8 @@ EXPORT_SYMBOL_GPL(nfs_zap_acl_cache);
 
 void nfs_invalidate_atime(struct inode *inode)
 {
+	if (nfs_have_delegated_atime(inode))
+		return;
 	spin_lock(&inode->i_lock);
 	nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATIME);
 	spin_unlock(&inode->i_lock);
@@ -491,6 +492,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 			inode->i_fop = NFS_SB(sb)->nfs_client->rpc_ops->file_ops;
 			inode->i_data.a_ops = &nfs_file_aops;
 			nfs_inode_init_regular(nfsi);
+			mapping_set_large_folios(inode->i_mapping);
 		} else if (S_ISDIR(inode->i_mode)) {
 			inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->dir_inode_ops;
 			inode->i_fop = &nfs_dir_operations;
@@ -604,6 +606,55 @@ out_no_inode:
 }
 EXPORT_SYMBOL_GPL(nfs_fhget);
 
+static void
+nfs_fattr_fixup_delegated(struct inode *inode, struct nfs_fattr *fattr)
+{
+	unsigned long cache_validity = NFS_I(inode)->cache_validity;
+
+	if (nfs_have_delegated_mtime(inode)) {
+		if (!(cache_validity & NFS_INO_INVALID_CTIME))
+			fattr->valid &= ~(NFS_ATTR_FATTR_PRECTIME |
+					  NFS_ATTR_FATTR_CTIME);
+
+		if (!(cache_validity & NFS_INO_INVALID_MTIME))
+			fattr->valid &= ~(NFS_ATTR_FATTR_PREMTIME |
+					  NFS_ATTR_FATTR_MTIME);
+
+		if (!(cache_validity & NFS_INO_INVALID_ATIME))
+			fattr->valid &= ~NFS_ATTR_FATTR_ATIME;
+	} else if (nfs_have_delegated_atime(inode)) {
+		if (!(cache_validity & NFS_INO_INVALID_ATIME))
+			fattr->valid &= ~NFS_ATTR_FATTR_ATIME;
+	}
+}
+
+void nfs_update_delegated_atime(struct inode *inode)
+{
+	spin_lock(&inode->i_lock);
+	if (nfs_have_delegated_atime(inode)) {
+		inode_update_timestamps(inode, S_ATIME);
+		NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_ATIME;
+	}
+	spin_unlock(&inode->i_lock);
+}
+
+void nfs_update_delegated_mtime_locked(struct inode *inode)
+{
+	if (nfs_have_delegated_mtime(inode)) {
+		inode_update_timestamps(inode, S_CTIME | S_MTIME);
+		NFS_I(inode)->cache_validity &= ~(NFS_INO_INVALID_CTIME |
+						  NFS_INO_INVALID_MTIME);
+	}
+}
+
+void nfs_update_delegated_mtime(struct inode *inode)
+{
+	spin_lock(&inode->i_lock);
+	nfs_update_delegated_mtime_locked(inode);
+	spin_unlock(&inode->i_lock);
+}
+EXPORT_SYMBOL_GPL(nfs_update_delegated_mtime);
+
 #define NFS_VALID_ATTRS (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_SIZE|ATTR_ATIME|ATTR_ATIME_SET|ATTR_MTIME|ATTR_MTIME_SET|ATTR_FILE|ATTR_OPEN)
 
 int
@@ -631,6 +682,17 @@ nfs_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 			attr->ia_valid &= ~ATTR_SIZE;
 	}
 
+	if (nfs_have_delegated_mtime(inode)) {
+		if (attr->ia_valid & ATTR_MTIME) {
+			nfs_update_delegated_mtime(inode);
+			attr->ia_valid &= ~ATTR_MTIME;
+		}
+		if (attr->ia_valid & ATTR_ATIME) {
+			nfs_update_delegated_atime(inode);
+			attr->ia_valid &= ~ATTR_ATIME;
+		}
+	}
+
 	/* Optimization: if the end result is no change, don't RPC */
 	if (((attr->ia_valid & NFS_VALID_ATTRS) & ~(ATTR_FILE|ATTR_OPEN)) == 0)
 		return 0;
@@ -686,6 +748,7 @@ static int nfs_vmtruncate(struct inode * inode, loff_t offset)
 
 	spin_unlock(&inode->i_lock);
 	truncate_pagecache(inode, offset);
+	nfs_update_delegated_mtime_locked(inode);
 	spin_lock(&inode->i_lock);
 out:
 	return err;
@@ -709,8 +772,9 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr,
 	spin_lock(&inode->i_lock);
 	NFS_I(inode)->attr_gencount = fattr->gencount;
 	if ((attr->ia_valid & ATTR_SIZE) != 0) {
-		nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME |
+		if (!nfs_have_delegated_mtime(inode))
-						     NFS_INO_INVALID_BLOCKS);
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_BLOCKS);
 		nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC);
 		nfs_vmtruncate(inode, attr->ia_size);
 	}
@@ -856,8 +920,12 @@ int nfs_getattr(struct mnt_idmap *idmap, const struct path *path,
 
 	/* Flush out writes to the server in order to update c/mtime/version.  */
 	if ((request_mask & (STATX_CTIME | STATX_MTIME | STATX_CHANGE_COOKIE)) &&
-	    S_ISREG(inode->i_mode))
+	    S_ISREG(inode->i_mode)) {
-		filemap_write_and_wait(inode->i_mapping);
+		if (nfs_have_delegated_mtime(inode))
+			filemap_fdatawrite(inode->i_mapping);
+		else
+			filemap_write_and_wait(inode->i_mapping);
+	}
 
 	/*
 	 * We may force a getattr if the user cares about atime.
@@ -1012,7 +1080,7 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 	if (!is_sync)
 		return;
 	inode = d_inode(ctx->dentry);
-	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
+	if (nfs_have_read_or_write_delegation(inode))
 		return;
 	nfsi = NFS_I(inode);
 	if (inode->i_mapping->nrpages == 0)
@@ -1482,7 +1550,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat
 	unsigned long invalid = 0;
 	struct timespec64 ts;
 
-	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
+	if (nfs_have_delegated_attributes(inode))
 		return 0;
 
 	if (!(fattr->valid & NFS_ATTR_FATTR_FILEID)) {
@@ -2118,6 +2186,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 	 */
 	nfsi->read_cache_jiffies = fattr->time_start;
 
+	/* Fix up any delegated attributes in the struct nfs_fattr */
+	nfs_fattr_fixup_delegated(inode, fattr);
+
 	save_cache_validity = nfsi->cache_validity;
 	nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR
 			| NFS_INO_INVALID_ATIME
@@ -2538,6 +2609,7 @@ static void __exit exit_nfs_fs(void)
 
 /* Not quite true; I just maintain it */
 MODULE_AUTHOR("Olaf Kirch <okir@monad.swb.de>");
+MODULE_DESCRIPTION("NFS client support");
 MODULE_LICENSE("GPL");
 module_param(enable_ino64, bool, 0644);
 

fs/nfs/internal.h

@@ -785,7 +785,7 @@ static inline void nfs_folio_mark_unstable(struct folio *folio,
 					   struct nfs_commit_info *cinfo)
 {
 	if (folio && !cinfo->dreq) {
-		struct inode *inode = folio_file_mapping(folio)->host;
+		struct inode *inode = folio->mapping->host;
 		long nr = folio_nr_pages(folio);
 
 		/* This page is really still in write-back - just that the
@@ -797,34 +797,15 @@ static inline void nfs_folio_mark_unstable(struct folio *folio,
 	}
 }
 
-/*
- * Determine the number of bytes of data the page contains
- */
-static inline
-unsigned int nfs_page_length(struct page *page)
-{
-	loff_t i_size = i_size_read(page_file_mapping(page)->host);
-
-	if (i_size > 0) {
-		pgoff_t index = page_index(page);
-		pgoff_t end_index = (i_size - 1) >> PAGE_SHIFT;
-		if (index < end_index)
-			return PAGE_SIZE;
-		if (index == end_index)
-			return ((i_size - 1) & ~PAGE_MASK) + 1;
-	}
-	return 0;
-}
-
 /*
  * Determine the number of bytes of data the page contains
  */
 static inline size_t nfs_folio_length(struct folio *folio)
 {
-	loff_t i_size = i_size_read(folio_file_mapping(folio)->host);
+	loff_t i_size = i_size_read(folio->mapping->host);
 
 	if (i_size > 0) {
-		pgoff_t index = folio_index(folio) >> folio_order(folio);
+		pgoff_t index = folio->index >> folio_order(folio);
 		pgoff_t end_index = (i_size - 1) >> folio_shift(folio);
 		if (index < end_index)
 			return folio_size(folio);

fs/nfs/mount_clnt.c

@@ -128,11 +128,6 @@ struct mountres {
 	rpc_authflavor_t *auth_flavors;
 };
 
-struct mnt_fhstatus {
-	u32 status;
-	struct nfs_fh *fh;
-};
-
 /**
  * nfs_mount - Obtain an NFS file handle for the given host and path
  * @info: pointer to mount request arguments

fs/nfs/nfs2super.c

@@ -26,6 +26,7 @@ static void __exit exit_nfs_v2(void)
 	unregister_nfs_version(&nfs_v2);
 }
 
+MODULE_DESCRIPTION("NFSv2 client support");
 MODULE_LICENSE("GPL");
 
 module_init(init_nfs_v2);

fs/nfs/nfs3proc.c

@@ -979,11 +979,18 @@ nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
 	return status;
 }
 
-static int nfs3_have_delegation(struct inode *inode, fmode_t flags)
+static int nfs3_have_delegation(struct inode *inode, fmode_t type, int flags)
 {
 	return 0;
 }
 
+static int nfs3_return_delegation(struct inode *inode)
+{
+	if (S_ISREG(inode->i_mode))
+		nfs_wb_all(inode);
+	return 0;
+}
+
 static const struct inode_operations nfs3_dir_inode_operations = {
 	.create		= nfs_create,
 	.atomic_open	= nfs_atomic_open_v23,
@@ -1062,6 +1069,7 @@ const struct nfs_rpc_ops nfs_v3_clientops = {
 	.clear_acl_cache = forget_all_cached_acls,
 	.close_context	= nfs_close_context,
 	.have_delegation = nfs3_have_delegation,
+	.return_delegation = nfs3_return_delegation,
 	.alloc_client	= nfs_alloc_client,
 	.init_client	= nfs_init_client,
 	.free_client	= nfs_free_client,

fs/nfs/nfs3super.c

@@ -27,6 +27,7 @@ static void __exit exit_nfs_v3(void)
 	unregister_nfs_version(&nfs_v3);
 }
 
+MODULE_DESCRIPTION("NFSv3 client support");
 MODULE_LICENSE("GPL");
 
 module_init(init_nfs_v3);

fs/nfs/nfs4_fs.h

@@ -67,7 +67,8 @@ struct nfs4_minor_version_ops {
 	void	(*free_lock_state)(struct nfs_server *,
 			struct nfs4_lock_state *);
 	int	(*test_and_free_expired)(struct nfs_server *,
-			nfs4_stateid *, const struct cred *);
+					 const nfs4_stateid *,
+					 const struct cred *);
 	struct nfs_seqid *
 		(*alloc_seqid)(struct nfs_seqid_counter *, gfp_t);
 	void	(*session_trunk)(struct rpc_clnt *clnt,

fs/nfs/nfs4client.c

@@ -231,9 +231,8 @@ struct nfs_client *nfs4_alloc_client(const struct nfs_client_initdata *cl_init)
 		__set_bit(NFS_CS_INFINITE_SLOTS, &clp->cl_flags);
 	__set_bit(NFS_CS_DISCRTRY, &clp->cl_flags);
 	__set_bit(NFS_CS_NO_RETRANS_TIMEOUT, &clp->cl_flags);
-
+	if (test_bit(NFS_CS_PNFS, &cl_init->init_flags))
-	if (test_bit(NFS_CS_DS, &cl_init->init_flags))
+		__set_bit(NFS_CS_PNFS, &clp->cl_flags);
-		__set_bit(NFS_CS_DS, &clp->cl_flags);
 	/*
 	 * Set up the connection to the server before we add add to the
 	 * global list.
@@ -1013,7 +1012,6 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv,
 	if (mds_srv->flags & NFS_MOUNT_NORESVPORT)
 		__set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags);
 
-	__set_bit(NFS_CS_DS, &cl_init.init_flags);
 	__set_bit(NFS_CS_PNFS, &cl_init.init_flags);
 	cl_init.max_connect = NFS_MAX_TRANSPORTS;
 	/*

fs/nfs/nfs4proc.c

@@ -103,10 +103,10 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp,
 		const struct cred *cred,
 		struct nfs4_slot *slot,
 		bool is_privileged);
-static int nfs41_test_stateid(struct nfs_server *, nfs4_stateid *,
+static int nfs41_test_stateid(struct nfs_server *, const nfs4_stateid *,
-		const struct cred *);
+			      const struct cred *);
 static int nfs41_free_stateid(struct nfs_server *, const nfs4_stateid *,
-		const struct cred *, bool);
+			      const struct cred *, bool);
 #endif
 
 #ifdef CONFIG_NFS_V4_SECURITY_LABEL
@@ -293,7 +293,7 @@ static void nfs4_bitmap_copy_adjust(__u32 *dst, const __u32 *src,
 	unsigned long cache_validity;
 
 	memcpy(dst, src, NFS4_BITMASK_SZ*sizeof(*dst));
-	if (!inode || !nfs4_have_delegation(inode, FMODE_READ))
+	if (!inode || !nfs_have_read_or_write_delegation(inode))
 		return;
 
 	cache_validity = READ_ONCE(NFS_I(inode)->cache_validity) | flags;
@@ -310,6 +310,18 @@ static void nfs4_bitmap_copy_adjust(__u32 *dst, const __u32 *src,
 		dst[1] &= ~FATTR4_WORD1_MODE;
 	if (!(cache_validity & NFS_INO_INVALID_OTHER))
 		dst[1] &= ~(FATTR4_WORD1_OWNER | FATTR4_WORD1_OWNER_GROUP);
+
+	if (nfs_have_delegated_mtime(inode)) {
+		if (!(cache_validity & NFS_INO_INVALID_ATIME))
+			dst[1] &= ~FATTR4_WORD1_TIME_ACCESS;
+		if (!(cache_validity & NFS_INO_INVALID_MTIME))
+			dst[1] &= ~FATTR4_WORD1_TIME_MODIFY;
+		if (!(cache_validity & NFS_INO_INVALID_CTIME))
+			dst[1] &= ~FATTR4_WORD1_TIME_METADATA;
+	} else if (nfs_have_delegated_atime(inode)) {
+		if (!(cache_validity & NFS_INO_INVALID_ATIME))
+			dst[1] &= ~FATTR4_WORD1_TIME_ACCESS;
+	}
 }
 
 static void nfs4_setup_readdir(u64 cookie, __be32 *verifier, struct dentry *dentry,
@@ -1245,7 +1257,8 @@ nfs4_update_changeattr_locked(struct inode *inode,
 	struct nfs_inode *nfsi = NFS_I(inode);
 	u64 change_attr = inode_peek_iversion_raw(inode);
 
-	cache_validity |= NFS_INO_INVALID_CTIME | NFS_INO_INVALID_MTIME;
+	if (!nfs_have_delegated_mtime(inode))
+		cache_validity |= NFS_INO_INVALID_CTIME | NFS_INO_INVALID_MTIME;
 	if (S_ISDIR(inode->i_mode))
 		cache_validity |= NFS_INO_INVALID_DATA;
 
@@ -1264,7 +1277,7 @@ nfs4_update_changeattr_locked(struct inode *inode,
 		if (S_ISDIR(inode->i_mode))
 			nfs_force_lookup_revalidate(inode);
 
-		if (!NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
+		if (!nfs_have_delegated_attributes(inode))
 			cache_validity |=
 				NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL |
 				NFS_INO_INVALID_SIZE | NFS_INO_INVALID_OTHER |
@@ -1320,8 +1333,7 @@ static fmode_t _nfs4_ctx_to_openmode(const struct nfs_open_context *ctx)
 }
 
 static u32
-nfs4_map_atomic_open_share(struct nfs_server *server,
+nfs4_fmode_to_share_access(fmode_t fmode)
-		fmode_t fmode, int openflags)
 {
 	u32 res = 0;
 
@@ -1335,11 +1347,27 @@ nfs4_map_atomic_open_share(struct nfs_server *server,
 	case FMODE_READ|FMODE_WRITE:
 		res = NFS4_SHARE_ACCESS_BOTH;
 	}
+	return res;
+}
+
+static u32
+nfs4_map_atomic_open_share(struct nfs_server *server,
+		fmode_t fmode, int openflags)
+{
+	u32 res = nfs4_fmode_to_share_access(fmode);
+
 	if (!(server->caps & NFS_CAP_ATOMIC_OPEN_V1))
 		goto out;
 	/* Want no delegation if we're using O_DIRECT */
-	if (openflags & O_DIRECT)
+	if (openflags & O_DIRECT) {
 		res |= NFS4_SHARE_WANT_NO_DELEG;
+		goto out;
+	}
+	/* res |= NFS4_SHARE_WANT_NO_PREFERENCE; */
+	if (server->caps & NFS_CAP_DELEGTIME)
+		res |= NFS4_SHARE_WANT_DELEG_TIMESTAMPS;
+	if (server->caps & NFS_CAP_OPEN_XOR)
+		res |= NFS4_SHARE_WANT_OPEN_XOR_DELEGATION;
 out:
 	return res;
 }
@@ -1954,44 +1982,41 @@ out_return_state:
 }
 
 static void
-nfs4_opendata_check_deleg(struct nfs4_opendata *data, struct nfs4_state *state)
+nfs4_process_delegation(struct inode *inode, const struct cred *cred,
+			enum open_claim_type4 claim,
+			const struct nfs4_open_delegation *delegation)
 {
-	struct nfs_client *clp = NFS_SERVER(state->inode)->nfs_client;
+	switch (delegation->open_delegation_type) {
-	struct nfs_delegation *delegation;
+	case NFS4_OPEN_DELEGATE_READ:
-	int delegation_flags = 0;
+	case NFS4_OPEN_DELEGATE_WRITE:
-
+	case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
-	rcu_read_lock();
+	case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
-	delegation = rcu_dereference(NFS_I(state->inode)->delegation);
-	if (delegation)
-		delegation_flags = delegation->flags;
-	rcu_read_unlock();
-	switch (data->o_arg.claim) {
-	default:
 		break;
+	default:
+		return;
+	}
+	switch (claim) {
 	case NFS4_OPEN_CLAIM_DELEGATE_CUR:
 	case NFS4_OPEN_CLAIM_DELEG_CUR_FH:
 		pr_err_ratelimited("NFS: Broken NFSv4 server %s is "
 				   "returning a delegation for "
 				   "OPEN(CLAIM_DELEGATE_CUR)\n",
-				   clp->cl_hostname);
+				   NFS_SERVER(inode)->nfs_client->cl_hostname);
-		return;
+		break;
+	case NFS4_OPEN_CLAIM_PREVIOUS:
+		nfs_inode_reclaim_delegation(inode, cred, delegation->type,
+					     &delegation->stateid,
+					     delegation->pagemod_limit,
+					     delegation->open_delegation_type);
+		break;
+	default:
+		nfs_inode_set_delegation(inode, cred, delegation->type,
+					 &delegation->stateid,
+					 delegation->pagemod_limit,
+					 delegation->open_delegation_type);
 	}
-	if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0)
+	if (delegation->do_recall)
-		nfs_inode_set_delegation(state->inode,
+		nfs_async_inode_return_delegation(inode, &delegation->stateid);
-				data->owner->so_cred,
-				data->o_res.delegation_type,
-				&data->o_res.delegation,
-				data->o_res.pagemod_limit);
-	else
-		nfs_inode_reclaim_delegation(state->inode,
-				data->owner->so_cred,
-				data->o_res.delegation_type,
-				&data->o_res.delegation,
-				data->o_res.pagemod_limit);
-
-	if (data->o_res.do_recall)
-		nfs_async_inode_return_delegation(state->inode,
-						  &data->o_res.delegation);
 }
 
 /*
@@ -2015,11 +2040,16 @@ _nfs4_opendata_reclaim_to_nfs4_state(struct nfs4_opendata *data)
 	if (ret)
 		return ERR_PTR(ret);
 
