Merge tag 'nfs-for-4.18-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

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

Stable fixes:

- Fix a 1-byte stack overflow in nfs_idmap_read_and_verify_message

- Fix a hang due to incorrect error returns in rpcrdma_convert_iovs()

- Revert an incorrect change to the NFSv4.1 callback channel

- Fix a bug in the NFSv4.1 sequence error handling

Features and optimisations:

- Support for piggybacking a LAYOUTGET operation to the OPEN compound

- RDMA performance enhancements to deal with transport congestion

- Add proper SPDX tags for NetApp-contributed RDMA source

- Do not request delegated file attributes (size+change) from the
server

- Optimise away a GETATTR in the lookup revalidate code when doing
NFSv4 OPEN

- Optimise away unnecessary lookups for rename targets

- Misc performance improvements when freeing NFSv4 delegations

Bugfixes and cleanups:

- Try to fail quickly if proto=rdma

- Clean up RDMA receive trace points

- Fix sillyrename to return the delegation when appropriate

- Misc attribute revalidation fixes

- Immediately clear the pNFS layout on a file when the server returns
ESTALE

- Return NFS4ERR_DELAY when delegation/layout recalls fail due to
igrab()

- Fix the client behaviour on NFS4ERR_SEQ_FALSE_RETRY"

* tag 'nfs-for-4.18-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (80 commits)
skip LAYOUTRETURN if layout is invalid
NFSv4.1: Fix the client behaviour on NFS4ERR_SEQ_FALSE_RETRY
NFSv4: Fix a typo in nfs41_sequence_process
NFSv4: Revert commit 5f83d86cf531d ("NFSv4.x: Fix wraparound issues..")
NFSv4: Return NFS4ERR_DELAY when a layout recall fails due to igrab()
NFSv4: Return NFS4ERR_DELAY when a delegation recall fails due to igrab()
NFSv4.0: Remove transport protocol name from non-UCS client ID
NFSv4.0: Remove cl_ipaddr from non-UCS client ID
NFSv4: Fix a compiler warning when CONFIG_NFS_V4_1 is undefined
NFS: Filter cache invalidation when holding a delegation
NFS: Ignore NFS_INO_REVAL_FORCED in nfs_check_inode_attributes()
NFS: Improve caching while holding a delegation
NFS: Fix attribute revalidation
NFS: fix up nfs_setattr_update_inode
NFSv4: Ensure the inode is clean when we set a delegation
NFSv4: Ignore NFS_INO_REVAL_FORCED in nfs4_proc_access
NFSv4: Don't ask for delegated attributes when adding a hard link
NFSv4: Don't ask for delegated attributes when revalidating the inode
NFS: Pass the inode down to the getattr() callback
NFSv4: Don't request size+change attribute if they are delegated to us
...

fs/nfs/callback_proc.c

···
 		rpc_peeraddr2str(cps->clp->cl_rpcclient, RPC_DISPLAY_ADDR));
 
 	inode = nfs_delegation_find_inode(cps->clp, &args->fh);
-	if (inode == NULL) {
+	if (IS_ERR(inode)) {
+		if (inode == ERR_PTR(-EAGAIN))
+			res->status = htonl(NFS4ERR_DELAY);
 		trace_nfs4_cb_getattr(cps->clp, &args->fh, NULL,
 				-ntohl(res->status));
 		goto out;
···
 
 	res = htonl(NFS4ERR_BADHANDLE);
 	inode = nfs_delegation_find_inode(cps->clp, &args->fh);
-	if (inode == NULL) {
+	if (IS_ERR(inode)) {
+		if (inode == ERR_PTR(-EAGAIN))
+			res = htonl(NFS4ERR_DELAY);
 		trace_nfs4_cb_recall(cps->clp, &args->fh, NULL,
 				&args->stateid, -ntohl(res));
 		goto out;
···
 	struct inode *inode;
 	struct pnfs_layout_hdr *lo;
 
-restart:
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
 		list_for_each_entry(lo, &server->layouts, plh_layouts) {
 			if (stateid != NULL &&
···
 				continue;
 			inode = igrab(lo->plh_inode);
 			if (!inode)
-				continue;
+				return ERR_PTR(-EAGAIN);
 			if (!nfs_sb_active(inode->i_sb)) {
 				rcu_read_unlock();
 				spin_unlock(&clp->cl_lock);
 				iput(inode);
 				spin_lock(&clp->cl_lock);
 				rcu_read_lock();
-				goto restart;
+				return ERR_PTR(-EAGAIN);
 			}
 			return inode;
 		}
 	}
 
-	return NULL;
+	return ERR_PTR(-ENOENT);
 }
 
 /*
···
 	struct inode *inode;
 	struct pnfs_layout_hdr *lo;
 
-restart:
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
 		list_for_each_entry(lo, &server->layouts, plh_layouts) {
 			nfsi = NFS_I(lo->plh_inode);
···
 				continue;
 			inode = igrab(lo->plh_inode);
 			if (!inode)
-				continue;
+				return ERR_PTR(-EAGAIN);
 			if (!nfs_sb_active(inode->i_sb)) {
 				rcu_read_unlock();
 				spin_unlock(&clp->cl_lock);
 				iput(inode);
 				spin_lock(&clp->cl_lock);
 				rcu_read_lock();
-				goto restart;
+				return ERR_PTR(-EAGAIN);
 			}
 			return inode;
 		}
 	}
 
-	return NULL;
+	return ERR_PTR(-ENOENT);
 }
 
 static struct inode *nfs_layout_find_inode(struct nfs_client *clp,
···
 	spin_lock(&clp->cl_lock);
 	rcu_read_lock();
 	inode = nfs_layout_find_inode_by_stateid(clp, stateid);
-	if (!inode)
+	if (inode == ERR_PTR(-ENOENT))
 		inode = nfs_layout_find_inode_by_fh(clp, fh);
 	rcu_read_unlock();
 	spin_unlock(&clp->cl_lock);
···
 	LIST_HEAD(free_me_list);
 
 	ino = nfs_layout_find_inode(clp, &args->cbl_fh, &args->cbl_stateid);
-	if (!ino)
-		goto out;
+	if (IS_ERR(ino)) {
+		if (ino == ERR_PTR(-EAGAIN))
+			rv = NFS4ERR_DELAY;
+		goto out_noput;
+	}
 
 	pnfs_layoutcommit_inode(ino, false);
 
···
 	nfs_commit_inode(ino, 0);
 	pnfs_put_layout_hdr(lo);
 out:
+	nfs_iput_and_deactive(ino);
+out_noput:
 	trace_nfs4_cb_layoutrecall_file(clp, &args->cbl_fh, ino,
 			&args->cbl_stateid, -rv);
-	nfs_iput_and_deactive(ino);
 	return rv;
 }
 
···
 static u32 do_callback_layoutrecall(struct nfs_client *clp,
 				    struct cb_layoutrecallargs *args)
 {
+	write_seqcount_begin(&clp->cl_callback_count);
+	write_seqcount_end(&clp->cl_callback_count);
 	if (args->cbl_recall_type == RETURN_FILE)
 		return initiate_file_draining(clp, args);
 	return initiate_bulk_draining(clp, args);
···
 		return htonl(NFS4ERR_SEQ_FALSE_RETRY);
 	}
 
-	/* Wraparound */
-	if (unlikely(slot->seq_nr == 0xFFFFFFFFU)) {
-		if (args->csa_sequenceid == 1)
-			return htonl(NFS4_OK);
-	} else if (likely(args->csa_sequenceid == slot->seq_nr + 1))
+	/* Note: wraparound relies on seq_nr being of type u32 */
+	if (likely(args->csa_sequenceid == slot->seq_nr + 1))
 		return htonl(NFS4_OK);
 
 	/* Misordered request */

fs/nfs/client.c

···
 	}
 
 	if (!(fattr->valid & NFS_ATTR_FATTR)) {
-		error = nfs_mod->rpc_ops->getattr(server, mount_info->mntfh, fattr, NULL);
+		error = nfs_mod->rpc_ops->getattr(server, mount_info->mntfh,
+				fattr, NULL, NULL);
 		if (error < 0) {
 			dprintk("nfs_create_server: getattr error = %d\n", -error);
 			goto error;

fs/nfs/delegation.c

···
 
 	trace_nfs4_set_delegation(inode, type);
 
+	spin_lock(&inode->i_lock);
+	if (NFS_I(inode)->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME))
+		NFS_I(inode)->cache_validity |= NFS_INO_REVAL_FORCED;
+	spin_unlock(&inode->i_lock);
 out:
 	spin_unlock(&clp->cl_lock);
 	if (delegation != NULL)
···
 int nfs_client_return_marked_delegations(struct nfs_client *clp)
 {
 	struct nfs_delegation *delegation;
+	struct nfs_delegation *prev;
 	struct nfs_server *server;
 	struct inode *inode;
+	struct inode *place_holder = NULL;
+	struct nfs_delegation *place_holder_deleg = NULL;
 	int err = 0;
 
 restart:
+	/*
+	 * To avoid quadratic looping we hold a reference
+	 * to an inode place_holder.  Each time we restart, we
+	 * list nfs_servers from the server of that inode, and
+	 * delegation in the server from the delegations of that
+	 * inode.
+	 * prev is an RCU-protected pointer to a delegation which
+	 * wasn't marked for return and might be a good choice for
+	 * the next place_holder.
+	 */
 	rcu_read_lock();
-	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		list_for_each_entry_rcu(delegation, &server->delegations,
-								super_list) {
-			if (!nfs_delegation_need_return(delegation))
+	prev = NULL;
+	if (place_holder)
+		server = NFS_SERVER(place_holder);
+	else
+		server = list_entry_rcu(clp->cl_superblocks.next,
+					struct nfs_server, client_link);
+	list_for_each_entry_from_rcu(server, &clp->cl_superblocks, client_link) {
+		delegation = NULL;
+		if (place_holder && server == NFS_SERVER(place_holder))
+			delegation = rcu_dereference(NFS_I(place_holder)->delegation);
+		if (!delegation || delegation != place_holder_deleg)
+			delegation = list_entry_rcu(server->delegations.next,
+						    struct nfs_delegation, super_list);
+		list_for_each_entry_from_rcu(delegation, &server->delegations, super_list) {
+			struct inode *to_put = NULL;
+
+			if (!nfs_delegation_need_return(delegation)) {
+				prev = delegation;
 				continue;
+			}
 			if (!nfs_sb_active(server->super))
-				continue;
+				break; /* continue in outer loop */
+
+			if (prev) {
+				struct inode *tmp;
+
+				tmp = nfs_delegation_grab_inode(prev);
+				if (tmp) {
+					to_put = place_holder;
+					place_holder = tmp;
+					place_holder_deleg = prev;
+				}
+			}
+
 			inode = nfs_delegation_grab_inode(delegation);
 			if (inode == NULL) {
 				rcu_read_unlock();
+				if (to_put)
+					iput(to_put);
 				nfs_sb_deactive(server->super);
 				goto restart;
 			}
 			delegation = nfs_start_delegation_return_locked(NFS_I(inode));
 			rcu_read_unlock();
 
+			if (to_put)
+				iput(to_put);
+
 			err = nfs_end_delegation_return(inode, delegation, 0);
 			iput(inode);
 			nfs_sb_deactive(server->super);
+			cond_resched();
 			if (!err)
 				goto restart;
 			set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
+			if (place_holder)
+				iput(place_holder);
 			return err;
 		}
 	}
 	rcu_read_unlock();
+	if (place_holder)
+		iput(place_holder);
 	return 0;
 }
 
···
 		if (delegation->inode != NULL &&
 		    nfs_compare_fh(fhandle, &NFS_I(delegation->inode)->fh) == 0) {
 			res = igrab(delegation->inode);
+			spin_unlock(&delegation->lock);
+			if (res != NULL)
+				return res;
+			return ERR_PTR(-EAGAIN);
 		}
 		spin_unlock(&delegation->lock);
-		if (res != NULL)
-			break;
 	}
-	return res;
+	return ERR_PTR(-ENOENT);
 }
 
 /**
···
 					const struct nfs_fh *fhandle)
 {
 	struct nfs_server *server;
-	struct inode *res = NULL;
+	struct inode *res;
 
 	rcu_read_lock();
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
 		res = nfs_delegation_find_inode_server(server, fhandle);
-		if (res != NULL)
-			break;
+		if (res != ERR_PTR(-ENOENT))
+			return res;
 	}
 	rcu_read_unlock();
-	return res;
+	return ERR_PTR(-ENOENT);
 }
 
 static void nfs_delegation_mark_reclaim_server(struct nfs_server *server)
···
 						&delegation->flags) == 0)
 				continue;
 			if (!nfs_sb_active(server->super))
-				continue;
+				break; /* continue in outer loop */
 			inode = nfs_delegation_grab_inode(delegation);
 			if (inode == NULL) {
 				rcu_read_unlock();
···
 			}
 			iput(inode);
 			nfs_sb_deactive(server->super);
+			cond_resched();
 			goto restart;
 		}
 	}
···
 						&delegation->flags) == 0)
 				continue;
 			if (!nfs_sb_active(server->super))
-				continue;
+				break; /* continue in outer loop */
 			inode = nfs_delegation_grab_inode(delegation);
 			if (inode == NULL) {
 				rcu_read_unlock();
···
 			}
 			iput(inode);
 			nfs_sb_deactive(server->super);
+			cond_resched();
 			goto restart;
 		}
 	}

fs/nfs/dir.c

···
 
 	if (IS_AUTOMOUNT(inode))
 		return 0;
+
+	if (flags & LOOKUP_OPEN) {
+		switch (inode->i_mode & S_IFMT) {
+		case S_IFREG:
+			/* A NFSv4 OPEN will revalidate later */
+			if (server->caps & NFS_CAP_ATOMIC_OPEN)
+				goto out;
+			/* Fallthrough */
+		case S_IFDIR:
+			if (server->flags & NFS_MOUNT_NOCTO)
+				break;
+			/* NFS close-to-open cache consistency validation */
+			goto out_force;
+		}
+	}
+
 	/* VFS wants an on-the-wire revalidation */
 	if (flags & LOOKUP_REVAL)
-		goto out_force;
-	/* This is an open(2) */
-	if ((flags & LOOKUP_OPEN) && !(server->flags & NFS_MOUNT_NOCTO) &&
-	    (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
 		goto out_force;
 out:
 	return (inode->i_nlink == 0) ? -ENOENT : 0;
···
  *
  * If LOOKUP_RCU prevents us from performing a full check, return 1
  * suggesting a reval is needed.
+ *
+ * Note that when creating a new file, or looking up a rename target,
+ * then it shouldn't be necessary to revalidate a negative dentry.
  */
 static inline
 int nfs_neg_need_reval(struct inode *dir, struct dentry *dentry,
 		       unsigned int flags)
 {
-	/* Don't revalidate a negative dentry if we're creating a new file */
-	if (flags & LOOKUP_CREATE)
+	if (flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
 		return 0;
 	if (NFS_SERVER(dir)->flags & NFS_MOUNT_LOOKUP_CACHE_NONEG)
 		return 1;
···
 		goto out_set_verifier;
 
 	/* Force a full look up iff the parent directory has changed */
-	if (!nfs_is_exclusive_create(dir, flags) &&
+	if (!(flags & (LOOKUP_EXCL | LOOKUP_REVAL)) &&
 	    nfs_check_verifier(dir, dentry, flags & LOOKUP_RCU)) {
 		error = nfs_lookup_verify_inode(inode, flags);
 		if (error) {
···
 {
 	spin_lock(&inode->i_lock);
 	/* drop the inode if we're reasonably sure this is the last link */
-	if (inode->i_nlink == 1)
-		clear_nlink(inode);
+	if (inode->i_nlink > 0)
+		drop_nlink(inode);
+	NFS_I(inode)->attr_gencount = nfs_inc_attr_generation_counter();
 	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
 		| NFS_INO_INVALID_CTIME
-		| NFS_INO_INVALID_OTHER;
+		| NFS_INO_INVALID_OTHER
+		| NFS_INO_REVAL_FORCED;
 	spin_unlock(&inode->i_lock);
 }
 
···
 	 * If we're doing an exclusive create, optimize away the lookup
 	 * but don't hash the dentry.
 	 */
-	if (nfs_is_exclusive_create(dir, flags))
+	if (nfs_is_exclusive_create(dir, flags) || flags & LOOKUP_RENAME_TARGET)
 		return NULL;
 
 	res = ERR_PTR(-ENOMEM);
···
 	nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 	if (!(fattr->valid & NFS_ATTR_FATTR)) {
 		struct nfs_server *server = NFS_SB(dentry->d_sb);
-		error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr, NULL);
+		error = server->nfs_client->rpc_ops->getattr(server, fhandle,
+				fattr, NULL, NULL);
 		if (error < 0)
 			goto out_error;
 	}
···
 	} else
 		error = task->tk_status;
 	rpc_put_task(task);
-	nfs_mark_for_revalidate(old_inode);
+	/* Ensure the inode attributes are revalidated */
+	if (error == 0) {
+		spin_lock(&old_inode->i_lock);
+		NFS_I(old_inode)->attr_gencount = nfs_inc_attr_generation_counter();
+		NFS_I(old_inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+			| NFS_INO_INVALID_CTIME
+			| NFS_INO_REVAL_FORCED;
+		spin_unlock(&old_inode->i_lock);
+	}
 out:
 	if (rehash)
 		d_rehash(rehash);

fs/nfs/export.c

···
 	}
 
 	rpc_ops = NFS_SB(sb)->nfs_client->rpc_ops;
-	ret = rpc_ops->getattr(NFS_SB(sb), server_fh, fattr, label);
+	ret = rpc_ops->getattr(NFS_SB(sb), server_fh, fattr, label, NULL);
 	if (ret) {
 		dprintk("%s: getattr failed %d\n", __func__, ret);
 		dentry = ERR_PTR(ret);

fs/nfs/flexfilelayout/flexfilelayout.c

···
 	.id			= LAYOUT_FLEX_FILES,
 	.name			= "LAYOUT_FLEX_FILES",
 	.owner			= THIS_MODULE,
+	.flags			= PNFS_LAYOUTGET_ON_OPEN,
 	.set_layoutdriver	= ff_layout_set_layoutdriver,
 	.alloc_layout_hdr	= ff_layout_alloc_layout_hdr,
 	.free_layout_hdr	= ff_layout_free_layout_hdr,

fs/nfs/inode.c

···
 static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
-	bool have_delegation = nfs_have_delegated_attributes(inode);
+	bool have_delegation = NFS_PROTO(inode)->have_delegation(inode, FMODE_READ);
 
-	if (have_delegation)
-		flags &= ~(NFS_INO_INVALID_CHANGE|NFS_INO_REVAL_PAGECACHE);
+	if (have_delegation) {
+		if (!(flags & NFS_INO_REVAL_FORCED))
+			flags &= ~NFS_INO_INVALID_OTHER;
+		flags &= ~(NFS_INO_INVALID_CHANGE
+				| NFS_INO_INVALID_SIZE
+				| NFS_INO_REVAL_PAGECACHE);
+	}
+
 	if (inode->i_mapping->nrpages == 0)
 		flags &= ~NFS_INO_INVALID_DATA;
 	nfsi->cache_validity |= flags;
···
 		/* We can't support update_atime(), since the server will reset it */
 		inode->i_flags |= S_NOATIME|S_NOCMTIME;
 		inode->i_mode = fattr->mode;
+		nfsi->cache_validity = 0;
 		if ((fattr->valid & NFS_ATTR_FATTR_MODE) == 0
 				&& nfs_server_capable(inode, NFS_CAP_MODE))
 			nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
···
 			 */
 			inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used);
 		}
+
+		if (nfsi->cache_validity != 0)
+			nfsi->cache_validity |= NFS_INO_REVAL_FORCED;
 
 		nfs_setsecurity(inode, fattr, label);
 
