39 files changed, 4011 insertions, 5240 deletions

diff --git a/fs/Kconfig b/fs/Kconfig
index 3c4886b849f..075c9997ddc 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -2019,7 +2019,7 @@ config CODA_FS_OLD_API
 config AFS_FS
 	tristate "Andrew File System support (AFS) (EXPERIMENTAL)"
 	depends on INET && EXPERIMENTAL
-	select RXRPC
+	select AF_RXRPC
 	help
 	  If you say Y here, you will get an experimental Andrew File System
 	  driver. It currently only supports unsecured read-only AFS access.
@@ -2028,6 +2028,17 @@ config AFS_FS
 
 	  If unsure, say N.
 
+config AFS_DEBUG
+	bool "AFS dynamic debugging"
+	depends on AFS_FS
+	help
+	  Say Y here to make runtime controllable debugging messages appear.
+
+	  See <file:Documentation/filesystems/afs.txt> for more information.
+
+	  If unsure, say N.
+
+
 config RXRPC
 	tristate
 
diff --git a/fs/afs/Makefile b/fs/afs/Makefile
index 8e719737967..66bdc219ccd 100644
--- a/fs/afs/Makefile
+++ b/fs/afs/Makefile
@@ -10,12 +10,11 @@ kafs-objs := \
 	file.o \
 	fsclient.o \
 	inode.o \
-	kafsasyncd.o \
-	kafstimod.o \
 	main.o \
 	misc.o \
 	mntpt.o \
 	proc.o \
+	rxrpc.o \
 	server.o \
 	super.o \
 	vlclient.o \
diff --git a/fs/afs/types.h b/fs/afs/afs.h
index db2b5dc9ff4..b9d2d2ceaf4 100644
--- a/fs/afs/types.h
+++ b/fs/afs/afs.h
@@ -1,6 +1,6 @@
-/* AFS types
+/* AFS common types
  *
- * Copyright (C) 2002 Red Hat, Inc. All Rights Reserved.
+ * Copyright (C) 2002, 2007 Red Hat, Inc. All Rights Reserved.
  * Written by David Howells (dhowells@redhat.com)
  *
  * This program is free software; you can redistribute it and/or
@@ -9,10 +9,10 @@
  * 2 of the License, or (at your option) any later version.
  */
 
-#ifndef AFS_TYPES_H
-#define AFS_TYPES_H
+#ifndef AFS_H
+#define AFS_H
 
-#include <rxrpc/types.h>
+#include <linux/in.h>
 
 typedef unsigned			afs_volid_t;
 typedef unsigned			afs_vnodeid_t;
@@ -31,9 +31,6 @@ typedef enum {
 	AFS_FTYPE_SYMLINK	= 3,
 } afs_file_type_t;
 
-struct afs_cell;
-struct afs_vnode;
-
 /*
  * AFS file identifier
  */
@@ -54,14 +51,13 @@ typedef enum {
 } afs_callback_type_t;
 
 struct afs_callback {
-	struct afs_server	*server;	/* server that made the promise */
 	struct afs_fid		fid;		/* file identifier */
 	unsigned		version;	/* callback version */
 	unsigned		expiry;		/* time at which expires */
 	afs_callback_type_t	type;		/* type of callback */
 };
 
-#define AFSCBMAX 50
+#define AFSCBMAX 50	/* maximum callbacks transferred per bulk op */
 
 /*
  * AFS volume information
@@ -70,7 +66,7 @@ struct afs_volume_info {
 	afs_volid_t		vid;		/* volume ID */
 	afs_voltype_t		type;		/* type of this volume */
 	afs_volid_t		type_vids[5];	/* volume ID's for possible types for this vol */
-	
+
 	/* list of fileservers serving this volume */
 	size_t			nservers;	/* number of entries used in servers[] */
 	struct {
@@ -88,7 +84,7 @@ struct afs_file_status {
 	afs_file_type_t		type;		/* file type */
 	unsigned		nlink;		/* link count */
 	size_t			size;		/* file size */
-	afs_dataversion_t	version;	/* current data version */
+	afs_dataversion_t	data_version;	/* current data version */
 	unsigned		author;		/* author ID */
 	unsigned		owner;		/* owner ID */
 	unsigned		caller_access;	/* access rights for authenticated caller */
@@ -106,4 +102,4 @@ struct afs_volsync {
 	time_t			creation;	/* volume creation time */
 };
 