-	if (data->o_res.delegation_type != 0)
+	nfs4_process_delegation(state->inode,
-		nfs4_opendata_check_deleg(data, state);
+				data->owner->so_cred,
+				data->o_arg.claim,
+				&data->o_res.delegation);
 
-	if (!update_open_stateid(state, &data->o_res.stateid,
+	if (!(data->o_res.rflags & NFS4_OPEN_RESULT_NO_OPEN_STATEID)) {
-				NULL, data->o_arg.fmode))
+		if (!update_open_stateid(state, &data->o_res.stateid,
+					 NULL, data->o_arg.fmode))
+			return ERR_PTR(-EAGAIN);
+	} else if (!update_open_stateid(state, NULL, NULL, data->o_arg.fmode))
 		return ERR_PTR(-EAGAIN);
 	refcount_inc(&state->count);
 
@@ -2083,10 +2113,18 @@ _nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data)
 	if (IS_ERR(state))
 		goto out;
 
-	if (data->o_res.delegation_type != 0)
+	nfs4_process_delegation(state->inode,
-		nfs4_opendata_check_deleg(data, state);
+				data->owner->so_cred,
-	if (!update_open_stateid(state, &data->o_res.stateid,
+				data->o_arg.claim,
-				NULL, data->o_arg.fmode)) {
+				&data->o_res.delegation);
+
+	if (!(data->o_res.rflags & NFS4_OPEN_RESULT_NO_OPEN_STATEID)) {
+		if (!update_open_stateid(state, &data->o_res.stateid,
+					 NULL, data->o_arg.fmode)) {
+			nfs4_put_open_state(state);
+			state = ERR_PTR(-EAGAIN);
+		}
+	} else if (!update_open_stateid(state, NULL, NULL, data->o_arg.fmode)) {
 		nfs4_put_open_state(state);
 		state = ERR_PTR(-EAGAIN);
 	}
@@ -2222,7 +2260,7 @@ static int _nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state
 {
 	struct nfs_delegation *delegation;
 	struct nfs4_opendata *opendata;
-	fmode_t delegation_type = 0;
+	u32 delegation_type = NFS4_OPEN_DELEGATE_NONE;
 	int status;
 
 	opendata = nfs4_open_recoverdata_alloc(ctx, state,
@@ -2231,8 +2269,20 @@ static int _nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state
 		return PTR_ERR(opendata);
 	rcu_read_lock();
 	delegation = rcu_dereference(NFS_I(state->inode)->delegation);
-	if (delegation != NULL && test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) != 0)
+	if (delegation != NULL && test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) != 0) {
-		delegation_type = delegation->type;
+		switch(delegation->type) {
+		case FMODE_READ:
+			delegation_type = NFS4_OPEN_DELEGATE_READ;
+			if (test_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags))
+				delegation_type = NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG;
+			break;
+		case FMODE_WRITE:
+		case FMODE_READ|FMODE_WRITE:
+			delegation_type = NFS4_OPEN_DELEGATE_WRITE;
+			if (test_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags))
+				delegation_type = NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG;
+		}
+	}
 	rcu_read_unlock();
 	opendata->o_arg.u.delegation_type = delegation_type;
 	status = nfs4_open_recover(opendata, state);
@@ -2825,16 +2875,16 @@ static int nfs40_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st
 }
 
 static int nfs40_test_and_free_expired_stateid(struct nfs_server *server,
-		nfs4_stateid *stateid,
+					       const nfs4_stateid *stateid,
-		const struct cred *cred)
+					       const struct cred *cred)
 {
 	return -NFS4ERR_BAD_STATEID;
 }
 
 #if defined(CONFIG_NFS_V4_1)
 static int nfs41_test_and_free_expired_stateid(struct nfs_server *server,
-		nfs4_stateid *stateid,
+					       const nfs4_stateid *stateid,
-		const struct cred *cred)
+					       const struct cred *cred)
 {
 	int status;
 
@@ -3111,7 +3161,7 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,
 	case NFS4_OPEN_CLAIM_DELEGATE_PREV:
 		if (!opendata->rpc_done)
 			break;
-		if (opendata->o_res.delegation_type != 0)
+		if (opendata->o_res.delegation.type != 0)
 			dir_verifier = nfs_save_change_attribute(dir);
 		nfs_set_verifier(dentry, dir_verifier);
 	}
@@ -3394,7 +3444,8 @@ static int nfs4_do_setattr(struct inode *inode, const struct cred *cred,
 		.inode = inode,
 		.stateid = &arg.stateid,
 	};
-	unsigned long adjust_flags = NFS_INO_INVALID_CHANGE;
+	unsigned long adjust_flags = NFS_INO_INVALID_CHANGE |
+				     NFS_INO_INVALID_CTIME;
 	int err;
 
 	if (sattr->ia_valid & (ATTR_MODE | ATTR_KILL_SUID | ATTR_KILL_SGID))
@@ -3700,7 +3751,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 
 	if (calldata->arg.fmode == 0 || calldata->arg.fmode == FMODE_READ) {
 		/* Close-to-open cache consistency revalidation */
-		if (!nfs4_have_delegation(inode, FMODE_READ)) {
+		if (!nfs4_have_delegation(inode, FMODE_READ, 0)) {
 			nfs4_bitmask_set(calldata->arg.bitmask_store,
 					 server->cache_consistency_bitmask,
 					 inode, 0);
@@ -3710,8 +3761,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 	}
 
 	calldata->arg.share_access =
-		nfs4_map_atomic_open_share(NFS_SERVER(inode),
+		nfs4_fmode_to_share_access(calldata->arg.fmode);
-				calldata->arg.fmode, 0);
 
 	if (calldata->res.fattr == NULL)
 		calldata->arg.bitmask = NULL;
@@ -3852,11 +3902,14 @@ static void nfs4_close_context(struct nfs_open_context *ctx, int is_sync)
 
 #define FATTR4_WORD1_NFS40_MASK (2*FATTR4_WORD1_MOUNTED_ON_FILEID - 1UL)
 #define FATTR4_WORD2_NFS41_MASK (2*FATTR4_WORD2_SUPPATTR_EXCLCREAT - 1UL)
-#define FATTR4_WORD2_NFS42_MASK (2*FATTR4_WORD2_XATTR_SUPPORT - 1UL)
+#define FATTR4_WORD2_NFS42_MASK (2*FATTR4_WORD2_OPEN_ARGUMENTS - 1UL)
 
 static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
 {
-	u32 bitmask[3] = {}, minorversion = server->nfs_client->cl_minorversion;
+	u32 minorversion = server->nfs_client->cl_minorversion;
+	u32 bitmask[3] = {
+		[0] = FATTR4_WORD0_SUPPORTED_ATTRS,
+	};
 	struct nfs4_server_caps_arg args = {
 		.fhandle = fhandle,
 		.bitmask = bitmask,
@@ -3882,6 +3935,14 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
 
 	status = nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0);
 	if (status == 0) {
+		bitmask[0] = (FATTR4_WORD0_SUPPORTED_ATTRS |
+			      FATTR4_WORD0_FH_EXPIRE_TYPE |
+			      FATTR4_WORD0_LINK_SUPPORT |
+			      FATTR4_WORD0_SYMLINK_SUPPORT |
+			      FATTR4_WORD0_ACLSUPPORT |
+			      FATTR4_WORD0_CASE_INSENSITIVE |
+			      FATTR4_WORD0_CASE_PRESERVING) &
+			     res.attr_bitmask[0];
 		/* Sanity check the server answers */
 		switch (minorversion) {
 		case 0:
@@ -3890,9 +3951,14 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
 			break;
 		case 1:
 			res.attr_bitmask[2] &= FATTR4_WORD2_NFS41_MASK;
+			bitmask[2] = FATTR4_WORD2_SUPPATTR_EXCLCREAT &
+				     res.attr_bitmask[2];
 			break;
 		case 2:
 			res.attr_bitmask[2] &= FATTR4_WORD2_NFS42_MASK;
+			bitmask[2] = (FATTR4_WORD2_SUPPATTR_EXCLCREAT |
+				      FATTR4_WORD2_OPEN_ARGUMENTS) &
+				     res.attr_bitmask[2];
 		}
 		memcpy(server->attr_bitmask, res.attr_bitmask, sizeof(server->attr_bitmask));
 		server->caps &= ~(NFS_CAP_ACLS | NFS_CAP_HARDLINKS |
@@ -3915,6 +3981,8 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
 #endif
 		if (res.attr_bitmask[0] & FATTR4_WORD0_FS_LOCATIONS)
 			server->caps |= NFS_CAP_FS_LOCATIONS;
+		if (res.attr_bitmask[2] & FATTR4_WORD2_TIME_DELEG_MODIFY)
+			server->caps |= NFS_CAP_DELEGTIME;
 		if (!(res.attr_bitmask[0] & FATTR4_WORD0_FILEID))
 			server->fattr_valid &= ~NFS_ATTR_FATTR_FILEID;
 		if (!(res.attr_bitmask[1] & FATTR4_WORD1_MODE))
@@ -3939,6 +4007,10 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
 				sizeof(server->attr_bitmask));
 		server->attr_bitmask_nl[2] &= ~FATTR4_WORD2_SECURITY_LABEL;
 
+		if (res.open_caps.oa_share_access_want[0] &
+		    NFS4_SHARE_WANT_OPEN_XOR_DELEGATION)
+			server->caps |= NFS_CAP_OPEN_XOR;
+
 		memcpy(server->cache_consistency_bitmask, res.attr_bitmask, sizeof(server->cache_consistency_bitmask));
 		server->cache_consistency_bitmask[0] &= FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE;
 		server->cache_consistency_bitmask[1] &= FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY;
@@ -4638,7 +4710,7 @@ static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry
 	};
 	int status = 0;
 
-	if (!nfs4_have_delegation(inode, FMODE_READ)) {
+	if (!nfs4_have_delegation(inode, FMODE_READ, 0)) {
 		res.fattr = nfs_alloc_fattr();
 		if (res.fattr == NULL)
 			return -ENOMEM;
@@ -4956,8 +5028,9 @@ static int _nfs4_proc_link(struct inode *inode, struct inode *dir, const struct
 		goto out;
 
 	nfs4_inode_make_writeable(inode);
-	nfs4_bitmap_copy_adjust(bitmask, nfs4_bitmask(server, res.fattr->label), inode,
+	nfs4_bitmap_copy_adjust(bitmask, nfs4_bitmask(server, res.fattr->label),
-				NFS_INO_INVALID_CHANGE);
+				inode,
+				NFS_INO_INVALID_CHANGE | NFS_INO_INVALID_CTIME);
 	status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
 	if (!status) {
 		nfs4_update_changeattr(dir, &res.cinfo, res.fattr->time_start,
@@ -5607,7 +5680,7 @@ bool nfs4_write_need_cache_consistency_data(struct nfs_pgio_header *hdr)
 	/* Otherwise, request attributes if and only if we don't hold
 	 * a delegation
 	 */
-	return nfs4_have_delegation(hdr->inode, FMODE_READ) == 0;
+	return nfs4_have_delegation(hdr->inode, FMODE_READ, 0) == 0;
 }
 
 void nfs4_bitmask_set(__u32 bitmask[], const __u32 src[],
@@ -6575,6 +6648,7 @@ struct nfs4_delegreturndata {
 		u32 roc_barrier;
 		bool roc;
 	} lr;
+	struct nfs4_delegattr sattr;
 	struct nfs_fattr fattr;
 	int rpc_status;
 	struct inode *inode;
@@ -6599,6 +6673,30 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
 			  &data->res.lr_ret) == -EAGAIN)
 		goto out_restart;
 
+	if (data->args.sattr_args && task->tk_status != 0) {
+		switch(data->res.sattr_ret) {
+		case 0:
+			data->args.sattr_args = NULL;
+			data->res.sattr_res = false;
+			break;
+		case -NFS4ERR_ADMIN_REVOKED:
+		case -NFS4ERR_DELEG_REVOKED:
+		case -NFS4ERR_EXPIRED:
+		case -NFS4ERR_BAD_STATEID:
+			/* Let the main handler below do stateid recovery */
+			break;
+		case -NFS4ERR_OLD_STATEID:
+			if (nfs4_refresh_delegation_stateid(&data->stateid,
+						data->inode))
+				goto out_restart;
+			fallthrough;
+		default:
+			data->args.sattr_args = NULL;
+			data->res.sattr_res = false;
+			goto out_restart;
+		}
+	}
+
 	switch (task->tk_status) {
 	case 0:
 		renew_lease(data->res.server, data->timestamp);
@@ -6692,7 +6790,10 @@ static const struct rpc_call_ops nfs4_delegreturn_ops = {
 	.rpc_release = nfs4_delegreturn_release,
 };
 
-static int _nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred, const nfs4_stateid *stateid, int issync)
+static int _nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
+				  const nfs4_stateid *stateid,
+				  struct nfs_delegation *delegation,
+				  int issync)
 {
 	struct nfs4_delegreturndata *data;
 	struct nfs_server *server = NFS_SERVER(inode);
@@ -6744,12 +6845,27 @@ static int _nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
 		}
 	}
 
+	if (delegation &&
+	    test_bit(NFS_DELEGATION_DELEGTIME, &delegation->flags)) {
+		if (delegation->type & FMODE_READ) {
+			data->sattr.atime = inode_get_atime(inode);
+			data->sattr.atime_set = true;
+		}
+		if (delegation->type & FMODE_WRITE) {
+			data->sattr.mtime = inode_get_mtime(inode);
+			data->sattr.mtime_set = true;
+		}
+		data->args.sattr_args = &data->sattr;
+		data->res.sattr_res = true;
+	}
+
 	if (!data->inode)
 		nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1,
 				   1);
 	else
 		nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1,
 				   0);
+
 	task_setup_data.callback_data = data;
 	msg.rpc_argp = &data->args;
 	msg.rpc_resp = &data->res;
@@ -6767,13 +6883,16 @@ out:
 	return status;
 }
 
-int nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred, const nfs4_stateid *stateid, int issync)
+int nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
+			  const nfs4_stateid *stateid,
+			  struct nfs_delegation *delegation, int issync)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs4_exception exception = { };
 	int err;
 	do {
-		err = _nfs4_proc_delegreturn(inode, cred, stateid, issync);
+		err = _nfs4_proc_delegreturn(inode, cred, stateid,
+					     delegation, issync);
 		trace_nfs4_delegreturn(inode, stateid, err);
 		switch (err) {
 			case -NFS4ERR_STALE_STATEID:
@@ -7629,10 +7748,10 @@ static int nfs4_add_lease(struct file *file, int arg, struct file_lease **lease,
 	int ret;
 
 	/* No delegation, no lease */
-	if (!nfs4_have_delegation(inode, type))
+	if (!nfs4_have_delegation(inode, type, 0))
 		return -EAGAIN;
 	ret = generic_setlease(file, arg, lease, priv);
-	if (ret || nfs4_have_delegation(inode, type))
+	if (ret || nfs4_have_delegation(inode, type, 0))
 		return ret;
 	/* We raced with a delegation return */
 	nfs4_delete_lease(file, priv);
@@ -8840,7 +8959,7 @@ nfs4_run_exchange_id(struct nfs_client *clp, const struct cred *cred,
 #ifdef CONFIG_NFS_V4_1_MIGRATION
 	calldata->args.flags |= EXCHGID4_FLAG_SUPP_MOVED_MIGR;
 #endif
-	if (test_bit(NFS_CS_DS, &clp->cl_flags))
+	if (test_bit(NFS_CS_PNFS, &clp->cl_flags))
 		calldata->args.flags |= EXCHGID4_FLAG_USE_PNFS_DS;
 	msg.rpc_argp = &calldata->args;
 	msg.rpc_resp = &calldata->res;
@@ -9854,6 +9973,11 @@ static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
 	if (!nfs41_sequence_process(task, &lrp->res.seq_res))
 		return;
 
+	if (task->tk_rpc_status == -ETIMEDOUT) {
+		lrp->rpc_status = -EAGAIN;
+		lrp->res.lrs_present = 0;
+		return;
+	}
 	/*
 	 * Was there an RPC level error? Assume the call succeeded,
 	 * and that we need to release the layout
@@ -9876,6 +10000,15 @@ static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
 		fallthrough;
 	case 0:
 		break;
+	case -NFS4ERR_BADSESSION:
+	case -NFS4ERR_DEADSESSION:
+	case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
+		nfs4_schedule_session_recovery(server->nfs_client->cl_session,
+					       task->tk_status);
+		lrp->res.lrs_present = 0;
+		lrp->rpc_status = -EAGAIN;
+		task->tk_status = 0;
+		break;
 	case -NFS4ERR_DELAY:
 		if (nfs4_async_handle_error(task, server, NULL, NULL) != -EAGAIN)
 			break;
@@ -9893,8 +10026,13 @@ static void nfs4_layoutreturn_release(void *calldata)
 	struct nfs4_layoutreturn *lrp = calldata;
 	struct pnfs_layout_hdr *lo = lrp->args.layout;
 
-	pnfs_layoutreturn_free_lsegs(lo, &lrp->args.stateid, &lrp->args.range,
+	if (lrp->rpc_status == 0 || !lrp->inode)
+		pnfs_layoutreturn_free_lsegs(
+			lo, &lrp->args.stateid, &lrp->args.range,
 			lrp->res.lrs_present ? &lrp->res.stateid : NULL);
+	else
+		pnfs_layoutreturn_retry_later(lo, &lrp->args.stateid,
+					      &lrp->args.range);
 	nfs4_sequence_free_slot(&lrp->res.seq_res);
 	if (lrp->ld_private.ops && lrp->ld_private.ops->free)
 		lrp->ld_private.ops->free(&lrp->ld_private);
@@ -9910,7 +10048,7 @@ static const struct rpc_call_ops nfs4_layoutreturn_call_ops = {
 	.rpc_release = nfs4_layoutreturn_release,
 };
 
-int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
+int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, unsigned int flags)
 {
 	struct rpc_task *task;
 	struct rpc_message msg = {
@@ -9933,7 +10071,7 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
 			&task_setup_data.rpc_client, &msg);
 
 	lrp->inode = nfs_igrab_and_active(lrp->args.inode);
-	if (!sync) {
+	if (flags & PNFS_FL_LAYOUTRETURN_ASYNC) {
 		if (!lrp->inode) {
 			nfs4_layoutreturn_release(lrp);
 			return -EAGAIN;
@@ -9941,6 +10079,8 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
 		task_setup_data.flags |= RPC_TASK_ASYNC;
 	}
 	if (!lrp->inode)
+		flags |= PNFS_FL_LAYOUTRETURN_PRIVILEGED;
+	if (flags & PNFS_FL_LAYOUTRETURN_PRIVILEGED)
 		nfs4_init_sequence(&lrp->args.seq_args, &lrp->res.seq_res, 1,
 				   1);
 	else
@@ -9949,7 +10089,7 @@ int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync)
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return PTR_ERR(task);
-	if (sync)
+	if (!(flags & PNFS_FL_LAYOUTRETURN_ASYNC))
 		status = task->tk_status;
 	trace_nfs4_layoutreturn(lrp->args.inode, &lrp->args.stateid, status);
 	dprintk("<-- %s status=%d\n", __func__, status);
@@ -10267,12 +10407,12 @@ out:
 }
 
 static int _nfs41_test_stateid(struct nfs_server *server,
-		nfs4_stateid *stateid,
+			       const nfs4_stateid *stateid,
-		const struct cred *cred)
+			       const struct cred *cred)
 {
 	int status;
 	struct nfs41_test_stateid_args args = {
-		.stateid = stateid,
+		.stateid = *stateid,
 	};
 	struct nfs41_test_stateid_res res;
 	struct rpc_message msg = {
@@ -10328,8 +10468,8 @@ static void nfs4_handle_delay_or_session_error(struct nfs_server *server,
  * failed or the state ID is not currently valid.
  */
 static int nfs41_test_stateid(struct nfs_server *server,
-		nfs4_stateid *stateid,
+			      const nfs4_stateid *stateid,
-		const struct cred *cred)
+			      const struct cred *cred)
 {
 	struct nfs4_exception exception = {
 		.interruptible = true,
@@ -10759,6 +10899,7 @@ const struct nfs_rpc_ops nfs_v4_clientops = {
 	.close_context  = nfs4_close_context,
 	.open_context	= nfs4_atomic_open,
 	.have_delegation = nfs4_have_delegation,
+	.return_delegation = nfs4_inode_return_delegation,
 	.alloc_client	= nfs4_alloc_client,
 	.init_client	= nfs4_init_client,
 	.free_client	= nfs4_free_client,

fs/nfs/nfs4state.c

@@ -1863,6 +1863,7 @@ static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp)
 
 	if (!nfs4_state_clear_reclaim_reboot(clp))
 		return;
+	pnfs_destroy_all_layouts(clp);
 	ops = clp->cl_mvops->reboot_recovery_ops;
 	cred = nfs4_get_clid_cred(clp);
 	err = nfs4_reclaim_complete(clp, ops, cred);
@@ -2068,7 +2069,6 @@ static int nfs4_establish_lease(struct nfs_client *clp)
 	put_cred(cred);
 	if (status != 0)
 		return status;
-	pnfs_destroy_all_layouts(clp);
 	return 0;
 }
 