···
 
 	spin_lock(&inode->i_lock);
 	NFS_I(inode)->attr_gencount = fattr->gencount;
-	nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE
-			| NFS_INO_INVALID_CTIME);
+	if ((attr->ia_valid & ATTR_SIZE) != 0) {
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
+		nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC);
+		nfs_vmtruncate(inode, attr->ia_size);
+	}
 	if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) {
+		NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_CTIME;
 		if ((attr->ia_valid & ATTR_MODE) != 0) {
 			int mode = attr->ia_mode & S_IALLUGO;
 			mode |= inode->i_mode & ~S_IALLUGO;
···
 			inode->i_uid = attr->ia_uid;
 		if ((attr->ia_valid & ATTR_GID) != 0)
 			inode->i_gid = attr->ia_gid;
+		if (fattr->valid & NFS_ATTR_FATTR_CTIME)
+			inode->i_ctime = fattr->ctime;
+		else
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE
+					| NFS_INO_INVALID_CTIME);
 		nfs_set_cache_invalid(inode, NFS_INO_INVALID_ACCESS
 				| NFS_INO_INVALID_ACL);
 	}
-	if ((attr->ia_valid & ATTR_SIZE) != 0) {
-		nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
-		nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC);
-		nfs_vmtruncate(inode, attr->ia_size);
+	if (attr->ia_valid & (ATTR_ATIME_SET|ATTR_ATIME)) {
+		NFS_I(inode)->cache_validity &= ~(NFS_INO_INVALID_ATIME
+				| NFS_INO_INVALID_CTIME);
+		if (fattr->valid & NFS_ATTR_FATTR_ATIME)
+			inode->i_atime = fattr->atime;
+		else if (attr->ia_valid & ATTR_ATIME_SET)
+			inode->i_atime = attr->ia_atime;
+		else
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATIME);
+
+		if (fattr->valid & NFS_ATTR_FATTR_CTIME)
+			inode->i_ctime = fattr->ctime;
+		else
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE
+					| NFS_INO_INVALID_CTIME);
+	}
+	if (attr->ia_valid & (ATTR_MTIME_SET|ATTR_MTIME)) {
+		NFS_I(inode)->cache_validity &= ~(NFS_INO_INVALID_MTIME
+				| NFS_INO_INVALID_CTIME);
+		if (fattr->valid & NFS_ATTR_FATTR_MTIME)
+			inode->i_mtime = fattr->mtime;
+		else if (attr->ia_valid & ATTR_MTIME_SET)
+			inode->i_mtime = attr->ia_mtime;
+		else
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
+
+		if (fattr->valid & NFS_ATTR_FATTR_CTIME)
+			inode->i_ctime = fattr->ctime;
+		else
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE
+					| NFS_INO_INVALID_CTIME);
 	}
 	if (fattr->valid)
 		nfs_update_inode(inode, fattr);
···
 		goto out;
 	}
 
-	status = NFS_PROTO(inode)->getattr(server, NFS_FH(inode), fattr, label);
+	status = NFS_PROTO(inode)->getattr(server, NFS_FH(inode), fattr,
+			label, inode);
 	if (status != 0) {
 		dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Lu) getattr failed, error=%d\n",
 			 inode->i_sb->s_id,
···
 	unsigned long invalid = 0;
 
 
-	if (nfs_have_delegated_attributes(inode))
+	if (NFS_PROTO(inode)->have_delegation(inode, FMODE_READ))
 		return 0;
+
 	/* Has the inode gone and changed behind our back? */
 	if ((fattr->valid & NFS_ATTR_FATTR_FILEID) && nfsi->fileid != fattr->fileid)
 		return -ESTALE;
···
 		invalid |= NFS_INO_INVALID_ATIME;
 
 	if (invalid != 0)
-		nfs_set_cache_invalid(inode, invalid | NFS_INO_REVAL_FORCED);
+		nfs_set_cache_invalid(inode, invalid);
 
 	nfsi->read_cache_jiffies = fattr->time_start;
 	return 0;
···
 	nfs_fattr_set_barrier(fattr);
 	status = nfs_post_op_update_inode_locked(inode, fattr,
 			NFS_INO_INVALID_CHANGE
-			| NFS_INO_INVALID_CTIME);
+			| NFS_INO_INVALID_CTIME
+			| NFS_INO_REVAL_FORCED);
 	spin_unlock(&inode->i_lock);
 
 	return status;
···
 	unsigned long save_cache_validity;
 	bool have_writers = nfs_file_has_buffered_writers(nfsi);
 	bool cache_revalidated = true;
+	bool attr_changed = false;
+	bool have_delegation;
 
 	dfprintk(VFS, "NFS: %s(%s/%lu fh_crc=0x%08x ct=%d info=0x%x)\n",
 			__func__, inode->i_sb->s_id, inode->i_ino,
···
 			!IS_AUTOMOUNT(inode))
 		server->fsid = fattr->fsid;
 
+	/* Save the delegation state before clearing cache_validity */
+	have_delegation = nfs_have_delegated_attributes(inode);
+
 	/*
 	 * Update the read time so we don't revalidate too often.
 	 */
···
 	/* More cache consistency checks */
 	if (fattr->valid & NFS_ATTR_FATTR_CHANGE) {
 		if (!inode_eq_iversion_raw(inode, fattr->change_attr)) {
-			dprintk("NFS: change_attr change on server for file %s/%ld\n",
-					inode->i_sb->s_id, inode->i_ino);
 			/* Could it be a race with writeback? */
-			if (!have_writers) {
-				invalid |= NFS_INO_INVALID_CHANGE
-					| NFS_INO_INVALID_DATA
+			if (!(have_writers || have_delegation)) {
+				invalid |= NFS_INO_INVALID_DATA
 					| NFS_INO_INVALID_ACCESS
 					| NFS_INO_INVALID_ACL;
 				/* Force revalidate of all attributes */
···
 					| NFS_INO_INVALID_OTHER;
 				if (S_ISDIR(inode->i_mode))
 					nfs_force_lookup_revalidate(inode);
+				dprintk("NFS: change_attr change on server for file %s/%ld\n",
+						inode->i_sb->s_id,
+						inode->i_ino);
 			}
 			inode_set_iversion_raw(inode, fattr->change_attr);
+			attr_changed = true;
 		}
 	} else {
 		nfsi->cache_validity |= save_cache_validity &
···
 	if (fattr->valid & NFS_ATTR_FATTR_SIZE) {
 		new_isize = nfs_size_to_loff_t(fattr->size);
 		cur_isize = i_size_read(inode);
-		if (new_isize != cur_isize) {
+		if (new_isize != cur_isize && !have_delegation) {
 			/* Do we perhaps have any outstanding writes, or has
 			 * the file grown beyond our last write? */
 			if (!nfs_have_writebacks(inode) || new_isize > cur_isize) {
 				i_size_write(inode, new_isize);
 				if (!have_writers)
 					invalid |= NFS_INO_INVALID_DATA;
+				attr_changed = true;
 			}
 			dprintk("NFS: isize change on server for file %s/%ld "
 					"(%Ld to %Ld)\n",
···
 			newmode |= fattr->mode & S_IALLUGO;
 			inode->i_mode = newmode;
 			invalid |= NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER;
+				| NFS_INO_INVALID_ACL;
+			attr_changed = true;
 		}
 	} else if (server->caps & NFS_CAP_MODE) {
 		nfsi->cache_validity |= save_cache_validity &
-				(NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER
+				(NFS_INO_INVALID_OTHER
 				| NFS_INO_REVAL_FORCED);
 		cache_revalidated = false;
 	}
···
 	if (fattr->valid & NFS_ATTR_FATTR_OWNER) {
 		if (!uid_eq(inode->i_uid, fattr->uid)) {
 			invalid |= NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER;
+				| NFS_INO_INVALID_ACL;
 			inode->i_uid = fattr->uid;
+			attr_changed = true;
 		}
 	} else if (server->caps & NFS_CAP_OWNER) {
 		nfsi->cache_validity |= save_cache_validity &
-				(NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER
+				(NFS_INO_INVALID_OTHER
 				| NFS_INO_REVAL_FORCED);
 		cache_revalidated = false;
 	}
···
 	if (fattr->valid & NFS_ATTR_FATTR_GROUP) {
 		if (!gid_eq(inode->i_gid, fattr->gid)) {
 			invalid |= NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER;
+				| NFS_INO_INVALID_ACL;
 			inode->i_gid = fattr->gid;
+			attr_changed = true;
 		}
 	} else if (server->caps & NFS_CAP_OWNER_GROUP) {
 		nfsi->cache_validity |= save_cache_validity &
-				(NFS_INO_INVALID_ACCESS
-				| NFS_INO_INVALID_ACL
-				| NFS_INO_INVALID_OTHER
+				(NFS_INO_INVALID_OTHER
 				| NFS_INO_REVAL_FORCED);
 		cache_revalidated = false;
 	}
 
 	if (fattr->valid & NFS_ATTR_FATTR_NLINK) {
 		if (inode->i_nlink != fattr->nlink) {
-			invalid |= NFS_INO_INVALID_OTHER;
 			if (S_ISDIR(inode->i_mode))
 				invalid |= NFS_INO_INVALID_DATA;
 			set_nlink(inode, fattr->nlink);
+			attr_changed = true;
 		}
 	} else if (server->caps & NFS_CAP_NLINK) {
 		nfsi->cache_validity |= save_cache_validity &
···
 		cache_revalidated = false;
 
 	/* Update attrtimeo value if we're out of the unstable period */
-	if (invalid & NFS_INO_INVALID_ATTR) {
+	if (attr_changed) {
 		invalid &= ~NFS_INO_INVALID_ATTR;
 		nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE);
 		nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
···
 	if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)
 				|| S_ISLNK(inode->i_mode)))
 		invalid &= ~NFS_INO_INVALID_DATA;
-	if (!NFS_PROTO(inode)->have_delegation(inode, FMODE_READ) ||
-			(save_cache_validity & NFS_INO_REVAL_FORCED))
-		nfs_set_cache_invalid(inode, invalid);
+	nfs_set_cache_invalid(inode, invalid);
 
 	return 0;
  out_err:

fs/nfs/nfs3proc.c

···
  */
 static int
 nfs3_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
-		struct nfs_fattr *fattr, struct nfs4_label *label)
+		struct nfs_fattr *fattr, struct nfs4_label *label,
+		struct inode *inode)
 {
 	struct rpc_message msg = {
 		.rpc_proc	= &nfs3_procedures[NFS3PROC_GETATTR],
···
 }
 
 static void
-nfs3_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
+nfs3_proc_unlink_setup(struct rpc_message *msg,
+		struct dentry *dentry,
+		struct inode *inode)
 {
 	msg->rpc_proc = &nfs3_procedures[NFS3PROC_REMOVE];
 }
···
 }
 
 static void nfs3_proc_write_setup(struct nfs_pgio_header *hdr,
-				  struct rpc_message *msg)
+				  struct rpc_message *msg,
+				  struct rpc_clnt **clnt)
 {
 	msg->rpc_proc = &nfs3_procedures[NFS3PROC_WRITE];
 }
···
 	return 0;
 }
 
-static void nfs3_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg)
+static void nfs3_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg,
+				   struct rpc_clnt **clnt)
 {
 	msg->rpc_proc = &nfs3_procedures[NFS3PROC_COMMIT];
 }

fs/nfs/nfs42proc.c

···
 	switch (task->tk_status) {
 	case 0:
 		break;
+	case -NFS4ERR_BADHANDLE:
+	case -ESTALE:
+		pnfs_destroy_layout(NFS_I(inode));
+		break;
 	case -NFS4ERR_EXPIRED:
 	case -NFS4ERR_ADMIN_REVOKED:
 	case -NFS4ERR_DELEG_REVOKED:
···
 		nfs42_layoutstat_release(data);
 		return -EAGAIN;
 	}
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0, 0);
 	task = rpc_run_task(&task_setup);
 	if (IS_ERR(task))
 		return PTR_ERR(task);

fs/nfs/nfs4_fs.h

···
 		struct rpc_cred *);
 };
 
+struct nfs4_opendata {
+	struct kref kref;
+	struct nfs_openargs o_arg;
+	struct nfs_openres o_res;
+	struct nfs_open_confirmargs c_arg;
+	struct nfs_open_confirmres c_res;
+	struct nfs4_string owner_name;
+	struct nfs4_string group_name;
+	struct nfs4_label *a_label;
+	struct nfs_fattr f_attr;
+	struct nfs4_label *f_label;
+	struct dentry *dir;
+	struct dentry *dentry;
+	struct nfs4_state_owner *owner;
+	struct nfs4_state *state;
+	struct iattr attrs;
+	struct nfs4_layoutget *lgp;
+	unsigned long timestamp;
+	bool rpc_done;
+	bool file_created;
+	bool is_recover;
+	bool cancelled;
+	int rpc_status;
+};
+
 struct nfs4_add_xprt_data {
 	struct nfs_client	*clp;
 	struct rpc_cred		*cred;
···
 extern int nfs4_call_sync(struct rpc_clnt *, struct nfs_server *,
 			  struct rpc_message *, struct nfs4_sequence_args *,
 			  struct nfs4_sequence_res *, int);
-extern void nfs4_init_sequence(struct nfs4_sequence_args *, struct nfs4_sequence_res *, int);
+extern void nfs4_init_sequence(struct nfs4_sequence_args *, struct nfs4_sequence_res *, int, int);
 extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *, struct nfs4_setclientid_res *);
 extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct nfs4_setclientid_res *arg, struct rpc_cred *);
 extern int nfs4_proc_get_rootfh(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *, bool);

fs/nfs/nfs4idmap.c

···
 	int id_len;
 	ssize_t ret;
 
-	id_len = snprintf(id_str, sizeof(id_str), "%u", id);
+	id_len = nfs_map_numeric_to_string(id, id_str, sizeof(id_str));
 	ret = nfs_idmap_get_key(id_str, id_len, type, buf, buflen, idmap);
 	if (ret < 0)
 		return -EINVAL;
···
 		if (strcmp(upcall->im_name, im->im_name) != 0)
 			break;
 		/* Note: here we store the NUL terminator too */
-		len = sprintf(id_str, "%d", im->im_id) + 1;
+		len = 1 + nfs_map_numeric_to_string(im->im_id, id_str,
+						    sizeof(id_str));
 		ret = nfs_idmap_instantiate(key, authkey, id_str, len);
 		break;
 	case IDMAP_CONV_IDTONAME:

fs/nfs/nfs4proc.c

···
 
 #define NFSDBG_FACILITY		NFSDBG_PROC
 
+#define NFS4_BITMASK_SZ		3
+
 #define NFS4_POLL_RETRY_MIN	(HZ/10)
 #define NFS4_POLL_RETRY_MAX	(15*HZ)
 
···
 	| ATTR_MTIME_SET)
 
 struct nfs4_opendata;
-static int _nfs4_proc_open(struct nfs4_opendata *data);
 static int _nfs4_recover_proc_open(struct nfs4_opendata *data);
 static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *);
 static void nfs_fixup_referral_attributes(struct nfs_fattr *fattr);
-static int nfs4_proc_getattr(struct nfs_server *, struct nfs_fh *, struct nfs_fattr *, struct nfs4_label *label);
-static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr, struct nfs4_label *label);
+static int nfs4_proc_getattr(struct nfs_server *, struct nfs_fh *, struct nfs_fattr *, struct nfs4_label *label, struct inode *inode);
+static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr, struct nfs4_label *label, struct inode *inode);
 static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 			    struct nfs_fattr *fattr, struct iattr *sattr,
 			    struct nfs_open_context *ctx, struct nfs4_label *ilabel,
···
 	| FATTR4_WORD1_MOUNTED_ON_FILEID,
 };
 
+static void nfs4_bitmap_copy_adjust(__u32 *dst, const __u32 *src,
+		struct inode *inode)
+{
+	unsigned long cache_validity;
+
+	memcpy(dst, src, NFS4_BITMASK_SZ*sizeof(*dst));
+	if (!inode || !nfs4_have_delegation(inode, FMODE_READ))
+		return;
+
+	cache_validity = READ_ONCE(NFS_I(inode)->cache_validity);
+	if (!(cache_validity & NFS_INO_REVAL_FORCED))
+		cache_validity &= ~(NFS_INO_INVALID_CHANGE
+				| NFS_INO_INVALID_SIZE);
+
+	if (!(cache_validity & NFS_INO_INVALID_SIZE))
+		dst[0] &= ~FATTR4_WORD0_SIZE;
+
+	if (!(cache_validity & NFS_INO_INVALID_CHANGE))
+		dst[0] &= ~FATTR4_WORD0_CHANGE;
+}
+
+static void nfs4_bitmap_copy_adjust_setattr(__u32 *dst,
+		const __u32 *src, struct inode *inode)
+{
+	nfs4_bitmap_copy_adjust(dst, src, inode);
+}
+
 static void nfs4_setup_readdir(u64 cookie, __be32 *verifier, struct dentry *dentry,
 		struct nfs4_readdir_arg *readdir)
 {
···
 	switch(errorcode) {
 		case 0:
 			return 0;
+		case -NFS4ERR_BADHANDLE:
+		case -ESTALE:
+			if (inode != NULL && S_ISREG(inode->i_mode))
+				pnfs_destroy_layout(NFS_I(inode));
+			break;
 		case -NFS4ERR_DELEG_REVOKED:
 		case -NFS4ERR_ADMIN_REVOKED:
 		case -NFS4ERR_EXPIRED:
···
 };
 
 void nfs4_init_sequence(struct nfs4_sequence_args *args,
-			struct nfs4_sequence_res *res, int cache_reply)
+			struct nfs4_sequence_res *res, int cache_reply,
+			int privileged)
 {
 	args->sa_slot = NULL;
 	args->sa_cache_this = cache_reply;
-	args->sa_privileged = 0;
+	args->sa_privileged = privileged;
 
 	res->sr_slot = NULL;
-}
-
-static void nfs4_set_sequence_privileged(struct nfs4_sequence_args *args)
-{
-	args->sa_privileged = 1;
 }
 
 static void nfs40_sequence_free_slot(struct nfs4_sequence_res *res)
···
 			slot->slot_nr,
 			slot->seq_nr);
 		goto out_retry;
+	case -NFS4ERR_RETRY_UNCACHED_REP:
+	case -NFS4ERR_SEQ_FALSE_RETRY:
+		/*
+		 * The server thinks we tried to replay a request.
+		 * Retry the call after bumping the sequence ID.
+		 */
+		goto retry_new_seq;
 	case -NFS4ERR_BADSLOT:
 		/*
 		 * The slot id we used was probably retired. Try again
 		 * using a different slot id.
 		 */
-		if (slot->seq_nr < slot->table->target_highest_slotid)
+		if (slot->slot_nr < slot->table->target_highest_slotid)
 			goto session_recover;
 		goto retry_nowait;
 	case -NFS4ERR_SEQ_MISORDERED:
···
 			slot->seq_nr = 1;
 			goto retry_nowait;
 		}
-		goto session_recover;
-	case -NFS4ERR_SEQ_FALSE_RETRY:
-		if (interrupted)
-			goto retry_new_seq;
 		goto session_recover;
 	default:
 		/* Just update the slot sequence no. */
···
 		   struct nfs4_sequence_res *res,
 		   int cache_reply)
 {
-	nfs4_init_sequence(args, res, cache_reply);
+	nfs4_init_sequence(args, res, cache_reply, 0);
 	return nfs4_call_sync_sequence(clnt, server, msg, args, res);
 }
 