-#endif /* AFS_TYPES_H */
+#endif /* AFS_H */
diff --git a/fs/afs/afs_cm.h b/fs/afs/afs_cm.h
new file mode 100644
index 00000000000..7c8e3d43c8e
--- /dev/null
+++ b/fs/afs/afs_cm.h
@@ -0,0 +1,28 @@
+/* AFS Cache Manager definitions
+ *
+ * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#ifndef AFS_CM_H
+#define AFS_CM_H
+
+#define AFS_CM_PORT		7001	/* AFS file server port */
+#define CM_SERVICE		1	/* AFS File Service ID */
+
+enum AFS_CM_Operations {
+	CBCallBack		= 204,	/* break callback promises */
+	CBInitCallBackState	= 205,	/* initialise callback state */
+	CBProbe			= 206,	/* probe client */
+	CBGetLock		= 207,	/* get contents of CM lock table */
+	CBGetCE			= 208,	/* get cache file description */
+	CBGetXStatsVersion	= 209,	/* get version of extended statistics */
+	CBGetXStats		= 210,	/* get contents of extended statistics data */
+};
+
+#endif /* AFS_FS_H */
diff --git a/fs/afs/errors.h b/fs/afs/afs_fs.h
index bcc0a3309e7..fd385954f21 100644
--- a/fs/afs/errors.h
+++ b/fs/afs/afs_fs.h
@@ -1,6 +1,6 @@
-/* AFS abort/error codes
+/* AFS File Service definitions
  *
- * Copyright (C) 2002 Red Hat, Inc. All Rights Reserved.
+ * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
  * Written by David Howells (dhowells@redhat.com)
  *
  * This program is free software; you can redistribute it and/or
@@ -9,15 +9,22 @@
  * 2 of the License, or (at your option) any later version.
  */
 
-#ifndef AFS_ERRORS_H
-#define AFS_ERRORS_H
+#ifndef AFS_FS_H
+#define AFS_FS_H
 
-#include "types.h"
+#define AFS_FS_PORT		7000	/* AFS file server port */
+#define FS_SERVICE		1	/* AFS File Service ID */
 
-/*
- * file server abort codes
- */
-typedef enum {
+enum AFS_FS_Operations {
+	FSFETCHSTATUS		= 132,	/* AFS Fetch file status */
+	FSFETCHDATA		= 130,	/* AFS Fetch file data */
+	FSGIVEUPCALLBACKS	= 147,	/* AFS Discard callback promises */
+	FSGETVOLUMEINFO		= 148,	/* AFS Get root volume information */
+	FSGETROOTVOLUME		= 151,	/* AFS Get root volume name */
+	FSLOOKUP		= 161	/* AFS lookup file in directory */
+};
+
+enum AFS_FS_Errors {
 	VSALVAGE	= 101,	/* volume needs salvaging */
 	VNOVNODE	= 102,	/* no such file/dir (vnode) */
 	VNOVOL		= 103,	/* no such volume or volume unavailable */
@@ -29,8 +36,6 @@ typedef enum {
 	VOVERQUOTA	= 109,	/* volume's maximum quota exceeded */
 	VBUSY		= 110,	/* volume is temporarily unavailable */
 	VMOVED		= 111,	/* volume moved to new server - ask this FS where */
-} afs_rxfs_abort_t;
-
-extern int afs_abort_to_error(int);
+};
 
-#endif /* AFS_ERRORS_H */
+#endif /* AFS_FS_H */
diff --git a/fs/afs/vlclient.h b/fs/afs/afs_vl.h
index 11dc10fe300..8bbefe009ed 100644
--- a/fs/afs/vlclient.h
+++ b/fs/afs/afs_vl.h
@@ -1,6 +1,6 @@
-/* Volume Location Service client interface
+/* AFS Volume Location Service client interface
  *
- * Copyright (C) 2002 Red Hat, Inc. All Rights Reserved.
+ * Copyright (C) 2002, 2007 Red Hat, Inc. All Rights Reserved.
  * Written by David Howells (dhowells@redhat.com)
  *
  * This program is free software; you can redistribute it and/or
@@ -9,10 +9,19 @@
  * 2 of the License, or (at your option) any later version.
  */
 
-#ifndef AFS_VLCLIENT_H
-#define AFS_VLCLIENT_H
+#ifndef AFS_VL_H
+#define AFS_VL_H
 
