afs: Overhaul volume and server record caching and fileserver rotation

The current code assumes that volumes and servers are per-cell and are
never shared, but this is not enforced, and, indeed, public cells do exist
that are aliases of each other.  Further, an organisation can, say, set up
a public cell and a private cell with overlapping, but not identical, sets
of servers.  The difference is purely in the database attached to the VL
servers.

The current code will malfunction if it sees a server in two cells as it
assumes global address -> server record mappings and that each server is in
just one cell.

Further, each server may have multiple addresses - and may have addresses
of different families (IPv4 and IPv6, say).

To this end, the following structural changes are made:

 (1) Server record management is overhauled:

     (a) Server records are made independent of cell.  The namespace keeps
     	 track of them, volume records have lists of them and each vnode
     	 has a server on which its callback interest currently resides.

     (b) The cell record no longer keeps a list of servers known to be in
     	 that cell.

     (c) The server records are now kept in a flat list because there's no
     	 single address to sort on.

     (d) Server records are now keyed by their UUID within the namespace.

     (e) The addresses for a server are obtained with the VL.GetAddrsU
     	 rather than with VL.GetEntryByName, using the server's UUID as a
     	 parameter.

     (f) Cached server records are garbage collected after a period of
     	 non-use and are counted out of existence before purging is allowed
     	 to complete.  This protects the work functions against rmmod.

     (g) The servers list is now in /proc/fs/afs/servers.

 (2) Volume record management is overhauled:

     (a) An RCU-replaceable server list is introduced.  This tracks both
     	 servers and their coresponding callback interests.

     (b) The superblock is now keyed on cell record and numeric volume ID.

     (c) The volume record is now tied to the superblock which mounts it,
     	 and is activated when mounted and deactivated when unmounted.
     	 This makes it easier to handle the cache cookie without causing a
     	 double-use in fscache.

     (d) The volume record is loaded from the VLDB using VL.GetEntryByNameU
     	 to get the server UUID list.

     (e) The volume name is updated if it is seen to have changed when the
     	 volume is updated (the update is keyed on the volume ID).

 (3) The vlocation record is got rid of and VLDB records are no longer
     cached.  Sufficient information is stored in the volume record, though
     an update to a volume record is now no longer shared between related
     volumes (volumes come in bundles of three: R/W, R/O and backup).

and the following procedural changes are made:

 (1) The fileserver cursor introduced previously is now fleshed out and
     used to iterate over fileservers and their addresses.

 (2) Volume status is checked during iteration, and the server list is
     replaced if a change is detected.

 (3) Server status is checked during iteration, and the address list is
     replaced if a change is detected.

 (4) The abort code is saved into the address list cursor and -ECONNABORTED
     returned in afs_make_call() if a remote abort happened rather than
     translating the abort into an error message.  This allows actions to
     be taken depending on the abort code more easily.

     (a) If a VMOVED abort is seen then this is handled by rechecking the
     	 volume and restarting the iteration.

     (b) If a VBUSY, VRESTARTING or VSALVAGING abort is seen then this is
         handled by sleeping for a short period and retrying and/or trying
         other servers that might serve that volume.  A message is also
         displayed once until the condition has cleared.

     (c) If a VOFFLINE abort is seen, then this is handled as VBUSY for the
     	 moment.

     (d) If a VNOVOL abort is seen, the volume is rechecked in the VLDB to
     	 see if it has been deleted; if not, the fileserver is probably
     	 indicating that the volume couldn't be attached and needs
     	 salvaging.

     (e) If statfs() sees one of these aborts, it does not sleep, but
     	 rather returns an error, so as not to block the umount program.

 (5) The fileserver iteration functions in vnode.c are now merged into
     their callers and more heavily macroised around the cursor.  vnode.c
     is removed.

 (6) Operations on a particular vnode are serialised on that vnode because
     the server will lock that vnode whilst it operates on it, so a second
     op sent will just have to wait.

 (7) Fileservers are probed with FS.GetCapabilities before being used.
     This is where service upgrade will be done.

 (8) A callback interest on a fileserver is set up before an FS operation
     is performed and passed through to afs_make_call() so that it can be
     set on the vnode if the operation returns a callback.  The callback
     interest is passed through to afs_iget() also so that it can be set
     there too.

In general, record updating is done on an as-needed basis when we try to
access servers, volumes or vnodes rather than offloading it to work items
and special threads.