@@ -2680,6 +2680,8 @@ static void nfs4_state_manager(struct nfs_client *clp)
 			section = "reclaim reboot";
 			status = nfs4_do_reclaim(clp,
 				clp->cl_mvops->reboot_recovery_ops);
+			if (status == 0)
+				status = pnfs_layout_handle_reboot(clp);
 			if (status == -EAGAIN)
 				continue;
 			if (status < 0)

fs/nfs/nfs4super.c

@@ -332,6 +332,7 @@ static void __exit exit_nfs_v4(void)
 	nfs_dns_resolver_destroy();
 }
 
+MODULE_DESCRIPTION("NFSv4 client support");
 MODULE_LICENSE("GPL");
 
 module_init(init_nfs_v4);

fs/nfs/nfs4trace.c

@@ -2,6 +2,8 @@
 /*
  * Copyright (c) 2013 Trond Myklebust <Trond.Myklebust@netapp.com>
  */
+#include <uapi/linux/pr.h>
+#include <linux/blkdev.h>
 #include <linux/nfs_fs.h>
 #include "nfs4_fs.h"
 #include "internal.h"
@@ -29,5 +31,10 @@ EXPORT_TRACEPOINT_SYMBOL_GPL(ff_layout_read_error);
 EXPORT_TRACEPOINT_SYMBOL_GPL(ff_layout_write_error);
 EXPORT_TRACEPOINT_SYMBOL_GPL(ff_layout_commit_error);
 
+EXPORT_TRACEPOINT_SYMBOL_GPL(bl_pr_key_reg);
+EXPORT_TRACEPOINT_SYMBOL_GPL(bl_pr_key_reg_err);
+EXPORT_TRACEPOINT_SYMBOL_GPL(bl_pr_key_unreg);
+EXPORT_TRACEPOINT_SYMBOL_GPL(bl_pr_key_unreg_err);
+
 EXPORT_TRACEPOINT_SYMBOL_GPL(fl_getdevinfo);
 #endif

fs/nfs/nfs4trace.h

@@ -2153,6 +2153,94 @@ TRACE_EVENT(ff_layout_commit_error,
 		)
 );
 
+DECLARE_EVENT_CLASS(pnfs_bl_pr_key_class,
+	TP_PROTO(
+		const struct block_device *bdev,
+		u64 key
+	),
+	TP_ARGS(bdev, key),
+	TP_STRUCT__entry(
+		__field(u64, key)
+		__field(dev_t, dev)
+		__string(device, bdev->bd_disk->disk_name)
+	),
+	TP_fast_assign(
+		__entry->key = key;
+		__entry->dev = bdev->bd_dev;
+		__assign_str(device);
+	),
+	TP_printk("dev=%d,%d (%s) key=0x%016llx",
+		MAJOR(__entry->dev), MINOR(__entry->dev),
+		__get_str(device), __entry->key
+	)
+);
+
+#define DEFINE_NFS4_BLOCK_PRKEY_EVENT(name) \
+	DEFINE_EVENT(pnfs_bl_pr_key_class, name, \
+		TP_PROTO( \
+			const struct block_device *bdev, \
+			u64 key \
+		), \
+		TP_ARGS(bdev, key))
+DEFINE_NFS4_BLOCK_PRKEY_EVENT(bl_pr_key_reg);
+DEFINE_NFS4_BLOCK_PRKEY_EVENT(bl_pr_key_unreg);
+
+/*
+ * From uapi/linux/pr.h
+ */
+TRACE_DEFINE_ENUM(PR_STS_SUCCESS);
+TRACE_DEFINE_ENUM(PR_STS_IOERR);
+TRACE_DEFINE_ENUM(PR_STS_RESERVATION_CONFLICT);
+TRACE_DEFINE_ENUM(PR_STS_RETRY_PATH_FAILURE);
+TRACE_DEFINE_ENUM(PR_STS_PATH_FAST_FAILED);
+TRACE_DEFINE_ENUM(PR_STS_PATH_FAILED);
+
+#define show_pr_status(x) \
+	__print_symbolic(x, \
+		{ PR_STS_SUCCESS,		"SUCCESS" }, \
+		{ PR_STS_IOERR,			"IOERR" }, \
+		{ PR_STS_RESERVATION_CONFLICT,	"RESERVATION_CONFLICT" }, \
+		{ PR_STS_RETRY_PATH_FAILURE,	"RETRY_PATH_FAILURE" }, \
+		{ PR_STS_PATH_FAST_FAILED,	"PATH_FAST_FAILED" }, \
+		{ PR_STS_PATH_FAILED,		"PATH_FAILED" })
+
+DECLARE_EVENT_CLASS(pnfs_bl_pr_key_err_class,
+	TP_PROTO(
+		const struct block_device *bdev,
+		u64 key,
+		int status
+	),
+	TP_ARGS(bdev, key, status),
+	TP_STRUCT__entry(
+		__field(u64, key)
+		__field(dev_t, dev)
+		__field(unsigned long, status)
+		__string(device, bdev->bd_disk->disk_name)
+	),
+	TP_fast_assign(
+		__entry->key = key;
+		__entry->dev = bdev->bd_dev;
+		__entry->status = status;
+		__assign_str(device);
+	),
+	TP_printk("dev=%d,%d (%s) key=0x%016llx status=%s",
+		MAJOR(__entry->dev), MINOR(__entry->dev),
+		__get_str(device), __entry->key,
+		show_pr_status(__entry->status)
+	)
+);
+
+#define DEFINE_NFS4_BLOCK_PRKEY_ERR_EVENT(name) \
+	DEFINE_EVENT(pnfs_bl_pr_key_err_class, name, \
+		TP_PROTO( \
+			const struct block_device *bdev, \
+			u64 key, \
+			int status \
+		), \
+		TP_ARGS(bdev, key, status))
+DEFINE_NFS4_BLOCK_PRKEY_ERR_EVENT(bl_pr_key_reg_err);
+DEFINE_NFS4_BLOCK_PRKEY_ERR_EVENT(bl_pr_key_unreg_err);
+
 #ifdef CONFIG_NFS_V4_2
 TRACE_DEFINE_ENUM(NFS4_CONTENT_DATA);
 TRACE_DEFINE_ENUM(NFS4_CONTENT_HOLE);

fs/nfs/nfs4xdr.c

@@ -224,6 +224,11 @@ static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req,
 				 encode_attrs_maxsz)
 #define decode_setattr_maxsz	(op_decode_hdr_maxsz + \
 				 nfs4_fattr_bitmap_maxsz)
+#define encode_delegattr_maxsz	(op_encode_hdr_maxsz + \
+				 encode_stateid_maxsz + \
+				nfs4_fattr_bitmap_maxsz + \
+				2*nfstime4_maxsz)
+#define decode_delegattr_maxsz	(decode_setattr_maxsz)
 #define encode_read_maxsz	(op_encode_hdr_maxsz + \
 				 encode_stateid_maxsz + 3)
 #define decode_read_maxsz	(op_decode_hdr_maxsz + 2 + pagepad_maxsz)
@@ -758,12 +763,14 @@ static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req,
 				encode_sequence_maxsz + \
 				encode_putfh_maxsz + \
 				encode_layoutreturn_maxsz + \
+				encode_delegattr_maxsz + \
 				encode_delegreturn_maxsz + \
 				encode_getattr_maxsz)
 #define NFS4_dec_delegreturn_sz (compound_decode_hdr_maxsz + \
 				decode_sequence_maxsz + \
 				decode_putfh_maxsz + \
 				decode_layoutreturn_maxsz + \
+				decode_delegattr_maxsz + \
 				decode_delegreturn_maxsz + \
 				decode_getattr_maxsz)
 #define NFS4_enc_getacl_sz	(compound_encode_hdr_maxsz + \
@@ -1060,9 +1067,10 @@ static void encode_nops(struct compound_hdr *hdr)
 	*hdr->nops_p = htonl(hdr->nops);
 }
 
-static void encode_nfs4_stateid(struct xdr_stream *xdr, const nfs4_stateid *stateid)
+static void encode_nfs4_stateid(struct xdr_stream *xdr,
+				const nfs4_stateid *stateid)
 {
-	encode_opaque_fixed(xdr, stateid, NFS4_STATEID_SIZE);
+	encode_opaque_fixed(xdr, stateid->data, NFS4_STATEID_SIZE);
 }
 
 static void encode_nfs4_verifier(struct xdr_stream *xdr, const nfs4_verifier *verf)
@@ -1468,20 +1476,18 @@ static void encode_opentype(struct xdr_stream *xdr, const struct nfs_openargs *a
 	}
 }
 
-static inline void encode_delegation_type(struct xdr_stream *xdr, fmode_t delegation_type)
+static inline void encode_delegation_type(struct xdr_stream *xdr, u32 delegation_type)
 {
 	__be32 *p;
 
 	p = reserve_space(xdr, 4);
 	switch (delegation_type) {
-	case 0:
+	case NFS4_OPEN_DELEGATE_NONE:
-		*p = cpu_to_be32(NFS4_OPEN_DELEGATE_NONE);
+	case NFS4_OPEN_DELEGATE_READ:
-		break;
+	case NFS4_OPEN_DELEGATE_WRITE:
-	case FMODE_READ:
+	case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
-		*p = cpu_to_be32(NFS4_OPEN_DELEGATE_READ);
+	case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
-		break;
+		*p = cpu_to_be32(delegation_type);
-	case FMODE_WRITE|FMODE_READ:
-		*p = cpu_to_be32(NFS4_OPEN_DELEGATE_WRITE);
 		break;
 	default:
 		BUG();
@@ -1497,7 +1503,7 @@ static inline void encode_claim_null(struct xdr_stream *xdr, const struct qstr *
 	encode_string(xdr, name->len, name->name);
 }
 
-static inline void encode_claim_previous(struct xdr_stream *xdr, fmode_t type)
+static inline void encode_claim_previous(struct xdr_stream *xdr, u32 type)
 {
 	__be32 *p;
 
@@ -1735,6 +1741,33 @@ static void encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs
 			server->attr_bitmask);
 }
 
+static void encode_delegattr(struct xdr_stream *xdr,
+		const nfs4_stateid *stateid,
+		const struct nfs4_delegattr *attr,
+		struct compound_hdr *hdr)
+{
+	uint32_t bitmap[3] = { 0 };
+	uint32_t len = 0;
+	__be32 *p;
+
+	encode_op_hdr(xdr, OP_SETATTR, encode_delegattr_maxsz, hdr);
+	encode_nfs4_stateid(xdr, stateid);
+	if (attr->atime_set) {
+		bitmap[2] |= FATTR4_WORD2_TIME_DELEG_ACCESS;
+		len += (nfstime4_maxsz << 2);
+	}
+	if (attr->mtime_set) {
+		bitmap[2] |= FATTR4_WORD2_TIME_DELEG_MODIFY;
+		len += (nfstime4_maxsz << 2);
+	}
+	xdr_encode_bitmap4(xdr, bitmap, ARRAY_SIZE(bitmap));
+	xdr_stream_encode_opaque_inline(xdr, (void **)&p, len);
+	if (bitmap[2] & FATTR4_WORD2_TIME_DELEG_ACCESS)
+		p = xdr_encode_nfstime4(p, &attr->atime);
+	if (bitmap[2] & FATTR4_WORD2_TIME_DELEG_MODIFY)
+		p = xdr_encode_nfstime4(p, &attr->mtime);
+}
+
 static void encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclientid *setclientid, struct compound_hdr *hdr)
 {
 	__be32 *p;
@@ -2105,7 +2138,7 @@ static void encode_test_stateid(struct xdr_stream *xdr,
 {
 	encode_op_hdr(xdr, OP_TEST_STATEID, decode_test_stateid_maxsz, hdr);
 	encode_uint32(xdr, 1);
-	encode_nfs4_stateid(xdr, args->stateid);
+	encode_nfs4_stateid(xdr, &args->stateid);
 }
 
 static void encode_free_stateid(struct xdr_stream *xdr,
@@ -2812,6 +2845,8 @@ static void nfs4_xdr_enc_delegreturn(struct rpc_rqst *req,
 	encode_putfh(xdr, args->fhandle, &hdr);
 	if (args->lr_args)
 		encode_layoutreturn(xdr, args->lr_args, &hdr);
+	if (args->sattr_args)
+		encode_delegattr(xdr, args->stateid, args->sattr_args, &hdr);
 	if (args->bitmask)
 		encode_getfattr(xdr, args->bitmask, &hdr);
 	encode_delegreturn(xdr, args->stateid, &hdr);
@@ -4303,6 +4338,28 @@ static int decode_attr_xattrsupport(struct xdr_stream *xdr, uint32_t *bitmap,
 	return 0;
 }
 
+static int decode_attr_open_arguments(struct xdr_stream *xdr, uint32_t *bitmap,
+		struct nfs4_open_caps *res)
+{
+	memset(res, 0, sizeof(*res));
+	if (unlikely(bitmap[2] & (FATTR4_WORD2_OPEN_ARGUMENTS - 1U)))
+		return -EIO;
+	if (likely(bitmap[2] & FATTR4_WORD2_OPEN_ARGUMENTS)) {
+		if (decode_bitmap4(xdr, res->oa_share_access, ARRAY_SIZE(res->oa_share_access)) < 0)
+			return -EIO;
+		if (decode_bitmap4(xdr, res->oa_share_deny, ARRAY_SIZE(res->oa_share_deny)) < 0)
+			return -EIO;
+		if (decode_bitmap4(xdr, res->oa_share_access_want, ARRAY_SIZE(res->oa_share_access_want)) < 0)
+			return -EIO;
+		if (decode_bitmap4(xdr, res->oa_open_claim, ARRAY_SIZE(res->oa_open_claim)) < 0)
+			return -EIO;
+		if (decode_bitmap4(xdr, res->oa_createmode, ARRAY_SIZE(res->oa_createmode)) < 0)
+			return -EIO;
+		bitmap[2] &= ~FATTR4_WORD2_OPEN_ARGUMENTS;
+	}
+	return 0;
+}
+
 static int verify_attr_len(struct xdr_stream *xdr, unsigned int savep, uint32_t attrlen)
 {
 	unsigned int attrwords = XDR_QUADLEN(attrlen);
@@ -4477,6 +4534,8 @@ static int decode_server_caps(struct xdr_stream *xdr, struct nfs4_server_caps_re
 	if ((status = decode_attr_exclcreat_supported(xdr, bitmap,
 				res->exclcreat_bitmask)) != 0)
 		goto xdr_error;
+	if ((status = decode_attr_open_arguments(xdr, bitmap, &res->open_caps)) != 0)
+		goto xdr_error;
 	status = verify_attr_len(xdr, savep, attrlen);
 xdr_error:
 	dprintk("%s: xdr returned %d!\n", __func__, -status);
@@ -5148,13 +5207,12 @@ static int decode_space_limit(struct xdr_stream *xdr,
 }
 
 static int decode_rw_delegation(struct xdr_stream *xdr,
-		uint32_t delegation_type,
+		struct nfs4_open_delegation *res)
-		struct nfs_openres *res)
 {
 	__be32 *p;
 	int status;
 
-	status = decode_delegation_stateid(xdr, &res->delegation);
+	status = decode_delegation_stateid(xdr, &res->stateid);
 	if (unlikely(status))
 		return status;
 	p = xdr_inline_decode(xdr, 4);
@@ -5162,52 +5220,57 @@ static int decode_rw_delegation(struct xdr_stream *xdr,
 		return -EIO;
 	res->do_recall = be32_to_cpup(p);
 
-	switch (delegation_type) {
+	switch (res->open_delegation_type) {
 	case NFS4_OPEN_DELEGATE_READ:
-		res->delegation_type = FMODE_READ;
+	case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
+		res->type = FMODE_READ;
 		break;
 	case NFS4_OPEN_DELEGATE_WRITE:
-		res->delegation_type = FMODE_WRITE|FMODE_READ;
+	case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
+		res->type = FMODE_WRITE|FMODE_READ;
 		if (decode_space_limit(xdr, &res->pagemod_limit) < 0)
 				return -EIO;
 	}
 	return decode_ace(xdr, NULL);
 }
 
-static int decode_no_delegation(struct xdr_stream *xdr, struct nfs_openres *res)
+static int decode_no_delegation(struct xdr_stream *xdr,
+		struct nfs4_open_delegation *res)
 {
 	__be32 *p;
-	uint32_t why_no_delegation;
 
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		return -EIO;
-	why_no_delegation = be32_to_cpup(p);
+	res->why_no_delegation = be32_to_cpup(p);
-	switch (why_no_delegation) {
+	switch (res->why_no_delegation) {
 		case WND4_CONTENTION:
 		case WND4_RESOURCE:
-			xdr_inline_decode(xdr, 4);
+			p = xdr_inline_decode(xdr, 4);
-			/* Ignore for now */
+			if (unlikely(!p))
+				return -EIO;
+			res->will_notify = be32_to_cpup(p);
 	}
 	return 0;
 }
 
-static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res)
+static int decode_delegation(struct xdr_stream *xdr,
+		struct nfs4_open_delegation *res)
 {
 	__be32 *p;
-	uint32_t delegation_type;
 
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		return -EIO;
-	delegation_type = be32_to_cpup(p);
+	res->open_delegation_type = be32_to_cpup(p);
-	res->delegation_type = 0;
+	switch (res->open_delegation_type) {
-	switch (delegation_type) {
 	case NFS4_OPEN_DELEGATE_NONE:
 		return 0;
 	case NFS4_OPEN_DELEGATE_READ:
 	case NFS4_OPEN_DELEGATE_WRITE:
-		return decode_rw_delegation(xdr, delegation_type, res);
+	case NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG:
+	case NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG:
+		return decode_rw_delegation(xdr, res);
 	case NFS4_OPEN_DELEGATE_NONE_EXT:
 		return decode_no_delegation(xdr, res);
 	}
@@ -5248,7 +5311,7 @@ static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res)
 	for (; i < NFS4_BITMAP_SIZE; i++)
 		res->attrset[i] = 0;
 
-	return decode_delegation(xdr, res);
+	return decode_delegation(xdr, &res->delegation);
 xdr_error:
 	dprintk("%s: Bitmap too large! Length = %u\n", __func__, bmlen);
 	return -EIO;
@@ -5480,6 +5543,11 @@ static int decode_setattr(struct xdr_stream *xdr)
 	return -EIO;
 }
 
+static int decode_delegattr(struct xdr_stream *xdr)
+{
+	return decode_setattr(xdr);
+}
+
 static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_setclientid_res *res)
 {
 	__be32 *p;
@@ -7052,6 +7120,12 @@ static int nfs4_xdr_dec_delegreturn(struct rpc_rqst *rqstp,
 		if (status)
 			goto out;
 	}
+	if (res->sattr_res) {
+		status = decode_delegattr(xdr);
+		res->sattr_ret = status;
+		if (status)
+			goto out;
+	}
 	if (res->fattr) {
 		status = decode_getfattr(xdr, res->fattr, res->server);
 		if (status != 0)

fs/nfs/nfstrace.h

@@ -939,10 +939,11 @@ TRACE_EVENT(nfs_sillyrename_unlink,
 DECLARE_EVENT_CLASS(nfs_folio_event,
 		TP_PROTO(
 			const struct inode *inode,
-			struct folio *folio
+			loff_t offset,
+			size_t count
 		),
 
-		TP_ARGS(inode, folio),
+		TP_ARGS(inode, offset, count),
 
 		TP_STRUCT__entry(
 			__field(dev_t, dev)
@@ -950,7 +951,7 @@ DECLARE_EVENT_CLASS(nfs_folio_event,
 			__field(u64, fileid)
 			__field(u64, version)
 			__field(loff_t, offset)
-			__field(u32, count)
+			__field(size_t, count)
 		),
 