-#include "types.h"
+#include "afs.h"
+
+#define AFS_VL_PORT		7003	/* volume location service port */
+#define VL_SERVICE		52	/* RxRPC service ID for the Volume Location service */
+
+enum AFSVL_Operations {
+	VLGETENTRYBYID		= 503,	/* AFS Get Cache Entry By ID operation ID */
+	VLGETENTRYBYNAME	= 504,	/* AFS Get Cache Entry By Name operation ID */
+	VLPROBE			= 514,	/* AFS Probe Volume Location Service operation ID */
+};
 
 enum AFSVL_Errors {
 	AFSVL_IDEXIST 		= 363520,	/* Volume Id entry exists in vl database */
@@ -40,14 +49,16 @@ enum AFSVL_Errors {
 	AFSVL_BADVOLOPER 	= 363542,	/* Bad volume operation code */
 	AFSVL_BADRELLOCKTYPE 	= 363543,	/* Bad release lock type */
 	AFSVL_RERELEASE 	= 363544,	/* Status report: last release was aborted */
-	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server °ag */
+	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server °ag */
 	AFSVL_PERM 		= 363546,	/* No permission access */
 	AFSVL_NOMEM 		= 363547,	/* malloc/realloc failed to alloc enough memory */
 };
 
-/* maps to "struct vldbentry" in vvl-spec.pdf */
+/*
+ * maps to "struct vldbentry" in vvl-spec.pdf
+ */
 struct afs_vldbentry {
-	char		name[65];		/* name of volume (including NUL char) */
+	char		name[65];		/* name of volume (with NUL char) */
 	afs_voltype_t	type;			/* volume type */
 	unsigned	num_servers;		/* num servers that hold instances of this vol */
 	unsigned	clone_id;		/* cloning ID */
@@ -70,16 +81,4 @@ struct afs_vldbentry {
 	} servers[8];
 };
 