Notes:

 (1) Pre AFS-3.4 servers are no longer supported, though this can be added
     back if necessary (AFS-3.4 was released in 1998).

 (2) VBUSY is retried forever for the moment at intervals of 1s.

 (3) /proc/fs/afs/<cell>/servers no longer exists.

Signed-off-by: David Howells <dhowells@redhat.com>

26 files changed, 2795 insertions, 2575 deletions

diff --git a/fs/afs/Makefile b/fs/afs/Makefile
index 192d476d7c76..45b7fc405fa6 100644
--- a/fs/afs/Makefile
+++ b/fs/afs/Makefile
@@ -24,11 +24,10 @@ kafs-objs := \
 	rxrpc.o \
 	security.o \
 	server.o \
+	server_list.o \
 	super.o \
 	netdevices.o \
 	vlclient.o \
-	vlocation.o \
-	vnode.o \
 	volume.o \
 	write.o \
 	xattr.o
diff --git a/fs/afs/addr_list.c b/fs/afs/addr_list.c
index ecb9c72aebd2..b91e59a77f0e 100644
--- a/fs/afs/addr_list.c
+++ b/fs/afs/addr_list.c
@@ -228,6 +228,37 @@ struct afs_addr_list *afs_dns_query(struct afs_cell *cell, time64_t *_expiry)
 }
 
 /*
+ * Merge an IPv4 entry into a fileserver address list.
+ */
+void afs_merge_fs_addr4(struct afs_addr_list *alist, __be32 xdr)
+{
+	struct sockaddr_in6 *a;
+	int i;
+
+	for (i = 0; i < alist->nr_ipv4; i++) {
+		a = &alist->addrs[i].transport.sin6;
+		if (xdr == a->sin6_addr.s6_addr32[3])
+			return;
+		if (xdr < a->sin6_addr.s6_addr32[3])
+			break;
+	}
+
+	if (i < alist->nr_addrs)
+		memmove(alist->addrs + i + 1,
+			alist->addrs + i,
+			sizeof(alist->addrs[0]) * (alist->nr_addrs - i));
+
+	a = &alist->addrs[i].transport.sin6;
+	a->sin6_port		  = htons(AFS_FS_PORT);
+	a->sin6_addr.s6_addr32[0] = 0;
+	a->sin6_addr.s6_addr32[1] = 0;
+	a->sin6_addr.s6_addr32[2] = htonl(0xffff);
+	a->sin6_addr.s6_addr32[3] = xdr;
+	alist->nr_ipv4++;
+	alist->nr_addrs++;
+}
+
+/*
  * Get an address to try.
  */
 bool afs_iterate_addresses(struct afs_addr_cursor *ac)
diff --git a/fs/afs/afs_fs.h b/fs/afs/afs_fs.h
index 05395d0f1941..d47b6d01e4c0 100644
--- a/fs/afs/afs_fs.h
+++ b/fs/afs/afs_fs.h
@@ -38,6 +38,7 @@ enum AFS_FS_Operations {
 	FSFETCHDATA64		= 65537, /* AFS Fetch file data */
 	FSSTOREDATA64		= 65538, /* AFS Store file data */
 	FSGIVEUPALLCALLBACKS	= 65539, /* AFS Give up all outstanding callbacks on a server */
+	FSGETCAPABILITIES	= 65540, /* Probe and get the capabilities of a fileserver */
 };
 
 enum AFS_FS_Errors {
diff --git a/fs/afs/afs_vl.h b/fs/afs/afs_vl.h
index 4eaa620992c8..6350b417aee9 100644
--- a/fs/afs/afs_vl.h
+++ b/fs/afs/afs_vl.h
@@ -88,4 +88,46 @@ struct afs_vldbentry {
 