 		TP_fast_assign(
@@ -960,13 +961,13 @@ DECLARE_EVENT_CLASS(nfs_folio_event,
 			__entry->fileid = nfsi->fileid;
 			__entry->fhandle = nfs_fhandle_hash(&nfsi->fh);
 			__entry->version = inode_peek_iversion_raw(inode);
-			__entry->offset = folio_file_pos(folio);
+			__entry->offset = offset,
-			__entry->count = nfs_folio_length(folio);
+			__entry->count = count;
 		),
 
 		TP_printk(
 			"fileid=%02x:%02x:%llu fhandle=0x%08x version=%llu "
-			"offset=%lld count=%u",
+			"offset=%lld count=%zu",
 			MAJOR(__entry->dev), MINOR(__entry->dev),
 			(unsigned long long)__entry->fileid,
 			__entry->fhandle, __entry->version,
@@ -978,18 +979,20 @@ DECLARE_EVENT_CLASS(nfs_folio_event,
 	DEFINE_EVENT(nfs_folio_event, name, \
 			TP_PROTO( \
 				const struct inode *inode, \
-				struct folio *folio \
+				loff_t offset, \
+				size_t count \
 			), \
-			TP_ARGS(inode, folio))
+			TP_ARGS(inode, offset, count))
 
 DECLARE_EVENT_CLASS(nfs_folio_event_done,
 		TP_PROTO(
 			const struct inode *inode,
-			struct folio *folio,
+			loff_t offset,
+			size_t count,
 			int ret
 		),
 
-		TP_ARGS(inode, folio, ret),
+		TP_ARGS(inode, offset, count, ret),
 
 		TP_STRUCT__entry(
 			__field(dev_t, dev)
@@ -998,7 +1001,7 @@ DECLARE_EVENT_CLASS(nfs_folio_event_done,
 			__field(u64, fileid)
 			__field(u64, version)
 			__field(loff_t, offset)
-			__field(u32, count)
+			__field(size_t, count)
 		),
 
 		TP_fast_assign(
@@ -1008,14 +1011,14 @@ DECLARE_EVENT_CLASS(nfs_folio_event_done,
 			__entry->fileid = nfsi->fileid;
 			__entry->fhandle = nfs_fhandle_hash(&nfsi->fh);
 			__entry->version = inode_peek_iversion_raw(inode);
-			__entry->offset = folio_file_pos(folio);
+			__entry->offset = offset,
-			__entry->count = nfs_folio_length(folio);
+			__entry->count = count,
 			__entry->ret = ret;
 		),
 
 		TP_printk(
 			"fileid=%02x:%02x:%llu fhandle=0x%08x version=%llu "
-			"offset=%lld count=%u ret=%d",
+			"offset=%lld count=%zu ret=%d",
 			MAJOR(__entry->dev), MINOR(__entry->dev),
 			(unsigned long long)__entry->fileid,
 			__entry->fhandle, __entry->version,
@@ -1027,10 +1030,11 @@ DECLARE_EVENT_CLASS(nfs_folio_event_done,
 	DEFINE_EVENT(nfs_folio_event_done, name, \
 			TP_PROTO( \
 				const struct inode *inode, \
-				struct folio *folio, \
+				loff_t offset, \
+				size_t count, \
 				int ret \
 			), \
-			TP_ARGS(inode, folio, ret))
+			TP_ARGS(inode, offset, count, ret))
 
 DEFINE_NFS_FOLIO_EVENT(nfs_aop_readpage);
 DEFINE_NFS_FOLIO_EVENT_DONE(nfs_aop_readpage_done);

fs/nfs/pagelist.c

@@ -187,102 +187,6 @@ nfs_async_iocounter_wait(struct rpc_task *task, struct nfs_lock_context *l_ctx)
 }
 EXPORT_SYMBOL_GPL(nfs_async_iocounter_wait);
 
-/*
- * nfs_page_lock_head_request - page lock the head of the page group
- * @req: any member of the page group
- */
-struct nfs_page *
-nfs_page_group_lock_head(struct nfs_page *req)
-{
-	struct nfs_page *head = req->wb_head;
-
-	while (!nfs_lock_request(head)) {
-		int ret = nfs_wait_on_request(head);
-		if (ret < 0)
-			return ERR_PTR(ret);
-	}
-	if (head != req)
-		kref_get(&head->wb_kref);
-	return head;
-}
-
-/*
- * nfs_unroll_locks -  unlock all newly locked reqs and wait on @req
- * @head: head request of page group, must be holding head lock
- * @req: request that couldn't lock and needs to wait on the req bit lock
- *
- * This is a helper function for nfs_lock_and_join_requests
- * returns 0 on success, < 0 on error.
- */
-static void
-nfs_unroll_locks(struct nfs_page *head, struct nfs_page *req)
-{
-	struct nfs_page *tmp;
-
-	/* relinquish all the locks successfully grabbed this run */
-	for (tmp = head->wb_this_page ; tmp != req; tmp = tmp->wb_this_page) {
-		if (!kref_read(&tmp->wb_kref))
-			continue;
-		nfs_unlock_and_release_request(tmp);
-	}
-}
-
-/*
- * nfs_page_group_lock_subreq -  try to lock a subrequest
- * @head: head request of page group
- * @subreq: request to lock
- *
- * This is a helper function for nfs_lock_and_join_requests which
- * must be called with the head request and page group both locked.
- * On error, it returns with the page group unlocked.
- */
-static int
-nfs_page_group_lock_subreq(struct nfs_page *head, struct nfs_page *subreq)
-{
-	int ret;
-
-	if (!kref_get_unless_zero(&subreq->wb_kref))
-		return 0;
-	while (!nfs_lock_request(subreq)) {
-		nfs_page_group_unlock(head);
-		ret = nfs_wait_on_request(subreq);
-		if (!ret)
-			ret = nfs_page_group_lock(head);
-		if (ret < 0) {
-			nfs_unroll_locks(head, subreq);
-			nfs_release_request(subreq);
-			return ret;
-		}
-	}
-	return 0;
-}
-
-/*
- * nfs_page_group_lock_subrequests -  try to lock the subrequests
- * @head: head request of page group
- *
- * This is a helper function for nfs_lock_and_join_requests which
- * must be called with the head request locked.
- */
-int nfs_page_group_lock_subrequests(struct nfs_page *head)
-{
-	struct nfs_page *subreq;
-	int ret;
-
-	ret = nfs_page_group_lock(head);
-	if (ret < 0)
-		return ret;
-	/* lock each request in the page group */
-	for (subreq = head->wb_this_page; subreq != head;
-			subreq = subreq->wb_this_page) {
-		ret = nfs_page_group_lock_subreq(head, subreq);
-		if (ret < 0)
-			return ret;
-	}
-	nfs_page_group_unlock(head);
-	return 0;
-}
-
 /*
  * nfs_page_set_headlock - set the request PG_HEADLOCK
  * @req: request that is to be locked
@@ -569,7 +473,7 @@ struct nfs_page *nfs_page_create_from_folio(struct nfs_open_context *ctx,
 
 	if (IS_ERR(l_ctx))
 		return ERR_CAST(l_ctx);
-	ret = nfs_page_create(l_ctx, offset, folio_index(folio), offset, count);
+	ret = nfs_page_create(l_ctx, offset, folio->index, offset, count);
 	if (!IS_ERR(ret)) {
 		nfs_page_assign_folio(ret, folio);
 		nfs_page_group_init(ret, NULL);
@@ -694,25 +598,6 @@ void nfs_release_request(struct nfs_page *req)
 }
 EXPORT_SYMBOL_GPL(nfs_release_request);
 
-/**
- * nfs_wait_on_request - Wait for a request to complete.
- * @req: request to wait upon.
- *
- * Interruptible by fatal signals only.
- * The user is responsible for holding a count on the request.
- */
-int
-nfs_wait_on_request(struct nfs_page *req)
-{
-	if (!test_bit(PG_BUSY, &req->wb_flags))
-		return 0;
-	set_bit(PG_CONTENDED2, &req->wb_flags);
-	smp_mb__after_atomic();
-	return wait_on_bit_io(&req->wb_flags, PG_BUSY,
-			      TASK_UNINTERRUPTIBLE);
-}
-EXPORT_SYMBOL_GPL(nfs_wait_on_request);
-
 /*
  * nfs_generic_pg_test - determine if requests can be coalesced
  * @desc: pointer to descriptor

fs/nfs/pnfs.c

@@ -61,6 +61,7 @@ static void pnfs_free_returned_lsegs(struct pnfs_layout_hdr *lo,
 		u32 seq);
 static bool pnfs_lseg_dec_and_remove_zero(struct pnfs_layout_segment *lseg,
 		                struct list_head *tmp_list);
+static int pnfs_layout_return_on_reboot(struct pnfs_layout_hdr *lo);
 
 /* Return the registered pnfs layout driver module matching given id */
 static struct pnfs_layoutdriver_type *
@@ -476,6 +477,18 @@ pnfs_mark_layout_stateid_invalid(struct pnfs_layout_hdr *lo,
 	return !list_empty(&lo->plh_segs);
 }
 
+static int pnfs_mark_layout_stateid_return(struct pnfs_layout_hdr *lo,
+					   struct list_head *lseg_list,
+					   enum pnfs_iomode iomode, u32 seq)
+{
+	struct pnfs_layout_range range = {
+		.iomode = iomode,
+		.length = NFS4_MAX_UINT64,
+	};
+
+	return pnfs_mark_matching_lsegs_return(lo, lseg_list, &range, seq);
+}
+
 static int
 pnfs_iomode_to_fail_bit(u32 iomode)
 {
@@ -846,8 +859,6 @@ pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
 			break;
 		inode = pnfs_grab_inode_layout_hdr(lo);
 		if (inode != NULL) {
-			if (test_and_clear_bit(NFS_LAYOUT_HASHED, &lo->plh_flags))
-				list_del_rcu(&lo->plh_layouts);
 			if (pnfs_layout_add_bulk_destroy_list(inode,
 						layout_list))
 				continue;
@@ -868,7 +879,7 @@ pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
 
 static int
 pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
-		bool is_bulk_recall)
+				   enum pnfs_layout_destroy_mode mode)
 {
 	struct pnfs_layout_hdr *lo;
 	struct inode *inode;
@@ -886,8 +897,11 @@ pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
 
 		spin_lock(&inode->i_lock);
 		list_del_init(&lo->plh_bulk_destroy);
-		if (pnfs_mark_layout_stateid_invalid(lo, &lseg_list)) {
+		if (mode == PNFS_LAYOUT_FILE_BULK_RETURN) {
-			if (is_bulk_recall)
+			pnfs_mark_layout_stateid_return(lo, &lseg_list,
+							IOMODE_ANY, 0);
+		} else if (pnfs_mark_layout_stateid_invalid(lo, &lseg_list)) {
+			if (mode == PNFS_LAYOUT_BULK_RETURN)
 				set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
 			ret = -EAGAIN;
 		}
@@ -901,10 +915,8 @@ pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
 	return ret;
 }
 
-int
+int pnfs_layout_destroy_byfsid(struct nfs_client *clp, struct nfs_fsid *fsid,
-pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
+			       enum pnfs_layout_destroy_mode mode)
-		struct nfs_fsid *fsid,
-		bool is_recall)
 {
 	struct nfs_server *server;
 	LIST_HEAD(layout_list);
@@ -923,33 +935,40 @@ restart:
 	rcu_read_unlock();
 	spin_unlock(&clp->cl_lock);
 
-	if (list_empty(&layout_list))
+	return pnfs_layout_free_bulk_destroy_list(&layout_list, mode);
-		return 0;
-	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
 }
 
-int
+static void pnfs_layout_build_destroy_list_byclient(struct nfs_client *clp,
-pnfs_destroy_layouts_byclid(struct nfs_client *clp,
+						    struct list_head *list)
-		bool is_recall)
 {
 	struct nfs_server *server;
-	LIST_HEAD(layout_list);
 
 	spin_lock(&clp->cl_lock);
 	rcu_read_lock();
 restart:
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		if (pnfs_layout_bulk_destroy_byserver_locked(clp,
+		if (pnfs_layout_bulk_destroy_byserver_locked(clp, server,
-					server,
+							     list) != 0)
-					&layout_list) != 0)
 			goto restart;
 	}
 	rcu_read_unlock();
 	spin_unlock(&clp->cl_lock);
+}
 
-	if (list_empty(&layout_list))
+static int pnfs_layout_do_destroy_byclid(struct nfs_client *clp,
-		return 0;
+					 struct list_head *list,
-	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
+					 enum pnfs_layout_destroy_mode mode)
+{
+	pnfs_layout_build_destroy_list_byclient(clp, list);
+	return pnfs_layout_free_bulk_destroy_list(list, mode);
+}
+
+int pnfs_layout_destroy_byclid(struct nfs_client *clp,
+			       enum pnfs_layout_destroy_mode mode)
+{
+	LIST_HEAD(layout_list);
+
+	return pnfs_layout_do_destroy_byclid(clp, &layout_list, mode);
 }
 
 /*
@@ -962,7 +981,68 @@ pnfs_destroy_all_layouts(struct nfs_client *clp)
 	nfs4_deviceid_mark_client_invalid(clp);
 	nfs4_deviceid_purge_client(clp);
 
-	pnfs_destroy_layouts_byclid(clp, false);
+	pnfs_layout_destroy_byclid(clp, PNFS_LAYOUT_INVALIDATE);
+}
+
+static void pnfs_layout_build_recover_list_byclient(struct nfs_client *clp,
+						    struct list_head *list)
+{
+	struct nfs_server *server;
+
+	spin_lock(&clp->cl_lock);
+	rcu_read_lock();
+restart:
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		if (!(server->caps & NFS_CAP_REBOOT_LAYOUTRETURN))
+			continue;
+		if (pnfs_layout_bulk_destroy_byserver_locked(clp, server,
+							     list) != 0)
+			goto restart;
+	}
+	rcu_read_unlock();
+	spin_unlock(&clp->cl_lock);
+}
+
+static int pnfs_layout_bulk_list_reboot(struct list_head *list)
+{
+	struct pnfs_layout_hdr *lo;
+	struct nfs_server *server;
+	int ret;
+
+	list_for_each_entry(lo, list, plh_bulk_destroy) {
+		server = NFS_SERVER(lo->plh_inode);
+		ret = pnfs_layout_return_on_reboot(lo);
+		switch (ret) {
+		case 0:
+			continue;
+		case -NFS4ERR_BAD_STATEID:
+			server->caps &= ~NFS_CAP_REBOOT_LAYOUTRETURN;
+			break;
+		case -NFS4ERR_NO_GRACE:
+			break;
+		default:
+			goto err;
+		}
+		break;
+	}
+	return 0;
+err:
+	return ret;
+}
+
+int pnfs_layout_handle_reboot(struct nfs_client *clp)
+{
+	LIST_HEAD(list);
+	int ret = 0, ret2;
+
+	pnfs_layout_build_recover_list_byclient(clp, &list);
+	if (!list_empty(&list))
+		ret = pnfs_layout_bulk_list_reboot(&list);
+	ret2 = pnfs_layout_do_destroy_byclid(clp, &list,
+					     PNFS_LAYOUT_INVALIDATE);
+	if (!ret)
+		ret = ret2;
+	return (ret == 0) ?  0 : -EAGAIN;
 }
 
 static void
@@ -1163,6 +1243,38 @@ static void pnfs_clear_layoutcommit(struct inode *inode,
 	}
 }
 
+static void
+pnfs_layoutreturn_retry_later_locked(struct pnfs_layout_hdr *lo,
+				     const nfs4_stateid *arg_stateid,
+				     const struct pnfs_layout_range *range)
+{
+	const struct pnfs_layout_segment *lseg;
+	u32 seq = be32_to_cpu(arg_stateid->seqid);
+
+	if (pnfs_layout_is_valid(lo) &&
+	    nfs4_stateid_match_other(&lo->plh_stateid, arg_stateid)) {
+		list_for_each_entry(lseg, &lo->plh_return_segs, pls_list) {
+			if (pnfs_seqid_is_newer(lseg->pls_seq, seq) ||
+			    !pnfs_should_free_range(&lseg->pls_range, range))
+				continue;
+			pnfs_set_plh_return_info(lo, range->iomode, seq);
+			break;
+		}
+	}
+}
+
+void pnfs_layoutreturn_retry_later(struct pnfs_layout_hdr *lo,
+				   const nfs4_stateid *arg_stateid,
+				   const struct pnfs_layout_range *range)
+{
+	struct inode *inode = lo->plh_inode;
+
+	spin_lock(&inode->i_lock);
+	pnfs_layoutreturn_retry_later_locked(lo, arg_stateid, range);
+	pnfs_clear_layoutreturn_waitbit(lo);
+	spin_unlock(&inode->i_lock);
+}
+
 void pnfs_layoutreturn_free_lsegs(struct pnfs_layout_hdr *lo,
 		const nfs4_stateid *arg_stateid,
 		const struct pnfs_layout_range *range,
@@ -1239,7 +1351,7 @@ pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo,
 		       const nfs4_stateid *stateid,
 		       const struct cred **pcred,
 		       enum pnfs_iomode iomode,
-		       bool sync)
+		       unsigned int flags)
 {
 	struct inode *ino = lo->plh_inode;
 	struct pnfs_layoutdriver_type *ld = NFS_SERVER(ino)->pnfs_curr_ld;
@@ -1266,33 +1378,21 @@ pnfs_send_layoutreturn(struct pnfs_layout_hdr *lo,
 	if (ld->prepare_layoutreturn)
 		ld->prepare_layoutreturn(&lrp->args);
 
-	status = nfs4_proc_layoutreturn(lrp, sync);
+	status = nfs4_proc_layoutreturn(lrp, flags);
 out:
 	dprintk("<-- %s status: %d\n", __func__, status);
 	return status;
 }
 
-static bool
-pnfs_layout_segments_returnable(struct pnfs_layout_hdr *lo,
-				enum pnfs_iomode iomode,
-				u32 seq)
-{
-	struct pnfs_layout_range recall_range = {
-		.length = NFS4_MAX_UINT64,
-		.iomode = iomode,
-	};
-	return pnfs_mark_matching_lsegs_return(lo, &lo->plh_return_segs,
-					       &recall_range, seq) != -EBUSY;
-}
-
 /* Return true if layoutreturn is needed */
 static bool
 pnfs_layout_need_return(struct pnfs_layout_hdr *lo)
 {
 	if (!test_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags))
 		return false;
-	return pnfs_layout_segments_returnable(lo, lo->plh_return_iomode,
+	return pnfs_mark_layout_stateid_return(lo, &lo->plh_return_segs,
-					       lo->plh_return_seq);
+					       lo->plh_return_iomode,
+					       lo->plh_return_seq) != EBUSY;
 }
 
 static void pnfs_layoutreturn_before_put_layout_hdr(struct pnfs_layout_hdr *lo)
@@ -1312,7 +1412,8 @@ static void pnfs_layoutreturn_before_put_layout_hdr(struct pnfs_layout_hdr *lo)
 		spin_unlock(&inode->i_lock);
 		if (send) {
 			/* Send an async layoutreturn so we dont deadlock */
-			pnfs_send_layoutreturn(lo, &stateid, &cred, iomode, false);
+			pnfs_send_layoutreturn(lo, &stateid, &cred, iomode,
+					       PNFS_FL_LAYOUTRETURN_ASYNC);
 		}
 	} else
 		spin_unlock(&inode->i_lock);
@@ -1379,7 +1480,8 @@ _pnfs_return_layout(struct inode *ino)
 	send = pnfs_prepare_layoutreturn(lo, &stateid, &cred, NULL);
 	spin_unlock(&ino->i_lock);
 	if (send)
-		status = pnfs_send_layoutreturn(lo, &stateid, &cred, IOMODE_ANY, true);
+		status = pnfs_send_layoutreturn(lo, &stateid, &cred, IOMODE_ANY,
+						0);
 out_wait_layoutreturn:
 	wait_on_bit(&lo->plh_flags, NFS_LAYOUT_RETURN, TASK_UNINTERRUPTIBLE);
 out_put_layout_hdr:
@@ -1417,6 +1519,24 @@ pnfs_commit_and_return_layout(struct inode *inode)
 	return ret;
 }
 