-extern int afs_rxvl_get_entry_by_name(struct afs_server *, const char *,
-				      unsigned, struct afs_cache_vlocation *);
-extern int afs_rxvl_get_entry_by_id(struct afs_server *, afs_volid_t,
-				    afs_voltype_t,
-				    struct afs_cache_vlocation *);
-
-extern int afs_rxvl_get_entry_by_id_async(struct afs_async_op *,
-					  afs_volid_t, afs_voltype_t);
-
-extern int afs_rxvl_get_entry_by_id_async2(struct afs_async_op *,
-					   struct afs_cache_vlocation *);
-
-#endif /* AFS_VLCLIENT_H */
+#endif /* AFS_VL_H */
diff --git a/fs/afs/cache.c b/fs/afs/cache.c
new file mode 100644
index 00000000000..de0d7de69ed
--- /dev/null
+++ b/fs/afs/cache.c
@@ -0,0 +1,256 @@
+/* AFS caching stuff
+ *
+ * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_cell_cache_match(void *target,
+						const void *entry);
+static void afs_cell_cache_update(void *source, void *entry);
+
+struct cachefs_index_def afs_cache_cell_index_def = {
+	.name			= "cell_ix",
+	.data_size		= sizeof(struct afs_cache_cell),
+	.keys[0]		= { CACHEFS_INDEX_KEYS_ASCIIZ, 64 },
+	.match			= afs_cell_cache_match,
+	.update			= afs_cell_cache_update,
+};
+#endif
+
+/*
+ * match a cell record obtained from the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_cell_cache_match(void *target,
+						const void *entry)
+{
+	const struct afs_cache_cell *ccell = entry;
+	struct afs_cell *cell = target;
+
+	_enter("{%s},{%s}", ccell->name, cell->name);
+
+	if (strncmp(ccell->name, cell->name, sizeof(ccell->name)) == 0) {
+		_leave(" = SUCCESS");
+		return CACHEFS_MATCH_SUCCESS;
+	}
+
+	_leave(" = FAILED");
+	return CACHEFS_MATCH_FAILED;
+}
+#endif
+
+/*
+ * update a cell record in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static void afs_cell_cache_update(void *source, void *entry)
+{
+	struct afs_cache_cell *ccell = entry;
+	struct afs_cell *cell = source;
+
+	_enter("%p,%p", source, entry);
+
+	strncpy(ccell->name, cell->name, sizeof(ccell->name));
+
+	memcpy(ccell->vl_servers,
+	       cell->vl_addrs,
+	       min(sizeof(ccell->vl_servers), sizeof(cell->vl_addrs)));
+
+}
+#endif
+
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_vlocation_cache_match(void *target,
+						     const void *entry);
+static void afs_vlocation_cache_update(void *source, void *entry);
+
+struct cachefs_index_def afs_vlocation_cache_index_def = {
+	.name		= "vldb",
+	.data_size	= sizeof(struct afs_cache_vlocation),
+	.keys[0]	= { CACHEFS_INDEX_KEYS_ASCIIZ, 64 },
+	.match		= afs_vlocation_cache_match,
+	.update		= afs_vlocation_cache_update,
+};
+#endif
+
+/*
+ * match a VLDB record stored in the cache
+ * - may also load target from entry
+ */
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_vlocation_cache_match(void *target,
+						     const void *entry)
+{
+	const struct afs_cache_vlocation *vldb = entry;
+	struct afs_vlocation *vlocation = target;
+
+	_enter("{%s},{%s}", vlocation->vldb.name, vldb->name);
+
+	if (strncmp(vlocation->vldb.name, vldb->name, sizeof(vldb->name)) == 0
+	    ) {
+		if (!vlocation->valid ||
+		    vlocation->vldb.rtime == vldb->rtime
+		    ) {
+			vlocation->vldb = *vldb;
+			vlocation->valid = 1;
+			_leave(" = SUCCESS [c->m]");
+			return CACHEFS_MATCH_SUCCESS;
+		} else if (memcmp(&vlocation->vldb, vldb, sizeof(*vldb)) != 0) {
+			/* delete if VIDs for this name differ */
+			if (memcmp(&vlocation->vldb.vid,
+				   &vldb->vid,
+				   sizeof(vldb->vid)) != 0) {
+				_leave(" = DELETE");
+				return CACHEFS_MATCH_SUCCESS_DELETE;
+			}
+
+			_leave(" = UPDATE");
+			return CACHEFS_MATCH_SUCCESS_UPDATE;
+		} else {
+			_leave(" = SUCCESS");
+			return CACHEFS_MATCH_SUCCESS;
+		}
+	}
+
+	_leave(" = FAILED");
+	return CACHEFS_MATCH_FAILED;
+}
+#endif
+
+/*
+ * update a VLDB record stored in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static void afs_vlocation_cache_update(void *source, void *entry)
+{
+	struct afs_cache_vlocation *vldb = entry;
+	struct afs_vlocation *vlocation = source;
+
+	_enter("");
+
+	*vldb = vlocation->vldb;
+}
+#endif
+
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_volume_cache_match(void *target,
+						  const void *entry);
+static void afs_volume_cache_update(void *source, void *entry);
+
+struct cachefs_index_def afs_volume_cache_index_def = {
+	.name		= "volume",
+	.data_size	= sizeof(struct afs_cache_vhash),
+	.keys[0]	= { CACHEFS_INDEX_KEYS_BIN, 1 },
+	.keys[1]	= { CACHEFS_INDEX_KEYS_BIN, 1 },
+	.match		= afs_volume_cache_match,
+	.update		= afs_volume_cache_update,
+};
+#endif
+
+/*
+ * match a volume hash record stored in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_volume_cache_match(void *target,
+						  const void *entry)
+{
+	const struct afs_cache_vhash *vhash = entry;
+	struct afs_volume *volume = target;
+
+	_enter("{%u},{%u}", volume->type, vhash->vtype);
+
+	if (volume->type == vhash->vtype) {
+		_leave(" = SUCCESS");
+		return CACHEFS_MATCH_SUCCESS;
+	}
+
+	_leave(" = FAILED");
+	return CACHEFS_MATCH_FAILED;
+}
+#endif
+
+/*
+ * update a volume hash record stored in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static void afs_volume_cache_update(void *source, void *entry)
+{
+	struct afs_cache_vhash *vhash = entry;
+	struct afs_volume *volume = source;
+
+	_enter("");
+
+	vhash->vtype = volume->type;
+}
+#endif
+
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_vnode_cache_match(void *target,
+						 const void *entry);
+static void afs_vnode_cache_update(void *source, void *entry);
+
+struct cachefs_index_def afs_vnode_cache_index_def = {
+	.name		= "vnode",
+	.data_size	= sizeof(struct afs_cache_vnode),
+	.keys[0]	= { CACHEFS_INDEX_KEYS_BIN, 4 },
+	.match		= afs_vnode_cache_match,
+	.update		= afs_vnode_cache_update,
+};
+#endif
+
+/*
+ * match a vnode record stored in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static cachefs_match_val_t afs_vnode_cache_match(void *target,
+						 const void *entry)
+{
+	const struct afs_cache_vnode *cvnode = entry;
+	struct afs_vnode *vnode = target;
+
+	_enter("{%x,%x,%Lx},{%x,%x,%Lx}",
+	       vnode->fid.vnode,
+	       vnode->fid.unique,
+	       vnode->status.version,
+	       cvnode->vnode_id,
+	       cvnode->vnode_unique,
+	       cvnode->data_version);
+
+	if (vnode->fid.vnode != cvnode->vnode_id) {
+		_leave(" = FAILED");
+		return CACHEFS_MATCH_FAILED;
+	}
+
+	if (vnode->fid.unique != cvnode->vnode_unique ||
+	    vnode->status.version != cvnode->data_version) {
+		_leave(" = DELETE");
+		return CACHEFS_MATCH_SUCCESS_DELETE;
+	}
+
+	_leave(" = SUCCESS");
+	return CACHEFS_MATCH_SUCCESS;
+}
+#endif
+
+/*
+ * update a vnode record stored in the cache
+ */
+#ifdef AFS_CACHING_SUPPORT
+static void afs_vnode_cache_update(void *source, void *entry)
+{
+	struct afs_cache_vnode *cvnode = entry;
+	struct afs_vnode *vnode = source;
+
+	_enter("");
+
+	cvnode->vnode_id	= vnode->fid.vnode;
+	cvnode->vnode_unique	= vnode->fid.unique;
+	cvnode->data_version	= vnode->status.version;
+}
+#endif
diff --git a/fs/afs/callback.c b/fs/afs/callback.c
index 26a48fea42f..61121554714 100644
--- a/fs/afs/callback.c
+++ b/fs/afs/callback.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2002 Red Hat, Inc. All rights reserved.
+ * Copyright (c) 2002, 2007 Red Hat, Inc. All rights reserved.
  *
  * This software may be freely redistributed under the terms of the
  * GNU General Public License.
@@ -16,83 +16,182 @@
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/init.h>
-#include "server.h"
-#include "vnode.h"
+#include <linux/circ_buf.h>
 #include "internal.h"
-#include "cmservice.h"
+
+unsigned afs_vnode_update_timeout = 10;
+
+#define afs_breakring_space(server) \
+	CIRC_SPACE((server)->cb_break_head, (server)->cb_break_tail,	\
+		   ARRAY_SIZE((server)->cb_break))
+
+//static void afs_callback_updater(struct work_struct *);
+
+static struct workqueue_struct *afs_callback_update_worker;
 