 #define AFS_VLDB_MAXNAMELEN 65
 
+
+struct afs_ListAddrByAttributes__xdr {
+	__be32			Mask;
+#define AFS_VLADDR_IPADDR	0x1	/* Match by ->ipaddr */
+#define AFS_VLADDR_INDEX	0x2	/* Match by ->index */
+#define AFS_VLADDR_UUID		0x4	/* Match by ->uuid */
+	__be32			ipaddr;
+	__be32			index;
+	__be32			spare;
+	struct afs_uuid__xdr	uuid;
+};
+
+struct afs_uvldbentry__xdr {
+	__be32			name[AFS_VLDB_MAXNAMELEN];
+	__be32			nServers;
+	struct afs_uuid__xdr	serverNumber[AFS_NMAXNSERVERS];
+	__be32			serverUnique[AFS_NMAXNSERVERS];
+	__be32			serverPartition[AFS_NMAXNSERVERS];
+	__be32			serverFlags[AFS_NMAXNSERVERS];
+	__be32			volumeId[AFS_MAXTYPES];
+	__be32			cloneId;
+	__be32			flags;
+	__be32			spares1;
+	__be32			spares2;
+	__be32			spares3;
+	__be32			spares4;
+	__be32			spares5;
+	__be32			spares6;
+	__be32			spares7;
+	__be32			spares8;
+	__be32			spares9;
+};
+
+struct afs_address_list {
+	refcount_t		usage;
+	unsigned int		version;
+	unsigned int		nr_addrs;
+	struct sockaddr_rxrpc	addrs[];
+};
+
+extern void afs_put_address_list(struct afs_address_list *alist);
+
 #endif /* AFS_VL_H */
diff --git a/fs/afs/callback.c b/fs/afs/callback.c
index 82f4c7a3b7b6..f4291b576054 100644
--- a/fs/afs/callback.c
+++ b/fs/afs/callback.c
@@ -26,10 +26,10 @@
  * - Called with volume->server_sem held.
  */
 int afs_register_server_cb_interest(struct afs_vnode *vnode,
-				    struct afs_cb_interest **ppcbi,
-				    struct afs_server *server)
+				    struct afs_server_entry *entry)
 {
-	struct afs_cb_interest *cbi = *ppcbi, *vcbi, *new, *x;
+	struct afs_cb_interest *cbi = entry->cb_interest, *vcbi, *new, *x;
+	struct afs_server *server = entry->server;
 
 again:
 	vcbi = vnode->cb_interest;
@@ -47,7 +47,7 @@ again:
 
 		if (!cbi && vcbi->server == server) {
 			afs_get_cb_interest(vcbi);
-			x = cmpxchg(ppcbi, cbi, vcbi);
+			x = cmpxchg(&entry->cb_interest, cbi, vcbi);
 			if (x != cbi) {
 				cbi = x;
 				afs_put_cb_interest(afs_v2net(vnode), vcbi);
@@ -72,7 +72,7 @@ again:
 		list_add_tail(&new->cb_link, &server->cb_interests);
 		write_unlock(&server->cb_break_lock);
 
-		x = cmpxchg(ppcbi, cbi, new);
+		x = cmpxchg(&entry->cb_interest, cbi, new);
 		if (x == cbi) {
 			cbi = new;
 		} else {
@@ -137,7 +137,7 @@ void afs_put_cb_interest(struct afs_net *net, struct afs_cb_interest *cbi)
  */
 void afs_init_callback_state(struct afs_server *server)
 {
-	if (!test_and_clear_bit(AFS_SERVER_NEW, &server->flags))
+	if (!test_and_clear_bit(AFS_SERVER_FL_NEW, &server->flags))
 		server->cb_s_break++;
 }
 
@@ -233,12 +233,12 @@ void afs_break_callbacks(struct afs_server *server, size_t count,
 /*
  * Clear the callback interests in a server list.
  */
-void afs_clear_callback_interests(struct afs_net *net, struct afs_volume *volume)
+void afs_clear_callback_interests(struct afs_net *net, struct afs_server_list *slist)
 {
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(volume->cb_interests); i++) {
-		afs_put_cb_interest(net, volume->cb_interests[i]);
-		volume->cb_interests[i] = NULL;
+	for (i = 0; i < slist->nr_servers; i++) {
+		afs_put_cb_interest(net, slist->servers[i].cb_interest);
+		slist->servers[i].cb_interest = NULL;
 	}
 }
diff --git a/fs/afs/cell.c b/fs/afs/cell.c
index a0e08d3a108c..1858c91169e4 100644
--- a/fs/afs/cell.c
+++ b/fs/afs/cell.c
@@ -146,13 +146,10 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net,
 
 	atomic_set(&cell->usage, 2);
 	INIT_WORK(&cell->manager, afs_manage_cell);
-	rwlock_init(&cell->servers_lock);
-	INIT_LIST_HEAD(&cell->servers);
-	init_rwsem(&cell->vl_sem);
-	INIT_LIST_HEAD(&cell->vl_list);
-	spin_lock_init(&cell->vl_lock);
 	cell->flags = ((1 << AFS_CELL_FL_NOT_READY) |
 		       (1 << AFS_CELL_FL_NO_LOOKUP_YET));
+	INIT_LIST_HEAD(&cell->proc_volumes);
+	rwlock_init(&cell->proc_lock);
 	rwlock_init(&cell->vl_addrs_lock);
 