+static int pnfs_layout_return_on_reboot(struct pnfs_layout_hdr *lo)
+{
+	struct inode *inode = lo->plh_inode;
+	const struct cred *cred;
+
+	spin_lock(&inode->i_lock);
+	if (!pnfs_layout_is_valid(lo)) {
+		spin_unlock(&inode->i_lock);
+		return 0;
+	}
+	cred = get_cred(lo->plh_lc_cred);
+	pnfs_get_layout_hdr(lo);
+	spin_unlock(&inode->i_lock);
+
+	return pnfs_send_layoutreturn(lo, &zero_stateid, &cred, IOMODE_ANY,
+				      PNFS_FL_LAYOUTRETURN_PRIVILEGED);
+}
+
 bool pnfs_roc(struct inode *ino,
 		struct nfs4_layoutreturn_args *args,
 		struct nfs4_layoutreturn_res *res,
@@ -1520,7 +1640,7 @@ out_noroc:
 		return true;
 	}
 	if (layoutreturn)
-		pnfs_send_layoutreturn(lo, &stateid, &lc_cred, iomode, true);
+		pnfs_send_layoutreturn(lo, &stateid, &lc_cred, iomode, 0);
 	pnfs_put_layout_hdr(lo);
 	return false;
 }
@@ -1570,8 +1690,7 @@ int pnfs_roc_done(struct rpc_task *task, struct nfs4_layoutreturn_args **argpp,
 }
 
 void pnfs_roc_release(struct nfs4_layoutreturn_args *args,
-		struct nfs4_layoutreturn_res *res,
+		      struct nfs4_layoutreturn_res *res, int ret)
-		int ret)
 {
 	struct pnfs_layout_hdr *lo = args->layout;
 	struct inode *inode = args->inode;
@@ -1579,11 +1698,13 @@ void pnfs_roc_release(struct nfs4_layoutreturn_args *args,
 	struct nfs4_xdr_opaque_data *ld_private = args->ld_private;
 
 	switch (ret) {
+	case -NFS4ERR_BADSESSION:
+	case -NFS4ERR_DEADSESSION:
+	case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
 	case -NFS4ERR_NOMATCHING_LAYOUT:
 		spin_lock(&inode->i_lock);
-		if (pnfs_layout_is_valid(lo) &&
+		pnfs_layoutreturn_retry_later_locked(lo, &args->stateid,
-		    nfs4_stateid_match_other(&args->stateid, &lo->plh_stateid))
+						     &args->range);
-			pnfs_set_plh_return_info(lo, args->range.iomode, 0);
 		pnfs_clear_layoutreturn_waitbit(lo);
 		spin_unlock(&inode->i_lock);
 		break;
@@ -2566,7 +2687,8 @@ pnfs_mark_layout_for_return(struct inode *inode,
 		return_now = pnfs_prepare_layoutreturn(lo, &stateid, &cred, &iomode);
 		spin_unlock(&inode->i_lock);
 		if (return_now)
-			pnfs_send_layoutreturn(lo, &stateid, &cred, iomode, false);
+			pnfs_send_layoutreturn(lo, &stateid, &cred, iomode,
+					       PNFS_FL_LAYOUTRETURN_ASYNC);
 	} else {
 		spin_unlock(&inode->i_lock);
 		nfs_commit_inode(inode, 0);
@@ -2682,7 +2804,8 @@ restart:
 		}
 		spin_unlock(&inode->i_lock);
 		rcu_read_unlock();
-		pnfs_send_layoutreturn(lo, &stateid, &cred, iomode, false);
+		pnfs_send_layoutreturn(lo, &stateid, &cred, iomode,
+				       PNFS_FL_LAYOUTRETURN_ASYNC);
 		pnfs_put_layout_hdr(lo);
 		cond_resched();
 		goto restart;

fs/nfs/pnfs.h

@@ -118,6 +118,12 @@ enum layoutdriver_policy_flags {
 	PNFS_LAYOUTGET_ON_OPEN		= 1 << 3,
 };
 
+enum pnfs_layout_destroy_mode {
+	PNFS_LAYOUT_INVALIDATE = 0,
+	PNFS_LAYOUT_BULK_RETURN,
+	PNFS_LAYOUT_FILE_BULK_RETURN,
+};
+
 struct nfs4_deviceid_node;
 
 /* Per-layout driver specific registration structure */
@@ -127,7 +133,6 @@ struct pnfs_layoutdriver_type {
 	const char *name;
 	struct module *owner;
 	unsigned flags;
-	unsigned max_deviceinfo_size;
 	unsigned max_layoutget_response;
 
 	int (*set_layoutdriver) (struct nfs_server *, const struct nfs_fh *);
@@ -193,8 +198,6 @@ struct pnfs_commit_ops {
 				  int max);
 	void (*recover_commit_reqs) (struct list_head *list,
 				     struct nfs_commit_info *cinfo);
-	struct nfs_page * (*search_commit_reqs)(struct nfs_commit_info *cinfo,
-						struct folio *folio);
 };
 
 struct pnfs_layout_hdr {
@@ -242,6 +245,9 @@ extern const struct pnfs_layoutdriver_type *pnfs_find_layoutdriver(u32 id);
 extern void pnfs_put_layoutdriver(const struct pnfs_layoutdriver_type *ld);
 
 /* nfs4proc.c */
+#define PNFS_FL_LAYOUTRETURN_ASYNC (1U << 0)
+#define PNFS_FL_LAYOUTRETURN_PRIVILEGED (1U << 1)
+
 extern size_t max_response_pages(struct nfs_server *server);
 extern int nfs4_proc_getdeviceinfo(struct nfs_server *server,
 				   struct pnfs_device *dev,
@@ -249,7 +255,8 @@ extern int nfs4_proc_getdeviceinfo(struct nfs_server *server,
 extern struct pnfs_layout_segment *
 nfs4_proc_layoutget(struct nfs4_layoutget *lgp,
 		    struct nfs4_exception *exception);
-extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync);
+extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp,
+				  unsigned int flags);
 
 /* pnfs.c */
 void pnfs_get_layout_hdr(struct pnfs_layout_hdr *lo);
@@ -273,11 +280,10 @@ void pnfs_free_lseg_list(struct list_head *tmp_list);
 void pnfs_destroy_layout(struct nfs_inode *);
 void pnfs_destroy_layout_final(struct nfs_inode *);
 void pnfs_destroy_all_layouts(struct nfs_client *);
-int pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
+int pnfs_layout_destroy_byfsid(struct nfs_client *clp, struct nfs_fsid *fsid,
-		struct nfs_fsid *fsid,
+			       enum pnfs_layout_destroy_mode mode);
-		bool is_recall);
+int pnfs_layout_destroy_byclid(struct nfs_client *clp,
-int pnfs_destroy_layouts_byclid(struct nfs_client *clp,
+			       enum pnfs_layout_destroy_mode mode);
-		bool is_recall);
 bool nfs4_layout_refresh_old_stateid(nfs4_stateid *dst,
 		struct pnfs_layout_range *dst_range,
 		struct inode *inode);
@@ -323,6 +329,9 @@ struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino,
 					       enum pnfs_iomode iomode,
 					       bool strict_iomode,
 					       gfp_t gfp_flags);
+void pnfs_layoutreturn_retry_later(struct pnfs_layout_hdr *lo,
+				   const nfs4_stateid *arg_stateid,
+				   const struct pnfs_layout_range *range);
 void pnfs_layoutreturn_free_lsegs(struct pnfs_layout_hdr *lo,
 		const nfs4_stateid *arg_stateid,
 		const struct pnfs_layout_range *range,
@@ -344,6 +353,7 @@ void pnfs_error_mark_layout_for_return(struct inode *inode,
 				       struct pnfs_layout_segment *lseg);
 void pnfs_layout_return_unused_byclid(struct nfs_client *clp,
 				      enum pnfs_iomode iomode);
+int pnfs_layout_handle_reboot(struct nfs_client *clp);
 
 /* nfs4_deviceid_flags */
 enum {
@@ -396,8 +406,6 @@ void pnfs_generic_prepare_to_resend_writes(struct nfs_commit_data *data);
 void pnfs_generic_rw_release(void *data);
 void pnfs_generic_recover_commit_reqs(struct list_head *dst,
 				      struct nfs_commit_info *cinfo);
-struct nfs_page *pnfs_generic_search_commit_reqs(struct nfs_commit_info *cinfo,
-						 struct folio *folio);
 int pnfs_generic_commit_pagelist(struct inode *inode,
 				 struct list_head *mds_pages,
 				 int how,
@@ -557,17 +565,6 @@ pnfs_recover_commit_reqs(struct list_head *head, struct nfs_commit_info *cinfo)
 		fl_cinfo->ops->recover_commit_reqs(head, cinfo);
 }
 
-static inline struct nfs_page *
-pnfs_search_commit_reqs(struct inode *inode, struct nfs_commit_info *cinfo,
-			struct folio *folio)
-{
-	struct pnfs_ds_commit_info *fl_cinfo = cinfo->ds;
-
-	if (!fl_cinfo->ops || !fl_cinfo->ops->search_commit_reqs)
-		return NULL;
-	return fl_cinfo->ops->search_commit_reqs(cinfo, folio);
-}
-
 /* Should the pNFS client commit and return the layout upon a setattr */
 static inline bool
 pnfs_ld_layoutret_on_setattr(struct inode *inode)
@@ -725,6 +722,11 @@ static inline void pnfs_destroy_layout_final(struct nfs_inode *nfsi)
 {
 }
 
+static inline int pnfs_layout_handle_reboot(struct nfs_client *clp)
+{
+	return 0;
+}
+
 static inline struct pnfs_layout_segment *
 pnfs_get_lseg(struct pnfs_layout_segment *lseg)
 {
@@ -864,13 +866,6 @@ pnfs_recover_commit_reqs(struct list_head *head, struct nfs_commit_info *cinfo)
 {
 }
 
-static inline struct nfs_page *
-pnfs_search_commit_reqs(struct inode *inode, struct nfs_commit_info *cinfo,
-			struct folio *folio)
-{
-	return NULL;
-}
-
 static inline int pnfs_layoutcommit_inode(struct inode *inode, bool sync)
 {
 	return 0;

fs/nfs/pnfs_dev.c

@@ -110,9 +110,6 @@ nfs4_get_device_info(struct nfs_server *server,
 	 * GETDEVICEINFO's maxcount
 	 */
 	max_resp_sz = server->nfs_client->cl_session->fc_attrs.max_resp_sz;
-	if (server->pnfs_curr_ld->max_deviceinfo_size &&
-	    server->pnfs_curr_ld->max_deviceinfo_size < max_resp_sz)
-		max_resp_sz = server->pnfs_curr_ld->max_deviceinfo_size;
 	max_pages = nfs_page_array_len(0, max_resp_sz);
 	dprintk("%s: server %p max_resp_sz %u max_pages %d\n",
 		__func__, server, max_resp_sz, max_pages);

fs/nfs/pnfs_nfs.c

@@ -351,53 +351,6 @@ void pnfs_generic_recover_commit_reqs(struct list_head *dst,
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_recover_commit_reqs);
 
-static struct nfs_page *
-pnfs_bucket_search_commit_reqs(struct pnfs_commit_bucket *buckets,
-			       unsigned int nbuckets, struct folio *folio)
-{
-	struct nfs_page *req;
-	struct pnfs_commit_bucket *b;
-	unsigned int i;
-
-	/* Linearly search the commit lists for each bucket until a matching
-	 * request is found */
-	for (i = 0, b = buckets; i < nbuckets; i++, b++) {
-		list_for_each_entry(req, &b->written, wb_list) {
-			if (nfs_page_to_folio(req) == folio)
-				return req->wb_head;
-		}
-		list_for_each_entry(req, &b->committing, wb_list) {
-			if (nfs_page_to_folio(req) == folio)
-				return req->wb_head;
-		}
-	}
-	return NULL;
-}
-
-/* pnfs_generic_search_commit_reqs - Search lists in @cinfo for the head request
- *				   for @folio
- * @cinfo - commit info for current inode
- * @folio - page to search for matching head request
- *
- * Return: the head request if one is found, otherwise %NULL.
- */
-struct nfs_page *pnfs_generic_search_commit_reqs(struct nfs_commit_info *cinfo,
-						 struct folio *folio)
-{
-	struct pnfs_ds_commit_info *fl_cinfo = cinfo->ds;
-	struct pnfs_commit_array *array;
-	struct nfs_page *req;
-
-	list_for_each_entry(array, &fl_cinfo->commits, cinfo_list) {
-		req = pnfs_bucket_search_commit_reqs(array->buckets,
-						     array->nbuckets, folio);
-		if (req)
-			return req;
-	}
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(pnfs_generic_search_commit_reqs);
-
 static struct pnfs_layout_segment *
 pnfs_bucket_get_committing(struct list_head *head,
 			   struct pnfs_commit_bucket *bucket,

fs/nfs/proc.c

@@ -687,11 +687,18 @@ out_einval:
 	return -EINVAL;
 }
 
-static int nfs_have_delegation(struct inode *inode, fmode_t flags)
+static int nfs_have_delegation(struct inode *inode, fmode_t type, int flags)
 {
 	return 0;
 }
 
+static int nfs_return_delegation(struct inode *inode)
+{
+	if (S_ISREG(inode->i_mode))
+		nfs_wb_all(inode);
+	return 0;
+}
+
 static const struct inode_operations nfs_dir_inode_operations = {
 	.create		= nfs_create,
 	.lookup		= nfs_lookup,
@@ -757,6 +764,7 @@ const struct nfs_rpc_ops nfs_v2_clientops = {
 	.lock_check_bounds = nfs_lock_check_bounds,
 	.close_context	= nfs_close_context,
 	.have_delegation = nfs_have_delegation,
+	.return_delegation = nfs_return_delegation,
 	.alloc_client	= nfs_alloc_client,
 	.init_client	= nfs_init_client,
 	.free_client	= nfs_free_client,

fs/nfs/read.c

@@ -28,6 +28,7 @@
 #include "fscache.h"
 #include "pnfs.h"
 #include "nfstrace.h"
+#include "delegation.h"
 
 #define NFSDBG_FACILITY		NFSDBG_PAGECACHE
 
@@ -286,7 +287,7 @@ int nfs_read_add_folio(struct nfs_pageio_descriptor *pgio,
 		       struct nfs_open_context *ctx,
 		       struct folio *folio)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 	struct nfs_server *server = NFS_SERVER(inode);
 	size_t fsize = folio_size(folio);
 	unsigned int rsize = server->rsize;
@@ -322,21 +323,57 @@ out:
 }
 
 /*
- * Read a page over NFS.
+ * Actually read a folio over the wire.
- * We read the page synchronously in the following case:
- *  -	The error flag is set for this page. This happens only when a
- *	previous async read operation failed.
  */
-int nfs_read_folio(struct file *file, struct folio *folio)
+static int nfs_do_read_folio(struct file *file, struct folio *folio)
 {
 	struct inode *inode = file_inode(file);
 	struct nfs_pageio_descriptor pgio;
 	struct nfs_open_context *ctx;
 	int ret;
 
-	trace_nfs_aop_readpage(inode, folio);
+	ctx = get_nfs_open_context(nfs_file_open_context(file));
+
+	xchg(&ctx->error, 0);
+	nfs_pageio_init_read(&pgio, inode, false,
+			     &nfs_async_read_completion_ops);
+
+	ret = nfs_read_add_folio(&pgio, ctx, folio);
+	if (ret)
+		goto out_put;
+
+	nfs_pageio_complete_read(&pgio);
+	nfs_update_delegated_atime(inode);
+	if (pgio.pg_error < 0) {
+		ret = pgio.pg_error;
+		goto out_put;
+	}
+
+	ret = folio_wait_locked_killable(folio);
+	if (!folio_test_uptodate(folio) && !ret)
+		ret = xchg(&ctx->error, 0);
+
+out_put:
+	put_nfs_open_context(ctx);
+	return ret;
+}
+
+/*
+ * Synchronously read a folio.
+ *
+ * This is not heavily used as most users to try an asynchronous
+ * large read through ->readahead first.
+ */
+int nfs_read_folio(struct file *file, struct folio *folio)
+{
+	struct inode *inode = file_inode(file);
+	loff_t pos = folio_pos(folio);
+	size_t len = folio_size(folio);
+	int ret;
+
+	trace_nfs_aop_readpage(inode, pos, len);
 	nfs_inc_stats(inode, NFSIOS_VFSREADPAGE);
-	task_io_account_read(folio_size(folio));
+	task_io_account_read(len);
 
 	/*
 	 * Try to flush any pending writes to the file..
@@ -356,30 +393,10 @@ int nfs_read_folio(struct file *file, struct folio *folio)
 		goto out_unlock;
 
 	ret = nfs_netfs_read_folio(file, folio);
-	if (!ret)
-		goto out;
-
-	ctx = get_nfs_open_context(nfs_file_open_context(file));
-
-	xchg(&ctx->error, 0);
-	nfs_pageio_init_read(&pgio, inode, false,
-			     &nfs_async_read_completion_ops);
-
-	ret = nfs_read_add_folio(&pgio, ctx, folio);
 	if (ret)
-		goto out_put;
+		ret = nfs_do_read_folio(file, folio);
-
-	nfs_pageio_complete_read(&pgio);
-	ret = pgio.pg_error < 0 ? pgio.pg_error : 0;
-	if (!ret) {
-		ret = folio_wait_locked_killable(folio);
-		if (!folio_test_uptodate(folio) && !ret)
-			ret = xchg(&ctx->error, 0);
-	}
-out_put:
-	put_nfs_open_context(ctx);
 out:
-	trace_nfs_aop_readpage_done(inode, folio, ret);
+	trace_nfs_aop_readpage_done(inode, pos, len, ret);
 	return ret;
 out_unlock:
 	folio_unlock(folio);
@@ -426,6 +443,7 @@ void nfs_readahead(struct readahead_control *ractl)
 	}
 
 	nfs_pageio_complete_read(&pgio);
+	nfs_update_delegated_atime(inode);
 
 	put_nfs_open_context(ctx);
 out:

fs/nfs/unlink.c

@@ -232,6 +232,8 @@ nfs_complete_unlink(struct dentry *dentry, struct inode *inode)
 	dentry->d_fsdata = NULL;
 	spin_unlock(&dentry->d_lock);
 
+	NFS_PROTO(inode)->return_delegation(inode);
+
 	if (NFS_STALE(inode) || !nfs_call_unlink(dentry, inode, data))
 		nfs_free_unlinkdata(data);
 }

fs/nfs/write.c

@@ -63,9 +63,6 @@ static void nfs_clear_request_commit(struct nfs_commit_info *cinfo,
 				     struct nfs_page *req);
 static void nfs_init_cinfo_from_inode(struct nfs_commit_info *cinfo,
 				      struct inode *inode);
-static struct nfs_page *
-nfs_page_search_commits_for_head_request_locked(struct nfs_inode *nfsi,
-						struct folio *folio);
 
 static struct kmem_cache *nfs_wdata_cachep;
 static mempool_t *nfs_wdata_mempool;
@@ -147,53 +144,23 @@ 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;
-}
-
-static struct nfs_page *nfs_folio_private_request(struct folio *folio)
-{
-	return folio_get_private(folio);
-}
-
 /**
- * nfs_folio_find_private_request - find head request associated with a folio
+ * nfs_folio_find_head_request - find head request associated with a folio
  * @folio: pointer to folio
  *
  * must be called while holding the inode lock.
  *
  * returns matching head request with reference held, or NULL if not found.
  */
-static struct nfs_page *nfs_folio_find_private_request(struct folio *folio)
+static struct nfs_page *nfs_folio_find_head_request(struct folio *folio)
 {
-	struct address_space *mapping = folio_file_mapping(folio);
+	struct address_space *mapping = folio->mapping;
 	struct nfs_page *req;
 
 	if (!folio_test_private(folio))
 		return NULL;
 	spin_lock(&mapping->i_private_lock);
-	req = nfs_folio_private_request(folio);
+	req = folio->private;
 	if (req) {
 		WARN_ON_ONCE(req->wb_head != req);
 		kref_get(&req->wb_kref);
@@ -202,84 +169,18 @@ static struct nfs_page *nfs_folio_find_private_request(struct folio *folio)
 	return req;
 }
 