 /*
  * allow the fileserver to request callback state (re-)initialisation
  */
-int SRXAFSCM_InitCallBackState(struct afs_server *server)
+void afs_init_callback_state(struct afs_server *server)
 {
-	struct list_head callbacks;
+	struct afs_vnode *vnode;
 
-	_enter("%p", server);
+	_enter("{%p}", server);
 
-	INIT_LIST_HEAD(&callbacks);
-
-	/* transfer the callback list from the server to a temp holding area */
 	spin_lock(&server->cb_lock);
 
-	list_add(&callbacks, &server->cb_promises);
-	list_del_init(&server->cb_promises);
+	/* kill all the promises on record from this server */
+	while (!RB_EMPTY_ROOT(&server->cb_promises)) {
+		vnode = rb_entry(server->cb_promises.rb_node,
+				 struct afs_vnode, cb_promise);
+		printk("\nUNPROMISE on %p\n", vnode);
+		rb_erase(&vnode->cb_promise, &server->cb_promises);
+		vnode->cb_promised = false;
+	}
 
-	/* munch our way through the list, grabbing the inode, dropping all the
-	 * locks and regetting them in the right order
-	 */
-	while (!list_empty(&callbacks)) {
-		struct afs_vnode *vnode;
-		struct inode *inode;
+	spin_unlock(&server->cb_lock);
+	_leave("");
+}
 