 	/* Fill in the VL server list if we were given a list of addresses to
diff --git a/fs/afs/dir.c b/fs/afs/dir.c
index 37083699a0df..53f3917440e7 100644
--- a/fs/afs/dir.c
+++ b/fs/afs/dir.c
@@ -553,7 +553,7 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry,
 	dentry->d_fsdata = (void *)(unsigned long) vnode->status.data_version;
 
 	/* instantiate the dentry */
-	inode = afs_iget(dir->i_sb, key, &fid, NULL, NULL);
+	inode = afs_iget(dir->i_sb, key, &fid, NULL, NULL, NULL);
 	key_put(key);
 	if (IS_ERR(inode)) {
 		_leave(" = %ld", PTR_ERR(inode));
@@ -741,20 +741,48 @@ static void afs_d_release(struct dentry *dentry)
 }
 
 /*
+ * Create a new inode for create/mkdir/symlink
+ */
+static void afs_vnode_new_inode(struct afs_fs_cursor *fc,
+				struct dentry *new_dentry,
+				struct afs_fid *newfid,
+				struct afs_file_status *newstatus,
+				struct afs_callback *newcb)
+{
+	struct inode *inode;
+
+	if (fc->ac.error < 0)
+		return;
+
+	inode = afs_iget(fc->vnode->vfs_inode.i_sb, fc->key,
+			 newfid, newstatus, newcb, fc->cbi);
+	if (IS_ERR(inode)) {
+		/* ENOMEM or EINTR at a really inconvenient time - just abandon
+		 * the new directory on the server.
+		 */
+		fc->ac.error = PTR_ERR(inode);
+		return;
+	}
+
+	d_instantiate(new_dentry, inode);
+	if (d_unhashed(new_dentry))
+		d_rehash(new_dentry);
+}
+
+/*
  * create a directory on an AFS filesystem
  */
 static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 {
-	struct afs_file_status status;
-	struct afs_callback cb;
-	struct afs_server *server;
-	struct afs_vnode *dvnode, *vnode;
-	struct afs_fid fid;
-	struct inode *inode;
+	struct afs_file_status newstatus;
+	struct afs_fs_cursor fc;
+	struct afs_callback newcb;
+	struct afs_vnode *dvnode = AFS_FS_I(dir);
+	struct afs_fid newfid;
 	struct key *key;
 	int ret;
 
-	dvnode = AFS_FS_I(dir);
+	mode |= S_IFDIR;
 
 	_enter("{%x:%u},{%pd},%ho",
 	       dvnode->fid.vid, dvnode->fid.vnode, dentry, mode);
@@ -765,40 +793,27 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 		goto error;
 	}
 
-	mode |= S_IFDIR;
-	ret = afs_vnode_create(dvnode, key, dentry->d_name.name,
-			       mode, &fid, &status, &cb, &server);
-	if (ret < 0)
-		goto mkdir_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			afs_fs_create(&fc, dentry->d_name.name, mode,
+				      &newfid, &newstatus, &newcb);
+		}
 
-	inode = afs_iget(dir->i_sb, key, &fid, &status, &cb);
-	if (IS_ERR(inode)) {
-		/* ENOMEM at a really inconvenient time - just abandon the new
-		 * directory on the server */
-		ret = PTR_ERR(inode);
-		goto iget_error;
+		afs_check_for_remote_deletion(&fc, fc.vnode);
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, &newcb);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret < 0)
+			goto error_key;
 	}
 
-	/* apply the status report we've got for the new vnode */
-	vnode = AFS_FS_I(inode);
-	spin_lock(&vnode->lock);
-	vnode->update_cnt++;
-	spin_unlock(&vnode->lock);
-	afs_vnode_finalise_status_update(vnode, server);
-	afs_put_server(afs_i2net(dir), server);
-
-	d_instantiate(dentry, inode);
-	if (d_unhashed(dentry)) {
-		_debug("not hashed");
-		d_rehash(dentry);
-	}
 	key_put(key);
 	_leave(" = 0");
 	return 0;
 
-iget_error:
-	afs_put_server(afs_i2net(dir), server);
-mkdir_error:
+error_key:
 	key_put(key);
 error:
 	d_drop(dentry);
@@ -807,16 +822,29 @@ error:
 }
 