-static struct nfs_page *nfs_folio_find_swap_request(struct folio *folio)
-{
-	struct inode *inode = folio_file_mapping(folio)->host;
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct nfs_page *req = NULL;
-	if (!folio_test_swapcache(folio))
-		return NULL;
-	mutex_lock(&nfsi->commit_mutex);
-	if (folio_test_swapcache(folio)) {
-		req = nfs_page_search_commits_for_head_request_locked(nfsi,
-								      folio);
-		if (req) {
-			WARN_ON_ONCE(req->wb_head != req);
-			kref_get(&req->wb_kref);
-		}
-	}
-	mutex_unlock(&nfsi->commit_mutex);
-	return req;
-}
-
-/**
- * nfs_folio_find_head_request - find head request associated with a folio
- * @folio: pointer to folio
- *
- * returns matching head request with reference held, or NULL if not found.
- */
-static struct nfs_page *nfs_folio_find_head_request(struct folio *folio)
-{
-	struct nfs_page *req;
-
-	req = nfs_folio_find_private_request(folio);
-	if (!req)
-		req = nfs_folio_find_swap_request(folio);
-	return req;
-}
-
-static struct nfs_page *nfs_folio_find_and_lock_request(struct folio *folio)
-{
-	struct inode *inode = folio_file_mapping(folio)->host;
-	struct nfs_page *req, *head;
-	int ret;
-
-	for (;;) {
-		req = nfs_folio_find_head_request(folio);
-		if (!req)
-			return req;
-		head = nfs_page_group_lock_head(req);
-		if (head != req)
-			nfs_release_request(req);
-		if (IS_ERR(head))
-			return head;
-		ret = nfs_cancel_remove_inode(head, inode);
-		if (ret < 0) {
-			nfs_unlock_and_release_request(head);
-			return ERR_PTR(ret);
-		}
-		/* Ensure that nobody removed the request before we locked it */
-		if (head == nfs_folio_private_request(folio))
-			break;
-		if (folio_test_swapcache(folio))
-			break;
-		nfs_unlock_and_release_request(head);
-	}
-	return head;
-}
-
 /* Adjust the file length if we're writing beyond the end */
 static void nfs_grow_file(struct folio *folio, unsigned int offset,
 			  unsigned int count)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 	loff_t end, i_size;
 	pgoff_t end_index;
 
 	spin_lock(&inode->i_lock);
 	i_size = i_size_read(inode);
 	end_index = ((i_size - 1) >> folio_shift(folio)) << folio_order(folio);
-	if (i_size > 0 && folio_index(folio) < end_index)
+	if (i_size > 0 && folio->index < end_index)
 		goto out;
 	end = folio_file_pos(folio) + (loff_t)offset + (loff_t)count;
 	if (i_size >= end)
@@ -289,6 +190,8 @@ static void nfs_grow_file(struct folio *folio, unsigned int offset,
 	NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_SIZE;
 	nfs_inc_stats(inode, NFSIOS_EXTENDWRITE);
 out:
+	/* Atomically update timestamps if they are delegated to us. */
+	nfs_update_delegated_mtime_locked(inode);
 	spin_unlock(&inode->i_lock);
 	nfs_fscache_invalidate(inode, 0);
 }
@@ -309,7 +212,7 @@ static void nfs_set_pageerror(struct address_space *mapping)
 
 static void nfs_mapping_set_error(struct folio *folio, int error)
 {
-	struct address_space *mapping = folio_file_mapping(folio);
+	struct address_space *mapping = folio->mapping;
 
 	filemap_set_wb_err(mapping, error);
 	if (mapping->host)
@@ -409,7 +312,7 @@ int nfs_congestion_kb;
 
 static void nfs_folio_set_writeback(struct folio *folio)
 {
-	struct nfs_server *nfss = NFS_SERVER(folio_file_mapping(folio)->host);
+	struct nfs_server *nfss = NFS_SERVER(folio->mapping->host);
 
 	folio_start_writeback(folio);
 	if (atomic_long_inc_return(&nfss->writeback) > NFS_CONGESTION_ON_THRESH)
@@ -418,12 +321,14 @@ static void nfs_folio_set_writeback(struct folio *folio)
 
 static void nfs_folio_end_writeback(struct folio *folio)
 {
-	struct nfs_server *nfss = NFS_SERVER(folio_file_mapping(folio)->host);
+	struct nfs_server *nfss = NFS_SERVER(folio->mapping->host);
 
 	folio_end_writeback(folio);
 	if (atomic_long_dec_return(&nfss->writeback) <
-	    NFS_CONGESTION_OFF_THRESH)
+	    NFS_CONGESTION_OFF_THRESH) {
 		nfss->write_congested = 0;
+		wake_up_all(&nfss->write_congestion_wait);
+	}
 }
 
 static void nfs_page_end_writeback(struct nfs_page *req)
@@ -547,6 +452,74 @@ void nfs_join_page_group(struct nfs_page *head, struct nfs_commit_info *cinfo,
 	nfs_destroy_unlinked_subrequests(destroy_list, head, inode);
 }
 
+/**
+ * nfs_wait_on_request - Wait for a request to complete.
+ * @req: request to wait upon.
+ *
+ * Interruptible by fatal signals only.
+ * The user is responsible for holding a count on the request.
+ */
+static int nfs_wait_on_request(struct nfs_page *req)
+{
+	if (!test_bit(PG_BUSY, &req->wb_flags))
+		return 0;
+	set_bit(PG_CONTENDED2, &req->wb_flags);
+	smp_mb__after_atomic();
+	return wait_on_bit_io(&req->wb_flags, PG_BUSY,
+			      TASK_UNINTERRUPTIBLE);
+}
+
+/*
+ * nfs_unroll_locks -  unlock all newly locked reqs and wait on @req
+ * @head: head request of page group, must be holding head lock
+ * @req: request that couldn't lock and needs to wait on the req bit lock
+ *
+ * This is a helper function for nfs_lock_and_join_requests
+ * returns 0 on success, < 0 on error.
+ */
+static void
+nfs_unroll_locks(struct nfs_page *head, struct nfs_page *req)
+{
+	struct nfs_page *tmp;
+
+	/* relinquish all the locks successfully grabbed this run */
+	for (tmp = head->wb_this_page ; tmp != req; tmp = tmp->wb_this_page) {
+		if (!kref_read(&tmp->wb_kref))
+			continue;
+		nfs_unlock_and_release_request(tmp);
+	}
+}
+
+/*
+ * nfs_page_group_lock_subreq -  try to lock a subrequest
+ * @head: head request of page group
+ * @subreq: request to lock
+ *
+ * This is a helper function for nfs_lock_and_join_requests which
+ * must be called with the head request and page group both locked.
+ * On error, it returns with the page group unlocked.
+ */
+static int
+nfs_page_group_lock_subreq(struct nfs_page *head, struct nfs_page *subreq)
+{
+	int ret;
+
+	if (!kref_get_unless_zero(&subreq->wb_kref))
+		return 0;
+	while (!nfs_lock_request(subreq)) {
+		nfs_page_group_unlock(head);
+		ret = nfs_wait_on_request(subreq);
+		if (!ret)
+			ret = nfs_page_group_lock(head);
+		if (ret < 0) {
+			nfs_unroll_locks(head, subreq);
+			nfs_release_request(subreq);
+			return ret;
+		}
+	}
+	return 0;
+}
+
 /*
  * nfs_lock_and_join_requests - join all subreqs to the head req
  * @folio: the folio used to lookup the "page group" of nfs_page structures
@@ -564,31 +537,75 @@ void nfs_join_page_group(struct nfs_page *head, struct nfs_commit_info *cinfo,
  */
 static struct nfs_page *nfs_lock_and_join_requests(struct folio *folio)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
-	struct nfs_page *head;
+	struct nfs_page *head, *subreq;
 	struct nfs_commit_info cinfo;
+	bool removed;
 	int ret;
 
-	nfs_init_cinfo_from_inode(&cinfo, inode);
 	/*
 	 * A reference is taken only on the head request which acts as a
 	 * reference to the whole page group - the group will not be destroyed
 	 * until the head reference is released.
 	 */
-	head = nfs_folio_find_and_lock_request(folio);
+retry:
-	if (IS_ERR_OR_NULL(head))
+	head = nfs_folio_find_head_request(folio);
-		return head;
+	if (!head)
+		return NULL;
 
-	/* lock each request in the page group */
+	while (!nfs_lock_request(head)) {
-	ret = nfs_page_group_lock_subrequests(head);
+		ret = nfs_wait_on_request(head);
-	if (ret < 0) {
+		if (ret < 0)
-		nfs_unlock_and_release_request(head);
+			return ERR_PTR(ret);
-		return ERR_PTR(ret);
 	}
 
-	nfs_join_page_group(head, &cinfo, inode);
+	/* Ensure that nobody removed the request before we locked it */
+	if (head != folio->private) {
+		nfs_unlock_and_release_request(head);
+		goto retry;
+	}
 
+	ret = nfs_page_group_lock(head);
+	if (ret < 0)
+		goto out_unlock;
+
+	removed = test_bit(PG_REMOVE, &head->wb_flags);
+
+	/* 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;
+	}
+
+	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;
+
+out_unlock:
+	nfs_unlock_and_release_request(head);
+	return ERR_PTR(ret);
 }
 
 static void nfs_write_error(struct nfs_page *req, int error)
@@ -640,7 +657,7 @@ static int nfs_page_async_flush(struct folio *folio,
 		nfs_redirty_request(req);
 		pgio->pg_error = 0;
 	} else
-		nfs_add_stats(folio_file_mapping(folio)->host,
+		nfs_add_stats(folio->mapping->host,
 			      NFSIOS_WRITEPAGES, 1);
 out:
 	return ret;
@@ -652,7 +669,7 @@ out_launder:
 static int nfs_do_writepage(struct folio *folio, struct writeback_control *wbc,
 			    struct nfs_pageio_descriptor *pgio)
 {
-	nfs_pageio_cond_complete(pgio, folio_index(folio));
+	nfs_pageio_cond_complete(pgio, folio->index);
 	return nfs_page_async_flush(folio, wbc, pgio);
 }
 
@@ -663,7 +680,7 @@ static int nfs_writepage_locked(struct folio *folio,
 				struct writeback_control *wbc)
 {
 	struct nfs_pageio_descriptor pgio;
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 	int err;
 
 	nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE);
@@ -697,12 +714,17 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 	struct nfs_pageio_descriptor pgio;
 	struct nfs_io_completion *ioc = NULL;
 	unsigned int mntflags = NFS_SERVER(inode)->flags;
+	struct nfs_server *nfss = NFS_SERVER(inode);
 	int priority = 0;
 	int err;
 
-	if (wbc->sync_mode == WB_SYNC_NONE &&
+	/* Wait with writeback until write congestion eases */
-	    NFS_SERVER(inode)->write_congested)
+	if (wbc->sync_mode == WB_SYNC_NONE && nfss->write_congested) {
-		return 0;
+		err = wait_event_killable(nfss->write_congestion_wait,
+					  nfss->write_congested == 0);
+		if (err)
+			return err;
+	}
 
 	nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES);
 
@@ -741,24 +763,17 @@ out_err:
 static void nfs_inode_add_request(struct nfs_page *req)
 {
 	struct folio *folio = nfs_page_to_folio(req);
-	struct address_space *mapping = folio_file_mapping(folio);
+	struct address_space *mapping = folio->mapping;
 	struct nfs_inode *nfsi = NFS_I(mapping->host);
 
 	WARN_ON_ONCE(req->wb_this_page != req);
 
 	/* Lock the request! */
 	nfs_lock_request(req);
-
-	/*
-	 * Swap-space should not get truncated. Hence no need to plug the race
-	 * with invalidate/truncate.
-	 */
 	spin_lock(&mapping->i_private_lock);
-	if (likely(!folio_test_swapcache(folio))) {
+	set_bit(PG_MAPPED, &req->wb_flags);
-		set_bit(PG_MAPPED, &req->wb_flags);
+	folio_set_private(folio);
-		folio_set_private(folio);
+	folio->private = req;
-		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
@@ -778,10 +793,10 @@ static void nfs_inode_remove_request(struct nfs_page *req)
 
 	if (nfs_page_group_sync_on_bit(req, PG_REMOVE)) {
 		struct folio *folio = nfs_page_to_folio(req->wb_head);
-		struct address_space *mapping = folio_file_mapping(folio);
+		struct address_space *mapping = folio->mapping;
 
 		spin_lock(&mapping->i_private_lock);
-		if (likely(folio && !folio_test_swapcache(folio))) {
+		if (likely(folio)) {
 			folio->private = NULL;
 			folio_clear_private(folio);
 			clear_bit(PG_MAPPED, &req->wb_head->wb_flags);
@@ -802,38 +817,6 @@ static void nfs_mark_request_dirty(struct nfs_page *req)
 		filemap_dirty_folio(folio_mapping(folio), folio);
 }
 
-/*
- * nfs_page_search_commits_for_head_request_locked
- *
- * Search through commit lists on @inode for the head request for @folio.
- * Must be called while holding the inode (which is cinfo) lock.
- *
- * Returns the head request if found, or NULL if not found.
- */
-static struct nfs_page *
-nfs_page_search_commits_for_head_request_locked(struct nfs_inode *nfsi,
-						struct folio *folio)
-{
-	struct nfs_page *freq, *t;
-	struct nfs_commit_info cinfo;
-	struct inode *inode = &nfsi->vfs_inode;
-
-	nfs_init_cinfo_from_inode(&cinfo, inode);
-
-	/* search through pnfs commit lists */
-	freq = pnfs_search_commit_reqs(inode, &cinfo, folio);
-	if (freq)
-		return freq->wb_head;
-
-	/* Linearly search the commit list for the correct request */
-	list_for_each_entry_safe(freq, t, &cinfo.mds->list, wb_list) {
-		if (nfs_page_to_folio(freq) == folio)
-			return freq->wb_head;
-	}
-
-	return NULL;
-}
-
 /**
  * nfs_request_add_commit_list_locked - add request to a commit list
  * @req: pointer to a struct nfs_page
@@ -940,7 +923,7 @@ static void nfs_folio_clear_commit(struct folio *folio)
 		long nr = folio_nr_pages(folio);
 
 		node_stat_mod_folio(folio, NR_WRITEBACK, -nr);
-		wb_stat_mod(&inode_to_bdi(folio_file_mapping(folio)->host)->wb,
+		wb_stat_mod(&inode_to_bdi(folio->mapping->host)->wb,
 			    WB_WRITEBACK, -nr);
 	}
 }
@@ -1125,7 +1108,7 @@ out_flushme:
 	 */
 	nfs_mark_request_dirty(req);
 	nfs_unlock_and_release_request(req);
-	error = nfs_wb_folio(folio_file_mapping(folio)->host, folio);
+	error = nfs_wb_folio(folio->mapping->host, folio);
 	return (error < 0) ? ERR_PTR(error) : NULL;
 }
 
@@ -1201,7 +1184,7 @@ int nfs_flush_incompatible(struct file *file, struct folio *folio)
 		nfs_release_request(req);
 		if (!do_flush)
 			return 0;
-		status = nfs_wb_folio(folio_file_mapping(folio)->host, folio);
+		status = nfs_wb_folio(folio->mapping->host, folio);
 	} while (status == 0);
 	return status;
 }
@@ -1275,7 +1258,7 @@ out:
  */
 static bool nfs_folio_write_uptodate(struct folio *folio, unsigned int pagelen)
 {
-	struct inode *inode = folio_file_mapping(folio)->host;
+	struct inode *inode = folio->mapping->host;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	if (nfs_have_delegated_attributes(inode))
@@ -1319,7 +1302,7 @@ static int nfs_can_extend_write(struct file *file, struct folio *folio,
 		return 0;
 	if (!nfs_folio_write_uptodate(folio, pagelen))
 		return 0;
-	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_WRITE))
+	if (nfs_have_write_delegation(inode))
 		return 1;
 	if (!flctx || (list_empty_careful(&flctx->flc_flock) &&
 		       list_empty_careful(&flctx->flc_posix)))
@@ -1353,7 +1336,7 @@ int nfs_update_folio(struct file *file, struct folio *folio,
 		     unsigned int offset, unsigned int count)
 {
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
-	struct address_space *mapping = folio_file_mapping(folio);
+	struct address_space *mapping = folio->mapping;
 	struct inode *inode = mapping->host;
 	unsigned int pagelen = nfs_folio_length(folio);
 	int		status = 0;
@@ -1367,8 +1350,12 @@ int nfs_update_folio(struct file *file, struct folio *folio,
 		goto out;
 
 	if (nfs_can_extend_write(file, folio, pagelen)) {
-		count = max(count + offset, pagelen);
+		unsigned int end = count + offset;
-		offset = 0;
+
+		offset = round_down(offset, PAGE_SIZE);
+		if (end < pagelen)
+			end = min(round_up(end, PAGE_SIZE), pagelen);
+		count = end - offset;
 	}
 
 	status = nfs_writepage_setup(ctx, folio, offset, count);
@@ -1513,6 +1500,13 @@ void nfs_writeback_update_inode(struct nfs_pgio_header *hdr)
 	struct nfs_fattr *fattr = &hdr->fattr;
 	struct inode *inode = hdr->inode;
 
+	if (nfs_have_delegated_mtime(inode)) {
+		spin_lock(&inode->i_lock);
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_BLOCKS);
+		spin_unlock(&inode->i_lock);
+		return;
+	}
+
 	spin_lock(&inode->i_lock);
 	nfs_writeback_check_extend(hdr, fattr);
 	nfs_post_op_update_inode_force_wcc_locked(inode, fattr);
@@ -1836,7 +1830,6 @@ static void nfs_commit_release_pages(struct nfs_commit_data *data)
 	struct nfs_page	*req;
 	int status = data->task.tk_status;
 	struct nfs_commit_info cinfo;
-	struct nfs_server *nfss;
 	struct folio *folio;
 
 	while (!list_empty(&data->pages)) {
@@ -1879,9 +1872,6 @@ static void nfs_commit_release_pages(struct nfs_commit_data *data)
 		/* Latency breaker */
 		cond_resched();
 	}
-	nfss = NFS_SERVER(data->inode);
-	if (atomic_long_read(&nfss->writeback) < NFS_CONGESTION_OFF_THRESH)
-		nfss->write_congested = 0;
 
 	nfs_init_cinfo(&cinfo, data->inode, data->dreq);
 	nfs_commit_end(cinfo.mds);
@@ -2072,17 +2062,17 @@ int nfs_wb_folio_cancel(struct inode *inode, struct folio *folio)
  */
 int nfs_wb_folio(struct inode *inode, struct folio *folio)
 {
-	loff_t range_start = folio_file_pos(folio);
+	loff_t range_start = folio_pos(folio);
-	loff_t range_end = range_start + (loff_t)folio_size(folio) - 1;
+	size_t len = folio_size(folio);
 	struct writeback_control wbc = {
 		.sync_mode = WB_SYNC_ALL,
 		.nr_to_write = 0,
 		.range_start = range_start,
-		.range_end = range_end,
+		.range_end = range_start + len - 1,
 	};
 	int ret;
 
-	trace_nfs_writeback_folio(inode, folio);
+	trace_nfs_writeback_folio(inode, range_start, len);
 
 	for (;;) {
 		folio_wait_writeback(folio);
@@ -2100,7 +2090,7 @@ int nfs_wb_folio(struct inode *inode, struct folio *folio)
 			goto out_error;
 	}
 out_error:
-	trace_nfs_writeback_folio_done(inode, folio, ret);
+	trace_nfs_writeback_folio_done(inode, range_start, len, ret);
 	return ret;
 }
 

fs/nfs_common/grace.c

@@ -139,6 +139,7 @@ exit_grace(void)
 }
 
 MODULE_AUTHOR("Jeff Layton <jlayton@primarydata.com>");
+MODULE_DESCRIPTION("NFS client and server infrastructure");
 MODULE_LICENSE("GPL");
 module_init(init_grace)
 module_exit(exit_grace)

fs/nfs_common/nfsacl.c

@@ -29,6 +29,7 @@
 #include <linux/nfs3.h>
 #include <linux/sort.h>
 
+MODULE_DESCRIPTION("NFS ACL support");
 MODULE_LICENSE("GPL");
 
 struct nfsacl_encode_desc {

include/linux/nfs4.h

@@ -367,6 +367,8 @@ enum open_delegation_type4 {
 	NFS4_OPEN_DELEGATE_READ = 1,
 	NFS4_OPEN_DELEGATE_WRITE = 2,
 	NFS4_OPEN_DELEGATE_NONE_EXT = 3, /* 4.1 */
+	NFS4_OPEN_DELEGATE_READ_ATTRS_DELEG = 4,
+	NFS4_OPEN_DELEGATE_WRITE_ATTRS_DELEG = 5,
 };
 
 enum why_no_delegation4 { /* new to v4.1 */
@@ -507,6 +509,12 @@ enum {
 	FATTR4_XATTR_SUPPORT		= 82,
 };
 