-		vnode = list_entry(callbacks.next, struct afs_vnode, cb_link);
-		list_del_init(&vnode->cb_link);
+/*
+ * handle the data invalidation side of a callback being broken
+ */
+void afs_broken_callback_work(struct work_struct *work)
+{
+	struct afs_vnode *vnode =
+		container_of(work, struct afs_vnode, cb_broken_work);
 
-		/* try and grab the inode - may fail */
-		inode = igrab(AFS_VNODE_TO_I(vnode));
-		if (inode) {
-			int release = 0;
+	_enter("");
 
-			spin_unlock(&server->cb_lock);
-			spin_lock(&vnode->lock);
+	if (test_bit(AFS_VNODE_DELETED, &vnode->flags))
+		return;
 
-			if (vnode->cb_server == server) {
-				vnode->cb_server = NULL;
-				afs_kafstimod_del_timer(&vnode->cb_timeout);
-				spin_lock(&afs_cb_hash_lock);
-				list_del_init(&vnode->cb_hash_link);
-				spin_unlock(&afs_cb_hash_lock);
-				release = 1;
-			}
+	/* we're only interested in dealing with a broken callback on *this*
+	 * vnode and only if no-one else has dealt with it yet */
+	if (!mutex_trylock(&vnode->cb_broken_lock))
+		return; /* someone else is dealing with it */
 
-			spin_unlock(&vnode->lock);
+	if (test_bit(AFS_VNODE_CB_BROKEN, &vnode->flags)) {
+		if (afs_vnode_fetch_status(vnode) < 0)
+			goto out;
 
-			iput(inode);
-			afs_put_server(server);
+		if (test_bit(AFS_VNODE_DELETED, &vnode->flags))
+			goto out;
 
-			spin_lock(&server->cb_lock);
+		/* if the vnode's data version number changed then its contents
+		 * are different */
+		if (test_and_clear_bit(AFS_VNODE_ZAP_DATA, &vnode->flags)) {
+			_debug("zap data");
+			invalidate_remote_inode(&vnode->vfs_inode);
 		}
 	}
 
-	spin_unlock(&server->cb_lock);
+out:
+	mutex_unlock(&vnode->cb_broken_lock);
 
-	_leave(" = 0");
-	return 0;
+	/* avoid the potential race whereby the mutex_trylock() in this
+	 * function happens again between the clear_bit() and the
+	 * mutex_unlock() */
+	if (test_bit(AFS_VNODE_CB_BROKEN, &vnode->flags)) {
+		_debug("requeue");
+		queue_work(afs_callback_update_worker, &vnode->cb_broken_work);
+	}
+	_leave("");
+}
+
+/*
+ * actually break a callback
+ */
+static void afs_break_callback(struct afs_server *server,
+			       struct afs_vnode *vnode)
+{
+	_enter("");
+
+	set_bit(AFS_VNODE_CB_BROKEN, &vnode->flags);
+
+	if (vnode->cb_promised) {
+		spin_lock(&vnode->lock);
+
+		_debug("break callback");
+
+		spin_lock(&server->cb_lock);
+		if (vnode->cb_promised) {
+			rb_erase(&vnode->cb_promise, &server->cb_promises);
+			vnode->cb_promised = false;
+		}
+		spin_unlock(&server->cb_lock);
+
+		queue_work(afs_callback_update_worker, &vnode->cb_broken_work);
+		spin_unlock(&vnode->lock);
+	}
+}
+
+/*
+ * allow the fileserver to explicitly break one callback
+ * - happens when
+ *   - the backing file is changed
+ *   - a lock is released
+ */
+static void afs_break_one_callback(struct afs_server *server,
+				   struct afs_fid *fid)
+{
+	struct afs_vnode *vnode;
+	struct rb_node *p;
+
+	_debug("find");
+	spin_lock(&server->fs_lock);
+	p = server->fs_vnodes.rb_node;
+	while (p) {
+		vnode = rb_entry(p, struct afs_vnode, server_rb);
+		if (fid->vid < vnode->fid.vid)
+			p = p->rb_left;
+		else if (fid->vid > vnode->fid.vid)
+			p = p->rb_right;
+		else if (fid->vnode < vnode->fid.vnode)
+			p = p->rb_left;
+		else if (fid->vnode > vnode->fid.vnode)
+			p = p->rb_right;
+		else if (fid->unique < vnode->fid.unique)
+			p = p->rb_left;
+		else if (fid->unique > vnode->fid.unique)
+			p = p->rb_right;
+		else
+			goto found;
+	}
+
+	/* not found so we just ignore it (it may have moved to another
+	 * server) */
+not_available:
+	_debug("not avail");
+	spin_unlock(&server->fs_lock);
+	_leave("");
+	return;
+
+found:
+	_debug("found");
+	ASSERTCMP(server, ==, vnode->server);
+
+	if (!igrab(AFS_VNODE_TO_I(vnode)))
+		goto not_available;
+	spin_unlock(&server->fs_lock);
+
+	afs_break_callback(server, vnode);
+	iput(&vnode->vfs_inode);
+	_leave("");
 }
 