 /*
+ * Remove a subdir from a directory.
+ */
+static void afs_dir_remove_subdir(struct dentry *dentry)
+{
+	if (d_really_is_positive(dentry)) {
+		struct afs_vnode *vnode = AFS_FS_I(d_inode(dentry));
+
+		clear_nlink(&vnode->vfs_inode);
+		set_bit(AFS_VNODE_DELETED, &vnode->flags);
+		clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags);
+	}
+}
+
+/*
  * remove a directory from an AFS filesystem
  */
 static int afs_rmdir(struct inode *dir, struct dentry *dentry)
 {
-	struct afs_vnode *dvnode, *vnode;
+	struct afs_fs_cursor fc;
+	struct afs_vnode *dvnode = AFS_FS_I(dir);
 	struct key *key;
 	int ret;
 
-	dvnode = AFS_FS_I(dir);
-
 	_enter("{%x:%u},{%pd}",
 	       dvnode->fid.vid, dvnode->fid.vnode, dentry);
 
@@ -826,45 +854,69 @@ static int afs_rmdir(struct inode *dir, struct dentry *dentry)
 		goto error;
 	}
 
-	ret = afs_vnode_remove(dvnode, key, dentry->d_name.name, true);
-	if (ret < 0)
-		goto rmdir_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			afs_fs_remove(&fc, dentry->d_name.name, true);
+		}
 
-	if (d_really_is_positive(dentry)) {
-		vnode = AFS_FS_I(d_inode(dentry));
-		clear_nlink(&vnode->vfs_inode);
-		set_bit(AFS_VNODE_DELETED, &vnode->flags);
-		clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags);
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret == 0)
+			afs_dir_remove_subdir(dentry);
 	}
 
 	key_put(key);
-	_leave(" = 0");
-	return 0;
-
-rmdir_error:
-	key_put(key);
 error:
-	_leave(" = %d", ret);
 	return ret;
 }
 
 /*
- * remove a file from an AFS filesystem
+ * Remove a link to a file or symlink from a directory.
+ *
+ * If the file was not deleted due to excess hard links, the fileserver will
+ * break the callback promise on the file - if it had one - before it returns
+ * to us, and if it was deleted, it won't
+ *
+ * However, if we didn't have a callback promise outstanding, or it was
+ * outstanding on a different server, then it won't break it either...
+ */
+static int afs_dir_remove_link(struct dentry *dentry, struct key *key)
+{
+	int ret = 0;
+
+	if (d_really_is_positive(dentry)) {
+		struct afs_vnode *vnode = AFS_FS_I(d_inode(dentry));
+
+		if (test_bit(AFS_VNODE_DELETED, &vnode->flags))
+			kdebug("AFS_VNODE_DELETED");
+		clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags);
+
+		ret = afs_validate(vnode, key);
+		if (ret == -ESTALE)
+			ret = 0;
+		_debug("nlink %d [val %d]", vnode->vfs_inode.i_nlink, ret);
+	}
+
+	return ret;
+}
+
+/*
+ * Remove a file or symlink from an AFS filesystem.
  */
 static int afs_unlink(struct inode *dir, struct dentry *dentry)
 {
-	struct afs_vnode *dvnode, *vnode;
+	struct afs_fs_cursor fc;
+	struct afs_vnode *dvnode = AFS_FS_I(dir), *vnode;
 	struct key *key;
 	int ret;
 
-	dvnode = AFS_FS_I(dir);
-
 	_enter("{%x:%u},{%pd}",
 	       dvnode->fid.vid, dvnode->fid.vnode, dentry);
 
-	ret = -ENAMETOOLONG;
 	if (dentry->d_name.len >= AFSNAMEMAX)
-		goto error;
+		return -ENAMETOOLONG;
 
 	key = afs_request_key(dvnode->volume->cell);
 	if (IS_ERR(key)) {
@@ -872,42 +924,28 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry)
 		goto error;
 	}
 
+	/* Try to make sure we have a callback promise on the victim. */
 	if (d_really_is_positive(dentry)) {
 		vnode = AFS_FS_I(d_inode(dentry));
-
-		/* make sure we have a callback promise on the victim */
 		ret = afs_validate(vnode, key);
 		if (ret < 0)
-			goto error;
+			goto error_key;
 	}
 
-	ret = afs_vnode_remove(dvnode, key, dentry->d_name.name, false);
-	if (ret < 0)
-		goto remove_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			afs_fs_remove(&fc, dentry->d_name.name, false);
+		}
 