···
 	nfs_fscache_invalidate(dir);
 	spin_unlock(&dir->i_lock);
 }
-
-struct nfs4_opendata {
-	struct kref kref;
-	struct nfs_openargs o_arg;
-	struct nfs_openres o_res;
-	struct nfs_open_confirmargs c_arg;
-	struct nfs_open_confirmres c_res;
-	struct nfs4_string owner_name;
-	struct nfs4_string group_name;
-	struct nfs4_label *a_label;
-	struct nfs_fattr f_attr;
-	struct nfs4_label *f_label;
-	struct dentry *dir;
-	struct dentry *dentry;
-	struct nfs4_state_owner *owner;
-	struct nfs4_state *state;
-	struct iattr attrs;
-	unsigned long timestamp;
-	bool rpc_done;
-	bool file_created;
-	bool is_recover;
-	bool cancelled;
-	int rpc_status;
-};
 
 struct nfs4_open_createattrs {
 	struct nfs4_label *label;
···
 			struct nfs4_opendata, kref);
 	struct super_block *sb = p->dentry->d_sb;
 
+	nfs4_lgopen_release(p->lgp);
 	nfs_free_seqid(p->o_arg.seqid);
 	nfs4_sequence_free_slot(&p->o_res.seq_res);
 	if (p->state != NULL)
···
 	};
 	int status;
 
-	nfs4_init_sequence(&data->c_arg.seq_args, &data->c_res.seq_res, 1);
+	nfs4_init_sequence(&data->c_arg.seq_args, &data->c_res.seq_res, 1,
+				data->is_recover);
 	kref_get(&data->kref);
 	data->rpc_done = false;
 	data->rpc_status = 0;
 	data->timestamp = jiffies;
-	if (data->is_recover)
-		nfs4_set_sequence_privileged(&data->c_arg.seq_args);
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return PTR_ERR(task);
···
 	.rpc_release = nfs4_open_release,
 };
 
-static int nfs4_run_open_task(struct nfs4_opendata *data, int isrecover)
+static int nfs4_run_open_task(struct nfs4_opendata *data,
+			      struct nfs_open_context *ctx)
 {
 	struct inode *dir = d_inode(data->dir);
 	struct nfs_server *server = NFS_SERVER(dir);
···
 	};
 	int status;
 
-	nfs4_init_sequence(&o_arg->seq_args, &o_res->seq_res, 1);
 	kref_get(&data->kref);
 	data->rpc_done = false;
 	data->rpc_status = 0;
 	data->cancelled = false;
 	data->is_recover = false;
-	if (isrecover) {
-		nfs4_set_sequence_privileged(&o_arg->seq_args);
+	if (!ctx) {
+		nfs4_init_sequence(&o_arg->seq_args, &o_res->seq_res, 1, 1);
 		data->is_recover = true;
+	} else {
+		nfs4_init_sequence(&o_arg->seq_args, &o_res->seq_res, 1, 0);
+		pnfs_lgopen_prepare(data, ctx);
 	}
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
···
 	struct nfs_openres *o_res = &data->o_res;
 	int status;
 
-	status = nfs4_run_open_task(data, 1);
+	status = nfs4_run_open_task(data, NULL);
 	if (status != 0 || !data->rpc_done)
 		return status;
 
···
 /*
  * Note: On error, nfs4_proc_open will free the struct nfs4_opendata
  */
-static int _nfs4_proc_open(struct nfs4_opendata *data)
+static int _nfs4_proc_open(struct nfs4_opendata *data,
+			   struct nfs_open_context *ctx)
 {
 	struct inode *dir = d_inode(data->dir);
 	struct nfs_server *server = NFS_SERVER(dir);
···
 	struct nfs_openres *o_res = &data->o_res;
 	int status;
 
-	status = nfs4_run_open_task(data, 0);
+	status = nfs4_run_open_task(data, ctx);
 	if (!data->rpc_done)
 		return status;
 	if (status != 0) {
···
 	}
 	if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) {
 		nfs4_sequence_free_slot(&o_res->seq_res);
-		nfs4_proc_getattr(server, &o_res->fh, o_res->f_attr, o_res->f_label);
+		nfs4_proc_getattr(server, &o_res->fh, o_res->f_attr,
+				o_res->f_label, NULL);
 	}
 	return 0;
 }
···
 
 	seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
 
-	ret = _nfs4_proc_open(opendata);
+	ret = _nfs4_proc_open(opendata, ctx);
 	if (ret != 0)
 		goto out;
 
-	state = nfs4_opendata_to_nfs4_state(opendata);
+	state = _nfs4_opendata_to_nfs4_state(opendata);
 	ret = PTR_ERR(state);
 	if (IS_ERR(state))
 		goto out;
···
 		nfs_inode_attach_open_context(ctx);
 		if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq))
 			nfs4_schedule_stateid_recovery(server, state);
+		else
+			pnfs_parse_lgopen(state->inode, opendata->lgp, ctx);
 	}
+
 out:
+	nfs4_sequence_free_slot(&opendata->o_res.seq_res);
 	return ret;
 }
 
···
 	};
 	struct rpc_cred *delegation_cred = NULL;
 	unsigned long timestamp = jiffies;
-	fmode_t fmode;
 	bool truncate;
 	int status;
 
···
 
 	/* Servers should only apply open mode checks for file size changes */
 	truncate = (arg->iap->ia_valid & ATTR_SIZE) ? true : false;
-	fmode = truncate ? FMODE_WRITE : FMODE_READ;
+	if (!truncate)
+		goto zero_stateid;
 
-	if (nfs4_copy_delegation_stateid(inode, fmode, &arg->stateid, &delegation_cred)) {
+	if (nfs4_copy_delegation_stateid(inode, FMODE_WRITE, &arg->stateid, &delegation_cred)) {
 		/* Use that stateid */
-	} else if (truncate && ctx != NULL) {
+	} else if (ctx != NULL) {
 		struct nfs_lock_context *l_ctx;
 		if (!nfs4_valid_open_stateid(ctx->state))
 			return -EBADF;
···
 		nfs_put_lock_context(l_ctx);
 		if (status == -EIO)
 			return -EBADF;
-	} else
+	} else {
+zero_stateid:
 		nfs4_stateid_copy(&arg->stateid, &zero_stateid);
+	}
 	if (delegation_cred)
 		msg.rpc_cred = delegation_cred;
 
···
 			   struct nfs4_label *olabel)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
+	__u32 bitmask[NFS4_BITMASK_SZ];
 	struct nfs4_state *state = ctx ? ctx->state : NULL;
 	struct nfs_setattrargs	arg = {
 		.fh		= NFS_FH(inode),
 		.iap		= sattr,
 		.server		= server,
-		.bitmask = server->attr_bitmask,
+		.bitmask = bitmask,
 		.label		= ilabel,
 	};
 	struct nfs_setattrres  res = {
···
 	};
 	int err;
 
-	arg.bitmask = nfs4_bitmask(server, ilabel);
-	if (ilabel)
-		arg.bitmask = nfs4_bitmask(server, olabel);
-
 	do {
+		nfs4_bitmap_copy_adjust_setattr(bitmask,
+				nfs4_bitmask(server, olabel),
+				inode);
+
 		err = _nfs4_do_setattr(inode, &arg, &res, cred, ctx);
 		switch (err) {
 		case -NFS4ERR_OPENMODE:
···
 	calldata = kzalloc(sizeof(*calldata), gfp_mask);
 	if (calldata == NULL)
 		goto out;
-	nfs4_init_sequence(&calldata->arg.seq_args, &calldata->res.seq_res, 1);
+	nfs4_init_sequence(&calldata->arg.seq_args, &calldata->res.seq_res, 1, 0);
 	calldata->inode = state->inode;
 	calldata->state = state;
 	calldata->arg.fh = NFS_FH(state->inode);
···
 	if (IS_ERR(label))
 		return PTR_ERR(label);
 
-	error = nfs4_proc_getattr(server, mntfh, fattr, label);
+	error = nfs4_proc_getattr(server, mntfh, fattr, label, NULL);
 	if (error < 0) {
 		dprintk("nfs4_get_root: getattr error = %d\n", -error);
 		goto err_free_label;
···
 }
 
 static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
-				struct nfs_fattr *fattr, struct nfs4_label *label)
+				struct nfs_fattr *fattr, struct nfs4_label *label,
+				struct inode *inode)
 {
+	__u32 bitmask[NFS4_BITMASK_SZ];
 	struct nfs4_getattr_arg args = {
 		.fh = fhandle,
-		.bitmask = server->attr_bitmask,
+		.bitmask = bitmask,
 	};
 	struct nfs4_getattr_res res = {
 		.fattr = fattr,
···
 		.rpc_resp = &res,
 	};
 
-	args.bitmask = nfs4_bitmask(server, label);
+	nfs4_bitmap_copy_adjust(bitmask, nfs4_bitmask(server, label), inode);
 
 	nfs_fattr_init(fattr);
 	return nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0);
 }
 
 static int nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
-				struct nfs_fattr *fattr, struct nfs4_label *label)
+				struct nfs_fattr *fattr, struct nfs4_label *label,
+				struct inode *inode)
 {
 	struct nfs4_exception exception = { };
 	int err;
 	do {
-		err = _nfs4_proc_getattr(server, fhandle, fattr, label);
+		err = _nfs4_proc_getattr(server, fhandle, fattr, label, inode);
 		trace_nfs4_getattr(server, fhandle, fattr, err);
 		err = nfs4_handle_exception(server, err,
 				&exception);
···
 	};
 	int status = 0;
 
-	if (!nfs_have_delegated_attributes(inode)) {
+	if (!nfs4_have_delegation(inode, FMODE_READ)) {
 		res.fattr = nfs_alloc_fattr();
 		if (res.fattr == NULL)
 			return -ENOMEM;
···
 	return err;
 }
 
-static void nfs4_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
+static void nfs4_proc_unlink_setup(struct rpc_message *msg,
+		struct dentry *dentry,
+		struct inode *inode)
 {
 	struct nfs_removeargs *args = msg->rpc_argp;
 	struct nfs_removeres *res = msg->rpc_resp;
-	struct inode *inode = d_inode(dentry);
 
 	res->server = NFS_SB(dentry->d_sb);
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_REMOVE];
-	nfs4_init_sequence(&args->seq_args, &res->seq_res, 1);
+	nfs4_init_sequence(&args->seq_args, &res->seq_res, 1, 0);
 
 	nfs_fattr_init(res->dir_attr);
 
···
 		nfs4_inode_return_delegation(new_inode);
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENAME];
 	res->server = NFS_SB(old_dentry->d_sb);
-	nfs4_init_sequence(&arg->seq_args, &res->seq_res, 1);
+	nfs4_init_sequence(&arg->seq_args, &res->seq_res, 1, 0);
 }
 
 static void nfs4_proc_rename_rpc_prepare(struct rpc_task *task, struct nfs_renamedata *data)
···
 static int _nfs4_proc_link(struct inode *inode, struct inode *dir, const struct qstr *name)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
+	__u32 bitmask[NFS4_BITMASK_SZ];
 	struct nfs4_link_arg arg = {
 		.fh     = NFS_FH(inode),
 		.dir_fh = NFS_FH(dir),
 		.name   = name,
-		.bitmask = server->attr_bitmask,
+		.bitmask = bitmask,
 	};
 	struct nfs4_link_res res = {
 		.server = server,
···
 		status = PTR_ERR(res.label);
 		goto out;
 	}
-	arg.bitmask = nfs4_bitmask(server, res.label);
 
 	nfs4_inode_make_writeable(inode);
+	nfs4_bitmap_copy_adjust_setattr(bitmask, nfs4_bitmask(server, res.label), inode);
 
 	status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
 	if (!status) {
···
 	if (!hdr->pgio_done_cb)
 		hdr->pgio_done_cb = nfs4_read_done_cb;
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
-	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 0);
+	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 0, 0);
 }
 
 static int nfs4_proc_pgio_rpc_prepare(struct rpc_task *task,
···
 }
 
 static void nfs4_proc_write_setup(struct nfs_pgio_header *hdr,
-				  struct rpc_message *msg)
+				  struct rpc_message *msg,
+				  struct rpc_clnt **clnt)
 {
 	struct nfs_server *server = NFS_SERVER(hdr->inode);
 
···
 	hdr->timestamp   = jiffies;
 
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_WRITE];
-	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 1);
+	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 1, 0);
+	nfs4_state_protect_write(server->nfs_client, clnt, msg, hdr);
 }
 
 static void nfs4_proc_commit_rpc_prepare(struct rpc_task *task, struct nfs_commit_data *data)
···
 	return data->commit_done_cb(task, data);
 }
 
-static void nfs4_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg)
+static void nfs4_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg,
+				   struct rpc_clnt **clnt)
 {
 	struct nfs_server *server = NFS_SERVER(data->inode);
 
···
 		data->commit_done_cb = nfs4_commit_done_cb;
 	data->res.server = server;
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_COMMIT];
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1, 0);
+	nfs4_state_protect(server->nfs_client, NFS_SP4_MACH_CRED_COMMIT, clnt, msg);
 }
 
 struct nfs4_renewdata {
···
 	 */
 	spin_lock(&inode->i_lock);
 	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
-		| NFS_INO_INVALID_CTIME;
+		| NFS_INO_INVALID_CTIME
+		| NFS_INO_REVAL_FORCED;
 	spin_unlock(&inode->i_lock);
 	nfs_access_zap_cache(inode);
 	nfs_zap_acl_cache(inode);
···
 		return 0;
 
 	rcu_read_lock();
-	len = 14 + strlen(clp->cl_ipaddr) + 1 +
-		strlen(rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR)) +
+	len = 14 +
+		strlen(clp->cl_rpcclient->cl_nodename) +
 		1 +
-		strlen(rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO)) +
+		strlen(rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR)) +
 		1;
 	rcu_read_unlock();
-
+	if (nfs4_client_id_uniquifier[0] != '\0')
+		len += strlen(nfs4_client_id_uniquifier) + 1;
 	if (len > NFS4_OPAQUE_LIMIT + 1)
 		return -EINVAL;
 
···
 		return -ENOMEM;
 
 	rcu_read_lock();
-	scnprintf(str, len, "Linux NFSv4.0 %s/%s %s",
-			clp->cl_ipaddr,
-			rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR),
-			rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO));
+	if (nfs4_client_id_uniquifier[0] != '\0')
+		scnprintf(str, len, "Linux NFSv4.0 %s/%s/%s",
+			  clp->cl_rpcclient->cl_nodename,
+			  nfs4_client_id_uniquifier,
+			  rpc_peeraddr2str(clp->cl_rpcclient,
+					   RPC_DISPLAY_ADDR));
+	else
+		scnprintf(str, len, "Linux NFSv4.0 %s/%s",
+			  clp->cl_rpcclient->cl_nodename,
+			  rpc_peeraddr2str(clp->cl_rpcclient,
+					   RPC_DISPLAY_ADDR));
 	rcu_read_unlock();
 
 	clp->cl_owner_id = str;
···
 	data = kzalloc(sizeof(*data), GFP_NOFS);
 	if (data == NULL)
 		return -ENOMEM;
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1, 0);
 
 	nfs4_state_protect(server->nfs_client,
 			NFS_SP4_MACH_CRED_CLEANUP,
···
 		return ERR_PTR(-ENOMEM);
 	}
 
-	nfs4_init_sequence(&data->arg.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&data->arg.seq_args, &data->res.seq_res, 1, 0);
 	msg.rpc_argp = &data->arg;
 	msg.rpc_resp = &data->res;
 	task_setup_data.callback_data = data;
···
 	case 0:
 		renew_lease(NFS_SERVER(d_inode(data->ctx->dentry)),
 				data->timestamp);
-		if (data->arg.new_lock) {
+		if (data->arg.new_lock && !data->cancelled) {
 			data->fl.fl_flags &= ~(FL_SLEEP | FL_ACCESS);
-			if (locks_lock_inode_wait(lsp->ls_state->inode, &data->fl) < 0) {
-				rpc_restart_call_prepare(task);
+			if (locks_lock_inode_wait(lsp->ls_state->inode, &data->fl) < 0)
 				break;
-			}
 		}
+
 		if (data->arg.new_lock_owner != 0) {
 			nfs_confirm_seqid(&lsp->ls_seqid, 0);
 			nfs4_stateid_copy(&lsp->ls_stateid, &data->res.stateid);
 			set_bit(NFS_LOCK_INITIALIZED, &lsp->ls_flags);
-		} else if (!nfs4_update_lock_stateid(lsp, &data->res.stateid))
-			rpc_restart_call_prepare(task);
+			goto out_done;
+		} else if (nfs4_update_lock_stateid(lsp, &data->res.stateid))
+			goto out_done;
+
 		break;
 	case -NFS4ERR_BAD_STATEID:
 	case -NFS4ERR_OLD_STATEID:
 	case -NFS4ERR_STALE_STATEID:
 	case -NFS4ERR_EXPIRED:
 		if (data->arg.new_lock_owner != 0) {
-			if (!nfs4_stateid_match(&data->arg.open_stateid,
+			if (nfs4_stateid_match(&data->arg.open_stateid,
 						&lsp->ls_state->open_stateid))
-				rpc_restart_call_prepare(task);
-		} else if (!nfs4_stateid_match(&data->arg.lock_stateid,
+				goto out_done;
+		} else if (nfs4_stateid_match(&data->arg.lock_stateid,
 						&lsp->ls_stateid))
-				rpc_restart_call_prepare(task);
+				goto out_done;
 	}
+	if (!data->cancelled)
+		rpc_restart_call_prepare(task);
+out_done:
 	dprintk("%s: done, ret = %d!\n", __func__, data->rpc_status);
 }
 
···
 		return -ENOMEM;
 	if (IS_SETLKW(cmd))
 		data->arg.block = 1;
-	nfs4_init_sequence(&data->arg.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&data->arg.seq_args, &data->res.seq_res, 1,
+				recovery_type > NFS_LOCK_NEW);
 	msg.rpc_argp = &data->arg;
 	msg.rpc_resp = &data->res;
 	task_setup_data.callback_data = data;
 	if (recovery_type > NFS_LOCK_NEW) {
 		if (recovery_type == NFS_LOCK_RECLAIM)
 			data->arg.reclaim = NFS_LOCK_RECLAIM;
-		nfs4_set_sequence_privileged(&data->arg.seq_args);
 	} else
 		data->arg.new_lock = 1;
 	task = rpc_run_task(&task_setup_data);
···
 
 	msg.rpc_argp = &data->args;
 	msg.rpc_resp = &data->res;
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0, 0);
 	rpc_call_async(server->client, &msg, 0, &nfs4_release_lockowner_ops, data);
 }
 
···
 	locations->server = server;
 	locations->nlocations = 0;
 
-	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0);
-	nfs4_set_sequence_privileged(&args.seq_args);
+	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0, 1);
 	status = nfs4_call_sync_sequence(clnt, server, &msg,
 					&args.seq_args, &res.seq_res);
 	if (status)
···
 	locations->server = server;
 	locations->nlocations = 0;
 
-	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0);
-	nfs4_set_sequence_privileged(&args.seq_args);
+	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0, 1);
 	status = nfs4_call_sync_sequence(clnt, server, &msg,
 					&args.seq_args, &res.seq_res);
 	if (status == NFS4_OK &&
···
 	if (res.fh == NULL)
 		return -ENOMEM;
 
-	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0);
-	nfs4_set_sequence_privileged(&args.seq_args);
+	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0, 1);
 	status = nfs4_call_sync_sequence(clnt, server, &msg,
 						&args.seq_args, &res.seq_res);
 	nfs_free_fhandle(res.fh);
···
 	if (res.fh == NULL)
 		return -ENOMEM;
 