 /*
  * allow the fileserver to break callback promises
  */
-int SRXAFSCM_CallBack(struct afs_server *server, size_t count,
-		      struct afs_callback callbacks[])
+void afs_break_callbacks(struct afs_server *server, size_t count,
+			 struct afs_callback callbacks[])
 {
-	_enter("%p,%u,", server, count);
+	_enter("%p,%zu,", server, count);
 
-	for (; count > 0; callbacks++, count--) {
-		struct afs_vnode *vnode = NULL;
-		struct inode *inode = NULL;
-		int valid = 0;
+	ASSERT(server != NULL);
+	ASSERTCMP(count, <=, AFSCBMAX);
 
+	for (; count > 0; callbacks++, count--) {
 		_debug("- Fid { vl=%08x n=%u u=%u }  CB { v=%u x=%u t=%u }",
 		       callbacks->fid.vid,
 		       callbacks->fid.vnode,
@@ -101,66 +200,244 @@ int SRXAFSCM_CallBack(struct afs_server *server, size_t count,
 		       callbacks->expiry,
 		       callbacks->type
 		       );
+		afs_break_one_callback(server, &callbacks->fid);
+	}
+
+	_leave("");
+	return;
+}
 
-		/* find the inode for this fid */
-		spin_lock(&afs_cb_hash_lock);
+/*
+ * record the callback for breaking
+ * - the caller must hold server->cb_lock
+ */
+static void afs_do_give_up_callback(struct afs_server *server,
+				    struct afs_vnode *vnode)
+{
+	struct afs_callback *cb;
 
-		list_for_each_entry(vnode,
-				    &afs_cb_hash(server, &callbacks->fid),
-				    cb_hash_link) {
-			if (memcmp(&vnode->fid, &callbacks->fid,
-				   sizeof(struct afs_fid)) != 0)
-				continue;
+	_enter("%p,%p", server, vnode);
 
-			/* right vnode, but is it same server? */
-			if (vnode->cb_server != server)
-				break; /* no */
+	cb = &server->cb_break[server->cb_break_head];
+	cb->fid		= vnode->fid;
+	cb->version	= vnode->cb_version;
+	cb->expiry	= vnode->cb_expiry;
+	cb->type	= vnode->cb_type;
+	smp_wmb();
+	server->cb_break_head =
+		(server->cb_break_head + 1) &
+		(ARRAY_SIZE(server->cb_break) - 1);
 
-			/* try and nail the inode down */
-			inode = igrab(AFS_VNODE_TO_I(vnode));
-			break;
+	/* defer the breaking of callbacks to try and collect as many as
+	 * possible to ship in one operation */
+	switch (atomic_inc_return(&server->cb_break_n)) {
+	case 1 ... AFSCBMAX - 1:
+		queue_delayed_work(afs_callback_update_worker,
+				   &server->cb_break_work, HZ * 2);
+		break;
+	case AFSCBMAX:
+		afs_flush_callback_breaks(server);
+		break;
+	default:
+		break;
+	}
+
+	ASSERT(server->cb_promises.rb_node != NULL);
+	rb_erase(&vnode->cb_promise, &server->cb_promises);
+	vnode->cb_promised = false;
+	_leave("");
+}
+
+/*
+ * give up the callback registered for a vnode on the file server when the
+ * inode is being cleared
+ */
+void afs_give_up_callback(struct afs_vnode *vnode)
+{
+	struct afs_server *server = vnode->server;
+
+	DECLARE_WAITQUEUE(myself, current);
+
+	_enter("%d", vnode->cb_promised);
+
+	_debug("GIVE UP INODE %p", &vnode->vfs_inode);
+
+	if (!vnode->cb_promised) {
+		_leave(" [not promised]");
+		return;
+	}
+
+	ASSERT(server != NULL);