-	if (d_really_is_positive(dentry)) {
-		/* if the file wasn't deleted due to excess hard links, the
-		 * fileserver will break the callback promise on the file - if
-		 * it had one - before it returns to us, and if it was deleted,
-		 * it won't
-		 *
-		 * however, if we didn't have a callback promise outstanding,
-		 * or it was outstanding on a different server, then it won't
-		 * break it either...
-		 */
-		vnode = AFS_FS_I(d_inode(dentry));
-		if (test_bit(AFS_VNODE_DELETED, &vnode->flags))
-			_debug("AFS_VNODE_DELETED");
-		clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags);
-		ret = afs_validate(vnode, key);
-		_debug("nlink %d [val %d]", vnode->vfs_inode.i_nlink, ret);
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret == 0)
+			ret = afs_dir_remove_link(dentry, key);
 	}
 
-	key_put(key);
-	_leave(" = 0");
-	return 0;
-
-remove_error:
+error_key:
 	key_put(key);
 error:
 	_leave(" = %d", ret);
@@ -920,60 +958,50 @@ error:
 static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
 		      bool excl)
 {
-	struct afs_file_status status;
-	struct afs_callback cb;
-	struct afs_server *server;
-	struct afs_vnode *dvnode, *vnode;
-	struct afs_fid fid;
-	struct inode *inode;
+	struct afs_fs_cursor fc;
+	struct afs_file_status newstatus;
+	struct afs_callback newcb;
+	struct afs_vnode *dvnode = dvnode = AFS_FS_I(dir);
+	struct afs_fid newfid;
 	struct key *key;
 	int ret;
 
-	dvnode = AFS_FS_I(dir);
+	mode |= S_IFREG;
 
 	_enter("{%x:%u},{%pd},%ho,",
 	       dvnode->fid.vid, dvnode->fid.vnode, dentry, mode);
 
+	ret = -ENAMETOOLONG;
+	if (dentry->d_name.len >= AFSNAMEMAX)
+		goto error;
+
 	key = afs_request_key(dvnode->volume->cell);
 	if (IS_ERR(key)) {
 		ret = PTR_ERR(key);
 		goto error;
 	}
 
-	mode |= S_IFREG;
-	ret = afs_vnode_create(dvnode, key, dentry->d_name.name,
-			       mode, &fid, &status, &cb, &server);
-	if (ret < 0)
-		goto create_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			afs_fs_create(&fc, dentry->d_name.name, mode,
+				      &newfid, &newstatus, &newcb);
+		}
 
-	inode = afs_iget(dir->i_sb, key, &fid, &status, &cb);
-	if (IS_ERR(inode)) {
-		/* ENOMEM at a really inconvenient time - just abandon the new
-		 * directory on the server */
-		ret = PTR_ERR(inode);
-		goto iget_error;
+		afs_check_for_remote_deletion(&fc, fc.vnode);
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, &newcb);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret < 0)
+			goto error_key;
 	}
 
-	/* apply the status report we've got for the new vnode */
-	vnode = AFS_FS_I(inode);
-	spin_lock(&vnode->lock);
-	vnode->update_cnt++;
-	spin_unlock(&vnode->lock);
-	afs_vnode_finalise_status_update(vnode, server);
-	afs_put_server(afs_i2net(dir), server);
-
-	d_instantiate(dentry, inode);
-	if (d_unhashed(dentry)) {
-		_debug("not hashed");
-		d_rehash(dentry);
-	}
 	key_put(key);
 	_leave(" = 0");
 	return 0;
 
-iget_error:
-	afs_put_server(afs_i2net(dir), server);
-create_error:
+error_key:
 	key_put(key);
 error:
 	d_drop(dentry);
@@ -987,6 +1015,7 @@ error:
 static int afs_link(struct dentry *from, struct inode *dir,
 		    struct dentry *dentry)
 {
+	struct afs_fs_cursor fc;
 	struct afs_vnode *dvnode, *vnode;
 	struct key *key;
 	int ret;
@@ -999,23 +1028,45 @@ static int afs_link(struct dentry *from, struct inode *dir,
 	       dvnode->fid.vid, dvnode->fid.vnode,
 	       dentry);
 
+	ret = -ENAMETOOLONG;
+	if (dentry->d_name.len >= AFSNAMEMAX)
+		goto error;
+
 	key = afs_request_key(dvnode->volume->cell);
 	if (IS_ERR(key)) {
 		ret = PTR_ERR(key);
 		goto error;
 	}
 