+enum {
+	FATTR4_TIME_DELEG_ACCESS	= 84,
+	FATTR4_TIME_DELEG_MODIFY	= 85,
+	FATTR4_OPEN_ARGUMENTS		= 86,
+};
+
 /*
  * The following internal definitions enable processing the above
  * attribute bits within 32-bit word boundaries.
@@ -586,6 +594,9 @@ enum {
 #define FATTR4_WORD2_SECURITY_LABEL     BIT(FATTR4_SEC_LABEL - 64)
 #define FATTR4_WORD2_MODE_UMASK		BIT(FATTR4_MODE_UMASK - 64)
 #define FATTR4_WORD2_XATTR_SUPPORT	BIT(FATTR4_XATTR_SUPPORT - 64)
+#define FATTR4_WORD2_TIME_DELEG_ACCESS	BIT(FATTR4_TIME_DELEG_ACCESS - 64)
+#define FATTR4_WORD2_TIME_DELEG_MODIFY	BIT(FATTR4_TIME_DELEG_MODIFY - 64)
+#define FATTR4_WORD2_OPEN_ARGUMENTS	BIT(FATTR4_OPEN_ARGUMENTS - 64)
 
 /* MDS threshold bitmap bits */
 #define THRESHOLD_RD                    (1UL << 0)

include/linux/nfs_fs_sb.h

@@ -140,6 +140,7 @@ struct nfs_server {
 	struct rpc_clnt *	client_acl;	/* ACL RPC client handle */
 	struct nlm_host		*nlm_host;	/* NLM client handle */
 	struct nfs_iostats __percpu *io_stats;	/* I/O statistics */
+	wait_queue_head_t	write_congestion_wait;	/* wait until write congestion eases */
 	atomic_long_t		writeback;	/* number of writeback pages */
 	unsigned int		write_congested;/* flag set when writeback gets too high */
 	unsigned int		flags;		/* various flags */
@@ -278,6 +279,9 @@ struct nfs_server {
 #define NFS_CAP_LGOPEN		(1U << 5)
 #define NFS_CAP_CASE_INSENSITIVE	(1U << 6)
 #define NFS_CAP_CASE_PRESERVING	(1U << 7)
+#define NFS_CAP_REBOOT_LAYOUTRETURN	(1U << 8)
+#define NFS_CAP_OPEN_XOR	(1U << 12)
+#define NFS_CAP_DELEGTIME	(1U << 13)
 #define NFS_CAP_POSIX_LOCK	(1U << 14)
 #define NFS_CAP_UIDGID_NOMAP	(1U << 15)
 #define NFS_CAP_STATEID_NFSV41	(1U << 16)

include/linux/nfs_page.h

@@ -152,11 +152,8 @@ extern	void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *, pgoff_t);
 extern size_t nfs_generic_pg_test(struct nfs_pageio_descriptor *desc,
 				struct nfs_page *prev,
 				struct nfs_page *req);
-extern  int nfs_wait_on_request(struct nfs_page *);
 extern	void nfs_unlock_request(struct nfs_page *req);
 extern	void nfs_unlock_and_release_request(struct nfs_page *);
-extern	struct nfs_page *nfs_page_group_lock_head(struct nfs_page *req);
-extern	int nfs_page_group_lock_subrequests(struct nfs_page *head);
 extern void nfs_join_page_group(struct nfs_page *head,
 				struct nfs_commit_info *cinfo,
 				struct inode *inode);
@@ -208,8 +205,8 @@ static inline struct inode *nfs_page_to_inode(const struct nfs_page *req)
 	struct folio *folio = nfs_page_to_folio(req);
 
 	if (folio == NULL)
-		return page_file_mapping(req->wb_page)->host;
+		return req->wb_page->mapping->host;
-	return folio_file_mapping(folio)->host;
+	return folio->mapping->host;
 }
 
 /**

include/linux/nfs_xdr.h

@@ -449,6 +449,22 @@ struct stateowner_id {
 	__u32	uniquifier;
 };
 
+struct nfs4_open_delegation {
+	__u32 open_delegation_type;
+	union {
+		struct {
+			fmode_t			type;
+			__u32			do_recall;
+			nfs4_stateid		stateid;
+			unsigned long		pagemod_limit;
+		};
+		struct {
+			__u32			why_no_delegation;
+			__u32			will_notify;
+		};
+	};
+};
+
 /*
  * Arguments to the open call.
  */
@@ -468,7 +484,7 @@ struct nfs_openargs {
 			nfs4_verifier   verifier; /* EXCLUSIVE */
 		};
 		nfs4_stateid	delegation;		/* CLAIM_DELEGATE_CUR */
-		fmode_t		delegation_type;	/* CLAIM_PREVIOUS */
+		__u32		delegation_type;	/* CLAIM_PREVIOUS */
 	} u;
 	const struct qstr *	name;
 	const struct nfs_server *server;	 /* Needed for ID mapping */
@@ -490,13 +506,10 @@ struct nfs_openres {
 	struct nfs_fattr *      f_attr;
 	struct nfs_seqid *	seqid;
 	const struct nfs_server *server;
-	fmode_t			delegation_type;
-	nfs4_stateid		delegation;
-	unsigned long		pagemod_limit;
-	__u32			do_recall;
 	__u32			attrset[NFS4_BITMAP_SIZE];
 	struct nfs4_string	*owner;
 	struct nfs4_string	*group_owner;
+	struct nfs4_open_delegation	delegation;
 	__u32			access_request;
 	__u32			access_supported;
 	__u32			access_result;
@@ -609,6 +622,13 @@ struct nfs_release_lockowner_res {
 	struct nfs4_sequence_res	seq_res;
 };
 
+struct nfs4_delegattr {
+	struct timespec64	atime;
+	struct timespec64	mtime;
+	bool			atime_set;
+	bool			mtime_set;
+};
+
 struct nfs4_delegreturnargs {
 	struct nfs4_sequence_args	seq_args;
 	const struct nfs_fh *fhandle;
@@ -616,6 +636,7 @@ struct nfs4_delegreturnargs {
 	const u32 *bitmask;
 	u32 bitmask_store[NFS_BITMASK_SZ];
 	struct nfs4_layoutreturn_args *lr_args;
+	struct nfs4_delegattr *sattr_args;
 };
 
 struct nfs4_delegreturnres {
@@ -624,6 +645,8 @@ struct nfs4_delegreturnres {
 	struct nfs_server *server;
 	struct nfs4_layoutreturn_res *lr_res;
 	int lr_ret;
+	bool sattr_res;
+	int sattr_ret;
 };
 
 /*
@@ -1190,6 +1213,14 @@ struct nfs4_statfs_res {
 	struct nfs_fsstat	       *fsstat;
 };
 
+struct nfs4_open_caps {
+	u32				oa_share_access[1];
+	u32				oa_share_deny[1];
+	u32				oa_share_access_want[1];
+	u32				oa_open_claim[1];
+	u32				oa_createmode[1];
+};
+
 struct nfs4_server_caps_arg {
 	struct nfs4_sequence_args	seq_args;
 	struct nfs_fh		       *fhandle;
@@ -1206,6 +1237,7 @@ struct nfs4_server_caps_res {
 	u32				fh_expire_type;
 	u32				case_insensitive;
 	u32				case_preserving;
+	struct nfs4_open_caps		open_caps;
 };
 
 #define NFS4_PATHNAME_MAXCOMPONENTS 512
@@ -1406,7 +1438,7 @@ struct nfs41_secinfo_no_name_args {
 
 struct nfs41_test_stateid_args {
 	struct nfs4_sequence_args	seq_args;
-	nfs4_stateid			*stateid;
+	nfs4_stateid			stateid;
 };
 
 struct nfs41_test_stateid_res {
@@ -1807,7 +1839,8 @@ struct nfs_rpc_ops {
 				int open_flags,
 				struct iattr *iattr,
 				int *);
-	int (*have_delegation)(struct inode *, fmode_t);
+	int (*have_delegation)(struct inode *, fmode_t, int);
+	int (*return_delegation)(struct inode *);
 	struct nfs_client *(*alloc_client) (const struct nfs_client_initdata *);
 	struct nfs_client *(*init_client) (struct nfs_client *,
 				const struct nfs_client_initdata *);

include/linux/sunrpc/rdma_rn.h

@@ -0,0 +1,27 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * * Copyright (c) 2024, Oracle and/or its affiliates.
+ */
+
+#ifndef _LINUX_SUNRPC_RDMA_RN_H
+#define _LINUX_SUNRPC_RDMA_RN_H
+
+#include <rdma/ib_verbs.h>
+
+/**
+ * rpcrdma_notification - request removal notification
+ */
+struct rpcrdma_notification {
+	void			(*rn_done)(struct rpcrdma_notification *rn);
+	u32			rn_index;
+};
+
+int rpcrdma_rn_register(struct ib_device *device,
+			struct rpcrdma_notification *rn,
+			void (*done)(struct rpcrdma_notification *rn));
+void rpcrdma_rn_unregister(struct ib_device *device,
+			   struct rpcrdma_notification *rn);
+int rpcrdma_ib_client_register(void);
+void rpcrdma_ib_client_unregister(void);
+
+#endif /* _LINUX_SUNRPC_RDMA_RN_H */

include/trace/events/rpcgss.h

@@ -54,7 +54,7 @@ TRACE_DEFINE_ENUM(GSS_S_UNSEQ_TOKEN);
 TRACE_DEFINE_ENUM(GSS_S_GAP_TOKEN);
 
 #define show_gss_status(x)						\
-	__print_flags(x, "|",						\
+	__print_symbolic(x, 						\
 		{ GSS_S_BAD_MECH, "GSS_S_BAD_MECH" },			\
 		{ GSS_S_BAD_NAME, "GSS_S_BAD_NAME" },			\
 		{ GSS_S_BAD_NAMETYPE, "GSS_S_BAD_NAMETYPE" },		\

include/trace/events/rpcrdma.h

@@ -669,6 +669,29 @@ TRACE_EVENT(xprtrdma_inline_thresh,
 DEFINE_CONN_EVENT(connect);
 DEFINE_CONN_EVENT(disconnect);
 
+TRACE_EVENT(xprtrdma_device_removal,
+	TP_PROTO(
+		const struct rdma_cm_id *id
+	),
+
+	TP_ARGS(id),
+
+	TP_STRUCT__entry(
+		__string(name, id->device->name)
+		__array(unsigned char, addr, sizeof(struct sockaddr_in6))
+	),
+
+	TP_fast_assign(
+		__assign_str(name);
+		memcpy(__entry->addr, &id->route.addr.dst_addr,
+		       sizeof(struct sockaddr_in6));
+	),
+
+	TP_printk("device %s to be removed, disconnecting %pISpc\n",
+		__get_str(name), __entry->addr
+	)
+);
+
 DEFINE_RXPRT_EVENT(xprtrdma_op_inject_dsc);
 
 TRACE_EVENT(xprtrdma_op_connect,
@@ -2220,6 +2243,40 @@ TRACE_EVENT(svcrdma_sq_post_err,
 	)
 );
 
+DECLARE_EVENT_CLASS(rpcrdma_client_device_class,
+	TP_PROTO(
+		const struct ib_device *device
+	),
+
+	TP_ARGS(device),
+
+	TP_STRUCT__entry(
+		__string(name, device->name)
+	),
+
+	TP_fast_assign(
+		__assign_str(name);
+	),
+
+	TP_printk("device=%s",
+		__get_str(name)
+	)
+);
+
+#define DEFINE_CLIENT_DEVICE_EVENT(name)				\
+	DEFINE_EVENT(rpcrdma_client_device_class, name,			\
+		TP_PROTO(						\
+			const struct ib_device *device			\
+		),							\
+		TP_ARGS(device)						\
+	)
+
+DEFINE_CLIENT_DEVICE_EVENT(rpcrdma_client_completion);
+DEFINE_CLIENT_DEVICE_EVENT(rpcrdma_client_add_one);
+DEFINE_CLIENT_DEVICE_EVENT(rpcrdma_client_remove_one);
+DEFINE_CLIENT_DEVICE_EVENT(rpcrdma_client_wait_on);
+DEFINE_CLIENT_DEVICE_EVENT(rpcrdma_client_remove_one_done);
+
 #endif /* _TRACE_RPCRDMA_H */
 
 #include <trace/define_trace.h>

include/uapi/linux/nfs4.h

@@ -46,6 +46,7 @@
 #define NFS4_OPEN_RESULT_CONFIRM		0x0002
 #define NFS4_OPEN_RESULT_LOCKTYPE_POSIX		0x0004
 #define NFS4_OPEN_RESULT_PRESERVE_UNLINKED	0x0008
+#define NFS4_OPEN_RESULT_NO_OPEN_STATEID	0x0010
 #define NFS4_OPEN_RESULT_MAY_NOTIFY_LOCK	0x0020
 
 #define NFS4_SHARE_ACCESS_MASK	0x000F
@@ -69,6 +70,9 @@
 #define NFS4_SHARE_SIGNAL_DELEG_WHEN_RESRC_AVAIL	0x10000
 #define NFS4_SHARE_PUSH_DELEG_WHEN_UNCONTENDED		0x20000
 
+#define NFS4_SHARE_WANT_DELEG_TIMESTAMPS		0x100000
+#define NFS4_SHARE_WANT_OPEN_XOR_DELEGATION		0x200000
+
 #define NFS4_CDFC4_FORE	0x1
 #define NFS4_CDFC4_BACK 0x2
 #define NFS4_CDFC4_BOTH 0x3

mm/filemap.c

@@ -3982,21 +3982,24 @@ ssize_t generic_perform_write(struct kiocb *iocb, struct iov_iter *i)
 	loff_t pos = iocb->ki_pos;
 	struct address_space *mapping = file->f_mapping;
 	const struct address_space_operations *a_ops = mapping->a_ops;
+	size_t chunk = mapping_max_folio_size(mapping);
 	long status = 0;
 	ssize_t written = 0;
 
 	do {
 		struct page *page;
-		unsigned long offset;	/* Offset into pagecache page */
+		struct folio *folio;
-		unsigned long bytes;	/* Bytes to write to page */
+		size_t offset;		/* Offset into folio */
+		size_t bytes;		/* Bytes to write to folio */
 		size_t copied;		/* Bytes copied from user */
 		void *fsdata = NULL;
 
-		offset = (pos & (PAGE_SIZE - 1));
+		bytes = iov_iter_count(i);
-		bytes = min_t(unsigned long, PAGE_SIZE - offset,
+retry:
-						iov_iter_count(i));
+		offset = pos & (chunk - 1);
+		bytes = min(chunk - offset, bytes);
+		balance_dirty_pages_ratelimited(mapping);
 
-again:
 		/*
 		 * Bring in the user page that we will copy from _first_.
 		 * Otherwise there's a nasty deadlock on copying from the
@@ -4018,11 +4021,16 @@ again:
 		if (unlikely(status < 0))
 			break;
 
-		if (mapping_writably_mapped(mapping))
+		folio = page_folio(page);
-			flush_dcache_page(page);
+		offset = offset_in_folio(folio, pos);
+		if (bytes > folio_size(folio) - offset)
+			bytes = folio_size(folio) - offset;
 
-		copied = copy_page_from_iter_atomic(page, offset, bytes, i);
+		if (mapping_writably_mapped(mapping))
-		flush_dcache_page(page);
+			flush_dcache_folio(folio);
+
+		copied = copy_folio_from_iter_atomic(folio, offset, bytes, i);
+		flush_dcache_folio(folio);
 
 		status = a_ops->write_end(file, mapping, pos, bytes, copied,
 						page, fsdata);
@@ -4040,14 +4048,16 @@ again:
 			 * halfway through, might be a race with munmap,
 			 * might be severe memory pressure.
 			 */
-			if (copied)
+			if (chunk > PAGE_SIZE)
+				chunk /= 2;
+			if (copied) {
 				bytes = copied;
-			goto again;
+				goto retry;
+			}
+		} else {
+			pos += status;
+			written += status;
 		}
-		pos += status;
-		written += status;
-
-		balance_dirty_pages_ratelimited(mapping);
 	} while (iov_iter_count(i));
 
 	if (!written)

net/sunrpc/clnt.c

@@ -2326,12 +2326,13 @@ call_transmit_status(struct rpc_task *task)
 		task->tk_action = call_transmit;
 		task->tk_status = 0;
 		break;
-	case -ECONNREFUSED:
 	case -EHOSTDOWN:
 	case -ENETDOWN:
 	case -EHOSTUNREACH:
 	case -ENETUNREACH:
 	case -EPERM:
+		break;
+	case -ECONNREFUSED:
 		if (RPC_IS_SOFTCONN(task)) {
 			if (!task->tk_msg.rpc_proc->p_proc)
 				trace_xprt_ping(task->tk_xprt,

net/sunrpc/sched.c

@@ -369,8 +369,10 @@ static void rpc_make_runnable(struct workqueue_struct *wq,
 	if (RPC_IS_ASYNC(task)) {
 		INIT_WORK(&task->u.tk_work, rpc_async_schedule);
 		queue_work(wq, &task->u.tk_work);
-	} else
+	} else {
+		smp_mb__after_atomic();
 		wake_up_bit(&task->tk_runstate, RPC_TASK_QUEUED);
+	}
 }
 
 /*

net/sunrpc/xprtrdma/Makefile

@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: GPL-2.0
 obj-$(CONFIG_SUNRPC_XPRT_RDMA) += rpcrdma.o
 
-rpcrdma-y := transport.o rpc_rdma.o verbs.o frwr_ops.o \
+rpcrdma-y := transport.o rpc_rdma.o verbs.o frwr_ops.o ib_client.o \
 	svc_rdma.o svc_rdma_backchannel.o svc_rdma_transport.o \
 	svc_rdma_sendto.o svc_rdma_recvfrom.o svc_rdma_rw.o \
 	svc_rdma_pcl.o module.o

net/sunrpc/xprtrdma/frwr_ops.c

@@ -54,7 +54,7 @@ static void frwr_cid_init(struct rpcrdma_ep *ep,
 	cid->ci_completion_id = mr->mr_ibmr->res.id;
 }
 
-static void frwr_mr_unmap(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr *mr)
+static void frwr_mr_unmap(struct rpcrdma_mr *mr)
 {
 	if (mr->mr_device) {
 		trace_xprtrdma_mr_unmap(mr);
@@ -73,7 +73,7 @@ void frwr_mr_release(struct rpcrdma_mr *mr)
 {
 	int rc;
 
-	frwr_mr_unmap(mr->mr_xprt, mr);
+	frwr_mr_unmap(mr);
 
 	rc = ib_dereg_mr(mr->mr_ibmr);
 	if (rc)
@@ -84,7 +84,7 @@ void frwr_mr_release(struct rpcrdma_mr *mr)
 
 static void frwr_mr_put(struct rpcrdma_mr *mr)
 {
-	frwr_mr_unmap(mr->mr_xprt, mr);
+	frwr_mr_unmap(mr);
 
 	/* The MR is returned to the req's MR free list instead
 	 * of to the xprt's MR free list. No spinlock is needed.
@@ -92,7 +92,8 @@ static void frwr_mr_put(struct rpcrdma_mr *mr)
 	rpcrdma_mr_push(mr, &mr->mr_req->rl_free_mrs);
 }
 