-	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0);
-	nfs4_set_sequence_privileged(&args.seq_args);
+	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0, 1);
 	status = nfs4_call_sync_sequence(clnt, server, &msg,
 						&args.seq_args, &res.seq_res);
 	nfs_free_fhandle(res.fh);
···
 	};
 	int status;
 
-	nfs4_init_sequence(&args.la_seq_args, &res.lr_seq_res, 0);
-	nfs4_set_sequence_privileged(&args.la_seq_args);
+	nfs4_init_sequence(&args.la_seq_args, &res.lr_seq_res, 0, 1);
 	task = rpc_run_task(&task_setup);
 
 	if (IS_ERR(task))
···
 	calldata = kzalloc(sizeof(*calldata), GFP_NOFS);
 	if (calldata == NULL)
 		goto out_put_clp;
-	nfs4_init_sequence(&calldata->args, &calldata->res, 0);
+	nfs4_init_sequence(&calldata->args, &calldata->res, 0, is_privileged);
 	nfs4_sequence_attach_slot(&calldata->args, &calldata->res, slot);
-	if (is_privileged)
-		nfs4_set_sequence_privileged(&calldata->args);
 	msg.rpc_argp = &calldata->args;
 	msg.rpc_resp = &calldata->res;
 	calldata->clp = clp;
···
 	calldata->clp = clp;
 	calldata->arg.one_fs = 0;
 
-	nfs4_init_sequence(&calldata->arg.seq_args, &calldata->res.seq_res, 0);
-	nfs4_set_sequence_privileged(&calldata->arg.seq_args);
+	nfs4_init_sequence(&calldata->arg.seq_args, &calldata->res.seq_res, 0, 1);
 	msg.rpc_argp = &calldata->arg;
 	msg.rpc_resp = &calldata->res;
 	task_setup_data.callback_data = calldata;
···
 	return status;
 }
 
-static size_t max_response_pages(struct nfs_server *server)
+size_t max_response_pages(struct nfs_server *server)
 {
 	u32 max_resp_sz = server->nfs_client->cl_session->fc_attrs.max_resp_sz;
 	return nfs_page_array_len(0, max_resp_sz);
 }
 
-static void nfs4_free_pages(struct page **pages, size_t size)
-{
-	int i;
-
-	if (!pages)
-		return;
-
-	for (i = 0; i < size; i++) {
-		if (!pages[i])
-			break;
-		__free_page(pages[i]);
-	}
-	kfree(pages);
-}
-
-static struct page **nfs4_alloc_pages(size_t size, gfp_t gfp_flags)
-{
-	struct page **pages;
-	int i;
-
-	pages = kcalloc(size, sizeof(struct page *), gfp_flags);
-	if (!pages) {
-		dprintk("%s: can't alloc array of %zu pages\n", __func__, size);
-		return NULL;
-	}
-
-	for (i = 0; i < size; i++) {
-		pages[i] = alloc_page(gfp_flags);
-		if (!pages[i]) {
-			dprintk("%s: failed to allocate page\n", __func__);
-			nfs4_free_pages(pages, size);
-			return NULL;
-		}
-	}
-
-	return pages;
-}
-
 static void nfs4_layoutget_release(void *calldata)
 {
 	struct nfs4_layoutget *lgp = calldata;
-	struct inode *inode = lgp->args.inode;
-	struct nfs_server *server = NFS_SERVER(inode);
-	size_t max_pages = max_response_pages(server);
 
 	dprintk("--> %s\n", __func__);
 	nfs4_sequence_free_slot(&lgp->res.seq_res);
-	nfs4_free_pages(lgp->args.layout.pages, max_pages);
-	pnfs_put_layout_hdr(NFS_I(inode)->layout);
-	put_nfs_open_context(lgp->args.ctx);
-	kfree(calldata);
+	pnfs_layoutget_free(lgp);
 	dprintk("<-- %s\n", __func__);
 }
 
···
 };
 
 struct pnfs_layout_segment *
-nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout, gfp_t gfp_flags)
+nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout)
 {
 	struct inode *inode = lgp->args.inode;
 	struct nfs_server *server = NFS_SERVER(inode);
-	size_t max_pages = max_response_pages(server);
 	struct rpc_task *task;
 	struct rpc_message msg = {
 		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LAYOUTGET],
···
 	/* nfs4_layoutget_release calls pnfs_put_layout_hdr */
 	pnfs_get_layout_hdr(NFS_I(inode)->layout);
 
-	lgp->args.layout.pages = nfs4_alloc_pages(max_pages, gfp_flags);
-	if (!lgp->args.layout.pages) {
-		nfs4_layoutget_release(lgp);
-		return ERR_PTR(-ENOMEM);
-	}
-	lgp->args.layout.pglen = max_pages * PAGE_SIZE;
-
-	lgp->res.layoutp = &lgp->args.layout;
-	lgp->res.seq_res.sr_slot = NULL;
-	nfs4_init_sequence(&lgp->args.seq_args, &lgp->res.seq_res, 0);
+	nfs4_init_sequence(&lgp->args.seq_args, &lgp->res.seq_res, 0, 0);
 
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
···
 		}
 		task_setup_data.flags |= RPC_TASK_ASYNC;
 	}
-	nfs4_init_sequence(&lrp->args.seq_args, &lrp->res.seq_res, 1);
+	nfs4_init_sequence(&lrp->args.seq_args, &lrp->res.seq_res, 1, 0);
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return PTR_ERR(task);
···
 		}
 		task_setup_data.flags = RPC_TASK_ASYNC;
 	}
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1, 0);
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return PTR_ERR(task);
···
 		&rpc_client, &msg);
 
 	dprintk("NFS call  test_stateid %p\n", stateid);
-	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0);
-	nfs4_set_sequence_privileged(&args.seq_args);
+	nfs4_init_sequence(&args.seq_args, &res.seq_res, 0, 1);
 	status = nfs4_call_sync_sequence(rpc_client, server, &msg,
 			&args.seq_args, &res.seq_res);
 	if (status != NFS_OK) {
···
 	.rpc_release = nfs41_free_stateid_release,
 };
 
-static struct rpc_task *_nfs41_free_stateid(struct nfs_server *server,
+/**
+ * nfs41_free_stateid - perform a FREE_STATEID operation
+ *
+ * @server: server / transport on which to perform the operation
+ * @stateid: state ID to release
+ * @cred: credential
+ * @is_recovery: set to true if this call needs to be privileged
+ *
+ * Note: this function is always asynchronous.
+ */
+static int nfs41_free_stateid(struct nfs_server *server,
 		const nfs4_stateid *stateid,
 		struct rpc_cred *cred,
 		bool privileged)
···
 		.flags = RPC_TASK_ASYNC,
 	};
 	struct nfs_free_stateid_data *data;
+	struct rpc_task *task;
 
 	nfs4_state_protect(server->nfs_client, NFS_SP4_MACH_CRED_STATEID,
 		&task_setup.rpc_client, &msg);
···
 	dprintk("NFS call  free_stateid %p\n", stateid);
 	data = kmalloc(sizeof(*data), GFP_NOFS);
 	if (!data)
-		return ERR_PTR(-ENOMEM);
+		return -ENOMEM;
 	data->server = server;
 	nfs4_stateid_copy(&data->args.stateid, stateid);
 
···
 
 	msg.rpc_argp = &data->args;
 	msg.rpc_resp = &data->res;
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
-	if (privileged)
-		nfs4_set_sequence_privileged(&data->args.seq_args);
-
-	return rpc_run_task(&task_setup);
-}
-
-/**
- * nfs41_free_stateid - perform a FREE_STATEID operation
- *
- * @server: server / transport on which to perform the operation
- * @stateid: state ID to release
- * @cred: credential
- * @is_recovery: set to true if this call needs to be privileged
- *
- * Note: this function is always asynchronous.
- */
-static int nfs41_free_stateid(struct nfs_server *server,
-		const nfs4_stateid *stateid,
-		struct rpc_cred *cred,
-		bool is_recovery)
-{
-	struct rpc_task *task;
-
-	task = _nfs41_free_stateid(server, stateid, cred, is_recovery);
+	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1, privileged);
+	task = rpc_run_task(&task_setup);
 	if (IS_ERR(task))
 		return PTR_ERR(task);
 	rpc_put_task(task);
···
 		| NFS_CAP_ATOMIC_OPEN
 		| NFS_CAP_POSIX_LOCK
 		| NFS_CAP_STATEID_NFSV41
-		| NFS_CAP_ATOMIC_OPEN_V1,
+		| NFS_CAP_ATOMIC_OPEN_V1
+		| NFS_CAP_LGOPEN,
 	.init_client = nfs41_init_client,
 	.shutdown_client = nfs41_shutdown_client,
 	.match_stateid = nfs41_match_stateid,
···
 		| NFS_CAP_POSIX_LOCK
 		| NFS_CAP_STATEID_NFSV41
 		| NFS_CAP_ATOMIC_OPEN_V1
+		| NFS_CAP_LGOPEN
 		| NFS_CAP_ALLOCATE
 		| NFS_CAP_COPY
 		| NFS_CAP_DEALLOCATE

fs/nfs/nfs4state.c

···
 	.type = NFS4_INVALID_STATEID_TYPE,
 };
 
+const nfs4_stateid current_stateid = {
+	{
+		/* Funky initialiser keeps older gcc versions happy */
+		.data = { 0x0, 0x0, 0x0, 0x1, 0 },
+	},
+	.type = NFS4_SPECIAL_STATEID_TYPE,
+};
+
 static DEFINE_MUTEX(nfs_clid_init_mutex);
 
 int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred)

fs/nfs/nfs4xdr.c

···
 /* Mapping from NFS error code to "errno" error code. */
 #define errno_NFSERR_IO		EIO
 
+struct compound_hdr;
 static int nfs4_stat_to_errno(int);
+static void encode_layoutget(struct xdr_stream *xdr,
+			     const struct nfs4_layoutget_args *args,
+			     struct compound_hdr *hdr);
+static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req,
+			     struct nfs4_layoutget_res *res);
 
 /* NFSv4 COMPOUND tags are only wanted for debugging purposes */
 #ifdef DEBUG
···
 #define decode_sequence_maxsz	0
 #define encode_layoutreturn_maxsz 0
 #define decode_layoutreturn_maxsz 0
+#define encode_layoutget_maxsz	0
+#define decode_layoutget_maxsz	0
 #endif /* CONFIG_NFS_V4_1 */
 
 #define NFS4_enc_compound_sz	(1024)  /* XXX: large enough? */
···
 				encode_open_maxsz + \
 				encode_access_maxsz + \
 				encode_getfh_maxsz + \
-				encode_getattr_maxsz)
+				encode_getattr_maxsz + \
+				encode_layoutget_maxsz)
 #define NFS4_dec_open_sz        (compound_decode_hdr_maxsz + \
 				decode_sequence_maxsz + \
 				decode_putfh_maxsz + \
 				decode_open_maxsz + \
 				decode_access_maxsz + \
 				decode_getfh_maxsz + \
-				decode_getattr_maxsz)
+				decode_getattr_maxsz + \
+				decode_layoutget_maxsz)
 #define NFS4_enc_open_confirm_sz \
 				(compound_encode_hdr_maxsz + \
 				 encode_putfh_maxsz + \
···
 					encode_putfh_maxsz + \
 					encode_open_maxsz + \
 					encode_access_maxsz + \
-					encode_getattr_maxsz)
+					encode_getattr_maxsz + \
+					encode_layoutget_maxsz)
 #define NFS4_dec_open_noattr_sz	(compound_decode_hdr_maxsz + \
 					decode_sequence_maxsz + \
 					decode_putfh_maxsz + \
 					decode_open_maxsz + \
 					decode_access_maxsz + \
-					decode_getattr_maxsz)
+					decode_getattr_maxsz + \
+					decode_layoutget_maxsz)
 #define NFS4_enc_open_downgrade_sz \
 				(compound_encode_hdr_maxsz + \
 				 encode_sequence_maxsz + \
···
 		    struct compound_hdr *hdr)
 {
 }
+
+static void
+encode_layoutget(struct xdr_stream *xdr,
+		      const struct nfs4_layoutget_args *args,
+		      struct compound_hdr *hdr)
+{
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 /*
···
 	if (args->access)
 		encode_access(xdr, args->access, &hdr);
 	encode_getfattr_open(xdr, args->bitmask, args->open_bitmap, &hdr);
+	if (args->lg_args) {
+		encode_layoutget(xdr, args->lg_args, &hdr);
+		xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2,
+				 args->lg_args->layout.pages,
+				 0, args->lg_args->layout.pglen);
+	}
 	encode_nops(&hdr);
 }
 
···
 	if (args->access)
 		encode_access(xdr, args->access, &hdr);
 	encode_getfattr_open(xdr, args->bitmask, args->open_bitmap, &hdr);
+	if (args->lg_args) {
+		encode_layoutget(xdr, args->lg_args, &hdr);
+		xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2,
+				 args->lg_args->layout.pages,
+				 0, args->lg_args->layout.pglen);
+	}
 	encode_nops(&hdr);
 }
 
···
 
 	status = decode_op_hdr(xdr, OP_LAYOUTGET);
 	if (status)
-		return status;
+		goto out;
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(!p))
 		goto out_overflow;
···
 	if (!layout_count) {
 		dprintk("%s: server responded with empty layout array\n",
 			__func__);
-		return -EINVAL;
+		status = -EINVAL;
+		goto out;
 	}
 
 	p = xdr_inline_decode(xdr, 28);
···
 		dprintk("NFS: server cheating in layoutget reply: "
 				"layout len %u > recvd %u\n",
 				res->layoutp->len, recvd);
-		return -EINVAL;
+		status = -EINVAL;
+		goto out;
 	}
 
 	if (layout_count > 1) {
···
 			__func__, layout_count);
 	}
 
-	return 0;
+out:
+	res->status = status;
+	return status;
 out_overflow:
 	print_overflow_msg(__func__, xdr);
-	return -EIO;
+	status = -EIO;
+	goto out;
 }
 
 static int decode_layoutreturn(struct xdr_stream *xdr,
···
 {
 	return 0;
 }
+
+static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req,
+			    struct nfs4_layoutget_res *res)
+{
+	return 0;
+}
+
 #endif /* CONFIG_NFS_V4_1 */
 
 /*
···
 	if (res->access_request)
 		decode_access(xdr, &res->access_supported, &res->access_result);
 	decode_getfattr_label(xdr, res->f_attr, res->f_label, res->server);
+	if (res->lg_res)
+		decode_layoutget(xdr, rqstp, res->lg_res);
 out:
 	return status;
 }
···
 	if (res->access_request)
 		decode_access(xdr, &res->access_supported, &res->access_result);
 	decode_getfattr(xdr, res->f_attr, res->server);
+	if (res->lg_res)
+		decode_layoutget(xdr, rqstp, res->lg_res);
 out:
 	return status;
 }

fs/nfs/pnfs.c

···
 #include "nfs4trace.h"
 #include "delegation.h"
 #include "nfs42.h"
+#include "nfs4_fs.h"
 
 #define NFSDBG_FACILITY		NFSDBG_PNFS
 #define PNFS_LAYOUTGET_RETRY_TIMEOUT (120*HZ)
···
 		test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
 }
 
-/*
- * Get layout from server.
- *    for now, assume that whole file layouts are requested.
- *    arg->offset: 0
- *    arg->length: all ones
- */
-static struct pnfs_layout_segment *
-send_layoutget(struct pnfs_layout_hdr *lo,
-	   struct nfs_open_context *ctx,
-	   nfs4_stateid *stateid,
-	   const struct pnfs_layout_range *range,
-	   long *timeout, gfp_t gfp_flags)
+static struct nfs_server *
+pnfs_find_server(struct inode *inode, struct nfs_open_context *ctx)
 {
-	struct inode *ino = lo->plh_inode;
-	struct nfs_server *server = NFS_SERVER(ino);
+	struct nfs_server *server;
+
+	if (inode) {
+		server = NFS_SERVER(inode);
+	} else {
+		struct dentry *parent_dir = dget_parent(ctx->dentry);
+		server = NFS_SERVER(parent_dir->d_inode);
+		dput(parent_dir);
+	}
+	return server;
+}
+
+static void nfs4_free_pages(struct page **pages, size_t size)
+{
+	int i;
+
+	if (!pages)
+		return;
+
+	for (i = 0; i < size; i++) {
+		if (!pages[i])
+			break;
+		__free_page(pages[i]);
+	}
+	kfree(pages);
+}
+
+static struct page **nfs4_alloc_pages(size_t size, gfp_t gfp_flags)
+{
+	struct page **pages;
+	int i;
+
+	pages = kcalloc(size, sizeof(struct page *), gfp_flags);
+	if (!pages) {
+		dprintk("%s: can't alloc array of %zu pages\n", __func__, size);
+		return NULL;
+	}
+
+	for (i = 0; i < size; i++) {
+		pages[i] = alloc_page(gfp_flags);
+		if (!pages[i]) {
+			dprintk("%s: failed to allocate page\n", __func__);
+			nfs4_free_pages(pages, size);
+			return NULL;
+		}
+	}
+
+	return pages;
+}
+
+static struct nfs4_layoutget *
+pnfs_alloc_init_layoutget_args(struct inode *ino,
+	   struct nfs_open_context *ctx,
+	   const nfs4_stateid *stateid,
+	   const struct pnfs_layout_range *range,
+	   gfp_t gfp_flags)
+{
+	struct nfs_server *server = pnfs_find_server(ino, ctx);
+	size_t max_pages = max_response_pages(server);
 	struct nfs4_layoutget *lgp;
-	loff_t i_size;
 
 	dprintk("--> %s\n", __func__);
 
-	/*
-	 * Synchronously retrieve layout information from server and
-	 * store in lseg. If we race with a concurrent seqid morphing
-	 * op, then re-send the LAYOUTGET.
-	 */
 	lgp = kzalloc(sizeof(*lgp), gfp_flags);
 	if (lgp == NULL)
-		return ERR_PTR(-ENOMEM);
+		return NULL;
 
-	i_size = i_size_read(ino);
+	lgp->args.layout.pages = nfs4_alloc_pages(max_pages, gfp_flags);
+	if (!lgp->args.layout.pages) {
+		kfree(lgp);
+		return NULL;
+	}
+	lgp->args.layout.pglen = max_pages * PAGE_SIZE;
+	lgp->res.layoutp = &lgp->args.layout;
+
+	/* Don't confuse uninitialised result and success */
+	lgp->res.status = -NFS4ERR_DELAY;
 
 	lgp->args.minlength = PAGE_SIZE;
 	if (lgp->args.minlength > range->length)
 		lgp->args.minlength = range->length;
-	if (range->iomode == IOMODE_READ) {
-		if (range->offset >= i_size)
-			lgp->args.minlength = 0;
-		else if (i_size - range->offset < lgp->args.minlength)
-			lgp->args.minlength = i_size - range->offset;
+	if (ino) {
+		loff_t i_size = i_size_read(ino);
+
+		if (range->iomode == IOMODE_READ) {
+			if (range->offset >= i_size)
+				lgp->args.minlength = 0;
+			else if (i_size - range->offset < lgp->args.minlength)
+				lgp->args.minlength = i_size - range->offset;
+		}
 	}
 	lgp->args.maxcount = PNFS_LAYOUT_MAXSIZE;
 	pnfs_copy_range(&lgp->args.range, range);
···
 	lgp->args.ctx = get_nfs_open_context(ctx);
 	nfs4_stateid_copy(&lgp->args.stateid, stateid);
 	lgp->gfp_flags = gfp_flags;
-	lgp->cred = lo->plh_lc_cred;
+	lgp->cred = get_rpccred(ctx->cred);
+	lgp->callback_count = raw_seqcount_begin(&server->nfs_client->cl_callback_count);
+	return lgp;
+}
 