-	ret = afs_vnode_link(dvnode, vnode, key, dentry->d_name.name);
-	if (ret < 0)
-		goto link_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		if (mutex_lock_interruptible_nested(&vnode->io_lock, 1) < 0) {
+			afs_end_vnode_operation(&fc);
+			return -ERESTARTSYS;
+		}
+
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			fc.cb_break_2 = vnode->cb_break + vnode->cb_s_break;
+			afs_fs_link(&fc, vnode, dentry->d_name.name);
+		}
+
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		afs_vnode_commit_status(&fc, vnode, fc.cb_break_2);
+		ihold(&vnode->vfs_inode);
+		d_instantiate(dentry, &vnode->vfs_inode);
+
+		mutex_unlock(&vnode->io_lock);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret < 0)
+			goto error_key;
+	}
 
-	ihold(&vnode->vfs_inode);
-	d_instantiate(dentry, &vnode->vfs_inode);
 	key_put(key);
 	_leave(" = 0");
 	return 0;
 
-link_error:
+error_key:
 	key_put(key);
 error:
 	d_drop(dentry);
@@ -1029,20 +1080,21 @@ error:
 static int afs_symlink(struct inode *dir, struct dentry *dentry,
 		       const char *content)
 {
-	struct afs_file_status status;
-	struct afs_server *server;
-	struct afs_vnode *dvnode, *vnode;
-	struct afs_fid fid;
-	struct inode *inode;
+	struct afs_fs_cursor fc;
+	struct afs_file_status newstatus;
+	struct afs_vnode *dvnode = AFS_FS_I(dir);
+	struct afs_fid newfid;
 	struct key *key;
 	int ret;
 
-	dvnode = AFS_FS_I(dir);
-
 	_enter("{%x:%u},{%pd},%s",
 	       dvnode->fid.vid, dvnode->fid.vnode, dentry,
 	       content);
 
+	ret = -ENAMETOOLONG;
+	if (dentry->d_name.len >= AFSNAMEMAX)
+		goto error;
+
 	ret = -EINVAL;
 	if (strlen(content) >= AFSPATHMAX)
 		goto error;
@@ -1053,39 +1105,27 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry,
 		goto error;
 	}
 
-	ret = afs_vnode_symlink(dvnode, key, dentry->d_name.name, content,
-				&fid, &status, &server);
-	if (ret < 0)
-		goto create_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, dvnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = dvnode->cb_break + dvnode->cb_s_break;
+			afs_fs_symlink(&fc, dentry->d_name.name, content,
+				       &newfid, &newstatus);
+		}
 
-	inode = afs_iget(dir->i_sb, key, &fid, &status, NULL);
-	if (IS_ERR(inode)) {
-		/* ENOMEM at a really inconvenient time - just abandon the new
-		 * directory on the server */
-		ret = PTR_ERR(inode);
-		goto iget_error;
+		afs_check_for_remote_deletion(&fc, fc.vnode);
+		afs_vnode_commit_status(&fc, dvnode, fc.cb_break);
+		afs_vnode_new_inode(&fc, dentry, &newfid, &newstatus, NULL);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret < 0)
+			goto error_key;
 	}
 
-	/* apply the status report we've got for the new vnode */
-	vnode = AFS_FS_I(inode);
-	spin_lock(&vnode->lock);
-	vnode->update_cnt++;
-	spin_unlock(&vnode->lock);
-	afs_vnode_finalise_status_update(vnode, server);
-	afs_put_server(afs_i2net(dir), server);
-
-	d_instantiate(dentry, inode);
-	if (d_unhashed(dentry)) {
-		_debug("not hashed");
-		d_rehash(dentry);
-	}
 	key_put(key);
 	_leave(" = 0");
 	return 0;
 
-iget_error:
-	afs_put_server(afs_i2net(dir), server);
-create_error:
+error_key:
 	key_put(key);
 error:
 	d_drop(dentry);
@@ -1100,6 +1140,7 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		      struct inode *new_dir, struct dentry *new_dentry,
 		      unsigned int flags)
 {
+	struct afs_fs_cursor fc;
 	struct afs_vnode *orig_dvnode, *new_dvnode, *vnode;
 	struct key *key;
 	int ret;
@@ -1123,16 +1164,35 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		goto error;
 	}
 