-/* frwr_reset - Place MRs back on the free list
+/**
+ * frwr_reset - Place MRs back on @req's free list
  * @req: request to reset
  *
  * Used after a failed marshal. For FRWR, this means the MRs

net/sunrpc/xprtrdma/ib_client.c

@@ -0,0 +1,181 @@
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
+/*
+ * Copyright (c) 2024 Oracle.  All rights reserved.
+ */
+
+/* #include <linux/module.h>
+#include <linux/slab.h> */
+#include <linux/xarray.h>
+#include <linux/types.h>
+#include <linux/kref.h>
+#include <linux/completion.h>
+
+#include <linux/sunrpc/svc_rdma.h>
+#include <linux/sunrpc/rdma_rn.h>
+
+#include "xprt_rdma.h"
+#include <trace/events/rpcrdma.h>
+
+/* Per-ib_device private data for rpcrdma */
+struct rpcrdma_device {
+	struct kref		rd_kref;
+	unsigned long		rd_flags;
+	struct ib_device	*rd_device;
+	struct xarray		rd_xa;
+	struct completion	rd_done;
+};
+
+#define RPCRDMA_RD_F_REMOVING	(0)
+
+static struct ib_client rpcrdma_ib_client;
+
+/*
+ * Listeners have no associated device, so we never register them.
+ * Note that ib_get_client_data() does not check if @device is
+ * NULL for us.
+ */
+static struct rpcrdma_device *rpcrdma_get_client_data(struct ib_device *device)
+{
+	if (!device)
+		return NULL;
+	return ib_get_client_data(device, &rpcrdma_ib_client);
+}
+
+/**
+ * rpcrdma_rn_register - register to get device removal notifications
+ * @device: device to monitor
+ * @rn: notification object that wishes to be notified
+ * @done: callback to notify caller of device removal
+ *
+ * Returns zero on success. The callback in rn_done is guaranteed
+ * to be invoked when the device is removed, unless this notification
+ * is unregistered first.
+ *
+ * On failure, a negative errno is returned.
+ */
+int rpcrdma_rn_register(struct ib_device *device,
+			struct rpcrdma_notification *rn,
+			void (*done)(struct rpcrdma_notification *rn))
+{
+	struct rpcrdma_device *rd = rpcrdma_get_client_data(device);
+
+	if (!rd || test_bit(RPCRDMA_RD_F_REMOVING, &rd->rd_flags))
+		return -ENETUNREACH;
+
+	kref_get(&rd->rd_kref);
+	if (xa_alloc(&rd->rd_xa, &rn->rn_index, rn, xa_limit_32b, GFP_KERNEL) < 0)
+		return -ENOMEM;
+	rn->rn_done = done;
+	return 0;
+}
+
+static void rpcrdma_rn_release(struct kref *kref)
+{
+	struct rpcrdma_device *rd = container_of(kref, struct rpcrdma_device,
+						 rd_kref);
+
+	trace_rpcrdma_client_completion(rd->rd_device);
+	complete(&rd->rd_done);
+}
+
+/**
+ * rpcrdma_rn_unregister - stop device removal notifications
+ * @device: monitored device
+ * @rn: notification object that no longer wishes to be notified
+ */
+void rpcrdma_rn_unregister(struct ib_device *device,
+			   struct rpcrdma_notification *rn)
+{
+	struct rpcrdma_device *rd = rpcrdma_get_client_data(device);
+
+	if (!rd)
+		return;
+
+	xa_erase(&rd->rd_xa, rn->rn_index);
+	kref_put(&rd->rd_kref, rpcrdma_rn_release);
+}
+
+/**
+ * rpcrdma_add_one - ib_client device insertion callback
+ * @device: device about to be inserted
+ *
+ * Returns zero on success. xprtrdma private data has been allocated
+ * for this device. On failure, a negative errno is returned.
+ */
+static int rpcrdma_add_one(struct ib_device *device)
+{
+	struct rpcrdma_device *rd;
+
+	rd = kzalloc(sizeof(*rd), GFP_KERNEL);
+	if (!rd)
+		return -ENOMEM;
+
+	kref_init(&rd->rd_kref);
+	xa_init_flags(&rd->rd_xa, XA_FLAGS_ALLOC1);
+	rd->rd_device = device;
+	init_completion(&rd->rd_done);
+	ib_set_client_data(device, &rpcrdma_ib_client, rd);
+
+	trace_rpcrdma_client_add_one(device);
+	return 0;
+}
+
+/**
+ * rpcrdma_remove_one - ib_client device removal callback
+ * @device: device about to be removed
+ * @client_data: this module's private per-device data
+ *
+ * Upon return, all transports associated with @device have divested
+ * themselves from IB hardware resources.
+ */
+static void rpcrdma_remove_one(struct ib_device *device,
+			       void *client_data)
+{
+	struct rpcrdma_device *rd = client_data;
+	struct rpcrdma_notification *rn;
+	unsigned long index;
+
+	trace_rpcrdma_client_remove_one(device);
+
+	set_bit(RPCRDMA_RD_F_REMOVING, &rd->rd_flags);
+	xa_for_each(&rd->rd_xa, index, rn)
+		rn->rn_done(rn);
+
+	/*
+	 * Wait only if there are still outstanding notification
+	 * registrants for this device.
+	 */
+	if (!refcount_dec_and_test(&rd->rd_kref.refcount)) {
+		trace_rpcrdma_client_wait_on(device);
+		wait_for_completion(&rd->rd_done);
+	}
+
+	trace_rpcrdma_client_remove_one_done(device);
+	kfree(rd);
+}
+
+static struct ib_client rpcrdma_ib_client = {
+	.name		= "rpcrdma",
+	.add		= rpcrdma_add_one,
+	.remove		= rpcrdma_remove_one,
+};
+
+/**
+ * rpcrdma_ib_client_unregister - unregister ib_client for xprtrdma
+ *
+ * cel: watch for orphaned rpcrdma_device objects on module unload
+ */
+void rpcrdma_ib_client_unregister(void)
+{
+	ib_unregister_client(&rpcrdma_ib_client);
+}
+
+/**
+ * rpcrdma_ib_client_register - register ib_client for rpcrdma
+ *
+ * Returns zero on success, or a negative errno.
+ */
+int rpcrdma_ib_client_register(void)
+{
+	return ib_register_client(&rpcrdma_ib_client);
+}

net/sunrpc/xprtrdma/module.c

@@ -11,6 +11,7 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/sunrpc/svc_rdma.h>
+#include <linux/sunrpc/rdma_rn.h>
 
 #include <asm/swab.h>
 
@@ -30,21 +31,32 @@ static void __exit rpc_rdma_cleanup(void)
 {
 	xprt_rdma_cleanup();
 	svc_rdma_cleanup();
+	rpcrdma_ib_client_unregister();
 }
 
 static int __init rpc_rdma_init(void)
 {
 	int rc;
 
+	rc = rpcrdma_ib_client_register();
+	if (rc)
+		goto out_rc;
+
 	rc = svc_rdma_init();
 	if (rc)
-		goto out;
+		goto out_ib_client;
 
 	rc = xprt_rdma_init();
 	if (rc)
-		svc_rdma_cleanup();
+		goto out_svc_rdma;
 
-out:
+	return 0;
+
+out_svc_rdma:
+	svc_rdma_cleanup();
+out_ib_client:
+	rpcrdma_ib_client_unregister();
+out_rc:
 	return rc;
 }
 

net/sunrpc/xprtrdma/rpc_rdma.c

@@ -1471,8 +1471,7 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *rep)
 		credits = 1;	/* don't deadlock */
 	else if (credits > r_xprt->rx_ep->re_max_requests)
 		credits = r_xprt->rx_ep->re_max_requests;
-	rpcrdma_post_recvs(r_xprt, credits + (buf->rb_bc_srv_max_requests << 1),
+	rpcrdma_post_recvs(r_xprt, credits + (buf->rb_bc_srv_max_requests << 1));
-			   false);
 	if (buf->rb_credits != credits)
 		rpcrdma_update_cwnd(r_xprt, credits);
 

net/sunrpc/xprtrdma/verbs.c

@@ -49,14 +49,14 @@
  *  o buffer memory
  */
 
+#include <linux/bitops.h>
 #include <linux/interrupt.h>
 #include <linux/slab.h>
 #include <linux/sunrpc/addr.h>
 #include <linux/sunrpc/svc_rdma.h>
 #include <linux/log2.h>
 
-#include <asm-generic/barrier.h>
+#include <asm/barrier.h>
-#include <asm/bitops.h>
 
 #include <rdma/ib_cm.h>
 
@@ -69,13 +69,15 @@ static void rpcrdma_sendctx_put_locked(struct rpcrdma_xprt *r_xprt,
 				       struct rpcrdma_sendctx *sc);
 static int rpcrdma_reqs_setup(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_reqs_reset(struct rpcrdma_xprt *r_xprt);
-static void rpcrdma_rep_destroy(struct rpcrdma_rep *rep);
 static void rpcrdma_reps_unmap(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_mrs_create(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_mrs_destroy(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_ep_get(struct rpcrdma_ep *ep);
 static int rpcrdma_ep_put(struct rpcrdma_ep *ep);
 static struct rpcrdma_regbuf *
+rpcrdma_regbuf_alloc_node(size_t size, enum dma_data_direction direction,
+			  int node);
+static struct rpcrdma_regbuf *
 rpcrdma_regbuf_alloc(size_t size, enum dma_data_direction direction);
 static void rpcrdma_regbuf_dma_unmap(struct rpcrdma_regbuf *rb);
 static void rpcrdma_regbuf_free(struct rpcrdma_regbuf *rb);
@@ -222,7 +224,6 @@ static void rpcrdma_update_cm_private(struct rpcrdma_ep *ep,
 static int
 rpcrdma_cm_event_handler(struct rdma_cm_id *id, struct rdma_cm_event *event)
 {
-	struct sockaddr *sap = (struct sockaddr *)&id->route.addr.dst_addr;
 	struct rpcrdma_ep *ep = id->context;
 
 	might_sleep();
@@ -241,14 +242,6 @@ rpcrdma_cm_event_handler(struct rdma_cm_id *id, struct rdma_cm_event *event)
 		ep->re_async_rc = -ENETUNREACH;
 		complete(&ep->re_done);
 		return 0;
-	case RDMA_CM_EVENT_DEVICE_REMOVAL:
-		pr_info("rpcrdma: removing device %s for %pISpc\n",
-			ep->re_id->device->name, sap);
-		switch (xchg(&ep->re_connect_status, -ENODEV)) {
-		case 0: goto wake_connect_worker;
-		case 1: goto disconnected;
-		}
-		return 0;
 	case RDMA_CM_EVENT_ADDR_CHANGE:
 		ep->re_connect_status = -ENODEV;
 		goto disconnected;
@@ -284,6 +277,14 @@ disconnected:
 	return 0;
 }
 
+static void rpcrdma_ep_removal_done(struct rpcrdma_notification *rn)
+{
+	struct rpcrdma_ep *ep = container_of(rn, struct rpcrdma_ep, re_rn);
+
+	trace_xprtrdma_device_removal(ep->re_id);
+	xprt_force_disconnect(ep->re_xprt);
+}
+
 static struct rdma_cm_id *rpcrdma_create_id(struct rpcrdma_xprt *r_xprt,
 					    struct rpcrdma_ep *ep)
 {
@@ -323,6 +324,10 @@ static struct rdma_cm_id *rpcrdma_create_id(struct rpcrdma_xprt *r_xprt,
 	if (rc)
 		goto out;
 
+	rc = rpcrdma_rn_register(id->device, &ep->re_rn, rpcrdma_ep_removal_done);
+	if (rc)
+		goto out;
+
 	return id;
 
 out:
@@ -350,6 +355,8 @@ static void rpcrdma_ep_destroy(struct kref *kref)
 		ib_dealloc_pd(ep->re_pd);
 	ep->re_pd = NULL;
 
+	rpcrdma_rn_unregister(ep->re_id->device, &ep->re_rn);
+
 	kfree(ep);
 	module_put(THIS_MODULE);
 }
@@ -505,7 +512,7 @@ int rpcrdma_xprt_connect(struct rpcrdma_xprt *r_xprt)
 	 * outstanding Receives.
 	 */
 	rpcrdma_ep_get(ep);
-	rpcrdma_post_recvs(r_xprt, 1, true);
+	rpcrdma_post_recvs(r_xprt, 1);
 
 	rc = rdma_connect(ep->re_id, &ep->re_remote_cma);
 	if (rc)
@@ -897,6 +904,8 @@ static int rpcrdma_reqs_setup(struct rpcrdma_xprt *r_xprt)
 
 static void rpcrdma_req_reset(struct rpcrdma_req *req)
 {
+	struct rpcrdma_mr *mr;
+
 	/* Credits are valid for only one connection */
 	req->rl_slot.rq_cong = 0;
 
@@ -906,7 +915,19 @@ static void rpcrdma_req_reset(struct rpcrdma_req *req)
 	rpcrdma_regbuf_dma_unmap(req->rl_sendbuf);
 	rpcrdma_regbuf_dma_unmap(req->rl_recvbuf);
 
-	frwr_reset(req);
+	/* The verbs consumer can't know the state of an MR on the
+	 * req->rl_registered list unless a successful completion
+	 * has occurred, so they cannot be re-used.
+	 */
+	while ((mr = rpcrdma_mr_pop(&req->rl_registered))) {
+		struct rpcrdma_buffer *buf = &mr->mr_xprt->rx_buf;
+
+		spin_lock(&buf->rb_lock);
+		list_del(&mr->mr_all);
+		spin_unlock(&buf->rb_lock);
+
+		frwr_mr_release(mr);
+	}
 }
 
 /* ASSUMPTION: the rb_allreqs list is stable for the duration,
@@ -924,18 +945,20 @@ static void rpcrdma_reqs_reset(struct rpcrdma_xprt *r_xprt)
 }
 
 static noinline
-struct rpcrdma_rep *rpcrdma_rep_create(struct rpcrdma_xprt *r_xprt,
+struct rpcrdma_rep *rpcrdma_rep_create(struct rpcrdma_xprt *r_xprt)
-				       bool temp)
 {
 	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
+	struct rpcrdma_ep *ep = r_xprt->rx_ep;
+	struct ib_device *device = ep->re_id->device;
 	struct rpcrdma_rep *rep;
 
 	rep = kzalloc(sizeof(*rep), XPRTRDMA_GFP_FLAGS);
 	if (rep == NULL)
 		goto out;
 
-	rep->rr_rdmabuf = rpcrdma_regbuf_alloc(r_xprt->rx_ep->re_inline_recv,
+	rep->rr_rdmabuf = rpcrdma_regbuf_alloc_node(ep->re_inline_recv,
-					       DMA_FROM_DEVICE);
+						    DMA_FROM_DEVICE,
+						    ibdev_to_node(device));
 	if (!rep->rr_rdmabuf)
 		goto out_free;
 
@@ -950,7 +973,6 @@ struct rpcrdma_rep *rpcrdma_rep_create(struct rpcrdma_xprt *r_xprt,
 	rep->rr_recv_wr.wr_cqe = &rep->rr_cqe;
 	rep->rr_recv_wr.sg_list = &rep->rr_rdmabuf->rg_iov;
 	rep->rr_recv_wr.num_sge = 1;
-	rep->rr_temp = temp;
 
 	spin_lock(&buf->rb_lock);
 	list_add(&rep->rr_all, &buf->rb_all_reps);
@@ -969,17 +991,6 @@ static void rpcrdma_rep_free(struct rpcrdma_rep *rep)
 	kfree(rep);
 }
 
-static void rpcrdma_rep_destroy(struct rpcrdma_rep *rep)
-{
-	struct rpcrdma_buffer *buf = &rep->rr_rxprt->rx_buf;
-
-	spin_lock(&buf->rb_lock);
-	list_del(&rep->rr_all);
-	spin_unlock(&buf->rb_lock);
-
-	rpcrdma_rep_free(rep);
-}
-
 static struct rpcrdma_rep *rpcrdma_rep_get_locked(struct rpcrdma_buffer *buf)
 {
 	struct llist_node *node;
@@ -1011,10 +1022,8 @@ static void rpcrdma_reps_unmap(struct rpcrdma_xprt *r_xprt)
 	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
 	struct rpcrdma_rep *rep;
 
-	list_for_each_entry(rep, &buf->rb_all_reps, rr_all) {
+	list_for_each_entry(rep, &buf->rb_all_reps, rr_all)
 		rpcrdma_regbuf_dma_unmap(rep->rr_rdmabuf);
-		rep->rr_temp = true;	/* Mark this rep for destruction */
-	}
 }
 
 static void rpcrdma_reps_destroy(struct rpcrdma_buffer *buf)
@@ -1231,14 +1240,15 @@ void rpcrdma_buffer_put(struct rpcrdma_buffer *buffers, struct rpcrdma_req *req)
  * or Replies they may be registered externally via frwr_map.
  */
 static struct rpcrdma_regbuf *
-rpcrdma_regbuf_alloc(size_t size, enum dma_data_direction direction)
+rpcrdma_regbuf_alloc_node(size_t size, enum dma_data_direction direction,
+			  int node)
 {
 	struct rpcrdma_regbuf *rb;
 
-	rb = kmalloc(sizeof(*rb), XPRTRDMA_GFP_FLAGS);
+	rb = kmalloc_node(sizeof(*rb), XPRTRDMA_GFP_FLAGS, node);
 	if (!rb)
 		return NULL;
-	rb->rg_data = kmalloc(size, XPRTRDMA_GFP_FLAGS);
+	rb->rg_data = kmalloc_node(size, XPRTRDMA_GFP_FLAGS, node);
 	if (!rb->rg_data) {
 		kfree(rb);
 		return NULL;
@@ -1250,6 +1260,12 @@ rpcrdma_regbuf_alloc(size_t size, enum dma_data_direction direction)
 	return rb;
 }
 
+static struct rpcrdma_regbuf *
+rpcrdma_regbuf_alloc(size_t size, enum dma_data_direction direction)
+{
+	return rpcrdma_regbuf_alloc_node(size, direction, NUMA_NO_NODE);
+}
+
 /**
  * rpcrdma_regbuf_realloc - re-allocate a SEND/RECV buffer
  * @rb: regbuf to reallocate
@@ -1327,10 +1343,9 @@ static void rpcrdma_regbuf_free(struct rpcrdma_regbuf *rb)
  * rpcrdma_post_recvs - Refill the Receive Queue
  * @r_xprt: controlling transport instance
  * @needed: current credit grant
- * @temp: mark Receive buffers to be deleted after one use
  *
  */
-void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed, bool temp)
+void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed)
 {
 	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
 	struct rpcrdma_ep *ep = r_xprt->rx_ep;
@@ -1344,8 +1359,7 @@ void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed, bool temp)
 	if (likely(ep->re_receive_count > needed))
 		goto out;
 	needed -= ep->re_receive_count;
-	if (!temp)
+	needed += RPCRDMA_MAX_RECV_BATCH;
-		needed += RPCRDMA_MAX_RECV_BATCH;
 
 	if (atomic_inc_return(&ep->re_receiving) > 1)
 		goto out;
@@ -1354,12 +1368,8 @@ void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed, bool temp)
 	wr = NULL;
 	while (needed) {
 		rep = rpcrdma_rep_get_locked(buf);
-		if (rep && rep->rr_temp) {
-			rpcrdma_rep_destroy(rep);
-			continue;
-		}
 		if (!rep)
-			rep = rpcrdma_rep_create(r_xprt, temp);
+			rep = rpcrdma_rep_create(r_xprt);
 		if (!rep)
 			break;
 		if (!rpcrdma_regbuf_dma_map(r_xprt, rep->rr_rdmabuf)) {

net/sunrpc/xprtrdma/xprt_rdma.h

@@ -56,6 +56,7 @@
 #include <linux/sunrpc/rpc_rdma_cid.h> 	/* completion IDs */
 #include <linux/sunrpc/rpc_rdma.h> 	/* RPC/RDMA protocol */
 #include <linux/sunrpc/xprtrdma.h> 	/* xprt parameters */
+#include <linux/sunrpc/rdma_rn.h>	/* removal notifications */
 
 #define RDMA_RESOLVE_TIMEOUT	(5000)	/* 5 seconds */
 #define RDMA_CONNECT_RETRY_MAX	(2)	/* retries if no listener backlog */
@@ -92,6 +93,7 @@ struct rpcrdma_ep {
 	struct rpcrdma_connect_private
 				re_cm_private;
 	struct rdma_conn_param	re_remote_cma;
+	struct rpcrdma_notification	re_rn;
 	int			re_receive_count;
 	unsigned int		re_max_requests; /* depends on device */
 	unsigned int		re_inline_send;	/* negotiated */
@@ -198,7 +200,6 @@ struct rpcrdma_rep {
 	__be32			rr_proc;
 	int			rr_wc_flags;
 	u32			rr_inv_rkey;
-	bool			rr_temp;
 	struct rpcrdma_regbuf	*rr_rdmabuf;
 	struct rpcrdma_xprt	*rr_rxprt;
 	struct rpc_rqst		*rr_rqst;
@@ -466,7 +467,7 @@ void rpcrdma_flush_disconnect(struct rpcrdma_xprt *r_xprt, struct ib_wc *wc);
 int rpcrdma_xprt_connect(struct rpcrdma_xprt *r_xprt);
 void rpcrdma_xprt_disconnect(struct rpcrdma_xprt *r_xprt);
 
-void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed, bool temp);
+void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, int needed);
 
 /*
  * Buffer calls - xprtrdma/verbs.c