-	return nfs4_proc_layoutget(lgp, timeout, gfp_flags);
+void pnfs_layoutget_free(struct nfs4_layoutget *lgp)
+{
+	size_t max_pages = lgp->args.layout.pglen / PAGE_SIZE;
+
+	nfs4_free_pages(lgp->args.layout.pages, max_pages);
+	if (lgp->args.inode)
+		pnfs_put_layout_hdr(NFS_I(lgp->args.inode)->layout);
+	put_rpccred(lgp->cred);
+	put_nfs_open_context(lgp->args.ctx);
+	kfree(lgp);
 }
 
 static void pnfs_clear_layoutcommit(struct inode *inode,
···
 	LIST_HEAD(tmp_list);
 	nfs4_stateid stateid;
 	int status = 0;
-	bool send;
+	bool send, valid_layout;
 
 	dprintk("NFS: %s for inode %lu\n", __func__, ino->i_ino);
 
···
 			goto out_put_layout_hdr;
 		spin_lock(&ino->i_lock);
 	}
+	valid_layout = pnfs_layout_is_valid(lo);
 	pnfs_clear_layoutcommit(ino, &tmp_list);
 	pnfs_mark_matching_lsegs_invalid(lo, &tmp_list, NULL, 0);
 
···
 	}
 
 	/* Don't send a LAYOUTRETURN if list was initially empty */
-	if (!test_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags)) {
+	if (!test_bit(NFS_LAYOUT_RETURN_REQUESTED, &lo->plh_flags) ||
+			!valid_layout) {
 		spin_unlock(&ino->i_lock);
 		dprintk("NFS: %s no layout segments to return\n", __func__);
 		goto out_put_layout_hdr;
···
 	wake_up_bit(bitlock, NFS_LAYOUT_FIRST_LAYOUTGET);
 }
 
+static void _add_to_server_list(struct pnfs_layout_hdr *lo,
+				struct nfs_server *server)
+{
+	if (list_empty(&lo->plh_layouts)) {
+		struct nfs_client *clp = server->nfs_client;
+
+		/* The lo must be on the clp list if there is any
+		 * chance of a CB_LAYOUTRECALL(FILE) coming in.
+		 */
+		spin_lock(&clp->cl_lock);
+		if (list_empty(&lo->plh_layouts))
+			list_add_tail(&lo->plh_layouts, &server->layouts);
+		spin_unlock(&clp->cl_lock);
+	}
+}
+
 /*
  * Layout segment is retreived from the server if not cached.
  * The appropriate layout segment is referenced and returned to the caller.
···
 	struct nfs_client *clp = server->nfs_client;
 	struct pnfs_layout_hdr *lo = NULL;
 	struct pnfs_layout_segment *lseg = NULL;
+	struct nfs4_layoutget *lgp;
 	nfs4_stateid stateid;
 	long timeout = 0;
 	unsigned long giveup = jiffies + (clp->cl_lease_time << 1);
···
 	atomic_inc(&lo->plh_outstanding);
 	spin_unlock(&ino->i_lock);
 
-	if (list_empty(&lo->plh_layouts)) {
-		/* The lo must be on the clp list if there is any
-		 * chance of a CB_LAYOUTRECALL(FILE) coming in.
-		 */
-		spin_lock(&clp->cl_lock);
-		if (list_empty(&lo->plh_layouts))
-			list_add_tail(&lo->plh_layouts, &server->layouts);
-		spin_unlock(&clp->cl_lock);
-	}
+	_add_to_server_list(lo, server);
 
 	pg_offset = arg.offset & ~PAGE_MASK;
 	if (pg_offset) {
···
 	if (arg.length != NFS4_MAX_UINT64)
 		arg.length = PAGE_ALIGN(arg.length);
 
-	lseg = send_layoutget(lo, ctx, &stateid, &arg, &timeout, gfp_flags);
+	lgp = pnfs_alloc_init_layoutget_args(ino, ctx, &stateid, &arg, gfp_flags);
+	if (!lgp) {
+		trace_pnfs_update_layout(ino, pos, count, iomode, lo, NULL,
+					 PNFS_UPDATE_LAYOUT_NOMEM);
+		atomic_dec(&lo->plh_outstanding);
+		goto out_put_layout_hdr;
+	}
+
+	lseg = nfs4_proc_layoutget(lgp, &timeout);
 	trace_pnfs_update_layout(ino, pos, count, iomode, lo, lseg,
 				 PNFS_UPDATE_LAYOUT_SEND_LAYOUTGET);
 	atomic_dec(&lo->plh_outstanding);
···
 	return true;
 }
 
+static struct pnfs_layout_hdr *
+_pnfs_grab_empty_layout(struct inode *ino, struct nfs_open_context *ctx)
+{
+	struct pnfs_layout_hdr *lo;
+
+	spin_lock(&ino->i_lock);
+	lo = pnfs_find_alloc_layout(ino, ctx, GFP_KERNEL);
+	if (!lo)
+		goto out_unlock;
+	if (!test_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags))
+		goto out_unlock;
+	if (test_bit(NFS_LAYOUT_RETURN, &lo->plh_flags))
+		goto out_unlock;
+	if (pnfs_layoutgets_blocked(lo))
+		goto out_unlock;
+	if (test_and_set_bit(NFS_LAYOUT_FIRST_LAYOUTGET, &lo->plh_flags))
+		goto out_unlock;
+	atomic_inc(&lo->plh_outstanding);
+	spin_unlock(&ino->i_lock);
+	_add_to_server_list(lo, NFS_SERVER(ino));
+	return lo;
+
+out_unlock:
+	spin_unlock(&ino->i_lock);
+	pnfs_put_layout_hdr(lo);
+	return NULL;
+}
+
+extern const nfs4_stateid current_stateid;
+
+static void _lgopen_prepare_attached(struct nfs4_opendata *data,
+				     struct nfs_open_context *ctx)
+{
+	struct inode *ino = data->dentry->d_inode;
+	struct pnfs_layout_range rng = {
+		.iomode = (data->o_arg.fmode & FMODE_WRITE) ?
+			  IOMODE_RW: IOMODE_READ,
+		.offset = 0,
+		.length = NFS4_MAX_UINT64,
+	};
+	struct nfs4_layoutget *lgp;
+	struct pnfs_layout_hdr *lo;
+
+	/* Heuristic: don't send layoutget if we have cached data */
+	if (rng.iomode == IOMODE_READ &&
+	   (i_size_read(ino) == 0 || ino->i_mapping->nrpages != 0))
+		return;
+
+	lo = _pnfs_grab_empty_layout(ino, ctx);
+	if (!lo)
+		return;
+	lgp = pnfs_alloc_init_layoutget_args(ino, ctx, &current_stateid,
+					     &rng, GFP_KERNEL);
+	if (!lgp) {
+		pnfs_clear_first_layoutget(lo);
+		pnfs_put_layout_hdr(lo);
+		return;
+	}
+	data->lgp = lgp;
+	data->o_arg.lg_args = &lgp->args;
+	data->o_res.lg_res = &lgp->res;
+}
+
+static void _lgopen_prepare_floating(struct nfs4_opendata *data,
+				     struct nfs_open_context *ctx)
+{
+	struct pnfs_layout_range rng = {
+		.iomode = (data->o_arg.fmode & FMODE_WRITE) ?
+			  IOMODE_RW: IOMODE_READ,
+		.offset = 0,
+		.length = NFS4_MAX_UINT64,
+	};
+	struct nfs4_layoutget *lgp;
+
+	lgp = pnfs_alloc_init_layoutget_args(NULL, ctx, &current_stateid,
+					     &rng, GFP_KERNEL);
+	if (!lgp)
+		return;
+	data->lgp = lgp;
+	data->o_arg.lg_args = &lgp->args;
+	data->o_res.lg_res = &lgp->res;
+}
+
+void pnfs_lgopen_prepare(struct nfs4_opendata *data,
+			 struct nfs_open_context *ctx)
+{
+	struct nfs_server *server = NFS_SERVER(data->dir->d_inode);
+
+	if (!(pnfs_enabled_sb(server) &&
+	      server->pnfs_curr_ld->flags & PNFS_LAYOUTGET_ON_OPEN))
+		return;
+	/* Could check on max_ops, but currently hardcoded high enough */
+	if (!nfs_server_capable(data->dir->d_inode, NFS_CAP_LGOPEN))
+		return;
+	if (data->state)
+		_lgopen_prepare_attached(data, ctx);
+	else
+		_lgopen_prepare_floating(data, ctx);
+}
+
+void pnfs_parse_lgopen(struct inode *ino, struct nfs4_layoutget *lgp,
+		       struct nfs_open_context *ctx)
+{
+	struct pnfs_layout_hdr *lo;
+	struct pnfs_layout_segment *lseg;
+	struct nfs_server *srv = NFS_SERVER(ino);
+	u32 iomode;
+
+	if (!lgp)
+		return;
+	dprintk("%s: entered with status %i\n", __func__, lgp->res.status);
+	if (lgp->res.status) {
+		switch (lgp->res.status) {
+		default:
+			break;
+		/*
+		 * Halt lgopen attempts if the server doesn't recognise
+		 * the "current stateid" value, the layout type, or the
+		 * layoutget operation as being valid.
+		 * Also if it complains about too many ops in the compound
+		 * or of the request/reply being too big.
+		 */
+		case -NFS4ERR_BAD_STATEID:
+		case -NFS4ERR_NOTSUPP:
+		case -NFS4ERR_REP_TOO_BIG:
+		case -NFS4ERR_REP_TOO_BIG_TO_CACHE:
+		case -NFS4ERR_REQ_TOO_BIG:
+		case -NFS4ERR_TOO_MANY_OPS:
+		case -NFS4ERR_UNKNOWN_LAYOUTTYPE:
+			srv->caps &= ~NFS_CAP_LGOPEN;
+		}
+		return;
+	}
+	if (!lgp->args.inode) {
+		lo = _pnfs_grab_empty_layout(ino, ctx);
+		if (!lo)
+			return;
+		lgp->args.inode = ino;
+	} else
+		lo = NFS_I(lgp->args.inode)->layout;
+
+	if (read_seqcount_retry(&srv->nfs_client->cl_callback_count,
+				lgp->callback_count))
+		return;
+	lseg = pnfs_layout_process(lgp);
+	if (!IS_ERR(lseg)) {
+		iomode = lgp->args.range.iomode;
+		pnfs_layout_clear_fail_bit(lo, pnfs_iomode_to_fail_bit(iomode));
+		pnfs_put_lseg(lseg);
+	}
+}
+
+void nfs4_lgopen_release(struct nfs4_layoutget *lgp)
+{
+	if (lgp != NULL) {
+		struct inode *inode = lgp->args.inode;
+		if (inode) {
+			struct pnfs_layout_hdr *lo = NFS_I(inode)->layout;
+			atomic_dec(&lo->plh_outstanding);
+			pnfs_clear_first_layoutget(lo);
+		}
+		pnfs_layoutget_free(lgp);
+	}
+}
+
 struct pnfs_layout_segment *
 pnfs_layout_process(struct nfs4_layoutget *lgp)
 {
···
 	spin_unlock(&ino->i_lock);
 	lseg->pls_layout = lo;
 	NFS_SERVER(ino)->pnfs_curr_ld->free_lseg(lseg);
-	if (!pnfs_layout_is_valid(lo))
-		nfs_commit_inode(ino, 0);
 	return ERR_PTR(-EAGAIN);
 }
 

fs/nfs/pnfs.h

···
 #include <linux/nfs_page.h>
 #include <linux/workqueue.h>
 
+struct nfs4_opendata;
+
 enum {
 	NFS_LSEG_VALID = 0,	/* cleared when lseg is recalled/returned */
 	NFS_LSEG_ROC,		/* roc bit received from server */
···
 	PNFS_LAYOUTRET_ON_SETATTR	= 1 << 0,
 	PNFS_LAYOUTRET_ON_ERROR		= 1 << 1,
 	PNFS_READ_WHOLE_PAGE		= 1 << 2,
+	PNFS_LAYOUTGET_ON_OPEN		= 1 << 3,
 };
 
 struct nfs4_deviceid_node;
···
 extern void pnfs_unregister_layoutdriver(struct pnfs_layoutdriver_type *);
 
 /* nfs4proc.c */
+extern size_t max_response_pages(struct nfs_server *server);
 extern int nfs4_proc_getdeviceinfo(struct nfs_server *server,
 				   struct pnfs_device *dev,
 				   struct rpc_cred *cred);
-extern struct pnfs_layout_segment* nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout, gfp_t gfp_flags);
+extern struct pnfs_layout_segment* nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout);
 extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp, bool sync);
 
 /* pnfs.c */
···
 			    struct nfs_page *prev, struct nfs_page *req);
 void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg);
 struct pnfs_layout_segment *pnfs_layout_process(struct nfs4_layoutget *lgp);
+void pnfs_layoutget_free(struct nfs4_layoutget *lgp);
 void pnfs_free_lseg_list(struct list_head *tmp_list);
 void pnfs_destroy_layout(struct nfs_inode *);
 void pnfs_destroy_all_layouts(struct nfs_client *);
···
 				     struct pnfs_layout_segment *lseg,
 				     struct nfs_commit_info *cinfo,
 				     u32 ds_commit_idx);
+void pnfs_lgopen_prepare(struct nfs4_opendata *data,
+			 struct nfs_open_context *ctx);
+void pnfs_parse_lgopen(struct inode *ino, struct nfs4_layoutget *lgp,
+		       struct nfs_open_context *ctx);
+void nfs4_lgopen_release(struct nfs4_layoutget *lgp);
 
 static inline bool nfs_have_layout(struct inode *inode)
 {
···
 {
 	return false;
 }
+
+static inline void pnfs_lgopen_prepare(struct nfs4_opendata *data,
+		struct nfs_open_context *ctx)
+{
+}
+
+static inline void pnfs_parse_lgopen(struct inode *ino,
+		struct nfs4_layoutget *lgp,
+		struct nfs_open_context *ctx)
+{
+}
+
+static inline void nfs4_lgopen_release(struct nfs4_layoutget *lgp)
+{
+}
+
 #endif /* CONFIG_NFS_V4_1 */
 
 #if IS_ENABLED(CONFIG_NFS_V4_2)

fs/nfs/proc.c

···
  */
 static int
 nfs_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
-		struct nfs_fattr *fattr, struct nfs4_label *label)
+		struct nfs_fattr *fattr, struct nfs4_label *label,
+		struct inode *inode)
 {
 	struct rpc_message msg = {
 		.rpc_proc	= &nfs_procedures[NFSPROC_GETATTR],
···
 }
 
 static void
-nfs_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
+nfs_proc_unlink_setup(struct rpc_message *msg,
+		struct dentry *dentry,
+		struct inode *inode)
 {
 	msg->rpc_proc = &nfs_procedures[NFSPROC_REMOVE];
 }
···
 }
 
 static void nfs_proc_write_setup(struct nfs_pgio_header *hdr,
-				 struct rpc_message *msg)
+				 struct rpc_message *msg,
+				 struct rpc_clnt **clnt)
 {
 	/* Note: NFSv2 ignores @stable and always uses NFS_FILE_SYNC */
 	hdr->args.stable = NFS_FILE_SYNC;
···
 }
 
 static void
-nfs_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg)
+nfs_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg,
+			struct rpc_clnt **clnt)
 {
 	BUG();
 }

fs/nfs/unlink.c

···
 	.rpc_call_prepare = nfs_unlink_prepare,
 };
 
-static void nfs_do_call_unlink(struct nfs_unlinkdata *data)
+static void nfs_do_call_unlink(struct inode *inode, struct nfs_unlinkdata *data)
 {
 	struct rpc_message msg = {
 		.rpc_argp = &data->args,
···
 	data->args.fh = NFS_FH(dir);
 	nfs_fattr_init(data->res.dir_attr);
 
-	NFS_PROTO(dir)->unlink_setup(&msg, data->dentry);
+	NFS_PROTO(dir)->unlink_setup(&msg, data->dentry, inode);
 
 	task_setup_data.rpc_client = NFS_CLIENT(dir);
 	task = rpc_run_task(&task_setup_data);
···
 		rpc_put_task_async(task);
 }
 
-static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data)
+static int nfs_call_unlink(struct dentry *dentry, struct inode *inode, struct nfs_unlinkdata *data)
 {
 	struct inode *dir = d_inode(dentry->d_parent);
 	struct dentry *alias;
···
 		return ret;
 	}
 	data->dentry = alias;
-	nfs_do_call_unlink(data);
+	nfs_do_call_unlink(inode, data);
 	return 1;
 }
 
···
 	dentry->d_fsdata = NULL;
 	spin_unlock(&dentry->d_lock);
 
-	if (NFS_STALE(inode) || !nfs_call_unlink(dentry, data))
+	if (NFS_STALE(inode) || !nfs_call_unlink(dentry, inode, data))
 		nfs_free_unlinkdata(data);
 }
 
···
 	unsigned char silly[SILLYNAME_LEN + 1];
 	unsigned long long fileid;
 	struct dentry *sdentry;
+	struct inode *inode = d_inode(dentry);
 	struct rpc_task *task;
 	int            error = -EBUSY;
 
···
 			goto out;
 	} while (d_inode(sdentry) != NULL); /* need negative lookup */
 
+	ihold(inode);
+
 	/* queue unlink first. Can't do this from rpc_release as it
 	 * has to allocate memory
 	 */
···
 	case 0:
 		/* The rename succeeded */
 		nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
+		spin_lock(&inode->i_lock);
+		NFS_I(inode)->attr_gencount = nfs_inc_attr_generation_counter();
+		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+			| NFS_INO_INVALID_CTIME
+			| NFS_INO_REVAL_FORCED;
+		spin_unlock(&inode->i_lock);
 		d_move(dentry, sdentry);
 		break;
 	case -ERESTARTSYS:
···
 	}
 	rpc_put_task(task);
 out_dput:
+	iput(inode);
 	dput(sdentry);
 out:
 	return error;

fs/nfs/write.c

···
 	int priority = flush_task_priority(how);
 
 	task_setup_data->priority = priority;
-	rpc_ops->write_setup(hdr, msg);
+	rpc_ops->write_setup(hdr, msg, &task_setup_data->rpc_client);
 	trace_nfs_initiate_write(hdr->inode, hdr->io_start, hdr->good_bytes,
 				 hdr->args.stable);
-
-	nfs4_state_protect_write(NFS_SERVER(hdr->inode)->nfs_client,
-				 &task_setup_data->rpc_client, msg, hdr);
 }
 
 /* If a nfs_flush_* function fails, it should remove reqs from @head and
···
 		.priority = priority,
 	};
 	/* Set up the initial task struct.  */
-	nfs_ops->commit_setup(data, &msg);
+	nfs_ops->commit_setup(data, &msg, &task_setup_data.rpc_client);
 	trace_nfs_initiate_commit(data);
 
 	dprintk("NFS: initiated commit call\n");
-
-	nfs4_state_protect(NFS_SERVER(data->inode)->nfs_client,
-		NFS_SP4_MACH_CRED_COMMIT, &task_setup_data.rpc_client, &msg);
 