-	ret = afs_vnode_rename(orig_dvnode, new_dvnode, key,
-			       old_dentry->d_name.name,
-			       new_dentry->d_name.name);
-	if (ret < 0)
-		goto rename_error;
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, orig_dvnode, key)) {
+		if (orig_dvnode != new_dvnode) {
+			if (mutex_lock_interruptible_nested(&new_dvnode->io_lock, 1) < 0) {
+				afs_end_vnode_operation(&fc);
+				return -ERESTARTSYS;
+			}
+		}
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = orig_dvnode->cb_break + orig_dvnode->cb_s_break;
+			fc.cb_break_2 = new_dvnode->cb_break + new_dvnode->cb_s_break;
+			afs_fs_rename(&fc, old_dentry->d_name.name,
+				      new_dvnode, new_dentry->d_name.name);
+		}
+
+		afs_vnode_commit_status(&fc, orig_dvnode, fc.cb_break);
+		afs_vnode_commit_status(&fc, new_dvnode, fc.cb_break_2);
+		if (orig_dvnode != new_dvnode)
+			mutex_unlock(&new_dvnode->io_lock);
+		ret = afs_end_vnode_operation(&fc);
+		if (ret < 0)
+			goto error_key;
+	}
+
 	key_put(key);
 	_leave(" = 0");
 	return 0;
 
-rename_error:
+error_key:
 	key_put(key);
 error:
 	d_drop(new_dentry);
diff --git a/fs/afs/file.c b/fs/afs/file.c
index 08f9f0c5dfac..1f26ac9f816d 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -138,6 +138,37 @@ static void afs_file_readpage_read_complete(struct page *page,
 #endif
 
 /*
+ * Fetch file data from the volume.
+ */
+int afs_fetch_data(struct afs_vnode *vnode, struct key *key, struct afs_read *desc)
+{
+	struct afs_fs_cursor fc;
+	int ret;
+
+	_enter("%s{%x:%u.%u},%x,,,",
+	       vnode->volume->name,
+	       vnode->fid.vid,
+	       vnode->fid.vnode,
+	       vnode->fid.unique,
+	       key_serial(key));
+
+	ret = -ERESTARTSYS;
+	if (afs_begin_vnode_operation(&fc, vnode, key)) {
+		while (afs_select_fileserver(&fc)) {
+			fc.cb_break = vnode->cb_break + vnode->cb_s_break;
+			afs_fs_fetch_data(&fc, desc);
+		}
+
+		afs_check_for_remote_deletion(&fc, fc.vnode);
+		afs_vnode_commit_status(&fc, vnode, fc.cb_break);
+		ret = afs_end_vnode_operation(&fc);
+	}
+
+	_leave(" = %d", ret);
+	return ret;
+}
+
+/*
  * read page from file, directory or symlink, given a key to use
  */
 int afs_page_filler(void *data, struct page *page)
@@ -199,7 +230,7 @@ int afs_page_filler(void *data, struct page *page)
 
 		/* read the contents of the file from the server into the
 		 * page */
-		ret = afs_vnode_fetch_data(vnode, key, req);
+		ret = afs_fetch_data(vnode, key, req);
 		afs_put_read(req);
 		if (ret < 0) {
 			if (ret == -ENOENT) {
@@ -264,7 +295,7 @@ static int afs_readpage(struct file *file, struct page *page)
 		ret = afs_page_filler(key, page);
 	} else {
 		struct inode *inode = page->mapping->host;
-		key = afs_request_key(AFS_FS_S(inode->i_sb)->volume->cell);
+		key = afs_request_key(AFS_FS_S(inode->i_sb)->cell);
 		if (IS_ERR(key)) {
 			ret = PTR_ERR(key);
 		} else {
@@ -369,7 +400,7 @@ static int afs_readpages_one(struct file *file, struct address_space *mapping,
 		return 0;
 	}
 
-	ret = afs_vnode_fetch_data(vnode, key, req);
+	ret = afs_fetch_data(vnode, key, req);
 	if (ret < 0)
 		goto error;
 
diff --git a/fs/afs/flock.c b/fs/afs/flock.c
index 2b31ea58c50c..77b0a4606efd 100644
--- a/fs/afs/flock.c
+++ b/fs/afs/flock.c
@@ -68,6 +68,100 @@ static void afs_grant_locks(struct afs_vnode *vnode, struct file_lock *fl)
 }
 
 /*
+ * Get a lock on a file
+ */
+static int afs_set_lock(struct afs_vnode *vnode, struct key *key,