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))

include/linux/nfs_fs_sb.h

···
 struct nfs_client {
 	refcount_t		cl_count;
 	atomic_t		cl_mds_count;
+	seqcount_t		cl_callback_count;
 	int			cl_cons_state;	/* current construction state (-ve: init error) */
 #define NFS_CS_READY		0		/* ready to be used */
 #define NFS_CS_INITING		1		/* busy initialising */
···
 #define NFS_CAP_ACLS		(1U << 3)
 #define NFS_CAP_ATOMIC_OPEN	(1U << 4)
 /* #define NFS_CAP_CHANGE_ATTR	(1U << 5) */
+#define NFS_CAP_LGOPEN		(1U << 5)
 #define NFS_CAP_FILEID		(1U << 6)
 #define NFS_CAP_MODE		(1U << 7)
 #define NFS_CAP_NLINK		(1U << 8)

include/linux/nfs_xdr.h

···
 
 struct nfs4_layoutget_res {
 	struct nfs4_sequence_res seq_res;
+	int status;
 	__u32 return_on_close;
 	struct pnfs_layout_range range;
 	__u32 type;
···
 	struct nfs4_layoutget_args args;
 	struct nfs4_layoutget_res res;
 	struct rpc_cred *cred;
+	unsigned callback_count;
 	gfp_t gfp_flags;
 };
 
···
 	enum createmode4	createmode;
 	const struct nfs4_label *label;
 	umode_t			umask;
+	struct nfs4_layoutget_args *lg_args;
 };
 
 struct nfs_openres {
···
 	__u32			access_request;
 	__u32			access_supported;
 	__u32			access_result;
+	struct nfs4_layoutget_res *lg_res;
 };
 
 /*
···
 	struct dentry *(*try_mount) (int, const char *, struct nfs_mount_info *,
 				     struct nfs_subversion *);
 	int	(*getattr) (struct nfs_server *, struct nfs_fh *,
-			    struct nfs_fattr *, struct nfs4_label *);
+			    struct nfs_fattr *, struct nfs4_label *,
+			    struct inode *);
 	int	(*setattr) (struct dentry *, struct nfs_fattr *,
 			    struct iattr *);
 	int	(*lookup)  (struct inode *, const struct qstr *,
···
 	int	(*create)  (struct inode *, struct dentry *,
 			    struct iattr *, int);
 	int	(*remove)  (struct inode *, struct dentry *);
-	void	(*unlink_setup)  (struct rpc_message *, struct dentry *);
+	void	(*unlink_setup)  (struct rpc_message *, struct dentry *, struct inode *);
 	void	(*unlink_rpc_prepare) (struct rpc_task *, struct nfs_unlinkdata *);
 	int	(*unlink_done) (struct rpc_task *, struct inode *);
 	void	(*rename_setup)  (struct rpc_message *msg,
···
 				    struct nfs_pgio_header *);
 	void	(*read_setup)(struct nfs_pgio_header *, struct rpc_message *);
 	int	(*read_done)(struct rpc_task *, struct nfs_pgio_header *);
-	void	(*write_setup)(struct nfs_pgio_header *, struct rpc_message *);
+	void	(*write_setup)(struct nfs_pgio_header *, struct rpc_message *,
+				struct rpc_clnt **);
 	int	(*write_done)(struct rpc_task *, struct nfs_pgio_header *);
-	void	(*commit_setup) (struct nfs_commit_data *, struct rpc_message *);
+	void	(*commit_setup) (struct nfs_commit_data *, struct rpc_message *,
+				struct rpc_clnt **);
 	void	(*commit_rpc_prepare)(struct rpc_task *, struct nfs_commit_data *);
 	int	(*commit_done) (struct rpc_task *, struct nfs_commit_data *);
 	int	(*lock)(struct file *, int, struct file_lock *);

include/linux/rculist.h

···
 	     pos = list_entry_rcu(pos->member.next, typeof(*pos), member))
 
 /**
+ * list_for_each_entry_from_rcu - iterate over a list from current point
+ * @pos:	the type * to use as a loop cursor.
+ * @head:	the head for your list.
+ * @member:	the name of the list_node within the struct.
+ *
+ * Iterate over the tail of a list starting from a given position,
+ * which must have been in the list when the RCU read lock was taken.
+ */
+#define list_for_each_entry_from_rcu(pos, head, member)			\
+	for (; &(pos)->member != (head);					\
+		pos = list_entry_rcu(pos->member.next, typeof(*(pos)), member))
+
+/**
  * hlist_del_rcu - deletes entry from hash list without re-initialization
  * @n: the element to delete from the hash list.
  *

include/linux/sunrpc/rpc_rdma.h

···
+/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
 /*
  * Copyright (c) 2015-2017 Oracle. All rights reserved.
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.

include/linux/sunrpc/xprt.h

···
 	void (*rq_release_snd_buf)(struct rpc_rqst *); /* release rq_enc_pages */
 	struct list_head	rq_list;
 
-	void			*rq_xprtdata;	/* Per-xprt private data */
 	void			*rq_buffer;	/* Call XDR encode buffer */
 	size_t			rq_callsize;
 	void			*rq_rbuffer;	/* Reply XDR decode buffer */
···
 	int		(*reserve_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
 	void		(*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
 	void		(*alloc_slot)(struct rpc_xprt *xprt, struct rpc_task *task);
+	void		(*free_slot)(struct rpc_xprt *xprt,
+				     struct rpc_rqst *req);
 	void		(*rpcbind)(struct rpc_task *task);
 	void		(*set_port)(struct rpc_xprt *xprt, unsigned short port);
 	void		(*connect)(struct rpc_xprt *xprt, struct rpc_task *task);
···
 struct rpc_xprt		*xprt_create_transport(struct xprt_create *args);
 void			xprt_connect(struct rpc_task *task);
 void			xprt_reserve(struct rpc_task *task);
+void			xprt_request_init(struct rpc_task *task);
 void			xprt_retry_reserve(struct rpc_task *task);
 int			xprt_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task);
 int			xprt_reserve_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_alloc_slot(struct rpc_xprt *xprt, struct rpc_task *task);
+void			xprt_free_slot(struct rpc_xprt *xprt,
+				       struct rpc_rqst *req);
 void			xprt_lock_and_alloc_slot(struct rpc_xprt *xprt, struct rpc_task *task);
 bool			xprt_prepare_transmit(struct rpc_task *task);
 void			xprt_transmit(struct rpc_task *task);

include/linux/sunrpc/xprtrdma.h

···
+/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
 /*
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
  *

include/trace/events/rpcrdma.h

···
 
 TRACE_EVENT(xprtrdma_post_recv,
 	TP_PROTO(
-		const struct rpcrdma_rep *rep,
-		int status
+		const struct ib_cqe *cqe
 	),
 
-	TP_ARGS(rep, status),
+	TP_ARGS(cqe),
 
 	TP_STRUCT__entry(
-		__field(const void *, rep)
-		__field(int, status)
+		__field(const void *, cqe)
 	),
 
 	TP_fast_assign(
-		__entry->rep = rep;
-		__entry->status = status;
+		__entry->cqe = cqe;
 	),
 
-	TP_printk("rep=%p status=%d",
-		__entry->rep, __entry->status
+	TP_printk("cqe=%p",
+		__entry->cqe
+	)
+);
+
+TRACE_EVENT(xprtrdma_post_recvs,
+	TP_PROTO(
+		const struct rpcrdma_xprt *r_xprt,
+		unsigned int count,
+		int status
+	),
+
+	TP_ARGS(r_xprt, count, status),
+
+	TP_STRUCT__entry(
+		__field(const void *, r_xprt)
+		__field(unsigned int, count)
+		__field(int, status)
+		__field(int, posted)
+		__string(addr, rpcrdma_addrstr(r_xprt))
+		__string(port, rpcrdma_portstr(r_xprt))
+	),
+
+	TP_fast_assign(
+		__entry->r_xprt = r_xprt;
+		__entry->count = count;
+		__entry->status = status;
+		__entry->posted = r_xprt->rx_buf.rb_posted_receives;
+		__assign_str(addr, rpcrdma_addrstr(r_xprt));
+		__assign_str(port, rpcrdma_portstr(r_xprt));
+	),
+
+	TP_printk("peer=[%s]:%s r_xprt=%p: %u new recvs, %d active (rc %d)",
+		__get_str(addr), __get_str(port), __entry->r_xprt,
+		__entry->count, __entry->posted, __entry->status
 	)
 );
 
···
 
 TRACE_EVENT(xprtrdma_wc_receive,
 	TP_PROTO(
-		const struct rpcrdma_rep *rep,
 		const struct ib_wc *wc
 	),
 
-	TP_ARGS(rep, wc),
+	TP_ARGS(wc),
 
 	TP_STRUCT__entry(
-		__field(const void *, rep)
-		__field(unsigned int, byte_len)
+		__field(const void *, cqe)
+		__field(u32, byte_len)
 		__field(unsigned int, status)
-		__field(unsigned int, vendor_err)
+		__field(u32, vendor_err)
 	),
 
 	TP_fast_assign(
-		__entry->rep = rep;
-		__entry->byte_len = wc->byte_len;
+		__entry->cqe = wc->wr_cqe;
 		__entry->status = wc->status;
-		__entry->vendor_err = __entry->status ? wc->vendor_err : 0;
+		if (wc->status) {
+			__entry->byte_len = 0;
+			__entry->vendor_err = wc->vendor_err;
+		} else {
+			__entry->byte_len = wc->byte_len;
+			__entry->vendor_err = 0;
+		}
 	),
 
-	TP_printk("rep=%p, %u bytes: %s (%u/0x%x)",
-		__entry->rep, __entry->byte_len,
+	TP_printk("cqe=%p %u bytes: %s (%u/0x%x)",
+		__entry->cqe, __entry->byte_len,
 		rdma_show_wc_status(__entry->status),
 		__entry->status, __entry->vendor_err
 	)
···
 DEFINE_FRWR_DONE_EVENT(xprtrdma_wc_li_wake);
 
 DEFINE_MR_EVENT(xprtrdma_localinv);
+DEFINE_MR_EVENT(xprtrdma_dma_map);
 DEFINE_MR_EVENT(xprtrdma_dma_unmap);
 DEFINE_MR_EVENT(xprtrdma_remoteinv);
 DEFINE_MR_EVENT(xprtrdma_recover_mr);
···
 		__field(unsigned int, task_id)
 		__field(unsigned int, client_id)
 		__field(const void *, req)
-		__field(const void *, rep)
 		__field(size_t, callsize)
 		__field(size_t, rcvsize)
 	),
···
 		__entry->task_id = task->tk_pid;
 		__entry->client_id = task->tk_client->cl_clid;
 		__entry->req = req;
-		__entry->rep = req ? req->rl_reply : NULL;
 		__entry->callsize = task->tk_rqstp->rq_callsize;
 		__entry->rcvsize = task->tk_rqstp->rq_rcvsize;
 	),
 
-	TP_printk("task:%u@%u req=%p rep=%p (%zu, %zu)",
+	TP_printk("task:%u@%u req=%p (%zu, %zu)",
 		__entry->task_id, __entry->client_id,
-		__entry->req, __entry->rep,
-		__entry->callsize, __entry->rcvsize
+		__entry->req, __entry->callsize, __entry->rcvsize
 	)
 );
 
···
 		__entry->req, __entry->rep
 	)
 );
-
-DEFINE_RXPRT_EVENT(xprtrdma_noreps);
 
 /**
  ** Callback events

net/sunrpc/clnt.c

···
 	task->tk_status = 0;
 	if (status >= 0) {
 		if (task->tk_rqstp) {
+			xprt_request_init(task);
 			task->tk_action = call_refresh;
 			return;
 		}

net/sunrpc/xprt.c

···
  * Local functions
  */
 static void	 xprt_init(struct rpc_xprt *xprt, struct net *net);
-static void	xprt_request_init(struct rpc_task *, struct rpc_xprt *);
+static __be32	xprt_alloc_xid(struct rpc_xprt *xprt);
 static void	xprt_connect_status(struct rpc_task *task);
 static int      __xprt_get_cong(struct rpc_xprt *, struct rpc_task *);
 static void     __xprt_put_cong(struct rpc_xprt *, struct rpc_rqst *);
···
 		task->tk_status = -EAGAIN;
 		goto out_unlock;
 	}
+	if (!bc_prealloc(req) && !req->rq_xmit_bytes_sent)
+		req->rq_xid = xprt_alloc_xid(xprt);
 	ret = true;
 out_unlock:
 	spin_unlock_bh(&xprt->transport_lock);
···
 out_init_req:
 	xprt->stat.max_slots = max_t(unsigned int, xprt->stat.max_slots,
 				     xprt->num_reqs);
+	spin_unlock(&xprt->reserve_lock);
+
 	task->tk_status = 0;
 	task->tk_rqstp = req;
-	xprt_request_init(task, xprt);
-	spin_unlock(&xprt->reserve_lock);
 }
 EXPORT_SYMBOL_GPL(xprt_alloc_slot);
 
···
 }
 EXPORT_SYMBOL_GPL(xprt_lock_and_alloc_slot);
 
-static void xprt_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
+void xprt_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
 {
 	spin_lock(&xprt->reserve_lock);
 	if (!xprt_dynamic_free_slot(xprt, req)) {
···
 	xprt_wake_up_backlog(xprt);
 	spin_unlock(&xprt->reserve_lock);
 }
+EXPORT_SYMBOL_GPL(xprt_free_slot);
 
 static void xprt_free_all_slots(struct rpc_xprt *xprt)
 {
···
 	xprt->xid = prandom_u32();
 }
 
-static void xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt)
+void xprt_request_init(struct rpc_task *task)
 {
+	struct rpc_xprt *xprt = task->tk_xprt;
 	struct rpc_rqst	*req = task->tk_rqstp;
 
 	INIT_LIST_HEAD(&req->rq_list);
···
 	req->rq_task	= task;
 	req->rq_xprt    = xprt;
 	req->rq_buffer  = NULL;
-	req->rq_xid     = xprt_alloc_xid(xprt);
 	req->rq_connect_cookie = xprt->connect_cookie - 1;
 	req->rq_bytes_sent = 0;
 	req->rq_snd_buf.len = 0;
···
 
 	dprintk("RPC: %5u release request %p\n", task->tk_pid, req);
 	if (likely(!bc_prealloc(req)))
-		xprt_free_slot(xprt, req);
+		xprt->ops->free_slot(xprt, req);
 	else
 		xprt_free_bc_request(req);
 }

net/sunrpc/xprtrdma/backchannel.c

···
 	spin_unlock(&buf->rb_reqslock);
 
 	rpcrdma_destroy_req(req);
-
-	kfree(rqst);
 }
 
-static int rpcrdma_bc_setup_rqst(struct rpcrdma_xprt *r_xprt,
-				 struct rpc_rqst *rqst)
+static int rpcrdma_bc_setup_reqs(struct rpcrdma_xprt *r_xprt,
+				 unsigned int count)
 {
-	struct rpcrdma_regbuf *rb;
-	struct rpcrdma_req *req;
-	size_t size;
+	struct rpc_xprt *xprt = &r_xprt->rx_xprt;
+	struct rpc_rqst *rqst;
+	unsigned int i;
 
-	req = rpcrdma_create_req(r_xprt);
-	if (IS_ERR(req))
-		return PTR_ERR(req);
+	for (i = 0; i < (count << 1); i++) {
+		struct rpcrdma_regbuf *rb;
+		struct rpcrdma_req *req;
+		size_t size;
 
-	size = r_xprt->rx_data.inline_rsize;
-	rb = rpcrdma_alloc_regbuf(size, DMA_TO_DEVICE, GFP_KERNEL);
-	if (IS_ERR(rb))
-		goto out_fail;
-	req->rl_sendbuf = rb;
-	xdr_buf_init(&rqst->rq_snd_buf, rb->rg_base,
-		     min_t(size_t, size, PAGE_SIZE));
-	rpcrdma_set_xprtdata(rqst, req);
+		req = rpcrdma_create_req(r_xprt);
+		if (IS_ERR(req))
+			return PTR_ERR(req);
+		rqst = &req->rl_slot;
+
+		rqst->rq_xprt = xprt;
+		INIT_LIST_HEAD(&rqst->rq_list);
+		INIT_LIST_HEAD(&rqst->rq_bc_list);
+		__set_bit(RPC_BC_PA_IN_USE, &rqst->rq_bc_pa_state);
+		spin_lock_bh(&xprt->bc_pa_lock);
+		list_add(&rqst->rq_bc_pa_list, &xprt->bc_pa_list);
+		spin_unlock_bh(&xprt->bc_pa_lock);
+
+		size = r_xprt->rx_data.inline_rsize;
+		rb = rpcrdma_alloc_regbuf(size, DMA_TO_DEVICE, GFP_KERNEL);
+		if (IS_ERR(rb))
+			goto out_fail;
+		req->rl_sendbuf = rb;
+		xdr_buf_init(&rqst->rq_snd_buf, rb->rg_base,
+			     min_t(size_t, size, PAGE_SIZE));
+	}
 	return 0;
 
 out_fail:
 	rpcrdma_bc_free_rqst(r_xprt, rqst);
 	return -ENOMEM;
-}
-
-/* Allocate and add receive buffers to the rpcrdma_buffer's
- * existing list of rep's. These are released when the
- * transport is destroyed.
- */
-static int rpcrdma_bc_setup_reps(struct rpcrdma_xprt *r_xprt,
-				 unsigned int count)
-{
-	int rc = 0;
-
-	while (count--) {
-		rc = rpcrdma_create_rep(r_xprt);
-		if (rc)
-			break;
-	}
-	return rc;
 }
 
 /**
···
 int xprt_rdma_bc_setup(struct rpc_xprt *xprt, unsigned int reqs)
 {
 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
-	struct rpcrdma_buffer *buffer = &r_xprt->rx_buf;
-	struct rpc_rqst *rqst;
-	unsigned int i;
 	int rc;
 
 	/* The backchannel reply path returns each rpc_rqst to the
···
 	if (reqs > RPCRDMA_BACKWARD_WRS >> 1)
 		goto out_err;
 
-	for (i = 0; i < (reqs << 1); i++) {
-		rqst = kzalloc(sizeof(*rqst), GFP_KERNEL);
-		if (!rqst)
-			goto out_free;
-
-		dprintk("RPC:       %s: new rqst %p\n", __func__, rqst);
-
-		rqst->rq_xprt = &r_xprt->rx_xprt;
-		INIT_LIST_HEAD(&rqst->rq_list);
-		INIT_LIST_HEAD(&rqst->rq_bc_list);
-		__set_bit(RPC_BC_PA_IN_USE, &rqst->rq_bc_pa_state);
-
-		if (rpcrdma_bc_setup_rqst(r_xprt, rqst))
-			goto out_free;
-
-		spin_lock_bh(&xprt->bc_pa_lock);
-		list_add(&rqst->rq_bc_pa_list, &xprt->bc_pa_list);
-		spin_unlock_bh(&xprt->bc_pa_lock);
-	}
-
-	rc = rpcrdma_bc_setup_reps(r_xprt, reqs);
+	rc = rpcrdma_bc_setup_reqs(r_xprt, reqs);
 	if (rc)
 		goto out_free;
 
-	rc = rpcrdma_ep_post_extra_recv(r_xprt, reqs);
-	if (rc)
-		goto out_free;
-
-	buffer->rb_bc_srv_max_requests = reqs;
+	r_xprt->rx_buf.rb_bc_srv_max_requests = reqs;
 	request_module("svcrdma");
 	trace_xprtrdma_cb_setup(r_xprt, reqs);
 	return 0;
···
 	if (rc < 0)
 		goto failed_marshal;
 
+	rpcrdma_post_recvs(r_xprt, true);
 	if (rpcrdma_ep_post(&r_xprt->rx_ia, &r_xprt->rx_ep, req))
 		goto drop_connection;
 	return 0;
···
  */
 void xprt_rdma_bc_free_rqst(struct rpc_rqst *rqst)
 {
+	struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
 	struct rpc_xprt *xprt = rqst->rq_xprt;
 
 	dprintk("RPC:       %s: freeing rqst %p (req %p)\n",
-		__func__, rqst, rpcr_to_rdmar(rqst));
+		__func__, rqst, req);
+
+	rpcrdma_recv_buffer_put(req->rl_reply);
+	req->rl_reply = NULL;
 
 	spin_lock_bh(&xprt->bc_pa_lock);
 	list_add_tail(&rqst->rq_bc_pa_list, &xprt->bc_pa_list);

net/sunrpc/xprtrdma/fmr_ops.c

···
 	fmr_op_release_mr(mr);
 }
 
+/* On success, sets:
+ *	ep->rep_attr.cap.max_send_wr
+ *	ep->rep_attr.cap.max_recv_wr
+ *	cdata->max_requests
+ *	ia->ri_max_segs
+ */
 static int
 fmr_op_open(struct rpcrdma_ia *ia, struct rpcrdma_ep *ep,
 	    struct rpcrdma_create_data_internal *cdata)
 {
+	int max_qp_wr;
+
+	max_qp_wr = ia->ri_device->attrs.max_qp_wr;
+	max_qp_wr -= RPCRDMA_BACKWARD_WRS;
+	max_qp_wr -= 1;
+	if (max_qp_wr < RPCRDMA_MIN_SLOT_TABLE)
+		return -ENOMEM;
+	if (cdata->max_requests > max_qp_wr)
+		cdata->max_requests = max_qp_wr;
+	ep->rep_attr.cap.max_send_wr = cdata->max_requests;
+	ep->rep_attr.cap.max_send_wr += RPCRDMA_BACKWARD_WRS;
+	ep->rep_attr.cap.max_send_wr += 1; /* for ib_drain_sq */
+	ep->rep_attr.cap.max_recv_wr = cdata->max_requests;
+	ep->rep_attr.cap.max_recv_wr += RPCRDMA_BACKWARD_WRS;
+	ep->rep_attr.cap.max_recv_wr += 1; /* for ib_drain_rq */
+
 	ia->ri_max_segs = max_t(unsigned int, 1, RPCRDMA_MAX_DATA_SEGS /
 				RPCRDMA_MAX_FMR_SGES);
 	return 0;
···
 				     mr->mr_sg, i, mr->mr_dir);
 	if (!mr->mr_nents)
 		goto out_dmamap_err;
+	trace_xprtrdma_dma_map(mr);
 
 	for (i = 0, dma_pages = mr->fmr.fm_physaddrs; i < mr->mr_nents; i++)
 		dma_pages[i] = sg_dma_address(&mr->mr_sg[i]);

net/sunrpc/xprtrdma/frwr_ops.c

···
 	frwr_op_release_mr(mr);
 }
 
+/* On success, sets:
+ *	ep->rep_attr.cap.max_send_wr
+ *	ep->rep_attr.cap.max_recv_wr
+ *	cdata->max_requests
+ *	ia->ri_max_segs
+ *
+ * And these FRWR-related fields:
+ *	ia->ri_max_frwr_depth
+ *	ia->ri_mrtype
+ */
 static int
 frwr_op_open(struct rpcrdma_ia *ia, struct rpcrdma_ep *ep,
 	     struct rpcrdma_create_data_internal *cdata)
 {
 	struct ib_device_attr *attrs = &ia->ri_device->attrs;
-	int depth, delta;
+	int max_qp_wr, depth, delta;
 
 	ia->ri_mrtype = IB_MR_TYPE_MEM_REG;
 	if (attrs->device_cap_flags & IB_DEVICE_SG_GAPS_REG)
···
 		} while (delta > 0);
 	}
 
-	ep->rep_attr.cap.max_send_wr *= depth;
-	if (ep->rep_attr.cap.max_send_wr > attrs->max_qp_wr) {
-		cdata->max_requests = attrs->max_qp_wr / depth;
+	max_qp_wr = ia->ri_device->attrs.max_qp_wr;
+	max_qp_wr -= RPCRDMA_BACKWARD_WRS;
+	max_qp_wr -= 1;
+	if (max_qp_wr < RPCRDMA_MIN_SLOT_TABLE)
+		return -ENOMEM;
+	if (cdata->max_requests > max_qp_wr)
+		cdata->max_requests = max_qp_wr;
+	ep->rep_attr.cap.max_send_wr = cdata->max_requests * depth;
+	if (ep->rep_attr.cap.max_send_wr > max_qp_wr) {
+		cdata->max_requests = max_qp_wr / depth;
 		if (!cdata->max_requests)
 			return -EINVAL;
 		ep->rep_attr.cap.max_send_wr = cdata->max_requests *
 					       depth;
 	}
+	ep->rep_attr.cap.max_send_wr += RPCRDMA_BACKWARD_WRS;
+	ep->rep_attr.cap.max_send_wr += 1; /* for ib_drain_sq */
+	ep->rep_attr.cap.max_recv_wr = cdata->max_requests;
+	ep->rep_attr.cap.max_recv_wr += RPCRDMA_BACKWARD_WRS;
+	ep->rep_attr.cap.max_recv_wr += 1; /* for ib_drain_rq */
 
 	ia->ri_max_segs = max_t(unsigned int, 1, RPCRDMA_MAX_DATA_SEGS /
 				ia->ri_max_frwr_depth);
···
 	mr->mr_nents = ib_dma_map_sg(ia->ri_device, mr->mr_sg, i, mr->mr_dir);
 	if (!mr->mr_nents)
 		goto out_dmamap_err;
+	trace_xprtrdma_dma_map(mr);
 
 	ibmr = frwr->fr_mr;
 	n = ib_map_mr_sg(ibmr, mr->mr_sg, mr->mr_nents, NULL, PAGE_SIZE);

net/sunrpc/xprtrdma/module.c

···
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
 /*
  * Copyright (c) 2015, 2017 Oracle.  All rights reserved.
  */

net/sunrpc/xprtrdma/rpc_rdma.c

···
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
 /*
  * Copyright (c) 2014-2017 Oracle.  All rights reserved.
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
···
 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
 # define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
-
-static const char transfertypes[][12] = {
-	"inline",	/* no chunks */
-	"read list",	/* some argument via rdma read */
-	"*read list",	/* entire request via rdma read */
-	"write list",	/* some result via rdma write */
-	"reply chunk"	/* entire reply via rdma write */
-};
 
 /* Returns size of largest RPC-over-RDMA header in a Call message
  *
···
 			 */
 			*ppages = alloc_page(GFP_ATOMIC);
 			if (!*ppages)
-				return -EAGAIN;
+				return -ENOBUFS;
 		}
 		seg->mr_page = *ppages;
 		seg->mr_offset = (char *)page_base;
···
 		seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
 						   false, &mr);
 		if (IS_ERR(seg))
-			goto out_maperr;
+			return PTR_ERR(seg);
 		rpcrdma_mr_push(mr, &req->rl_registered);
 
 		if (encode_read_segment(xdr, mr, pos) < 0)
···
 	} while (nsegs);
 
 	return 0;
-
-out_maperr:
-	if (PTR_ERR(seg) == -EAGAIN)
-		xprt_wait_for_buffer_space(rqst->rq_task, NULL);
-	return PTR_ERR(seg);
 }
 
 /* Register and XDR encode the Write list. Supports encoding a list
···
 		seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
 						   true, &mr);
 		if (IS_ERR(seg))
-			goto out_maperr;
+			return PTR_ERR(seg);
 		rpcrdma_mr_push(mr, &req->rl_registered);
 
 		if (encode_rdma_segment(xdr, mr) < 0)
···
 	*segcount = cpu_to_be32(nchunks);
 
 	return 0;
-
-out_maperr:
-	if (PTR_ERR(seg) == -EAGAIN)
-		xprt_wait_for_buffer_space(rqst->rq_task, NULL);
-	return PTR_ERR(seg);
 }
 
 /* Register and XDR encode the Reply chunk. Supports encoding an array
···
 		seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
 						   true, &mr);
 		if (IS_ERR(seg))
-			goto out_maperr;
+			return PTR_ERR(seg);
 		rpcrdma_mr_push(mr, &req->rl_registered);
 
 		if (encode_rdma_segment(xdr, mr) < 0)
···
 	*segcount = cpu_to_be32(nchunks);
 
 	return 0;
-
-out_maperr:
-	if (PTR_ERR(seg) == -EAGAIN)
-		xprt_wait_for_buffer_space(rqst->rq_task, NULL);
-	return PTR_ERR(seg);
 }
 
 /**
···
 {
 	req->rl_sendctx = rpcrdma_sendctx_get_locked(&r_xprt->rx_buf);
 	if (!req->rl_sendctx)
-		return -ENOBUFS;
+		return -EAGAIN;
 	req->rl_sendctx->sc_wr.num_sge = 0;
 	req->rl_sendctx->sc_unmap_count = 0;
 	req->rl_sendctx->sc_req = req;
···
 	return 0;
 
 out_err:
-	r_xprt->rx_stats.failed_marshal_count++;
+	switch (ret) {
+	case -EAGAIN:
+		xprt_wait_for_buffer_space(rqst->rq_task, NULL);
+		break;
+	case -ENOBUFS:
+		break;
+	default:
+		r_xprt->rx_stats.failed_marshal_count++;
+	}
 	return ret;
 }
 
···
 
 out_short:
 	pr_warn("RPC/RDMA short backward direction call\n");
-	if (rpcrdma_ep_post_recv(&r_xprt->rx_ia, rep))
-		xprt_disconnect_done(&r_xprt->rx_xprt);
 	return true;
 }
 #else	/* CONFIG_SUNRPC_BACKCHANNEL */
···
 	u32 credits;
 	__be32 *p;
 
+	--buf->rb_posted_receives;
+
 	if (rep->rr_hdrbuf.head[0].iov_len == 0)
 		goto out_badstatus;
 
+	/* Fixed transport header fields */
 	xdr_init_decode(&rep->rr_stream, &rep->rr_hdrbuf,
 			rep->rr_hdrbuf.head[0].iov_base);
-
-	/* Fixed transport header fields */
 	p = xdr_inline_decode(&rep->rr_stream, 4 * sizeof(*p));
 	if (unlikely(!p))
 		goto out_shortreply;
···
 
 	trace_xprtrdma_reply(rqst->rq_task, rep, req, credits);
 
+	rpcrdma_post_recvs(r_xprt, false);
 	queue_work(rpcrdma_receive_wq, &rep->rr_work);
-	return;
-
-out_badstatus:
-	rpcrdma_recv_buffer_put(rep);
-	if (r_xprt->rx_ep.rep_connected == 1) {
-		r_xprt->rx_ep.rep_connected = -EIO;
-		rpcrdma_conn_func(&r_xprt->rx_ep);
-	}
 	return;
 
 out_badversion:
···
  * receive buffer before returning.
  */
 repost:
-	r_xprt->rx_stats.bad_reply_count++;
-	if (rpcrdma_ep_post_recv(&r_xprt->rx_ia, rep))
-		rpcrdma_recv_buffer_put(rep);
+	rpcrdma_post_recvs(r_xprt, false);
+out_badstatus:
+	rpcrdma_recv_buffer_put(rep);
 }

net/sunrpc/xprtrdma/svc_rdma_backchannel.c

···
 	.reserve_xprt		= xprt_reserve_xprt_cong,
 	.release_xprt		= xprt_release_xprt_cong,
 	.alloc_slot		= xprt_alloc_slot,
+	.free_slot		= xprt_free_slot,
 	.release_request	= xprt_release_rqst_cong,
 	.buf_alloc		= xprt_rdma_bc_allocate,
 	.buf_free		= xprt_rdma_bc_free,

net/sunrpc/xprtrdma/transport.c

···
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
 /*
  * Copyright (c) 2014-2017 Oracle.  All rights reserved.
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
···
 		return ERR_PTR(-EBADF);
 	}
 
-	xprt = xprt_alloc(args->net, sizeof(struct rpcrdma_xprt),
-			xprt_rdma_slot_table_entries,
-			xprt_rdma_slot_table_entries);
+	xprt = xprt_alloc(args->net, sizeof(struct rpcrdma_xprt), 0, 0);
 	if (xprt == NULL) {
 		dprintk("RPC:       %s: couldn't allocate rpcrdma_xprt\n",
 			__func__);
···
 		xprt_set_bound(xprt);
 	xprt_rdma_format_addresses(xprt, sap);
 
-	cdata.max_requests = xprt->max_reqs;
+	cdata.max_requests = xprt_rdma_slot_table_entries;
 
 	cdata.rsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA write max */
 	cdata.wsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA read max */
···
 	}
 }
 
+/**
+ * xprt_rdma_alloc_slot - allocate an rpc_rqst
+ * @xprt: controlling RPC transport
+ * @task: RPC task requesting a fresh rpc_rqst
+ *
+ * tk_status values:
+ *	%0 if task->tk_rqstp points to a fresh rpc_rqst
+ *	%-EAGAIN if no rpc_rqst is available; queued on backlog
+ */
+static void
+xprt_rdma_alloc_slot(struct rpc_xprt *xprt, struct rpc_task *task)
+{
+	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
+	struct rpcrdma_req *req;
+
+	req = rpcrdma_buffer_get(&r_xprt->rx_buf);
+	if (!req)
+		goto out_sleep;
+	task->tk_rqstp = &req->rl_slot;
+	task->tk_status = 0;
+	return;
+
+out_sleep:
+	rpc_sleep_on(&xprt->backlog, task, NULL);
+	task->tk_status = -EAGAIN;
+}
+
+/**
+ * xprt_rdma_free_slot - release an rpc_rqst
+ * @xprt: controlling RPC transport
+ * @rqst: rpc_rqst to release
+ *
+ */
+static void
+xprt_rdma_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *rqst)
+{
+	memset(rqst, 0, sizeof(*rqst));
+	rpcrdma_buffer_put(rpcr_to_rdmar(rqst));
+	rpc_wake_up_next(&xprt->backlog);
+}
+
 static bool
 rpcrdma_get_sendbuf(struct rpcrdma_xprt *r_xprt, struct rpcrdma_req *req,
 		    size_t size, gfp_t flags)
···
 {
 	struct rpc_rqst *rqst = task->tk_rqstp;
 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(rqst->rq_xprt);
-	struct rpcrdma_req *req;
+	struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
 	gfp_t flags;
-
-	req = rpcrdma_buffer_get(&r_xprt->rx_buf);
-	if (req == NULL)
-		goto out_get;
 
 	flags = RPCRDMA_DEF_GFP;
 	if (RPC_IS_SWAPPER(task))
···
 	if (!rpcrdma_get_recvbuf(r_xprt, req, rqst->rq_rcvsize, flags))
 		goto out_fail;
 
-	rpcrdma_set_xprtdata(rqst, req);
 	rqst->rq_buffer = req->rl_sendbuf->rg_base;
 	rqst->rq_rbuffer = req->rl_recvbuf->rg_base;
 	trace_xprtrdma_allocate(task, req);
 	return 0;
 
 out_fail:
-	rpcrdma_buffer_put(req);
-out_get:
 	trace_xprtrdma_allocate(task, NULL);
 	return -ENOMEM;
 }
···
 	if (test_bit(RPCRDMA_REQ_F_PENDING, &req->rl_flags))
 		rpcrdma_release_rqst(r_xprt, req);
 	trace_xprtrdma_rpc_done(task, req);
-	rpcrdma_buffer_put(req);
 }
 
 /**
···
 	rc = rpcrdma_marshal_req(r_xprt, rqst);
 	if (rc < 0)
 		goto failed_marshal;
-
-	if (req->rl_reply == NULL) 		/* e.g. reconnection */
-		rpcrdma_recv_buffer_get(req);
 
 	/* Must suppress retransmit to maintain credits */
 	if (rqst->rq_connect_cookie == xprt->connect_cookie)
···
 static const struct rpc_xprt_ops xprt_rdma_procs = {
 	.reserve_xprt		= xprt_reserve_xprt_cong,
 	.release_xprt		= xprt_release_xprt_cong, /* sunrpc/xprt.c */
-	.alloc_slot		= xprt_alloc_slot,
+	.alloc_slot		= xprt_rdma_alloc_slot,
+	.free_slot		= xprt_rdma_free_slot,
 	.release_request	= xprt_release_rqst_cong,       /* ditto */
 	.set_retrans_timeout	= xprt_set_retrans_timeout_def, /* ditto */
 	.timer			= xprt_rdma_timer,

net/sunrpc/xprtrdma/verbs.c

···
+// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
 /*
  * Copyright (c) 2014-2017 Oracle.  All rights reserved.
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
···
 /*
  * internal functions
  */
+static void rpcrdma_sendctx_put_locked(struct rpcrdma_sendctx *sc);
 static void rpcrdma_mrs_create(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf);
+static int rpcrdma_create_rep(struct rpcrdma_xprt *r_xprt, bool temp);
 static void rpcrdma_dma_unmap_regbuf(struct rpcrdma_regbuf *rb);
 
 struct workqueue_struct *rpcrdma_receive_wq __read_mostly;
···
 					       rr_cqe);
 
 	/* WARNING: Only wr_id and status are reliable at this point */
-	trace_xprtrdma_wc_receive(rep, wc);
+	trace_xprtrdma_wc_receive(wc);
 	if (wc->status != IB_WC_SUCCESS)
 		goto out_fail;
 
···
 		complete(&ia->ri_done);
 		break;
 	case RDMA_CM_EVENT_ADDR_ERROR:
-		ia->ri_async_rc = -EHOSTUNREACH;
+		ia->ri_async_rc = -EPROTO;
 		complete(&ia->ri_done);
 		break;
 	case RDMA_CM_EVENT_ROUTE_ERROR:
···
 		connstate = -ENOTCONN;
 		goto connected;
 	case RDMA_CM_EVENT_UNREACHABLE:
-		connstate = -ENETDOWN;
+		connstate = -ENETUNREACH;
 		goto connected;
 	case RDMA_CM_EVENT_REJECTED:
 		dprintk("rpcrdma: connection to %s:%s rejected: %s\n",
···
 	init_completion(&ia->ri_done);
 	init_completion(&ia->ri_remove_done);
 
-	id = rdma_create_id(&init_net, rpcrdma_conn_upcall, xprt, RDMA_PS_TCP,
-			    IB_QPT_RC);
+	id = rdma_create_id(xprt->rx_xprt.xprt_net, rpcrdma_conn_upcall,
+			    xprt, RDMA_PS_TCP, IB_QPT_RC);
 	if (IS_ERR(id)) {
 		rc = PTR_ERR(id);
 		dprintk("RPC:       %s: rdma_create_id() failed %i\n",
···
 		  struct rpcrdma_create_data_internal *cdata)
 {
 	struct rpcrdma_connect_private *pmsg = &ep->rep_cm_private;
-	unsigned int max_qp_wr, max_sge;
 	struct ib_cq *sendcq, *recvcq;
+	unsigned int max_sge;
 	int rc;
 
 	max_sge = min_t(unsigned int, ia->ri_device->attrs.max_sge,
···
 	}
 	ia->ri_max_send_sges = max_sge;
 
-	if (ia->ri_device->attrs.max_qp_wr <= RPCRDMA_BACKWARD_WRS) {
-		dprintk("RPC:       %s: insufficient wqe's available\n",
-			__func__);
-		return -ENOMEM;
-	}
-	max_qp_wr = ia->ri_device->attrs.max_qp_wr - RPCRDMA_BACKWARD_WRS - 1;
-
-	/* check provider's send/recv wr limits */
-	if (cdata->max_requests > max_qp_wr)
-		cdata->max_requests = max_qp_wr;
+	rc = ia->ri_ops->ro_open(ia, ep, cdata);
+	if (rc)
+		return rc;
 
 	ep->rep_attr.event_handler = rpcrdma_qp_async_error_upcall;
 	ep->rep_attr.qp_context = ep;
 	ep->rep_attr.srq = NULL;
-	ep->rep_attr.cap.max_send_wr = cdata->max_requests;
-	ep->rep_attr.cap.max_send_wr += RPCRDMA_BACKWARD_WRS;
-	ep->rep_attr.cap.max_send_wr += 1;	/* drain cqe */
-	rc = ia->ri_ops->ro_open(ia, ep, cdata);
-	if (rc)
-		return rc;
-	ep->rep_attr.cap.max_recv_wr = cdata->max_requests;
-	ep->rep_attr.cap.max_recv_wr += RPCRDMA_BACKWARD_WRS;
-	ep->rep_attr.cap.max_recv_wr += 1;	/* drain cqe */
 	ep->rep_attr.cap.max_send_sge = max_sge;
 	ep->rep_attr.cap.max_recv_sge = 1;
 	ep->rep_attr.cap.max_inline_data = 0;
···
 {
 	struct rpcrdma_xprt *r_xprt = container_of(ia, struct rpcrdma_xprt,
 						   rx_ia);
-	unsigned int extras;
 	int rc;
 
 retry:
···
 	}
 
 	dprintk("RPC:       %s: connected\n", __func__);
-	extras = r_xprt->rx_buf.rb_bc_srv_max_requests;
-	if (extras)
-		rpcrdma_ep_post_extra_recv(r_xprt, extras);
+
+	rpcrdma_post_recvs(r_xprt, true);
 
 out:
 	if (rc)
···
 		sc->sc_xprt = r_xprt;
 		buf->rb_sc_ctxs[i] = sc;
 	}
+	buf->rb_flags = 0;
 
 	return 0;
 
···
 	 * completions recently. This is a sign the Send Queue is
 	 * backing up. Cause the caller to pause and try again.
 	 */
-	dprintk("RPC:       %s: empty sendctx queue\n", __func__);
+	set_bit(RPCRDMA_BUF_F_EMPTY_SCQ, &buf->rb_flags);
 	r_xprt = container_of(buf, struct rpcrdma_xprt, rx_buf);
 	r_xprt->rx_stats.empty_sendctx_q++;
 	return NULL;
···
  *
  * The caller serializes calls to this function (per rpcrdma_buffer).
  */
-void rpcrdma_sendctx_put_locked(struct rpcrdma_sendctx *sc)
+static void
+rpcrdma_sendctx_put_locked(struct rpcrdma_sendctx *sc)
 {
 	struct rpcrdma_buffer *buf = &sc->sc_xprt->rx_buf;
 	unsigned long next_tail;
···
 
 	/* Paired with READ_ONCE */
 	smp_store_release(&buf->rb_sc_tail, next_tail);
+
+	if (test_and_clear_bit(RPCRDMA_BUF_F_EMPTY_SCQ, &buf->rb_flags)) {
+		smp_mb__after_atomic();
+		xprt_write_space(&sc->sc_xprt->rx_xprt);
+	}
 }
 
 static void
···
 	return req;
 }
 
-/**
- * rpcrdma_create_rep - Allocate an rpcrdma_rep object
- * @r_xprt: controlling transport
- *
- * Returns 0 on success or a negative errno on failure.
- */
-int
-rpcrdma_create_rep(struct rpcrdma_xprt *r_xprt)
+static int
+rpcrdma_create_rep(struct rpcrdma_xprt *r_xprt, bool temp)
 {
 	struct rpcrdma_create_data_internal *cdata = &r_xprt->rx_data;
 	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
···
 	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_list, &buf->rb_recv_bufs);
···
 		list_add(&req->rl_list, &buf->rb_send_bufs);
 	}
 
+	buf->rb_posted_receives = 0;
 	INIT_LIST_HEAD(&buf->rb_recv_bufs);
-	for (i = 0; i <= buf->rb_max_requests; i++) {
-		rc = rpcrdma_create_rep(r_xprt);
-		if (rc)
-			goto out;
-	}
 
 	rc = rpcrdma_sendctxs_create(r_xprt);
 	if (rc)
···
 out:
 	rpcrdma_buffer_destroy(buf);
 	return rc;
-}
-
-static struct rpcrdma_req *
-rpcrdma_buffer_get_req_locked(struct rpcrdma_buffer *buf)
-{
-	struct rpcrdma_req *req;
-
-	req = list_first_entry(&buf->rb_send_bufs,
-			       struct rpcrdma_req, rl_list);
-	list_del_init(&req->rl_list);
-	return req;
-}
-
-static struct rpcrdma_rep *
-rpcrdma_buffer_get_rep_locked(struct rpcrdma_buffer *buf)
-{
-	struct rpcrdma_rep *rep;
-
-	rep = list_first_entry(&buf->rb_recv_bufs,
-			       struct rpcrdma_rep, rr_list);
-	list_del(&rep->rr_list);
-	return rep;
 }
 
 static void
···
 	while (!list_empty(&buf->rb_recv_bufs)) {
 		struct rpcrdma_rep *rep;
 
-		rep = rpcrdma_buffer_get_rep_locked(buf);
+		rep = list_first_entry(&buf->rb_recv_bufs,
+				       struct rpcrdma_rep, rr_list);
+		list_del(&rep->rr_list);
 		rpcrdma_destroy_rep(rep);
 	}
-	buf->rb_send_count = 0;
 
 	spin_lock(&buf->rb_reqslock);
 	while (!list_empty(&buf->rb_allreqs)) {
···
 		spin_lock(&buf->rb_reqslock);
 	}
 	spin_unlock(&buf->rb_reqslock);
-	buf->rb_recv_count = 0;
 
 	rpcrdma_mrs_destroy(buf);
 }
···
 	__rpcrdma_mr_put(&r_xprt->rx_buf, mr);
 }
 
-static struct rpcrdma_rep *
-rpcrdma_buffer_get_rep(struct rpcrdma_buffer *buffers)
-{
-	/* If an RPC previously completed without a reply (say, a
-	 * credential problem or a soft timeout occurs) then hold off
-	 * on supplying more Receive buffers until the number of new
-	 * pending RPCs catches up to the number of posted Receives.
-	 */
-	if (unlikely(buffers->rb_send_count < buffers->rb_recv_count))
-		return NULL;
-
-	if (unlikely(list_empty(&buffers->rb_recv_bufs)))
-		return NULL;
-	buffers->rb_recv_count++;
-	return rpcrdma_buffer_get_rep_locked(buffers);
-}
-
-/*
- * Get a set of request/reply buffers.
+/**
+ * rpcrdma_buffer_get - Get a request buffer
+ * @buffers: Buffer pool from which to obtain a buffer
  *
- * Reply buffer (if available) is attached to send buffer upon return.
+ * Returns a fresh rpcrdma_req, or NULL if none are available.
  */
 struct rpcrdma_req *
 rpcrdma_buffer_get(struct rpcrdma_buffer *buffers)
···
 	struct rpcrdma_req *req;
 
 	spin_lock(&buffers->rb_lock);
-	if (list_empty(&buffers->rb_send_bufs))
-		goto out_reqbuf;
-	buffers->rb_send_count++;
-	req = rpcrdma_buffer_get_req_locked(buffers);
-	req->rl_reply = rpcrdma_buffer_get_rep(buffers);
+	req = list_first_entry_or_null(&buffers->rb_send_bufs,
+				       struct rpcrdma_req, rl_list);
+	if (req)
+		list_del_init(&req->rl_list);
 	spin_unlock(&buffers->rb_lock);
-
 	return req;
-
-out_reqbuf:
-	spin_unlock(&buffers->rb_lock);
-	return NULL;
 }
 
-/*
- * Put request/reply buffers back into pool.
- * Pre-decrement counter/array index.
+/**
+ * rpcrdma_buffer_put - Put request/reply buffers back into pool
+ * @req: object to return
+ *
  */
 void
 rpcrdma_buffer_put(struct rpcrdma_req *req)
···
 	req->rl_reply = NULL;
 
 	spin_lock(&buffers->rb_lock);
-	buffers->rb_send_count--;
-	list_add_tail(&req->rl_list, &buffers->rb_send_bufs);
+	list_add(&req->rl_list, &buffers->rb_send_bufs);
 	if (rep) {
-		buffers->rb_recv_count--;
-		list_add_tail(&rep->rr_list, &buffers->rb_recv_bufs);
+		if (!rep->rr_temp) {
+			list_add(&rep->rr_list, &buffers->rb_recv_bufs);
+			rep = NULL;
+		}
 	}
 	spin_unlock(&buffers->rb_lock);
-}
-
-/*
- * Recover reply buffers from pool.
- * This happens when recovering from disconnect.
- */
-void
-rpcrdma_recv_buffer_get(struct rpcrdma_req *req)
-{
-	struct rpcrdma_buffer *buffers = req->rl_buffer;
-
-	spin_lock(&buffers->rb_lock);
-	req->rl_reply = rpcrdma_buffer_get_rep(buffers);
-	spin_unlock(&buffers->rb_lock);
+	if (rep)
+		rpcrdma_destroy_rep(rep);
 }
 
 /*
···
 {
 	struct rpcrdma_buffer *buffers = &rep->rr_rxprt->rx_buf;
 
-	spin_lock(&buffers->rb_lock);
-	buffers->rb_recv_count--;
-	list_add_tail(&rep->rr_list, &buffers->rb_recv_bufs);
-	spin_unlock(&buffers->rb_lock);
+	if (!rep->rr_temp) {
+		spin_lock(&buffers->rb_lock);
+		list_add(&rep->rr_list, &buffers->rb_recv_bufs);
+		spin_unlock(&buffers->rb_lock);
+	} else {
+		rpcrdma_destroy_rep(rep);
+	}
 }
 
 /**
···
 	struct ib_send_wr *send_wr = &req->rl_sendctx->sc_wr;
 	int rc;
 
-	if (req->rl_reply) {
-		rc = rpcrdma_ep_post_recv(ia, req->rl_reply);
-		if (rc)
-			return rc;
-		req->rl_reply = NULL;
-	}
-
 	if (!ep->rep_send_count ||
 	    test_bit(RPCRDMA_REQ_F_TX_RESOURCES, &req->rl_flags)) {
 		send_wr->send_flags |= IB_SEND_SIGNALED;
···
 	return 0;
 }
 
-int
-rpcrdma_ep_post_recv(struct rpcrdma_ia *ia,
-		     struct rpcrdma_rep *rep)
-{
-	struct ib_recv_wr *recv_wr_fail;
-	int rc;
-
-	if (!rpcrdma_dma_map_regbuf(ia, rep->rr_rdmabuf))
-		goto out_map;
-	rc = ib_post_recv(ia->ri_id->qp, &rep->rr_recv_wr, &recv_wr_fail);
-	trace_xprtrdma_post_recv(rep, rc);
-	if (rc)
-		return -ENOTCONN;
-	return 0;
-
-out_map:
-	pr_err("rpcrdma: failed to DMA map the Receive buffer\n");
-	return -EIO;
-}
-
 /**
- * rpcrdma_ep_post_extra_recv - Post buffers for incoming backchannel requests
- * @r_xprt: transport associated with these backchannel resources
- * @count: minimum number of incoming requests expected
+ * rpcrdma_post_recvs - Maybe post some Receive buffers
+ * @r_xprt: controlling transport
+ * @temp: when true, allocate temp rpcrdma_rep objects
  *
- * Returns zero if all requested buffers were posted, or a negative errno.
  */
-int
-rpcrdma_ep_post_extra_recv(struct rpcrdma_xprt *r_xprt, unsigned int count)
+void
+rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, bool temp)
 {
-	struct rpcrdma_buffer *buffers = &r_xprt->rx_buf;
-	struct rpcrdma_ia *ia = &r_xprt->rx_ia;
-	struct rpcrdma_rep *rep;
-	int rc;
+	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
+	struct ib_recv_wr *wr, *bad_wr;
+	int needed, count, rc;
 
-	while (count--) {
-		spin_lock(&buffers->rb_lock);
-		if (list_empty(&buffers->rb_recv_bufs))
-			goto out_reqbuf;
-		rep = rpcrdma_buffer_get_rep_locked(buffers);
-		spin_unlock(&buffers->rb_lock);
+	needed = buf->rb_credits + (buf->rb_bc_srv_max_requests << 1);
+	if (buf->rb_posted_receives > needed)
+		return;
+	needed -= buf->rb_posted_receives;
 
-		rc = rpcrdma_ep_post_recv(ia, rep);
-		if (rc)
-			goto out_rc;
+	count = 0;
+	wr = NULL;
+	while (needed) {
+		struct rpcrdma_regbuf *rb;
+		struct rpcrdma_rep *rep;
+
+		spin_lock(&buf->rb_lock);
+		rep = list_first_entry_or_null(&buf->rb_recv_bufs,
+					       struct rpcrdma_rep, rr_list);
+		if (likely(rep))
+			list_del(&rep->rr_list);
+		spin_unlock(&buf->rb_lock);
+		if (!rep) {
+			if (rpcrdma_create_rep(r_xprt, temp))
+				break;
+			continue;
+		}
+
+		rb = rep->rr_rdmabuf;
+		if (!rpcrdma_regbuf_is_mapped(rb)) {
+			if (!__rpcrdma_dma_map_regbuf(&r_xprt->rx_ia, rb)) {
+				rpcrdma_recv_buffer_put(rep);
+				break;
+			}
+		}
+
+		trace_xprtrdma_post_recv(rep->rr_recv_wr.wr_cqe);
+		rep->rr_recv_wr.next = wr;
+		wr = &rep->rr_recv_wr;
+		++count;
+		--needed;
 	}
+	if (!count)
+		return;
 
-	return 0;
+	rc = ib_post_recv(r_xprt->rx_ia.ri_id->qp, wr, &bad_wr);
+	if (rc) {
+		for (wr = bad_wr; wr; wr = wr->next) {
+			struct rpcrdma_rep *rep;
 
-out_reqbuf:
-	spin_unlock(&buffers->rb_lock);
-	trace_xprtrdma_noreps(r_xprt);
-	return -ENOMEM;
-
-out_rc:
-	rpcrdma_recv_buffer_put(rep);
-	return rc;
+			rep = container_of(wr, struct rpcrdma_rep, rr_recv_wr);
+			rpcrdma_recv_buffer_put(rep);
+			--count;
+		}
+	}
+	buf->rb_posted_receives += count;
+	trace_xprtrdma_post_recvs(r_xprt, count, rc);
 }

net/sunrpc/xprtrdma/xprt_rdma.h

···
+/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */
 /*
  * Copyright (c) 2014-2017 Oracle.  All rights reserved.
  * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
···
 	__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 work_struct	rr_work;
···
 struct rpcrdma_buffer;
 struct rpcrdma_req {
 	struct list_head	rl_list;
+	struct rpc_rqst		rl_slot;
 	struct rpcrdma_buffer	*rl_buffer;
 	struct rpcrdma_rep	*rl_reply;
 	struct xdr_stream	rl_stream;
···
 	RPCRDMA_REQ_F_TX_RESOURCES,
 };
 
-static inline void
-rpcrdma_set_xprtdata(struct rpc_rqst *rqst, struct rpcrdma_req *req)
-{
-	rqst->rq_xprtdata = req;
-}
-
 static inline struct rpcrdma_req *
 rpcr_to_rdmar(const struct rpc_rqst *rqst)
 {
-	return rqst->rq_xprtdata;
+	return container_of(rqst, struct rpcrdma_req, rl_slot);
 }
 
 static inline void
···
 	struct rpcrdma_sendctx	**rb_sc_ctxs;
 
 	spinlock_t		rb_lock;	/* protect buf lists */
-	int			rb_send_count, rb_recv_count;
 	struct list_head	rb_send_bufs;
 	struct list_head	rb_recv_bufs;
+	unsigned long		rb_flags;
 	u32			rb_max_requests;
 	u32			rb_credits;	/* most recent credit grant */
+	int			rb_posted_receives;
 
 	u32			rb_bc_srv_max_requests;
 	spinlock_t		rb_reqslock;	/* protect rb_allreqs */
···
 	struct delayed_work	rb_refresh_worker;
 };
 #define rdmab_to_ia(b) (&container_of((b), struct rpcrdma_xprt, rx_buf)->rx_ia)
+
+/* rb_flags */
+enum {
+	RPCRDMA_BUF_F_EMPTY_SCQ = 0,
+};
 
 /*
  * Internal structure for transport instance creation. This
···
 
 int rpcrdma_ep_post(struct rpcrdma_ia *, struct rpcrdma_ep *,
 				struct rpcrdma_req *);
-int rpcrdma_ep_post_recv(struct rpcrdma_ia *, struct rpcrdma_rep *);
+void rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, bool temp);
 
 /*
  * Buffer calls - xprtrdma/verbs.c
  */
 struct rpcrdma_req *rpcrdma_create_req(struct rpcrdma_xprt *);
 void rpcrdma_destroy_req(struct rpcrdma_req *);
-int rpcrdma_create_rep(struct rpcrdma_xprt *r_xprt);
 int rpcrdma_buffer_create(struct rpcrdma_xprt *);
 void rpcrdma_buffer_destroy(struct rpcrdma_buffer *);
 struct rpcrdma_sendctx *rpcrdma_sendctx_get_locked(struct rpcrdma_buffer *buf);
-void rpcrdma_sendctx_put_locked(struct rpcrdma_sendctx *sc);
 
 struct rpcrdma_mr *rpcrdma_mr_get(struct rpcrdma_xprt *r_xprt);
 void rpcrdma_mr_put(struct rpcrdma_mr *mr);
···
 
 struct rpcrdma_req *rpcrdma_buffer_get(struct rpcrdma_buffer *);
 void rpcrdma_buffer_put(struct rpcrdma_req *);
-void rpcrdma_recv_buffer_get(struct rpcrdma_req *);
 void rpcrdma_recv_buffer_put(struct rpcrdma_rep *);
 
 struct rpcrdma_regbuf *rpcrdma_alloc_regbuf(size_t, enum dma_data_direction,
···
 		return true;
 	return __rpcrdma_dma_map_regbuf(ia, rb);
 }
-
-int rpcrdma_ep_post_extra_recv(struct rpcrdma_xprt *, unsigned int);
 
 int rpcrdma_alloc_wq(void);
 void rpcrdma_destroy_wq(void);

net/sunrpc/xprtsock.c

···
 	.reserve_xprt		= xprt_reserve_xprt,
 	.release_xprt		= xs_tcp_release_xprt,
 	.alloc_slot		= xprt_alloc_slot,
+	.free_slot		= xprt_free_slot,
 	.rpcbind		= xs_local_rpcbind,
 	.set_port		= xs_local_set_port,
 	.connect		= xs_local_connect,
···
 	.reserve_xprt		= xprt_reserve_xprt_cong,
 	.release_xprt		= xprt_release_xprt_cong,
 	.alloc_slot		= xprt_alloc_slot,
+	.free_slot		= xprt_free_slot,
 	.rpcbind		= rpcb_getport_async,
 	.set_port		= xs_set_port,
 	.connect		= xs_connect,
···
 	.reserve_xprt		= xprt_reserve_xprt,
 	.release_xprt		= xs_tcp_release_xprt,
 	.alloc_slot		= xprt_lock_and_alloc_slot,
+	.free_slot		= xprt_free_slot,
 	.rpcbind		= rpcb_getport_async,
 	.set_port		= xs_set_port,
 	.connect		= xs_connect,
···
 	.reserve_xprt		= xprt_reserve_xprt,
 	.release_xprt		= xprt_release_xprt,
 	.alloc_slot		= xprt_alloc_slot,
+	.free_slot		= xprt_free_slot,
 	.buf_alloc		= bc_malloc,
 	.buf_free		= bc_free,
 	.send_request		= bc_send_request,